00:01:04 -!- Phantom__Hoover has quit (Quit: Leaving).
00:02:57 -!- Behold has changed nick to BeholdMyGlory.
00:03:19 -!- BeholdMyGlory has quit (Remote host closed the connection).
00:04:55 -!- BeholdMyGlory has joined.
00:05:35 <elliott> http://www.esolangs.org/wiki/ClearBF
00:05:40 <elliott> This project was a scholar project at ENSIAS - Morocco to build a compiler in C/Flex. The team is supervised by Mr. Karim Baîna and Mrs. Mounia ABIK. Members of the team are: [...]
00:05:52 <elliott> http://clearbf.wordpress.com/2011/01/20/clearbf-project-introduced-at-esolang/
00:05:56 <elliott> ClearBF Project introduced at Esolang !
00:05:56 <elliott> We have just announced the ClearBF Project in the Esolang official website. In fact, we’ve added a new page in the wiki to present our project.
00:12:31 <Sgeo> This sounds like the sort of thing I used to support
00:13:47 -!- cheater00 has joined.
00:16:40 -!- cheater99 has quit (Ping timeout: 240 seconds).
00:20:11 <Sgeo> Easy compilation to BF
00:20:12 -!- Wamanuz3 has quit (Remote host closed the connection).
00:20:29 <elliott> C->BF is interesting, Scheme->BF would be, anything else, naw :P
00:22:52 -!- Wamanuz has joined.
00:22:53 <Sgeo> There's a specific language with that goal in mind
00:23:22 <Sgeo> In connection with a brainfuck CPU iirc
00:23:56 <Sgeo> http://www.clifford.at/bfcpu/bfcomp.html
00:24:00 -!- poiuy_qwert has quit (Read error: Connection reset by peer).
00:25:36 -!- poiuy_qwert has joined.
00:31:59 -!- comex has changed nick to icefire.
00:43:59 -!- poiuy_qwert has quit (Read error: Connection reset by peer).
00:45:06 -!- poiuy_qwert has joined.
00:46:09 -!- BeholdMyGlory has quit (Remote host closed the connection).
00:46:41 -!- Wamanuz has quit (Remote host closed the connection).
00:48:08 -!- icefire has changed nick to comex.
00:51:15 <Sgeo> Maybe I should learn Lisp Flavored Erlang
00:54:54 <Sgeo> Or I should just learn Erlang
00:55:20 <Sgeo> How do I keep from being bored while looking at ooc?
00:56:07 <Sgeo> Better yet: Show me in what order I should read the guide
00:56:33 <elliott> make him feel the pain i feel
00:56:55 <Sgeo> elliott, surely you must have learned about all these languages somehow
00:57:07 <elliott> yes, reddit throws me at them.
00:57:20 <elliott> atomo for instance i looked at the front page once and decided it was shit.
01:00:12 <Sgeo> what about it?
01:00:40 -!- sftp has quit (Remote host closed the connection).
01:02:45 <nooga> i just looked at it
01:03:14 -!- sftp has joined.
01:09:10 -!- Mathnerd314 has joined.
01:15:21 <elliott> http://www.formauri.es/personal/pgimeno/misc/elpp.html
01:15:24 <elliott> Gregor: So what happened to WIKI!
01:15:34 <elliott> Exclamation marks in a project name, a sure sign of insanity.
01:16:40 <nooga> compilation to BF is completely useless
01:16:40 <Gregor> I turned "Wiki!" into Giki, maintained it for a few years, then it was superceded by Hackiki.
01:17:15 <nooga> since compiling BF to make programs run faster is pretty non trivial
01:18:03 <nooga> i should think about an array of primitive, super-fast hardware bf processors in FPGA
01:18:36 <nooga> then the compilers targeting brainfuck would gain some purpose
01:24:11 <elliott> Gregor: What language is Hackiki written in?
01:24:44 <Gregor> elliott: What little there is that is properly Hackiki is in PHP. The wiki software you actually see is Python.
01:24:57 <elliott> Gregor: PHP communicating with a chroot?
01:25:10 <Gregor> PHP does no communicating with a chroot.
01:25:19 <Gregor> plash does all the chrooting and shtuff.
01:25:32 <elliott> Gregor: PHP communicating with plash... nope, I'm still vomiting :P
01:25:49 <Gregor> exec("pola-run", "shtuff") OH NOSE
01:25:50 <elliott> Gregor: I swear plash is overkill >_>
01:26:02 <Gregor> Probably, but it's good fun 8-D
01:26:04 <elliott> Mostly due to the fact that, despite its huge amount of code, it's still primarily libc-only :P
01:26:36 <elliott> Which is like writing the cleverest, optimising Brainfuck interpreter ever... in sed, making it one of the slowest out there.
01:27:59 <elliott> Gregor: Also the Debian-specificity is lame, even if you're a Debian fan :P
01:28:14 <elliott> (Because "only works on Debian" = "YOU'RE GONNA HAVE TO DEAL WITH DEBIAN STUFF")
01:28:14 <Gregor> Yeah, that is kinda lame *shrugs*
01:28:36 <elliott> See botte has a Linux-portable layer for sandb...
01:28:55 <elliott> (In all srsness though, overriding syscalls isn't hard, see: anarchy golf, which has very good sandboxing :P)
01:29:57 <Gregor> I don't want to write something myself, and plash is trivially simple to use. There are other things that have their own system hacked in, but nothing so easily usable as plash that I've seen.
01:30:32 <elliott> Oh, certainly, it's not something you can fix without a project, I'm just saying that thankfully a project isn't too much work :P
01:31:00 <variable> pur-logicsolutions.com/.../White_Paper_7_Shortcuts_To_Lose_Your_Data_And_Probably_Your_Job_1_.pdf -->
01:31:17 <elliott> Gregor: OK, by "replacing syscalls" I didn't actually mean that.
01:31:26 <elliott> I meant overriding syscall().
01:31:45 <elliott> I forget how to handle statically-linked stuff :P
01:32:00 <Gregor> Do what plash does: Stick it in a chroot.
01:32:20 <elliott> Gregor: Yeah, but if you're happy relying on a chroot you can just use a chroot...
01:32:45 <Gregor> The chroot is just what it uses to set the lower-bound, I would NOT be happy with a chroot.
01:33:08 <elliott> Gregor: But you can reduce it to being just-a-chroot by calling syscalls directly or statically linking.
01:33:22 <elliott> Admittedly a rather empty one, but that's lame; you should be able to use a stock libc imo
01:33:33 <elliott> because plenty of stuff doesn't use libc :P
01:33:38 <elliott> OK, so not plenty of stuff, but...
01:33:45 <Gregor> elliott: Fine. Write the system that provides what I need doing that, and I'll use it.
01:33:55 <Gregor> Oh, doesn't exist? Actually a bitch to write? Well then I'll continue using plash.
01:33:58 <elliott> Gregor: I'm not criticising your choice, I'm thinking out aloud...
01:34:24 <elliott> I'm just sayin' that "I would NOT be happy with a chroot" is silly, because plash is trivially reducable to a chroot :P
01:34:45 <Gregor> It's reducible by attempting to be malicious.
01:34:55 <Gregor> 99.99999999% of programs will not see the empty chroot.
01:35:42 <elliott> Gregor: You mean benign programs aren't a security risk, but malicious programs are?
01:35:59 <elliott> I cannot think of a response :P
01:36:11 <Gregor> No, malicious programs can't do shit, but benign programs are happy.
01:37:05 <Gregor> Oh, I see what you're saying, I'm providing no greater protection than running the program in an empty chroot as a random user.
01:37:12 <Gregor> It's all convenience beyond that.
01:37:45 <elliott> Gregor: But there are programs that don't use libc... and programs you might want to link statically with another libc... etc.; OK, so they're rare, but plash is big enough that it should be platonically perfect :-P
01:37:54 <elliott> You could do it easily with a kernel module.
01:37:57 * Sgeo ponders Erlang web frameworks
01:38:07 <elliott> (Although writing to the syscall table is probably Unsupported I imagine it works.)
01:38:14 <Gregor> elliott: Doing it by a kernel module would be frikkin' awesome.
01:38:18 <Sgeo> <elliott> <some generic thing about being in pain>
01:38:28 <elliott> Gregor: Yeah, I am really tempted to do that now...
01:38:41 <elliott> The practice of replacing syscall table entries is frowned upon by
01:38:41 <elliott> Linus and the other kernel maintainers -- so much so that the
01:38:41 <elliott> sys_call_table symbol is no longer exported. This explains why your
01:38:41 <elliott> module can't find it. If you care to know more details, use google to
01:38:41 <elliott> find the (many and extensive) discussions about, for and against the
01:38:55 <elliott> Gregor: Woot, so it _is_ possible, just so frowned upon that I get to do fun hackery to manage to do it :-D
01:39:06 <elliott> [[1. Modify the kernel source to directly include your own system call
01:39:07 <elliott> 2. Modify the kernel source to export the sys_call_table symbol. (Of
01:39:07 <elliott> course, your modification won't ever make it into the official kernel
01:39:07 <elliott> source tree but if you're just trying to learn...)
01:39:07 <elliott> 3. Figure out how to dynamically locate the sys call table from your
01:39:09 <elliott> loadable module. (This is ugly and I don't recommend it, but it
01:39:11 <elliott> certainly seems feasible. Hint: where in kernel memory could you find a
01:39:13 <elliott> block of 230+ consecutive words, each of which contains a valid kernel
01:39:20 <elliott> Gregor: Since when has the Linux kernel policy been to stop people from shooting themselves in the foot???
01:39:40 <elliott> [[I am trying to develope recycle bin like thing in Linux.
01:39:41 <elliott> For that i need to override unlink sys call!.
01:39:41 <elliott> Anyways,It is working now!!
01:39:41 <elliott> > Anything that can be done by intercepting a system call can be done
01:39:41 <elliott> > another way. What are you trying to do?
01:39:41 <lambdabot> Not in scope: data constructor `Anything'Not in scope: `that'Not in scope: ...
01:39:42 <lambdabot> <no location info>: parse error on input `do'
01:39:46 <lambdabot> <no location info>: parse error on input `]'
01:39:47 <elliott> WORST REASON TO OVERRIDE SYSCALLS EVER
01:40:36 <elliott> Gregor: Hmm, couldn't you get most of the way by using FUSE, actually?
01:40:40 <Gregor> Time to see if SDL_mixer 1.2.11 does seamless looping of .ogg properly ...
01:40:49 <elliott> Network access can be restricted in other ways, inside a chroot not much else matters but the fs...
01:40:54 <elliott> At least that's what I'm thinking.
01:41:04 <Gregor> elliott: Plash doesn't handle networking.
01:41:12 <Gregor> elliott: Actually FUSE does sound like a nice solution.
01:41:33 <elliott> But yeah, I think FUSE would get you like 75% of what overriding syscalls would, without all the fuss.
01:41:44 <elliott> Since it's basically equivalent to overriding all filesystem syscalls...
01:41:47 <Gregor> Much more flexible too :P
01:42:15 <elliott> And the only syscalls are pretty much random local shit + filesystem + network + root-only shit that the unprivileged user can't use anyway :P
01:42:28 <elliott> Admittedly network would be nice to control...
01:42:45 <elliott> But there's not much control you could do anyway beyond blanket policies.
01:42:52 <Gregor> There is no netfuse :P
01:43:01 <elliott> Gregor: Err, how's that related?
01:43:04 <elliott> I mean controlling the network syscalls.
01:43:41 <Gregor> elliott: Well, a user-controllable tuntap (netfuse) + a net version of chroot = tada :P
01:44:11 <Gregor> YES YES YES SEAMLESS LOOPING
01:44:12 <elliott> Gregor: Hmm, I am confused. You mean create a virtual network device?
01:44:59 <elliott> Gregor: Just make the tuntap a named pipe or a FIFO or whatever the kids are calling it this day.
01:45:08 <elliott> Or even mount /dev as FUSE.
01:45:09 <Gregor> Yeah, I didn't quite think that one through :P
01:45:12 <elliott> And have it control your tuntap.
01:45:26 <Gregor> Anyway, you can't lock a process to a particular network device (AFAIK?)
01:45:45 <elliott> Gregor: Can't you "hide" eth0 somehow?
01:45:54 <elliott> Gregor: You know how you block network access in HackEgo?
01:45:57 <elliott> Just do it the same: firewall.
01:46:08 <elliott> Gregor: And you run the tuntap from /outside/.
01:46:33 <Gregor> So just use that instead of my HTTP proxy.
01:46:57 <elliott> Gregor: Plenty of legit non-HTTP connections to make :P
01:47:11 <elliott> Gregor: Hmm, what would the FUSE fs do, anyway, apart from protect certain files?
01:47:38 <Gregor> elliott: Plash is essentially a unionfs, which is nice since I can have a hackiki fs checked out to <wherever> always be at /hackiki
01:48:01 <Gregor> (A unionfs with its own security system of course)
01:48:14 <elliott> Gregor: Sure, it's just that I've just forgotten what plash actually protects again :P
01:48:17 <elliott> Help me out here, brainfart...
01:48:49 <Gregor> w.r.t. the filesystem, it's just about writability.
01:48:58 <Gregor> The rest comes from chroot+random-user.
01:49:45 <elliott> Gregor: Can't you just do that by setting the owner?
01:50:07 <Gregor> Yeah ... but that's global ...
01:50:35 <elliott> Gregor: Eh? What local permissions do you have in hackego/hackiki?
01:51:27 <Gregor> Pretty minor actually :P
01:51:41 <elliott> Gregor: Example? I'm really having trouble thinking here :P
01:51:52 <Gregor> Just read on /usr and friends, r/w to /hackiki which is really /tmp/<foo> and read/write to /tmp which is really /tmp/<bar>
01:52:27 <elliott> Gregor: So basically, it's unix permissions + mapping two directories to process-specific ones :P
01:52:35 <elliott> http://www.themonkeysyouordered.com/ OH MY GOD THIS IS THE BEST THING EVER.
01:52:39 <elliott> New Yorker cartoons with literal captions.
01:52:44 <elliott> http://www.kottke.org/plus/misc/images/literal-new-yorker.jpg
01:53:07 <Gregor> elliott: I used plash because it's easy, not because I needed all of it :P
01:53:26 <elliott> Gregor: Sure, sure, I just had this LAPSE OF JUDGEMENT where I considered THROWING IT ALL OUT :P
01:53:40 <Gregor> elliott: Note that although plash didn't invent empty chroots and random users, it sure makes them easy to use :P
01:54:04 <elliott> Gregor: Except that you don't NEED an empty chroot :P
01:54:05 <elliott> <Gregor> Just read on /usr and friends, r/w to /hackiki which is really /tmp/<foo> and read/write to /tmp which is really /tmp/<bar>
01:54:16 <elliott> The remapping is the only thing that doesn't fit into straight unix permissions there..
01:54:39 <Gregor> <Gregor> elliott: I used plash because it's easy, not because I needed all of it :P
01:54:56 <elliott> Gregor: Sure, sure, I realise that, you think I'm being antagonistic or questioning your approach or something
01:56:19 <Gregor> I am in no way stopping you from writing the simpler alternative :P
01:57:21 <Gregor> Watching really bad TV shows on Hulu: So awesome.
01:59:17 -!- myndzi has quit (Read error: Connection reset by peer).
02:03:00 -!- elliott has quit (Quit: Leaving).
02:06:54 <Sgeo> I think Erlang may make it too easy to write my own impure functions
02:24:16 -!- copumpkin has quit (Quit: Computer has gone to sleep.).
02:28:04 <Gregor> OK, SDL_mixer is actually pretty awesome.
02:28:10 <Gregor> For some reason I was convinced that it was really difficult to use.
02:28:18 <Gregor> But I just went from no music to perfect seamless looping in no time flat.
02:39:51 -!- azaq23 has quit (Quit: Leaving.).
02:51:32 -!- azaq23 has joined.
02:54:12 -!- azaq23 has quit (Client Quit).
03:04:00 -!- myndzi has joined.
03:11:40 -!- pikhq has quit (Ping timeout: 255 seconds).
03:22:20 -!- pikhq has joined.
03:24:34 -!- calamari has joined.
03:47:44 -!- calamari has quit (Quit: Leaving).
04:17:23 -!- benuphoenix has joined.
04:18:36 <benuphoenix> question: out of boredom, i decided to moniter the logs of my linode server. someone is trying to connect to it. what do i do?
04:19:45 -!- benuphoenix has quit (Client Quit).
04:43:03 <Mathnerd314> benuphoenix: hunt them down and **** with them until they stop
05:18:42 -!- sftp has quit (Remote host closed the connection).
05:19:11 -!- sftp has joined.
05:35:04 -!- copumpkin has joined.
05:53:13 -!- poiuy_qwert has quit (Ping timeout: 255 seconds).
05:54:18 -!- poiuy_qwert has joined.
06:37:40 -!- cheater00 has quit (Ping timeout: 240 seconds).
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:19:09 -!- Deewiant has quit (Ping timeout: 240 seconds).
08:19:16 -!- Deewiant has joined.
08:43:24 <Ilari> Apparently IANA is delaying it to create a media event...
08:57:03 <Ilari> Anyway, both main estimates are "today" now...
08:58:26 -!- poiuy_qwert has quit (Ping timeout: 240 seconds).
09:03:53 -!- poiuy_qwert has joined.
09:07:23 -!- cal153 has quit (Ping timeout: 240 seconds).
09:08:39 <pikhq> And APNIC is sitting at 1.67 /8s.
09:11:06 -!- cal153 has joined.
09:21:51 <Ilari> The new figures aren't out yet.
09:23:49 <Ilari> I think those figures appear in about 6 hours...
09:44:41 -!- GreaseMonkey has quit (Quit: ilua).
09:49:43 -!- myndzi has quit (Read error: Connection reset by peer).
09:50:17 -!- myndzi has joined.
10:48:50 -!- oerjan has joined.
11:05:54 -!- poiuy_qwert has quit (Quit: This computer has gone to sleep).
11:27:09 -!- cheater99 has joined.
12:00:41 -!- Tritonio has joined.
12:00:43 -!- cheater99 has quit (Ping timeout: 240 seconds).
12:07:47 -!- cheater99 has joined.
12:09:30 -!- azaq23 has joined.
12:27:30 <Ilari> That is, I think the figures update at about 15:15Z or so...
12:30:23 <Ilari> And then one will see if today APNIC allocated something like a /14 or something like a /9...
12:34:37 <Sgeo> "Warning: This document contains examples of bad code.
12:38:50 -!- Tritonio has quit (Quit: Leaving).
12:38:53 -!- j-invariant has joined.
12:43:05 -!- Tritonio has joined.
12:59:05 -!- cheater99 has quit (Ping timeout: 240 seconds).
12:59:28 -!- cheater99 has joined.
13:12:54 -!- Phantom_Hoover has joined.
13:17:49 <quintopia> the day everyone has their own IPv6 address is the end of the road for privacy. discuss
13:18:49 <j-invariant> quintopia: yeah might as well tattoo everyone with a barcode
13:19:01 <fizzie> Yes, just like everyone with just a single computer at home (with a single IP address) has no privacy nowadays.
13:21:59 <quintopia> friend of mine says there will be nowhere on earth that will legally allow anonymous proxies
13:23:23 -!- copumpkin has quit (Ping timeout: 250 seconds).
13:23:39 -!- Sgeo has quit (Read error: Connection reset by peer).
13:23:49 -!- copumpkin has joined.
13:24:48 -!- Sgeo has joined.
13:26:57 <Ilari> At least IPv6 has no digital signatures (at least outside IPSec). I heard some nutter say it has...
13:34:15 <fizzie> But IPSec's a mandatory part of IPv6, so it still counts.
13:34:44 <quintopia> so, raise your hand if you are jealous of Craig Rowin
13:41:52 <Ilari> NSA has sabotaged it enough... :->
13:43:46 <Ilari> (No, I'm not saying that there's a backdoor in the IPSec specs... There isn't.)
13:50:42 <Ilari> Putting backdoor in specs is quite stupid. There are better methods to sabotage it... :-)
14:01:04 <j-invariant> http://ninapaley.com/mimiandeunice/wp-content/uploads/2010/08/ME_160_Rivalrous1-640x199.png
14:28:39 -!- Phantom_Hoover has quit (Quit: Leaving).
14:29:07 -!- Phantom_Hoover has joined.
14:29:34 -!- poiuy_qwert has joined.
14:31:10 -!- j-invariant has quit (Quit: leaving).
14:38:55 -!- oerjan has quit (Quit: leaving).
14:38:57 -!- j-invariant has joined.
14:41:41 -!- poiuy_qwert has quit (Quit: This computer has gone to sleep).
14:55:29 <Gregor> Phantom_Hoover: Not one iota.
14:55:46 <Phantom_Hoover> Gregor, that is evidently why your trombute playing sucks,
14:57:41 <Ilari> IPv4 depletion mess: Now someone says that IANA does not plan to delay servicing the request (and thus main reason is APNIC somehow not sending the request).
14:58:51 <Ilari> Wow... Release date for DNF...
14:59:16 <Gregor> Phantom_Hoover: One of my great life plans is to make a series of instruments in which the note selection is by a keyboard, but the sound production is like the original instrument.
14:59:33 <Gregor> Phantom_Hoover: So a keyboard horn would be played by lip buzzing, but the note selected by a keyboard and the world's most complicated series of valves.
14:59:49 <Gregor> Kiol is probably the most impossible :P
15:00:47 <Phantom_Hoover> That's not vastly difficult, except for the difficulty of bowing it while playing the keyboard.
15:00:58 -!- Tritonio has quit (Quit: Leaving).
15:02:44 <Gregor> The problem, in my opinion, is mainly with making double-stops work in any way, and that essentially for it to make any physical sense you'd have to key with your right hand and bow with your /left/ hand.
15:05:15 <Gregor> (And bowing with your left hand is ... it's just wrong)
15:09:53 <Gregor> Phantom_Hoover: Yeah, but physically what does that do, and how do you bow it?
15:10:43 <Phantom_Hoover> (The pianoflute has issues with being either really tiny or several octaves too low to play.)
15:11:40 <Gregor> Phantom_Hoover: I think the trick to that one is that the flute is shorter than the keyboard, and the actual terminus to the flute proper is somewhere midway on the instrument.
15:12:53 <Gregor> ... transverse? You'd be holding up your right hand at a bit of an angle to get to the keyboard, and holding it with your left hand.
15:13:21 <Gregor> Alternatively it could go with the bass-flute route and be blown transverse, but then have an immediate bend and actually be held upright.
15:13:49 <Phantom_Hoover> Gregor, we appear to have come to an insurmountable disagreement.
15:13:58 -!- copumpkin has quit (Quit: Computer has gone to sleep.).
15:14:15 <Gregor> Phantom_Hoover: How is yours played :P
15:14:41 <Phantom_Hoover> Like a piano, except there's a mouthpiece to blow into.
15:14:50 <Gregor> Well that's a melodica.
15:14:59 <Gregor> It would be difficult-to-impossible to make that produce a flute-like sound.
15:15:24 <Gregor> Because it's not blown transverse ... you could make a recorderalike like that and have a fipple.
15:15:52 <Gregor> Or The World's Most Impossible Instrument, the keyboe, with a double-reed mouthpiece.
15:16:27 <Gregor> You could even go the clarinet or saxophone route with a single reed (which sound you'll get depending on the rest of the construction)
15:16:40 <Gregor> But for a flute ... that's really got to be transverse.
15:21:45 <variable> I have a question about regular languages. An article I'm reading defines them to be "those languages which can be matched in a single pass using a fixed amount of memory". Can someone give me an example of something which fails this?
15:22:12 <Phantom_Hoover> <Gregor> But for a flute ... that's really got to be transverse. ← yeah, that's the tricky bit.
15:29:10 <variable> could anyone please explain what http://oi56.tinypic.com/1qhocx.jpg means? Where does S2 go on B ?
15:30:07 -!- ais523 has joined.
15:30:29 <Phantom_Hoover> <variable> I have a question about regular languages. An article I'm reading defines them to be "those languages which can be matched in a single pass using a fixed amount of memory". Can someone give me an example of something which fails this? ← I'd parse that as "can be matched in finite time by an FSA"
15:30:45 <variable> j-invariant, how does it choose? randomly?
15:31:20 <variable> Phantom_Hoover, that makes a bit more sense
15:31:52 <j-invariant> variable: consider the string abbbbbba it can accept that
15:32:10 <j-invariant> variable: by going s0 -> s1 -> s2 -> s1 -> s2 -> ... -> s2 -> s3 -> s4
15:37:38 -!- oerjan has joined.
15:37:39 -!- ais523 has quit (Read error: Connection reset by peer).
15:38:30 -!- ais523 has joined.
15:42:36 -!- copumpkin has joined.
15:43:27 <j-invariant> If I write a program z = 3; w = f(x,y); f cannot destroy my value of z by accident
15:52:08 -!- elliott has joined.
15:52:13 <oerjan> IT'S THE FIIINAL COUNTDOOOWN
15:53:18 <oerjan> that was exactly what i expected it to be
15:53:29 <elliott> 02:13:44 <Sgeo> I think Erlang may make it too easy to write my own impure functions
15:53:37 <elliott> Sgeo: I thought you _liked_ that kind of shit.
15:53:46 <elliott> 04:25:30 <benuphoenix> question: out of boredom, i decided to moniter the logs of my linode server. someone is trying to connect to it. what do i do?
15:53:55 <elliott> set up a chroot and run ssh in it
15:53:59 <elliott> when they connect, they think they're root
15:54:04 <elliott> but every command does weird shit or mocks then
15:54:16 <elliott> I'm sorry Dave, I can't let you do that.
15:54:43 <elliott> 13:29:10 <quintopia> friend of mine says there will be nowhere on earth that will legally allow anonymous proxies
15:54:51 <elliott> quintopia: ah, Friend of Mine, the ultimate authority on everything
15:55:59 <oerjan> elliott: although i did briefly consider the opposite interpretation
15:56:16 <elliott> oerjan: using a car as ... a moustache?
15:56:35 <oerjan> a moustache shaped like a car
15:56:47 <oerjan> google seems to fail me
15:57:30 <elliott> 15:23:06 <Gregor> Or The World's Most Impossible Instrument, the keyboe, with a double-reed mouthpiece.
15:57:33 <elliott> Gregor: why does this not exist
15:57:46 <elliott> 15:28:59 <variable> I have a question about regular languages. An article I'm reading defines them to be "those languages which can be matched in a single pass using a fixed amount of memory". Can someone give me an example of something which fails this?
15:57:54 <elliott> variable: not commenting on whether that's a good definition
15:57:57 <elliott> variable: but consider a^nb^nc^n
15:58:05 <elliott> i.e. any number of as, the same number of bs, and the same number of cs
15:58:09 <elliott> you have to store the number of "a"s
15:58:14 <elliott> which is an unbounded natural
15:58:19 <elliott> and, therefore, takes arbitrary amounts of memory
15:58:25 <Gregor> That's not so much a definition as a consequence >_>
15:58:44 <elliott> 15:50:10 <j-invariant> it's forth stupid?
15:58:45 <elliott> 15:50:42 <j-invariant> If I write a program z = 3; w = f(x,y); f cannot destroy my value of z by accident
15:58:55 <elliott> j-invariant: in Haskell you can say "= undefined" why would you want that
15:59:00 <elliott> No Forther says that that is a feature :P
15:59:10 <elliott> You're treating that like it's a purported feature, not a caveat
16:01:11 <oerjan> variable: also matching parentheses is not regular
16:01:37 <elliott> oerjan: hey that means that you can't even implement finite-tape brainfuck on an FSA :D
16:01:41 <oerjan> *checking for proper matching of parentheses
16:01:46 <elliott> oerjan: because you can't parse every program
16:02:03 <elliott> oerjan: hmm, and yet finite-tape brainfuck isn't TC
16:02:08 <elliott> it's Turing-requiring but not Turing complete
16:02:39 <oerjan> elliott: parentheses matching doesn't require TC, just something within context-free
16:02:59 <elliott> well point is, it requires a more powerful automaton to run than it can actually "harness"
16:04:30 <oerjan> actually a^nb^n is enough for not being regular. a^nb^nc^n gives you not even context-free iirc
16:04:37 <oerjan> (stronger pumping lemma)
16:04:57 <elliott> hmm, someone needs to make a language based on a^nb^nc^n
16:05:03 <elliott> preferably, with almost no computational power :D
16:05:15 <oerjan> why does that ring a bell
16:06:25 <Ilari> a^n b^n c^n is context-sensitive but not context-free.
16:06:36 <Ilari> Same for a^n b^n c^n d^n
16:07:33 <oerjan> yeah adding letter doesn't help, context-free == needs at most linear memory
16:09:41 <Ilari> Fun task: Describe a language that is not even context-sensitive but is recognable by recursive algorithm.
16:10:02 <oerjan> and that's enough for most things you'd want to do in practice, even includes SAT (NP-complete)
16:10:50 <oerjan> Ilari: ML/Haskell type checking iirc
16:11:08 <oerjan> or possibly that's just on the border of unknown equality
16:12:59 -!- variable has quit (Quit: hit 60 second ping time - reconnecting).
16:13:09 <Ilari> oerjan: What's the complexity class of that (at least the bounds for the class)?
16:15:02 -!- invariable has joined.
16:15:19 <invariable> elliott, what does Turing-requiring mean?
16:15:32 <elliott> it's something i just made up and ignore me
16:16:17 <invariable> also a^nb^nc^n == a{n}b{n}c{n} in regex
16:16:34 <oerjan> Ilari: bah i'm having trouble googling it
16:16:38 <elliott> invariable: no that's not valid regex
16:16:49 <elliott> regex = REGular EXpressions, most regexp implementations are actually more powerful than regular languages
16:16:52 <elliott> but it's still mostly regular :)
16:17:15 <elliott> only perl regexes can do a^nb^nc^n i think
16:17:30 <ais523> elliott: discussion we had recently at work: not counting (??{}) and (?{}) (i.e. embedded Perl), are Perl regexen TC?
16:17:46 <ais523> I'm not entirely convinced Perl regexen /can/ do a^nb^nc^n
16:17:55 <ais523> (although they can definitely do a^nb^n, and regular expressions can't)
16:17:56 <elliott> ais523: I suspect you can do a^nb^nc^n with the same stuff you use to do nested parens
16:18:10 <ais523> elliott: you can't, nested parens can be done by a PDA, a^nb^nc^n can't
16:18:11 <invariable> elliott, when you say ^n you just mean repeating N times - right?
16:18:23 <ais523> invariable: yes, but the point is that it's the same n for each of them
16:18:23 <elliott> invariable: all Ns must be the same
16:18:33 <elliott> ais523: hmm, well you can do something similar: let a be (, b be (, and c be ))
16:18:35 <ais523> so aabbcc matches but aabbccc doesn't
16:18:41 <ais523> elliott: that checks that a+b = c
16:18:41 <elliott> then c's n must be a's n plus b's n divided by two
16:18:58 <elliott> ais523: which is close but not exact
16:19:03 <invariable> elliott, ah - so your defining N by the first pattern. I thought N was fixed.
16:19:06 <ais523> aha, I think you might be able to do it via zero-width assertions
16:19:08 <elliott> ais523: now, if you can somehow localise the checking that a and b are the same
16:19:17 <ais523> you could check that a+b = 2c and that a=b
16:19:21 <elliott> ais523: then that becomes a+a = b+b = 2c, and thus -- yup
16:19:28 <elliott> and thus 2a = 2b = 2c, ergo a=b=c
16:21:57 <oerjan> Ilari: ah, "An example of this phenomena is the complexity of Type Inference in ML which was shown. to be complete for EXPTIME in [HM]."
16:22:44 <elliott> ais523: I'm thinking you'd treat a as (, b as [ (just like ( but a different char), and c as ))
16:22:52 <Ilari> Ah, that's between context-sensitive and recursive.
16:22:58 <ais523> elliott: that doesn't work
16:23:11 <oerjan> Ilari: although i don't think it is known whether EXPTIME is larger than PSPACE so that _might_ not actually be an example
16:23:14 <ais523> a as (, b as [, c as [)
16:24:03 <oerjan> although the general beliefs would imply that it is
16:24:22 <elliott> ais523: right, i meant c as ))
16:24:28 <elliott> and b would be treated as ( for that stage
16:24:37 <ais523> that still doesn't prove that a=b, at all
16:24:49 <elliott> ais523: well, you'd do the first part by treating a as ( and b as )
16:25:05 -!- poiuy_qwert has joined.
16:25:20 <oerjan> Ilari: anything EXPSPACE-complete will certainly work, however
16:25:23 -!- cheater99 has quit (Ping timeout: 240 seconds).
16:25:31 -!- poiuy_qwert has quit (Remote host closed the connection).
16:25:36 -!- cheater00 has joined.
16:26:06 <oerjan> Ilari: "An example of an EXPSPACE-complete problem is the problem of recognizing whether two regular expressions represent different languages, ..."
16:27:29 -!- Wamanuz has joined.
16:30:16 <ais523> /^$|^(?=(?<x>a(?&x)?b)c+$)a+(?<y>b(?&y)?c)$/
16:30:20 <ais523> that took a bit of working out
16:30:23 <Ilari> Yes, PSPACE is strict subset of EXPSPACE, so EXPSPACE-complete problems are not in PSPACE.
16:30:23 <j-invariant> elliott: "Is it possible to write bug free programs in haskell" reply: "why would you want to do that"
16:30:27 <ais523> it could be golfed from there, I was trying to keep it readable
16:30:31 <elliott> j-invariant: I agree; who said that?
16:30:36 <elliott> ais523: now do it in cyclexa
16:30:46 * ais523 tries to remember Cyclexa syntax
16:31:08 <ais523> assuming I've remembered what : does correctly
16:31:19 <ais523> there is definitely an operator that does that, I might just have got the wrong character
16:31:20 <j-invariant> elliott: somone said " you can give formal proof your behavior" and I asked how and he said he didn't know
16:31:35 <elliott> j-invariant: Writing a bug-free program -- as in *writing* one -- is far, *far* more trouble than it is worth. Writing a buggy program, and then iteratively improving it, is far more productive and practical, 80% of the time, IMO.
16:31:40 <ais523> the Cyclex'as much simpler
16:31:47 <elliott> Of course formal verification is useful often, especially for data structure libraries and the like.
16:31:52 <elliott> XMonad has a few components formally proven, I believe.
16:31:56 <oerjan> Ilari: to be precise every context-sensitive language is in PSPACE, and some PSPACE-complete problems are, although not all since needing linear memory is not closed under polymial reductions
16:31:59 <ais523> as it has a "parse trees are equal" assertion
16:32:10 <j-invariant> "bug free software is by definition trivial"
16:32:18 <oerjan> *problems are context-sensitive
16:32:53 <ais523> elliott: the compilers I'm creating as part of my PhD are formally proven
16:32:59 <ais523> mostly because the proofs are more interesting than the compilers
16:32:59 <invariable> that reminds me of the famous knuth quote: "I have only proved this correct ...."
16:33:05 <ais523> in fact, they exist only in proof form atm
16:33:15 <ais523> and it isn't executable
16:33:16 <invariable> ais523, where are you doing your PhD ?
16:33:17 * oerjan thinks if he keeps adding *-corrections deciphering them might end up being NP-complete
16:33:29 <ais523> invariable: Birmingham University
16:33:30 <elliott> invariable: birmingham! ok i swear i will stop answering people's questions for them some day
16:33:54 <ais523> to be fair, you didn't answer my question for me
16:33:57 <ais523> because I'd already answered
16:33:59 <elliott> ais523: I typed it slowly to seem less creepy
16:34:04 <elliott> oerjan: I don't see why you'd think that
16:34:29 <j-invariant> elliott: anyone the sense I get form #haskell is don't bother writing correct programs
16:34:32 <ais523> oh look, P==NP has been proven again
16:34:40 <j-invariant> elliott: since they don't exist, your an idiot if you want a program that works, etc .
16:34:43 <ais523> I'd be more excited if P==NP and P!=NP weren't proved so often
16:34:44 <invariable> ais523, some day... :-) what is your thesis on?
16:34:46 <elliott> j-invariant: I don't agree.
16:35:06 <ais523> invariable: compiling software to hardware, via type systems
16:35:12 <elliott> j-invariant: But I believe writing mostly-correct programs and then iteratively improving them is far, far more productive in many, many cases than writing programs that are correct the first time.
16:35:18 <j-invariant> ais523: I don't know why so many people care, probably just because there is money on it
16:35:30 <elliott> because it's an important problem
16:35:32 <ais523> j-invariant: also because it's a really irritating problem, in a sense
16:35:32 <j-invariant> ais523: Nobody gave a shit with the BQP thing that was recently proved
16:35:37 <j-invariant> elliott: you're an important interesting problem
16:35:40 <ais523> if P does != NP, there's no obvious way to prove it at all
16:35:56 <ais523> I'd say the details of the proof would be more interesting than that for, say, Fermat's Last Theorem
16:36:00 <j-invariant> it's like "Hey im a complexity theorist when people prove HUGE resuts in the field"
16:36:15 <elliott> scott aaronson cares about everything!
16:36:22 * oerjan swats elliott -----###
16:36:26 <elliott> [[When people want to emphasize how pathetically far we are from proving P≠NP, they often use the following argument: for godsakes, we can’t even prove that NEXP-complete problems aren’t solvable by depth-3, polynomial-size circuits consisting entirely of mod 6 gates!
16:36:33 <ais523> at least the latest P == NP result is an algorithm for 3-SAT that actually has source available, so errors should be quite easy to find
16:36:39 <j-invariant> ais523: isn't that just because you understand computation much deeper than modular forms?
16:37:03 <j-invariant> personally, I know fine well that when it's proved I will not be able to understand a single word
16:37:12 <elliott> are there any interesting theorems that depend on fermat's last? :D
16:37:16 <j-invariant> I would have to spend decades studying this stuff to get to that level
16:37:21 <elliott> it seems completely useless, off the top of my head
16:37:28 <oerjan> ais523: the nice thing about P=NP proofs with algorithms is that it's easy to test them on actual hard cases
16:37:45 <elliott> oerjan: "Now we're just waiting for the A(G64,G64) constant to run out."
16:37:56 <ais523> ah, Slashdot has worked it out already, apparently it doesn't work in all cases
16:38:06 <ais523> so it's just an algorithm for solving special cases of 3SAT, which obviously doesn't prove P=NP
16:38:30 <elliott> ais523: slashdot, the home of complexity theorists
16:38:50 <ais523> elliott: the thing I love about Slashdot is, whatever the story, there will be exactly one comment by someone who really knows what they're doing that explains the whole thing
16:39:00 <ais523> and it's worth putting up with the other 500 or so comments of nonsense just to find that one
16:39:21 <ais523> (more than one person who knows what's really going on often reads Slashdot, but once one completely correct explanation is posted, the others don't bother)
16:39:27 <elliott> Sgeo will be so sad, Stargate Universe has been cancelled
16:40:12 <quintopia> elliott: it was an opinion of his obviously...i was asking for alternate opinions
16:40:14 <elliott> ais523: I'm surprised there isn't some super-secret invite-only version of Slashdot for just the people who post those kinds of comments :P (okay, not *very* surprised)
16:40:29 <elliott> quintopia: um i think it's a stupid opinion, everyone already has an IP address now, to some approximation
16:40:30 <ais523> elliott: the same people are the people who post all the junk on other stories, though
16:40:42 <elliott> you think making the two separate computers in your house have different IP addresses will change anything?
16:40:46 <ais523> elliott: that's obviously incorrect, there are more people in the world than there are IPv4 addresses
16:40:56 <elliott> ignoring things like Qatar's country-wide NAT, because they're edge-cases
16:40:58 <ais523> scary thought, isn't it?
16:40:59 <elliott> ais523: I mean people on the Internet
16:41:17 <ais523> now I'm wondering how Qatar /does/ NAT a whole country
16:41:23 <quintopia> elliott: i agree. but it can't be denied that there is some influence at work in the world to make piracy more difficult
16:41:27 <ais523> surely sometimes they have more than 65535 people trying to connect at once?
16:41:44 <elliott> quintopia: nice conspiracy, but IPv6 has been around for a LONG time
16:41:50 <elliott> before piracy fighting became the hot thing...
16:42:06 <elliott> RFC 2460 was published in 1998
16:42:08 <quintopia> this is not about IPv6. this is about anonymous proxies
16:42:15 <elliott> you said it in the context of ipv6
16:42:17 <ais523> what is this opinion? IPv4 being deliberately limited in order to prevent people torrenting?
16:42:28 <elliott> anyway, it is impossible to block all anonymous proxies
16:42:32 <coppro> IPv4 is far older than torrenting
16:42:33 <quintopia> well, i meant it as a completely different discussion
16:42:34 <elliott> because setting one up is trivial
16:42:47 <elliott> 13:25:00 <quintopia> the day everyone has their own IPv6 address is the end of the road for privacy. discuss
16:42:47 <elliott> 13:26:00 <j-invariant> quintopia: yeah might as well tattoo everyone with a barcode
16:42:54 <ais523> coppro: I know, but elliott implied that it was a stupid opinion, and I was trying to think up a suitably stupid opinion that fit in the context of the conversation
16:42:57 <elliott> first line is wrong, second line is wrong
16:43:10 <ais523> elliott: second line is arguably sarcasm
16:43:17 -!- copumpkin has quit (Ping timeout: 265 seconds).
16:43:24 <quintopia> the first line is also the opinion of the friend i disagree with
16:43:27 <ais523> there were only two lines
16:43:33 <elliott> quintopia: your friend is stupid :D
16:43:42 <elliott> <elliott> 13:25:00 <quintopia> the day everyone has their own IPv6 address is the end of the road for privacy. discuss
16:43:42 <elliott> <elliott> 13:26:00 <j-invariant> quintopia: yeah might as well tattoo everyone with a barcode
16:43:42 <elliott> <elliott> 13:26:11 <fizzie> Yes, just like everyone with just a single computer at home (with a single IP address) has no privacy nowadays.
16:43:51 <ais523> something seems wrong with your copy/paste
16:44:00 <elliott> <elliott> <elliott> 13:25:00 <quintopia> the day everyone has their own IPv6 address is the end of the road for privacy. discuss
16:44:00 <elliott> <elliott> <elliott> 13:26:00 <j-invariant> quintopia: yeah might as well tattoo everyone with a barcode
16:44:00 <elliott> <elliott> <elliott> 13:26:11 <fizzie> Yes, just like everyone with just a single computer at home (with a single IP address) has no privacy nowadays.
16:44:13 <elliott> ais523: you have some ignore or something
16:44:33 <ais523> elliott: herobrine only saw three
16:44:46 <ais523> and it's (4) that I'm missing
16:44:48 -!- copumpkin has joined.
16:44:51 <elliott> umm i wonder if i have some script that's broken
16:44:59 <elliott> * *** Message to #esoteric throttled due to flooding
16:45:17 <elliott> <elliott> <elliott> <elliott> 13:25:00 <quintopia> the day everyone has their own IPv6 address is the end of the road for privacy. discuss
16:45:20 <elliott> XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
16:45:24 <elliott> <elliott> <elliott> <elliott> 13:26:00 <j-invariant> quintopia: yeah might as well tattoo everyone with a barcode
16:45:26 <ais523> it makes more sense than the several minutes of fakelag that flooding used to give you
16:45:27 <elliott> XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
16:45:31 <elliott> <elliott> <elliott> <elliott> 13:26:11 <fizzie> Yes, just like everyone with just a single computer at home (with a single IP address) has no privacy nowadays.
16:45:34 <elliott> XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
16:45:38 <elliott> <elliott> <elliott> <elliott> 13:29:10 <quintopia> friend of mine says there will be nowhere on earth that will legally allow anonymous proxies
16:45:40 <elliott> XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
16:46:01 <ais523> the last one seems a little implausible just due to the instability of various countries
16:46:11 <ais523> Somalia is actually an anarchy IIRC, it legally allows anything because there are no laws
16:46:18 <elliott> no, somalia has a government now i think
16:46:37 <ais523> I think its populace would certainly want a government, at least
16:46:37 <elliott> ais523: also, that's a strange interpretation; it (was/is) locally ruled by mobs
16:46:45 <ais523> anarchy tends not to work too well in practice
16:46:52 <ais523> elliott: indeed, that's what happens in an anarchy
16:46:57 <ais523> I wouldn't call that law, though
16:47:07 <elliott> ais523: you're strawmanning anarchy here
16:47:11 <elliott> Somalia has never been an anarchy
16:47:26 <elliott> ais523: it's like telling a socialist "hey, go and live in soviet russia LOL if you think that's so great LOL"
16:47:28 <ais523> at what point does a mob become a government?
16:47:36 <elliott> when it eliminates all the other mobs
16:49:59 <ais523> going in the other direction, at what point does a government become a mob?
16:50:16 <ais523> ooh, idea: you know how you treated B Nomic as an esolang a while ago?
16:50:22 <ais523> I wonder if you could do that with real-world governments
16:50:22 <elliott> ais523: when other mobs appear that manage to override its power in some cases
16:50:39 <elliott> ais523: (arguably, governments are just a certain /type/ of mob, and so they never cease to be one)
16:53:38 <ais523> also, Duke Nukem Forever now has a specific release date (May 3)
16:54:05 <oerjan> afair the "mobs" in somalia weren't without law, it's just that their law is an even-more-than-average fucked up version of sharia
16:54:16 <elliott> right, somalia wasn't lawless, just fractured
16:54:31 <oerjan> (i.e. punishing women for _being_ raped)
16:54:42 <elliott> there _was_ that thing about their market being so free that they had the absolute lowest SMS prices in the world, but that's not quite worth the rest of the situation there :)
16:54:54 <elliott> (and also ignores all the /downsides/ of a totally unrestricted economy)
16:55:29 <elliott> I can't believe that in the US it costs to _receive_ a text message
16:55:55 <quintopia> i think that model is going out of style elliott
16:56:24 <ais523> in the UK, it never costs to receive a call or text
16:56:44 <ais523> except when the other person explicitly reverses the charges, in which case the operator phones you up and asks if you're willing to accept a reverse charge call
16:56:56 <quintopia> in the US, wireless companies are trolls and milk you for whatever money they can get
16:57:28 <ais523> (or in a few cases like the one we have at home, where in addition to our normal phone number, we have a reverse charge phone number where we pay for anyone phoning it; it's stupidly long to prevent anyone guessing it, and exists so that I can call home via payphones in emergencies)
16:57:51 <elliott> ais523: Couldn't you just reverse the charges in the normal way?
16:57:54 <ais523> elliott: in the UK it's less than in the US because there's lots of competition, although instead of giving people good deals, the wireless companies mostly just try to confuse the hell out of everyone
16:58:02 <elliott> I don't actually know how that works here though
16:58:08 <ais523> you have to contact the operator and ask
16:58:16 <ais523> and with an operator involves, things get more expensive
16:58:26 <elliott> Yeah, but it's for /emergencies/ :P
16:58:32 <ais523> I used to regularly phone home reverse-charge from school, not for emergencies at all
16:58:39 <ais523> the reverse-charge number was my usual method of contact
16:58:44 <ais523> because I didn't have a mobile, and didn't want one
16:59:03 <ais523> the number still works AFAIK, but I hardly ever end up needing to use a payphone nowadays, mostly because most of them no longer exist
16:59:16 <ais523> but I've needed to a couple of times in emergencies, and typed in the whole number and it still works
16:59:27 <ais523> (the number starts 0800, so it works from payphones too)
16:59:29 <elliott> So what is it, 0123456789876543210?
16:59:44 <ais523> nah, it's 23 digits long
16:59:57 <ais523> and at that length, unlikely to be guessed
17:00:25 <ais523> (the first 10 are the same for all reverse charge numbers, though, and follow a pattern; the other 13 provide the entropy)
17:02:28 -!- Oklopol has joined.
17:02:45 <elliott> http://yro.slashdot.org/story/11/01/20/2028252/Facebook-Images-To-Get-Expiration-Date err, no thank you
17:03:03 <elliott> prediction: X-Pire will never be implemented for Linux
17:03:04 <j-invariant> SECURITY EXCEPTION: Please type in your mothers madien name:
17:03:16 <elliott> causing Linux to be useless on Facebook (OK, this one is a stretch, it does not appear to be associated with facebook at all)
17:03:32 <ais523> wow that's a pretty stupid idea
17:03:39 <ais523> what's to stop people just saving the key with the image?
17:03:48 -!- Oklopol has changed nick to oklopol.
17:03:48 <elliott> ais523: worse is the Slashdot-quality reporting, which suggests that it's in any way affiliated with Facebook
17:04:03 -!- Wamanuz has quit (Ping timeout: 265 seconds).
17:04:14 <ais523> what I assumed that would mean would be that facebook would automatically delete images after a while
17:04:23 <j-invariant> elliott: what the hell that makes no sense
17:04:35 <ais523> which would actually be useful, but which would never happen because facebook users probably use it for the purpose of keeping the photos around forever
17:04:43 <elliott> It's not cryptography, it's CRYPTOLOGY! Like cryptography but less scientific!
17:04:48 <j-invariant> unless the "date" is actually a password that comes from facebook.org
17:04:50 <elliott> We used astrological measurements to invent the scheme!
17:05:18 <ais523> IIRC the Artemis Fowl books needed something like that for plot reasons, they invented a sort of virus that physically lived on the bits, being copied along with them, and degraded them over time
17:05:26 <elliott> ais523: thus solving piracy!
17:05:28 <ais523> which is good enough for fiction, I suppose
17:05:35 <ais523> and more plausible than most of the DRM schemes people come up with
17:06:02 <quintopia> sounds like something from discworld...
17:06:05 <ais523> at least it isn't obviously completely broken, the only problem is the physical impossibility because bits don't work like that
17:06:38 <elliott> but what happens if you use the analogue hole :)
17:06:41 <elliott> does it INFECT the result?
17:07:24 <ais523> elliott: the character in the book used the analog hole specifically to get a permanent copy before it degraded completely
17:07:28 <elliott> ais523: you're a pope, ais523
17:07:35 <elliott> G. is asking you if you're ready
17:07:47 <ais523> I should look up what that means
17:08:08 <j-invariant> Viewing these images requires the free X-Pire browser add-on. Currently only a version that works with Firefox is available. Those without the viewer will be unable to see any protected image.
17:08:08 <elliott> I wonder if you're a dictator
17:08:18 <j-invariant> how are these people able to call themselves "researchers"??
17:08:38 <ais523> in the book, the virus was used in order to transmit a message without the source being traceable, the idea being that the analog hole would save the video, but not the source information
17:08:47 <ais523> j-invariant: you can do research into engineering, I suppose
17:08:58 <j-invariant> ais523: it's not a research problem to write a stupid little plugin :/
17:09:11 <ais523> the research is probably the DRM algo
17:09:14 <j-invariant> this makes zero sense, they are using "research" to sound authoratative
17:09:22 <elliott> plenty of research involves engineering to demonstrate its effectiveness
17:09:22 <ais523> and the plugin just an implementation
17:09:35 <ais523> like in my PhD where the research is the algos for contructing the compiler, and the engineering is the compiler itself
17:09:46 <ais523> OK, that spam is pointless
17:09:47 <elliott> ais523: hmm, I vaguely recall that virus thing (I read the books, but many years ago)
17:09:56 <ais523> subject: xvff85; content: wq5
17:11:27 <ais523> hmm, the pope rules still don't actually do anythign
17:12:13 <ais523> also, I think I'm platonically ready
17:12:19 -!- poiuy_qwert has joined.
17:12:30 <ais523> popes are defined to be ready, that's what G. was referring to
17:12:55 <elliott> we have a policy that we are ready
17:13:11 <elliott> hmm, I think I'm going to read every single log of #esoteric, in order, from the first day to the last
17:13:32 -!- poiuy_qwert has quit (Client Quit).
17:14:01 <j-invariant> to be brutally honest, I don't *beleive* in timing attacks
17:14:24 <elliott> your belief is incorrect... depending on your definition of timing attacks
17:14:32 -!- pumpkin has joined.
17:14:37 <ais523> elliott: got a copy of the Vladivostok Telephone DIrectory handy, btw?
17:14:49 <j-invariant> oh wait a second TCP timing attack makes sense
17:15:02 <ais523> j-invariant: don't believe they can work? don't believe they exist? don't believe they're useful?
17:15:06 <ais523> don't believe anyone tries them?
17:15:13 <elliott> ais523: http://www.nomer.org/vladivostok/
17:15:14 <j-invariant> ais523: I did not belive anyone uses them successfully
17:15:46 -!- copumpkin has quit (Ping timeout: 246 seconds).
17:15:53 <ais523> elliott: it's now a requirement (not just a SHOULD) to read the vladivostok telephone directory in the first week of February
17:15:59 <ais523> because people were getting bored a couple of months ago
17:15:59 <j-invariant> if something takes 34 microseconds that means what? You would have to know my CPU and how much other stuff I was computing at the same time etc
17:16:14 <elliott> ais523: hmm, can I NoV you for that?
17:16:19 <ais523> perhaps we should change it back before the first week of February actually happens
17:16:34 <ais523> elliott: oh, I'm going to find a few phone numbers in Vladivostok and read those, it's close enough
17:16:39 <ais523> also, you should NoV Yally, e's the Pariah
17:16:49 <elliott> ais523: very insufficient, also, what does pariah do again?
17:17:06 <ais523> pariah wins if e goes long enough without accumulating Rests
17:17:09 <ais523> the idea is that everyone picks on them
17:17:18 <ais523> for even really minor rules breaches
17:17:30 <elliott> then i'll just NoV everyone
17:17:42 <ais523> you can't, NoVs are rate-limited
17:17:45 <elliott> (IMO, the burden of proof is on YOU to prove that you obtained and read the Vladivostok telephone directory :))
17:18:04 <ais523> elliott: except it isn't, the burden of proof is on the judge
17:18:12 <ais523> or, well, the person bringing the NoV, indiretly
17:18:20 <elliott> oh, I just mean that I'm not going to bother to /ask/ anyone before NoVing them
17:19:27 <ais523> hmm, this one's shorter: http://pozvoni.net.ua/eng/yadro/179
17:19:38 <ais523> and has a funky sidebar
17:20:10 <quintopia> i'm glad i am not participating in this nomic
17:20:38 <ais523> it isn't normally, just people were getting bored
17:21:10 <elliott> quintopia: but it's the longest-running Nomic by far!
17:21:19 <elliott> in fact, it's heading towards its 18th birthday
17:21:25 <elliott> although i wonder if it won't almost die before then
17:22:11 <quintopia> surely it's had winners in that time...
17:24:07 <elliott> quintopia: winning doesn't end the game
17:24:26 <ais523> I've won over ten times now, I think
17:24:42 -!- asiekierka has joined.
17:26:01 <elliott> 05.10.20: Lisp/Scheme program text
17:26:39 <elliott> we also have java, pascal and c++ programs as logs
17:27:01 <ais523> file(1) is pretty bad at identifying programming languages
17:27:50 <elliott> Hey, cpressey was here in '05.
17:28:14 <elliott> 12:42:06 <GregorR-L> Where's {^Raven^} :P
17:28:15 <elliott> 12:42:17 <GregorR-L> Whatever happened to {^Raven^}...
17:28:15 <elliott> 12:42:37 <Robdgreat> nevermore.
17:28:22 <elliott> 12:43:31 <GregorR-L> Maybe the two working ones exploded :-P
17:29:34 <ais523> now I'm trying to remember when I joined #esoteric
17:29:42 <ais523> I know I was active on the wiki first, and found the channel via the wiki
17:31:05 <elliott> 07.01.15:09:15:34 --- join: ais523 (n=chatzill@chillingi.eee.bham.ac.uk) joined #esoteric
17:31:05 <elliott> 07.01.15:09:16:16 <ais523> So there are people on #esoteric at the moment after all, then? I was monitoring the logs to see if anyone was online, but somehow I never seem to be online at the same time as other people...
17:31:10 -!- impomatic has joined.
17:31:19 <ais523> <bjourne> There are no 800 pound gorillas. Very obese gorillas held in captivity may top out at 600 lbs at most. A healthy, strong, alpha gorilla out in the wild would weigh no more than 350-400 lbs. 800 pound gorillas are pure fantasy.
17:31:22 <elliott> can i be the channel's official archivist, that would be nice
17:31:39 <ais523> elliott: ah, that hostname brings back memories
17:31:51 <elliott> 12:58:02 <EgoBot> ^AACTION is not!^A
17:31:58 <elliott> (those are literal ^As in the log)
17:32:09 <ais523> also, I've been here just over three years, it feels like longer
17:33:10 -!- invariable has quit (Remote host closed the connection).
17:33:13 <ais523> in other news, the video formats war is getting steadily more ridiculous
17:33:25 <elliott> 05:21:54 <ihope> I did Bitwise Cyclic Tag in Excel.
17:33:25 <elliott> 05:22:09 <ihope> And I don't mean a macro or anything.
17:33:26 <elliott> 05:22:21 <oklopol> with cell arithmetic?
17:33:27 <ais523> Microsoft released a plugin for Firefox on Windows that makes it do H.264 in the <video> tag
17:33:40 <ais523> elliott: it's doable with cell arithmetic, in both Excel and Works
17:33:55 <elliott> ais523: here's the standard Loper OS Malcontent Contrarian Opinion on the video format wars for you: http://www.loper-os.org/?p=309
17:34:09 <ais523> I used to spend a lot of time finding things on my Windows computer that were programmable, the lack of any sensible programming languages meant I couldn't use them
17:34:13 -!- variable has joined.
17:34:14 <elliott> worth a read but strangley irritating, like all loper posts :)
17:34:26 <ais523> wow, that one didn't try to set a cookie
17:35:00 <elliott> doesn't seem to have any tracking
17:35:36 -!- variable has quit (Remote host closed the connection).
17:35:48 <ais523> hmm, that blog is seriously advocating shipping executables rather than videos
17:35:55 <elliott> ais523: wait, "that blog"?
17:35:57 <elliott> _surely_ you know of Loper OS
17:35:59 -!- variable has joined.
17:36:29 <ais523> hmm, that's like the way my mother argues
17:36:42 <elliott> well because i link to the blog constantly and because it's been talked about at length in here
17:36:53 <ais523> I'm not sure if it's contrariness or disbelief
17:37:09 <elliott> Loper is the standard vapourware Lisp Machine revival OS with orthogonal persistence and all that
17:37:13 <ais523> elliott: well, that shows how rarely I click on links
17:37:21 <elliott> stanislav's our friend even if he is a bit barmy :)
17:37:42 <elliott> ais523: FWIW, he's basically right, if you consider it in the context of the rest of Loper
17:37:51 <ais523> elliott: Stallman would never agree to it
17:37:53 <elliott> in that the programs being shipped would be S-Expressions
17:38:02 <elliott> ais523: no, he isn't advocating shipping binaries
17:38:04 <ais523> he's been going on a crusade against closed-source JavaScript on webpages recently
17:38:07 <elliott> ais523: he's advocating shipping source which is executed directly
17:38:11 <ais523> on the basis that it's running on his computer
17:38:25 <elliott> ais523: Loper is an OS on top of a Lisp CPU emulated in x86-64, essentially
17:38:44 <elliott> ais523: what he's saying is more: the ability to access objects over the network transparently
17:38:56 <elliott> it's just that, obviously, objects can execute code
17:39:00 <ais523> this reminds me of my concept of an OS where everything is a function, even the files on disk
17:39:16 <ais523> so you have self-decompressing files that are indistinguishable from the unrolled version
17:39:22 <ais523> I think it'd be a security nightmare, though
17:39:39 <ais523> as if transparency was done properly, you could have, say, a file that expired a month later
17:39:45 <ais523> (although it could be copied up to that point)
17:39:46 <elliott> ais523: why a function? that's irrelevant to the concept
17:39:50 <ais523> or a file that was different on Tuesdays
17:39:55 <ais523> elliott: in the sense that it's opaque
17:39:57 <elliott> in fact, none of the lines you said make any sense at all, pretty much
17:40:00 <ais523> you can feed it arguments to get return values
17:40:05 <ais523> but you can't do anything else
17:40:27 <elliott> ais523: that's stupid, any decent os lets you view the code of functions
17:40:44 <ais523> elliott: it's obviously stupid, I didn't say it was a /good/ idea
17:40:46 <elliott> hmm, I'm going to mention the words "orthogonal persistence" now, and ais523 won't know what it means
17:40:49 <ais523> just that it was an idea
17:40:54 <ais523> elliott: yes I will, because you used to go on about it a lot
17:41:06 <ais523> I'm not sure if it was in that context or another one
17:41:11 <ais523> elliott: I haven't heard you go on about it recently
17:41:17 <elliott> I go on about @ all the time
17:41:22 <ais523> probably just missing each other online, or whatever
17:41:29 <ais523> you haven't even mentioned @ all that often
17:41:29 * elliott is still trying to process the fact that ais523 doesn't know what Loper is
17:41:46 <ais523> it didn't take you long to explain
17:41:47 <elliott> ais523: umm, do you know what LoseThos is?
17:41:59 <elliott> also, maybe not, but even Vorpal has read I think most of the posts :)
17:42:02 <ais523> it's an operating system, I forget the details beyond there
17:42:06 <elliott> *maybe not [take long to explain],
17:42:12 <elliott> you knew about LoseThos and not Loper
17:42:12 <ais523> I think I maybe knew what it was once
17:43:11 <elliott> that OS I managed to install once, yes!
17:43:39 <elliott> ais523: year-old proggit submission of the Loper OS blog: "Loper OS: every layer of abstraction by anyone else is crap"
17:43:43 <ais523> elliott: how many OSes are there which are large enough to be used by more than, say, 10 people, and that you've never installed?
17:43:58 <elliott> ais523: are there any OSes used by more than 10 people and less than a thousand?
17:44:06 <elliott> general-purpose ones, I mean
17:44:12 <ais523> is Plan 9 used by 1000 people?
17:44:28 <ais523> I have no idea how big it is
17:44:34 <elliott> it runs on supercomputers and did ...something at the 2000 olympics
17:44:42 <elliott> ok, one supercomputer, but it's IBM's
17:44:51 <elliott> and that probably has a lot of users
17:44:59 <elliott> the mailing list is fairly active
17:45:06 <elliott> let's say a few hundred users of plan 9, then
17:45:26 <elliott> memetech.com is still down :/
17:45:30 <j-invariant> I need strength not to rejoin haskell another day
17:45:57 <pumpkin> j-invariant: I can ban you if you need help :)
17:46:09 <elliott> YES BAN HIM BAN HIM FOREVER BAN HIM WITH FIRE MWAHAHAHAHAHAHAHA
17:46:36 <elliott> ais523: fwiw, @'s "ideal" model of How The Web Should Be is similar to that blog post :)
17:46:53 <elliott> ais523: in that, it's as simple as "you can transparently access objects over a network and they're cached locally"
17:46:58 -!- aloril_ has quit (Read error: Operation timed out).
17:47:07 <elliott> ais523: and then you'd have some kind of standard "document" type which can include hyperlinks to objects
17:47:13 <elliott> and also include other objects (such as images)
17:47:18 <elliott> you wouldn't even need that type locally
17:47:24 <elliott> since it'd come with the page, so to speak
17:47:37 <ais523> that wouldn't be the web, then, it'd be something else
17:47:41 <elliott> you could link to anything, even, say, an application (although @ has little concept of that)
17:47:47 <elliott> ais523: it wouldn't be The Web, but it'd be a Web
17:47:59 <elliott> ais523: heck, it's closer to the original conception of the Web than The Web is
17:48:09 <elliott> hyperlinked documents and resourecs
17:48:21 <elliott> oh, disturbingly, this'd actually do java web start /correctly/
17:48:21 <ais523> wasn't the original concept of the web to host FAQs to stop people repeatedly emailing duplicate questions?
17:48:31 <elliott> in that, you could just link to, say, a game
17:48:42 <ais523> elliott: the real problem with java web start is that it requires people who don't know what they're doing to do something that people who do know what they're doing repeatedly tell them not to do
17:48:44 <elliott> ais523: er, no, not that I am aware of
17:48:55 <elliott> it started out as research
17:48:55 -!- Wamanuz has joined.
17:49:07 <elliott> ais523: ok, managed to parse that :)
17:49:08 <ais523> elliott: well, suppose I send you a .jnlp link
17:49:20 <elliott> ais523: with @, every single thing runs in a total sandbox, pretty much
17:49:20 <ais523> in theory it's completely safe to download and run it
17:49:27 <pumpkin> j-invariant: that isn't an implication that I want you banned or not to return, by the way :P I don't think you should leave. But if you really don't want to return, I can help with that
17:49:44 <elliott> ais523: you can do this without it being a usability nightmare, because it doesn't share the stupid filesystem concept that holds back everything else's security
17:50:14 -!- aloril_ has joined.
17:50:17 <elliott> ais523: a random program being able to edit some other data you own, in @, is equivalent to, say, an OS that lets you run the C code "*0x834 = 9;"
17:50:21 <ais523> elliott: what to people use in order to store and find the equivalent of files?
17:50:29 <elliott> ais523: you don't, you just store objects
17:50:35 <ais523> I mean, how do you find them?
17:50:42 <elliott> ais523: by having an indexer run over them
17:50:48 <elliott> ais523: or by organising them into lists and sets
17:50:52 <elliott> ais523: or by organising them into lists and sets
17:50:58 <ais523> that's probably a little better
17:51:03 <elliott> what's wrong with indexing
17:51:11 <ais523> it assumes that the files in question have useful content
17:51:28 <elliott> you would store an image as an Image object
17:52:07 <ais523> but I mean, suppose I want to start writing a paper for a journal, and I haven't worked out what content I want yet, but the journal demands a particular format for submissions
17:52:16 <ais523> so I start off by creating a document that has that format, but no content
17:52:29 <ais523> how would an indexer know what that was meant to be, unless I told it explicitly?
17:52:42 <elliott> ais523: just title it that?
17:53:01 <elliott> ais523: I mean, you could even have a command "describe-object"
17:53:04 <ais523> but then it might get muddled with different documents that did have the content, but not the formatting
17:53:05 <elliott> so you'd do the equivalent of the Emacs:
17:53:16 <elliott> M-x describe-object RET journal paper 2010
17:53:33 <elliott> (if you say "but i can't think of a good name", well, what would you call the _file_? this doesn't introduce any new problems)
17:54:13 <ais523> what I'd normally do would be to organise things hierarchically, i.e. on a normal filesystem I'd have a directory that held together everything related to the subject
17:54:29 <elliott> say the subject was frobnicating
17:54:34 <ais523> and the start of the paper would just have a meaningless name like "paper" because the rest of the system gave the context
17:54:39 <elliott> I'd do "M-x describe-object RET frobnicating paper"
17:54:45 <elliott> ais523: sure, you could easily do that
17:54:57 <ais523> elliott: hmm, that goes at odds with your own opinions from a while ago
17:55:04 <elliott> ais523: just have a set object filled with pointers to relevant objects
17:55:11 <elliott> ais523: well, I'm trying to tell you how the system could work in the way you want it
17:55:13 <ais523> you said that you thought keyword-based systems were bad for web browser bookmarks
17:55:27 <elliott> ais523: this isn't keyword-based, it'd store the text you enter raw
17:55:32 <ais523> and preferred a relatively flat hierarchy
17:55:39 <elliott> ais523: I do, but evidently you don't
17:55:51 <elliott> and we're talking about how you'd use it, I believe
17:55:52 <ais523> elliott: it's /effectively/ keyword-based, as that's how you'd search the system
17:56:00 <ais523> elliott: actually, for web browsers, I just memorise URLs
17:56:18 <elliott> ais523: not really, you could easily stick a set called "relevant stuff" somewhere and e.g. have a pointer to the frobnicating set there
17:56:29 <ais523> and bookmark things only if I rarely access them
17:56:32 <elliott> ais523: this is effectively a regular filesystem but much much more flexible
17:56:35 <ais523> sort-of the other way round from most people
17:56:35 <elliott> because objects aren't "inside" directories
17:56:38 <elliott> directories are just sets of pointers
17:56:47 <ais523> elliott: hmm, didn't we discuss something like that for plan 10?
17:57:04 <elliott> ais523: plan 10 is just the implementor's wimpmode of @ times a billion
17:57:11 <ais523> besides, directories are just sets of pointers even on UNIX
17:57:13 <Ilari> http://inetcore.com/project/ipv4ec/en-us/index.html ... Allocation threshold fun (IANA hasn't allocated yet...)
17:57:28 <ais523> that's how hardlinks work
17:57:30 <elliott> ais523: (consider that @ implementation is currently blocked due to some very difficult compiler research)
17:57:40 <elliott> also, sure, but it's a pain to do things like this in unix :)
17:57:49 <ais523> hmm... I wonder if UNIX would be better if it were designed to work well with pervasive hardlinking?
17:58:01 <elliott> ais523: (as in, more research has to be done before I can know how to implement @ well)
17:58:13 <ais523> elliott: but my PHD's about something else
17:58:26 <elliott> ais523: well, that's your fault!
17:58:29 <ais523> (note: this is probably irrelevant)
17:59:35 <elliott> ais523: so clearly, your phd should be about lazy specialisers
17:59:48 <elliott> or if you want to do /my/ part of the job for me as well, lazy, parallel specialisers
18:00:03 <ais523> elliott: at least you'll be able to compile it into hardware when you're finished!
18:01:14 <elliott> 09:45:05 <ais523> Yes, I'm enjoying esolangs. I enjoyed the logs, too, before I had access to an IRC client.
18:01:29 * elliott tries to figure out what ais523-y reason made ais523 unable to obtain an IRC client
18:01:40 <ais523> elliott: to be precise, there was an IRC client installed, I just wasn't aware of it
18:01:45 <elliott> also, /me freaks out at the capitalisation
18:02:02 <ais523> anyway, /you/ try finding an IRC client for SunOS in 2007
18:02:17 <elliott> (related anecdote: I am pretty sure I once caused a friend to start using uppercase letters in IM, and then they /never stopped/ and have stayed like that ever since)
18:02:39 <ais523> also, we weren't allowed to install software on the machine, typical dangerous-executables stuff
18:02:45 <elliott> (this is more interesting when you consider that it seems to have unified /all/ their styles of textual messaging)
18:02:47 <ais523> although I compiled C-INTERCAL from source there
18:02:58 <elliott> ais523: couldn't you just write an irc client in perl or something?
18:03:15 <ais523> I wasn't aware that the format was so simple
18:03:20 <ais523> or of Perl's existence
18:03:27 <elliott> ...also, you didn't know perl existed in 2007?
18:03:40 <ais523> elliott: I'd spent my whole life on Windows
18:03:49 <ais523> and had pretty much got completely sick of it
18:03:56 <Vorpal> <elliott> also, maybe not, but even Vorpal has read I think most of the posts :) <-- yep, they are interesting though sometimes a bit ... hateful against the rest of the world? No not the right words but I can't think of any better.
18:03:57 <elliott> well, ok, but if you compiled C-INTERCAL... although i guess that was later on
18:04:00 <ais523> I found out I had a login on a SunOS box, as did everyone else, but I decided to actually use it
18:04:17 <Vorpal> elliott, ah yes that would fit it very well
18:04:22 <ais523> so I went and learnt UNIX, and preferred it to Windows
18:04:34 <elliott> I hope Loper and @ are engaged in bitter OS wars two decades from now
18:04:35 <ais523> hmm, I wonder how many other people here used UNIX before Linux?
18:04:59 <oerjan> <ais523> elliott: got a copy of the Vladivostok Telephone DIrectory handy, btw? <-- hey agora still remembers that?
18:05:06 <elliott> I started using Linux in late 2005 or early 2006
18:05:09 <elliott> oerjan: see, you should come back!
18:05:10 <ais523> then I tried to get an Internet connection, which I did via guessing names of browsers
18:05:21 <elliott> ais523: did you not know about ls?
18:05:31 <Vorpal> <oerjan> <ais523> elliott: got a copy of the Vladivostok Telephone DIrectory handy, btw? <-- hey agora still remembers that? <-- ???
18:05:32 <ais523> I did know about ls, but didn't realise you could point it at /usr/bin
18:05:45 <elliott> ais523: Windows brain damage at its finest.
18:05:54 <ais523> or indeed, that /usr/bin existed
18:06:02 <elliott> it would have been better if you thought /usr/bin existed, but was special
18:06:07 <elliott> like trying to do dir "Control Panel" in Windows
18:06:32 <ais523> oerjan: rule 1750/4: The first Agoran week each year which falls entirely in February is known as Read the Vladivostok Telephone Directory Week. During this time, Agorans SHALL read the Vladivostok Telephone Directory.
18:06:46 <elliott> hmm, is read the ruleset week gone?
18:06:51 <ais523> elliott: that actually works, it's just that its name is really messed up
18:06:51 <elliott> or is that in addition to?
18:07:24 <ais523> anyway, a bit of web searching seems to imply that Vladivostok doesn't actually /have/ a telephone directory
18:07:24 <Vorpal> ais523, did the guessing word?
18:07:42 <ais523> Vorpal: yes, "netscape" was the first one I found
18:07:45 <ais523> also, a really old version
18:07:55 <Vorpal> ais523, hah. It must have been ages ago to find that
18:07:57 <ais523> it once blanked the main page talk on Wikipedia
18:08:00 <Vorpal> ais523, on a *linux* system
18:08:04 <ais523> it just hadn't been updated since
18:08:13 <ais523> was Netscape ever released for Linux?
18:08:18 <ais523> (note: SunOS was obsolete in 2007)
18:08:31 <Vorpal> ais523, I think I used it back on redhat 5
18:08:34 <elliott> sunos was obsolete in 2007, ORLY
18:08:39 <pikhq> It used to be part of distros until Mozilla became usable.
18:09:04 <ais523> elliott: it was replaced by Solaris in 1992
18:09:14 <ais523> and end-of-lined in 1994
18:09:21 <ais523> according to Wikipedia, at least
18:09:27 <pikhq> And later versions were still offered for Linux just because there was no point in *not*.
18:09:31 <elliott> i remember talking to you when you were on SunOS/CDE, you're less weird now!
18:09:40 <pikhq> (take Mozilla, add a few patches, compile, and you've got Netscape)
18:09:49 <elliott> `addquote <ais523> yay CDE
18:09:53 <elliott> first time anyone has ever said that
18:10:08 <ais523> elliott: well, I only used xterms on it
18:10:14 <ais523> but it was nice having more than one up at a time
18:10:26 <pikhq> Ah, the *classic* use of X11: fancy terminal multiplexer.
18:10:27 <ais523> and the file manager, the one time I opened it by mistake, didn't look too awful
18:10:35 <ais523> I'm not entirely sure how to open it deliberately
18:11:23 <elliott> wow, oerjan only came in here in 2006
18:11:26 <ais523> elliott: imagine that the only way you have to open applications is something that looks and behaves vaguely like an OS X-style Dock, except it's completely inconsistent with everything and makes no sense
18:11:32 <elliott> 06.06.08:10:42:07 <lament> is Oerjan Johansen here?
18:11:32 <elliott> 06.06.12:12:33:20 <GregorR> Oerjan rawx 8-D
18:11:32 <elliott> 06.06.12:16:38:33 --- join: oerjan (n=oerjan@hagbart.nvg.ntnu.no) joined #esoteric
18:11:43 <elliott> ais523: so, OS X's Dock, then
18:12:02 <ais523> elliott: imagine each icon actually refers to more than one program
18:12:19 <Gregor> elliott: Why did I say that if it was his first time in the channel? :P
18:12:36 <ais523> and sometimes it opens a second copy Windows quick-launch style, rather than acting Dock-style
18:12:39 -!- BeholdMyGlory has joined.
18:12:41 <elliott> Gregor: Because he improved your glass. Page.
18:12:50 <Gregor> Who are the active users who are longer-term than me? lament (if he's to be considered active), ais523, ...?
18:12:57 <elliott> ais523 came in here in 2007
18:13:10 <elliott> lament is not even vaguely active. fizzie has been here for 2002, but only re-activated recently.
18:13:21 <elliott> So fizzie, despite being here all this time, is basically a returning oldbie :P
18:13:27 -!- Phantom__Hoover has joined.
18:13:51 <Gregor> jix is never active ...
18:14:07 -!- Phantom_Hoover has quit (Ping timeout: 240 seconds).
18:14:09 <elliott> 05.05.02:18:56:58 --- join: GregorR (~GregorR@c-24-21-138-66.hsd1.or.comcast.net) joined #esoteric
18:14:09 <elliott> 05.05.02:18:57:28 <GregorR> Does anybody want to try out my combination of BrainFuck and CoreWars? Essentially, you have two BrainFuckish programs running concurrently, and their data space is the opponent's program space.
18:14:10 <elliott> 05.05.02:18:57:41 <GregorR> The objective being to crash the opponent.
18:14:18 <elliott> so, now I have to list every name before 2005
18:14:24 <elliott> and we can pick out the active ones
18:14:28 <Gregor> I totally do not want to be the longest-term person ...
18:14:40 <Vorpal> <pikhq> Ah, the *classic* use of X11: fancy terminal multiplexer. <
18:14:52 <Vorpal> *<-- hey I use it that mostly
18:14:53 <Gregor> Phantom__Hoover: We were just discussing whether he's still "active"
18:14:58 <Vorpal> that and a handful of GUI programs
18:15:11 <Gregor> Phantom__Hoover: Or whether the fact that he was away for a long time makes him not really long-term (although I was too X-P )
18:15:29 <ais523> hmm, Gregor originally came here to advertise FYB?
18:15:41 <ais523> how old is EgoBot, btw?
18:15:58 <elliott> Gregor: OK, I've got the list up :P
18:16:02 <Gregor> I wrote it to one-up calamari, who doesn't get to count as active.
18:16:16 <Gregor> elliott: Up ... where?
18:16:41 <elliott> Well, catseye is there, which is cpressey, who came back for a while earlier.
18:16:48 <Phantom__Hoover> Gregor, I'm curious as to what made you question fizzie's current activeness.
18:16:56 <elliott> Phantom__Hoover: He didn't.
18:17:11 <Gregor> Phantom__Hoover: We were trying to decide whether it's continuous activeness or current activeness that counts.
18:17:13 <elliott> Gregor: dbc, he counts as active just because.
18:17:24 <Gregor> lol, I don't think he does :P
18:17:25 <elliott> fizzie is indeed on the list, obviously.
18:18:02 <oklopol> elliott was just linking some oklo from the olden days not so long ago
18:18:09 -!- FireFly has joined.
18:18:10 <elliott> Gregor: hcf is there, he was in here a whole once, to say that clog was here now
18:18:16 <oklopol> yeah i do recall why i came
18:18:21 <ais523> hmm, what are there more of, regular esolangers or pointless BF derivatives?
18:18:37 <ais523> elliott: what did you correct?
18:18:42 <ais523> I meant, people who studied esolangs
18:18:43 <elliott> Gregor: hmm "It`s_Puzzlet", I think puzzlet is lifthrasiir
18:18:43 <oklopol> i listed 50 bf derivatives once
18:18:49 <ais523> I didn't mean the esolangs themselves
18:18:56 <elliott> ais523: latter then I'd say
18:19:05 <elliott> Gregor: but lifthrasiir isn't too active :P
18:19:15 <elliott> Gregor: KEYMAKER AND KIPPLE THEY ARE VERY ACTIVE
18:19:23 <ais523> Phantom__Hoover: I'm comparing two integers
18:19:24 <elliott> Phantom__Hoover: they're both numbers
18:19:39 <elliott> Gregor: mooz and mtve... :P
18:19:40 <oklopol> integers form a complete lattice
18:19:49 <Gregor> elliott: Shaddap with totally inactive names X-P
18:20:07 <elliott> And nooga definitely counts as active, just not super-active :P
18:20:16 <elliott> hahaha this is so shameful for Gregor
18:20:19 <elliott> i think he has to kill himself now
18:20:21 <Gregor> How is it possible that I'm so long-term and I don't have any sort of operator privileges on the channel X-P
18:20:32 <elliott> Gregor: It's because you're really, really gay.
18:20:44 <Gregor> elliott: That's like the prerequisite for ops.
18:20:47 <elliott> Gregor: Also, by that logic nooga should be an op :P
18:20:53 <oerjan> <ais523> hmm, I wonder how many other people here used UNIX before Linux? <-- me definitely since i first used it back in '91-92 or so
18:20:58 <ais523> hmm, I actually admire this channel for not spontaneously giving me op privileges
18:21:04 <ais523> it seems to happen all too often these days
18:21:11 <elliott> ais523: the problem with giving you op privileges is that you'd use them even less than our other ops
18:21:26 <elliott> oerjan is currently on my naughty-list for disturbing the peaceful anarchy of the channel by banning shutup :D
18:21:27 <ais523> elliott: hey, I have used op privileges before
18:21:37 <oklopol> i used to have op privileges on ##algorithms or something
18:21:45 <elliott> ais523: it's not like we really need ops here, though
18:21:51 <ais523> banning spambots on #nomic (one #, on slashnet); and I once banned a troll from #interhcak
18:22:16 <elliott> ais523: the only time I'm aware of that they've actually helped here, other than banning spambots, was when Quadrescence and dixon hogged up the channel with trolling for an entire day
18:22:18 <Gregor> nooga: OH COME ON PLEAAAAAAAAAASE SAY SOMETHING D-8
18:22:25 <elliott> And Quad got unbanned the next day :P
18:22:47 <oklopol> Gregor: just wait a few years and i'll have been here longer than you
18:22:58 <elliott> in retrospect that wasn't even necessary, as communal readings of the Funge-98 spec do just a good job of getting rid of trolls
18:23:01 <ais523> oklopol: I don't think timespans work like that
18:23:01 <Gregor> oklopol: ... that ... follows logically ...
18:23:11 <elliott> i guess it's cpressey's calming influence
18:23:14 <Gregor> Unless I leave in a huff RIGHT NOW.
18:23:14 <elliott> oklopol: and when I'm older than you...
18:23:15 <oerjan> <Vorpal> <oerjan> <ais523> elliott: got a copy of the Vladivostok Telephone DIrectory handy, btw? <-- hey agora still remembers that? <-- ??? <-- it was an _old_ famous scam from the nomic preceding agora. in fact i think it referred to the 1957 phonebook or thereabouts
18:23:21 -!- variable has quit (Remote host closed the connection).
18:23:23 <ais523> `addquote <elliott> in retrospect that wasn't even necessary, as communal readings of the Funge-98 spec do just a good job of getting rid of trolls
18:23:24 <HackEgo> 273) <elliott> in retrospect that wasn't even necessary, as communal readings of the Funge-98 spec do just a good job of getting rid of trolls
18:23:38 <elliott> ais523: now you'll never be an op, you just endorsed rabid flooding!
18:23:48 <elliott> ais523: admittedly, of on-topic material
18:23:49 <ais523> that's not flooding, it's ontopic!
18:23:55 <oklopol> this one time elliott told me the funge 98 language is sooooooooooooooooooooooo complicated
18:24:00 <ais523> I know I got in trouble once for trying to learn python via bsmnt_bot
18:24:04 <Phantom__Hoover> <Gregor> elliott: That's like the prerequisite for ops. ← fizzie isn't gay, although now that you come to mention it lament *does* have a girl's name.
18:24:05 <ais523> even though I was writing a BF interp
18:24:10 <elliott> we should print out the funge-98 spec, to use as a doorstop
18:24:22 <ais523> it would make a pretty poor doorstop, it isn't long enough
18:24:24 <elliott> fizzie could be female and gay
18:24:32 <elliott> her hair is very long, after all!
18:24:40 <elliott> ais523: it is if you print it in 72pt type
18:24:46 <Phantom__Hoover> elliott, well, the photos of him don't exactly contradict that...
18:24:52 <ais523> elliott: heh, I was about to mention 72pt
18:25:04 <ais523> it's the largest font size in existence, obviously, because the dropdown boxes in Windows don't go higher
18:25:05 <Gregor> Now "Gregor" on the other hand.
18:25:08 <oklopol> my dad's also called heikki
18:25:11 <elliott> ais523: nuh uh, some of them go to 96
18:25:17 <ais523> some go much higher than that
18:25:24 <ais523> I think so they can advertise super-large font sizes
18:25:35 <ais523> (note: all these boxes just let you type arbitrary numbers in if you double-single-click them)
18:25:39 <elliott> Phantom__Hoover: she's his dad, obvs finland has a large transgender population
18:25:51 <elliott> ais523: you can't just use NUMBERS like that!
18:25:56 <ais523> there's a racing driver called heikki, who is also male
18:25:58 <elliott> Whoa, pasky ... I wonder if that's the same pasky as in git pasky ...
18:25:59 <oklopol> Phantom__Hoover: well i'm not completely sure about that, but i'm rather sure he's male
18:26:07 <ais523> elliott: alternatively, select it and hold down control-] (in Word)
18:26:21 <elliott> Gregor: {^Raven^}, very active
18:27:04 <elliott> 04.08.15:19:48:25 --- join: sbp (sbp@vorpal.notabug.com) joined #esoteric
18:27:04 <elliott> 04.08.15:19:48:31 --- part: sbp left #esoteric
18:27:16 <elliott> oklopol: it's from Jabberwocky
18:27:22 <elliott> Vorpal has just stolen it and ruined the poem for us forever
18:27:29 <elliott> because we can never again read it without being reminded of him
18:27:44 <ais523> elliott: how did your hyperignore filter not ignore that line?
18:27:46 <elliott> Gregor: Two months earlier, he was quoted:
18:27:47 <elliott> 04.06.09:14:58:59 <deltab> 215723Z #esp <sbp> http://www.ecs.soton.ac.uk/~sdh300/stuffage/bf.net/
18:27:47 <elliott> 04.06.09:14:58:59 <deltab> 215733Z #esp <sbp> best bit: [[[
18:27:49 <ais523> because it decided it was being used as a word?
18:27:51 <elliott> 04.06.09:15:00:43 <deltab> 220014Z #esp <sbp> hello #esoteric
18:27:54 <elliott> Maybe it took him two months to decide to join :P
18:27:59 <elliott> ais523: because I disabled it
18:28:06 <elliott> ais523: Vorpal isn't on ignore, anyway
18:28:09 <Phantom__Hoover> Did I mention that someone should totally make a language that naturally works on TMs?
18:28:15 <elliott> also, I /wrote/ that line, how could I ignore it?
18:28:17 <ais523> hmm, I'm going to get dinner
18:28:41 <ais523> I'll be back in a while (maybe over an hour); and although I'll leave the IRC client running, this connection is unreliable and I'll be surprised if I'm still connected by the time I get back
18:28:46 <oklopol> Phantom__Hoover: what does that mean
18:28:49 <elliott> Gregor: sexygirl153 was here before you
18:28:56 <Gregor> elliott: No, that's me.
18:29:08 <elliott> Who HAS TALKED IN HERE IN THE PAST CENTURY.
18:29:44 <elliott> cal153: I even grepped the logs to verify!
18:29:49 <elliott> You HAVE INDEED talked in the past century.
18:29:58 <cal153> iamcal = cal153 = sexygirl153
18:29:58 <Phantom__Hoover> oklopol, erm. That you could convert it into a TM without a thick abstraction layer.
18:30:01 <elliott> cal153: now, say 15 more lines, and you count as active enough for Gregor
18:30:02 -!- oklopol has changed nick to okl043.
18:30:06 <elliott> and therefore have superiority over him
18:30:29 <Phantom__Hoover> Vulgar version: you could write a compiler for it without tearing your hair out.
18:30:29 <elliott> slava. I wonder if it's /that/ slava
18:30:51 <okl043> Phantom__Hoover: you can directly compile brainfuck to tm
18:30:51 <elliott> Your line is SHOWING ITS LEGS!
18:30:57 <okl043> but you don't get all tm's this way
18:31:41 <impomatic> `addquote <testing> `addquote <testing123> Hmmm...
18:31:42 <HackEgo> 274) <testing> `addquote <testing123> Hmmm...
18:31:49 <Gregor> (Vulgar means "in the language of the common people")
18:32:08 <HackEgo> 34) <zzo38> I am not on the moon.
18:32:12 <Gregor> (Where "common people" is taken to mean "those inferiors")
18:32:20 <HackEgo> 42) <oklopol> i'm my dad's unborn sister
18:33:32 <oerjan> <oklopol> integers form a complete lattice <-- um, i see no top or bottom
18:33:47 <okl043> oerjan: that was implied
18:33:52 -!- okl043 has changed nick to oklopol.
18:34:10 <oklopol> that i meant the extended integers
18:34:45 <oklopol> implied by "<oklopol> that i meant the extended integers", that is
18:34:51 <elliott> theory, cal153 is actually mars
18:36:52 <HackEgo> 273) I wonder how the numbering system works.
18:37:08 <HackEgo> 92) <Sgeo|web> Where's the link to the log? <lament> THERE'S NO LOG. YOUR REQUEST IS SUSPICIOUS AND HAS BEEN LOGGED.
18:37:26 <cal153> i post every 779.96 days. you have discovered my secret
18:37:44 <Phantom__Hoover> (I really hope that the Ed apollo is referring to is Ed MacPherson.)
18:38:04 <elliott> OK, Adhamhahnajhnahnjanhanhanhn.
18:38:12 <elliott> How do you pronounce that name anyway?
18:38:42 -!- variable has joined.
18:38:49 <Phantom__Hoover> It's the CamelCase in "MacPherson" which annoys me: I've never seen it done that way. Ever.
18:39:35 <quintopia> considering that "Mac" is the prefix for "son of" and "Pherson" is the real name...
18:39:47 <fizzie> elliott: (As I've probably mentioned before) I keep getting ads from this German place (from where I bought a used laptop once) addressed to "Frau Kallasjoki", perhaps because "Heike" is a German female given name.
18:40:37 <oerjan> <Phantom__Hoover> [...] although now that you come to mention it lament *does* have a girl's name. <-- nikita isn't a female name in russian. or at least not primarily. i suspect something got lost in translation when it became popular in the west.
18:40:40 <oklopol> oerjan: also actually A(4, 4) is larger than any other integer, i learned this the other day when i was reading about this algo, it had complexity O(n a^-1(n)) = O(n a^-1(4))
18:41:06 <quintopia> everyone in the west wanted their daughters named after kruschev
18:41:09 <Gregor> ... surnames with multiple capital letters annoy Phantom__Hoover.
18:41:14 <quintopia> TO PROVE HOW NOT COMMUNIST THEY WERE
18:41:34 <elliott> Gregor: But Phantom__Hoover's name is a McFoo.
18:41:48 <oerjan> <elliott> Heikki, it's like, Helen! <-- incidentally "Kari" is a male name in finnish and a female one in norwegian
18:42:03 <elliott> or should I say Mr(s). Kari
18:42:14 <elliott> `addquote <oklopol> oerjan: also actually A(4, 4) is larger than any other integer, i learned this the other day when i was reading about this algo, it had complexity O(n a^-1(n)) = O(n a^-1(4))
18:42:15 <HackEgo> 273) <oklopol> oerjan: also actually A(4, 4) is larger than any other integer, i learned this the other day when i was reading about this algo, it had complexity O(n a^-1(n)) = O(n a^-1(4))
18:42:27 <HackEgo> 229) <elliott> Vorpal loves the sodomy. <Vorpal> elliott, sure why not
18:42:28 <HackEgo> 251) <Goosey> I can play crysis, but not minecraft?
18:42:30 <HackEgo> 164) <cheater99> you don't have an urethra, you're a girl.
18:42:33 <HackEgo> 222) <pikhq> Give me a beaver and I'll put it to work.
18:42:47 <elliott> that was insufficient quotes HackEgo
18:42:59 <oklopol> actually it was sufficient quotes
18:43:06 <HackEgo> 59) <Warrigal> I think hamsters cannot be inert.
18:43:10 <HackEgo> 159) <fungot> ais523: elf corpses are not considered expensive health food. but the most expensive.
18:43:11 <HackEgo> 119) <ais523> theory: some amused deity is making the laws of physics up as they go along
18:43:12 <HackEgo> 175) <fungot> pikhq: it was fragrant with the scent of abomination. hear a speech declaring a holy war, is the man insane? some idiot missionary gets himself killed, some man writes some gibberish about the shape of a dragon, wonse?"
18:43:17 <elliott> ais523: is this throttling thing new?
18:43:35 <HackEgo> 30) <Deewiant> ehird: There is no h in "honour"
18:43:36 <HackEgo> 41) <Dylan> kaelis: yes kaelis, but however will get the horses to wear knickers?
18:43:51 <HackEgo> 272) <elliott> in retrospect that wasn't even necessary, as communal readings of the Funge-98 spec do just a good job of getting rid of trolls
18:44:19 <HackEgo> 209) <fungot> ais523: my nose feels like a bad heuristic
18:44:26 <HackEgo> 236) <Vorpal> (had real world issues) <Vorpal> (to deal with) <ehird> Vorpal's pregnant. <Vorpal> yes
18:44:30 <Gregor> http://codu.org/music/vg/zee1.ogg Hey guys I'm linking this 'cuz it's actually used in ZEE now durpadurp.
18:44:39 <elliott> Gregor: Zee pirates are invaoeisjgdflkfgpdtsref
18:46:08 <Gregor> oklopol: No, not that one :P
18:46:24 <Gregor> oklopol: ZEE is an eternally-stalled game I'm eternally making.
18:46:31 <elliott> this is the worst song and Gregor is the worst person ever
18:46:37 <elliott> just kidding i hate Gregor's soul and children
18:47:08 <elliott> ugh where is that stupid language
18:47:14 <oerjan> <okl043> oerjan: that was implied <-- O KAY
18:47:17 <Gregor> Phantom__Hoover: Actually yes :P
18:47:21 <Gregor> Phantom__Hoover: But more eternally-stalled
18:47:43 <Gregor> Phantom__Hoover: The fact that it needs crazydata.
18:47:59 <Gregor> oklopol: http://codu.org/projects/zee/
18:48:31 <Gregor> ... -Enhance-Extrapolate
18:48:36 <oklopol> i'm not sure that's the correct music
18:48:45 <oklopol> by zoom i meant is it the thing
18:49:00 <oklopol> that you talked about when we were young
18:50:25 <Sgeo> elliott, that was random
18:50:38 <oklopol> hey speaking of young, some kinds asked me to buy some tobacco for them and i did, and then they were all likd "wow that guy's coool" when i told them i don't need their money
18:51:06 <elliott> `addquote <oklopol> hey speaking of young, some kinds asked me to buy some tobacco for them and i did, and then they were all likd "wow that guy's coool" when i told them i don't need their money
18:51:07 <HackEgo> 274) <oklopol> hey speaking of young, some kinds asked me to buy some tobacco for them and i did, and then they were all likd "wow that guy's coool" when i told them i don't need their money
18:51:08 <Sgeo> Claiming Stargate Universe was cancelled
18:51:24 <elliott> Sgeo: Syfy announced on December 16, 2010 that it would not be picking Stargate Universe up for a third season and that the Spring 2011 season would be the last to air on its channel.[6] There has been no word from MGM Television on this decision.
18:51:26 <oklopol> (i mean i actually heard them say that behind my back)
18:51:33 <fizzie> oklopol: You sure know how to make friends.
18:51:52 <elliott> the great thing is, one day we will all get oklopol arrested after he does something really, really terrible
18:51:58 <elliott> but until then we're just gonna laugh at pat him on the back
18:52:22 <elliott> like say oklopol have you raped anyone yet
18:52:45 <fizzie> "But I'm working on it"?
18:52:57 <Gregor> oklopol: Have you stopped raping people in the park yet?
18:53:10 <oklopol> Gregor: i don't recall stopping that, now
18:53:57 <oklopol> is buying kids tobacco bad? that was like my good deed for the day
18:54:04 <oklopol> i felt so good helping those kids
18:54:05 <quintopia> oklopol: have you reported your violence issues around children to the authorities?
18:54:50 <elliott> ais523: do you know if i can set ubuntu to install all updates automatically? i'm weird lik ethat
18:55:22 <oerjan> <oklopol> is buying kids tobacco bad? that was like my good deed for the day <-- oh no that's fine, the world needs all the help against overpopulation it can get
18:55:44 * elliott makes note to self - stay away from oerjan - murderer
18:55:56 <Sgeo> elliott, supposedly, the producers are looking for a way to keep it going
18:55:57 <Gregor> oerjan: This is also why we need more gays.
18:56:11 <elliott> Sgeo: That's because they really, really hate everyone.
18:56:12 <oklopol> yeah i also turned one of those boys gay
18:56:23 <elliott> "You're not going to escape our shitty show! MWAHHAHAHAHAHAHAA"
18:56:50 <quintopia> Gregor: easy way to ensure that is to discover which babies are at high risk for heterosexuality and smother them
18:57:09 <quintopia> infanticide has a long historical tradition of being effective population control
18:57:17 <elliott> or just have gay sex with everyone from a young age... uh
18:58:16 <quintopia> the first link in the topic is Herobrine's logs, yes?
18:58:41 <elliott> i'm only tampering with them if you believe I am
18:58:56 <elliott> mostly because editing http://208.78.103.223/2011-01-21.txt sounds like an exercise in pain!
18:59:25 <elliott> i'm going to import the clog logs into irc format sometime
18:59:32 <elliott> so that you can view all the past logs in formatted, UTC glory
18:59:33 <oerjan> elliott: NEEDS MOAR COLOR
18:59:46 <elliott> oerjan: Design suggestions welcome and likely to be ignored :P
19:00:26 <oerjan> i'm just nostalgic for old ircbrowse, is all. although not its speed.
19:00:50 <Gregor> This may come across a little bit effeminate, but cold whether does horrible things to my skin >_>
19:01:20 <elliott> Gregor: It destroys your FABULOSITY!
19:01:24 <oerjan> oh and maybe linkifying. not that irssi does that anyway.
19:01:27 <elliott> oerjan: we all miss ircbrowse :(
19:01:53 <elliott> oerjan: in my defence: http://sprunge.us/Cjff
19:02:03 <elliott> oerjan: editing this code is a real bitch because it's a horrible hack :D
19:02:10 <elliott> (yes, that /is/ the worst implementation of HTTP ever)
19:02:53 <Gregor> Phantom__Hoover: No. I have the beginnings of one level, and the engine is mostly in place ... data = difficult.
19:02:54 -!- pumpkin has quit (Quit: Computer has gone to sleep.).
19:02:59 <quintopia> gregor: it came across as idiotic, as anyone unable to spell "weather" must be
19:03:02 <Gregor> Basically I've resigned myself to getting all the data myself ...
19:03:14 <elliott> Gregor is both gay and stupid
19:03:33 <Gregor> I've been confusing a lot of homonyms recently ... as in, in the last couple years ...
19:03:36 <oklopol> Gregor is doing zee completely wrong, he's going to have actual pictures in it
19:03:43 <elliott> Gregor: It's Gaylzheimer's.
19:03:52 <elliott> It happens to people who are gay, because they are bad people and hate God.
19:03:54 <quintopia> Gregor: same problem, but at least i notice and correct myself immediately :P
19:04:04 <oklopol> ah yes, indeed, alan turing was gay and stupid
19:04:30 <Gregor> Phantom__Hoover: The idea is to find clues to solve mysteries within images, with the further caveat being that the "campaign" will have a running mystery through it, so there could be clues to either the mystery at-hand, or more global things.
19:04:39 <elliott> why are we talking about programming
19:04:46 <elliott> we can all go home now folks
19:04:49 <quintopia> `addquote <oklopol> ah yes, indeed, alan turing was gay and stupid
19:04:50 -!- elliott has left (?).
19:04:52 <Gregor> Phantom__Hoover: Basically, within an image, you zoom in, "enhance" to get better quality, then sometimes "extrapolate" to look through reflections etc.
19:04:58 <HackEgo> 275) <oklopol> ah yes, indeed, alan turing was gay and stupid
19:05:00 -!- elliott has joined.
19:05:09 <oklopol> Gregor: wait, there's more than one pic?
19:05:27 <oklopol> elliott: haha everyone was inside him on sunday
19:05:28 <elliott> OK, we can just talk about functional programming
19:05:29 <Gregor> oklopol: There's one starting-point picture per "case"
19:05:32 <Phantom__Hoover> (His sun supposedly dated Haskell Curry's daughter, which is completely crazy.)
19:05:33 <elliott> LC derivatives are OK, imperative languages are not
19:05:39 <Gregor> oklopol: And that can lead (via reflections etc) to other pictures.
19:05:44 <elliott> Phantom__Hoover: Yeah who has their own private sun?
19:05:50 <elliott> Phantom__Hoover: Also: BEST GENETIC PAIRING EVER
19:05:58 <oklopol> Gregor: okay one per case is maybe k
19:06:10 <quintopia> elliott: also, we can talk about the entscheidunsproblem since church basically proved that too (and first, yes?)
19:06:21 <ais523> wow, I am still connected as well
19:06:39 <Sgeo> elliott, I am learning Go.
19:06:39 <elliott> quintopia: I think so, so yes
19:06:51 <impomatic> Anyone have an idea how a functional programming game might work?
19:06:59 <elliott> impomatic: me and oerjan designed one once!
19:07:02 <ais523> <elliott> ais523: do you know if i can set ubuntu to install all updates automatically? i'm weird like that <-- there's a GUI setting for just security updates, but nothing's preventing you sticking an aptitude update aptitude upgrade in one of the startup scripts
19:07:05 <elliott> the idea was to predict elements of your opponent's list
19:07:09 <elliott> it was actually quite cool
19:07:12 <elliott> oerjan: <elliott> the idea was to predict elements of your opponent's list
19:07:20 <elliott> there was something about recognising your own function too but that sounds lamer
19:07:43 <oklopol> predict elements on your opponent's list how?
19:07:43 <elliott> impomatic: no, it was in-channel, if oerjan wakes up he'll remember some of it probably
19:07:49 <elliott> oklopol: i forget, I implemented it once
19:07:55 <elliott> it was like... ugh i forget
19:08:10 <impomatic> This is the only functional programming game I'm aware of http://pagesperso-systeme.lip6.fr/Christian.Queinnec/Papers/strugman.ps.gz
19:08:38 <elliott> ours was cool, help me wake oerjan up
19:08:53 <oklopol> i bet it was not actually very cool
19:08:54 <elliott> impomatic: it was a program-battling game obviously
19:08:57 <elliott> but i suppose you know that
19:09:10 <quintopia> 1) make sure your functional language is implemented on a program stack, 2) make the stack circular so the top of one stack grows towards the bottom of the other stack, and 3) see who can smash the other player's stack to their own purposes first. >_>
19:09:30 <ais523> then that's not really functional
19:10:43 <oklopol> poop your fucking ASS i'm a poop!
19:10:54 -!- cheater00 has quit (Ping timeout: 240 seconds).
19:11:06 <oklopol> i haven't been on the computer like all week
19:11:19 <oklopol> i don't think i can ever leave again
19:11:43 <Sgeo> elliott, so, I decided that the one true language is Java.
19:11:51 -!- cheater99 has joined.
19:12:16 <oklopol> elliott: i played a game of freecell just now
19:12:24 <elliott> oklopol: you cannot get free...from your cell
19:12:25 -!- variable has quit (Read error: Connection timed out).
19:12:36 <ais523> oklopol: which version?
19:12:36 <oklopol> elliott: yesterday, i ate some pizza
19:12:37 <Sgeo> ais523, I was trying to get a reaction from elliott
19:12:41 <ais523> the one that comes with Windows?
19:12:50 <oerjan> elliott: you had a function, which got either itself or an opponent's function as argument, and your task was to determine which it was
19:13:00 <ais523> apparently, freecell accepts -1 as a puzzle number, and it's impossible
19:13:04 <elliott> oerjan: the one involving lazy lists
19:13:12 <elliott> oerjan: and it was like, mutual or something
19:13:13 <ais523> everyone likes undocumented features!
19:13:23 <oklopol> ais523: which freecell? this?
19:13:43 <oklopol> elliott: btw, there's a mat on the floor, but my feet are not touching it as it is a bit too far away
19:14:01 <Sgeo> Does that mean it's impossible to play Freecell off the computer, if there are unwinnable configurations?
19:14:19 -!- variable has joined.
19:14:24 <elliott> oerjan: do you remember itttttttttttttttt
19:14:35 <oklopol> yeah Sgeo i didn't quite catch that either
19:15:07 <Sgeo> If sometimes it's impossible to win Freecell, that means, if you''re not on a computer, sometimes you'll have unwinnable games?
19:15:08 -!- ais523 has quit (Remote host closed the connection).
19:15:23 -!- ais523 has joined.
19:15:31 <oklopol> Sgeo: have you an idea what the game is?
19:15:31 * oerjan doesn't remember the list version
19:16:05 <quintopia> sgeo: you're saying that the computer selects games that are guaranteed winnable?
19:16:47 <Sgeo> "All of the 32,000 Microsoft deals except for number 11982 are solvable."
19:16:58 <oklopol> so basically they made them manually
19:17:07 <oklopol> and had every game checked a few times
19:17:08 <oerjan> <ais523> apparently, freecell accepts -1 as a puzzle number, and it's impossible <-- there is also exactly one or... what Sgeo said
19:17:26 <ais523> oerjan: yes, I knew there was exactly one of the main 32000 that was impossible, but I wasn't sure which
19:17:53 <ais523> I suspect Freecell games are overwhelmingly likely to be possible, and 11982 was just chance
19:18:00 <ais523> (whereas -1 was deliberately selected to be impossible)
19:18:06 <oklopol> was just going to say what ais523 said
19:18:07 <ais523> quintopia: Solitaire on Windows doesn't have puzzle numbers
19:18:20 <ais523> (and has more than a 1/32000 chance of being impossible)
19:18:25 <oklopol> solitaire is usually not winnable
19:18:35 <oklopol> or i have completely misunderstood it
19:18:45 <ais523> klondike is winnable more often than not if you don't have a restriction on how many times through the deck you can go
19:19:11 <oklopol> winnable as in, there is a winning strategy?
19:19:21 <oklopol> that is, is there a computer program that usually wins
19:19:23 <ais523> as in some sequence of moves wins
19:19:42 <oerjan> solitaire is not perfect information like freecell is
19:19:46 <ais523> normally, the move sequence is reasonably obvious though, so I imagine a computer program could get above 50-50 with infinite deck recycle
19:19:55 <ais523> oerjan: *klondike, there are loads of versions of solitaire
19:20:07 <oklopol> and solitaire is also a synonym for klondike
19:20:07 <ais523> and aisleriot, the Linux version of solitaire, has loads of them
19:20:11 <ais523> I used to like playing Gold Mine
19:20:37 <quintopia> the windows klondike program is just called "solitaire" ... so i blame microsoft for that
19:20:45 <oerjan> ais523: well i meant the windows included game
19:20:51 <quintopia> the linux standard "solitaire" lets you pick the game rules
19:22:31 <ais523> I'm not sure if aisleriot is gnome or cross-toolkit
19:22:34 <fizzie> I don't think there's any "standard" solitaire. (The GnomeGames version is called "Aisleriot", it does "over eight different" variants.)
19:22:36 <ais523> but it does more than just let you pick the rules
19:22:44 <ais523> fizzie: "over eight"? it's a lot more than that
19:23:22 <fizzie> Shouldn't have been all "oh that bit is so short I won't bother actually copy-pasting for a quote".
19:23:26 -!- ais523 has quit (Read error: Connection reset by peer).
19:23:29 <quintopia> fizzie: standard as in "ubiquitous"
19:23:40 <fizzie> KDE's definitely not "ubiquitous" either.
19:23:46 -!- ais523 has joined.
19:24:17 <quintopia> fizzie: but for folks using it, kpatience is more likely than any other solitaire
19:24:32 <quintopia> aka, it is ubiquitous among kde users
19:24:44 <fizzie> Welllll, maybe, but most you can say is it's the "KDE standard solitaire", not "Linux standard".
19:25:22 <Phantom__Hoover> It's the Windows solitaire, so that makes it far more widespread.
19:25:29 <quintopia> which is that microsoft decided that solitaire and klondike are synonymous
19:25:33 <fizzie> Yes, I just wanted to nitpick.
19:25:38 <ais523> also, Gnome is probably more popular than KDE at the moment because of Ubuntu
19:25:57 <quintopia> ais523: i wouldn't downplay the presence of kubuntu
19:26:16 <quintopia> seems to be very thriving from what i can tell
19:26:18 <fizzie> "Everything from favorites like Freecell and Klondike through to the hopelessly pointless Clock Patience."; the last sounds so very enticing.
19:26:26 <fizzie> "This is a game of zero skill and is a purely mechanical process. The chances of winning are 1 in 13."
19:26:52 <ais523> fizzie: I've seen people play Clock Patience in RL, with a real pack of cards
19:26:55 <oerjan> <quintopia> which is that microsoft decided that solitaire and klondike are synonymous <-- obviously that's just history - it was once the _only_ solitaire game included with windows
19:27:05 <ais523> and yes, the chance of winning is exactly 1 in 13, and you only have one legal move at each turn
19:27:18 <fizzie> ais523: Maybe some sort of a card fetishist could have fun doing it.
19:27:53 <fizzie> I used to know one "only one legal move" sort of a "game" too, though it was more complicated than that.
19:28:16 <fizzie> All the sheep in Scotland are black.
19:28:30 <oerjan> <ais523> fizzie: I've seen people play Clock Patience in RL, with a real pack of cards <-- i'm pretty sure i've done it :D
19:29:13 <Gregor> I had to look it up to see what it actually is.
19:29:18 <Gregor> "Hopelessly pointless" indeed.
19:29:46 <fizzie> Wikipedia's "solitaire card games" list isn't indexed by metadata like "shape of playfield", I can't find that one I used to know.
19:30:11 <oklopol> "<ais523> fizzie: I've seen people play Clock Patience in RL, with a real pack of cards" <<< i used to
19:31:17 <oklopol> so maybe it's a game for those who like mathematics
19:32:43 <ais523> it might be for people who are used to really interesting and difficult problems just wanting to do something mindless
19:32:59 -!- variable has quit (Remote host closed the connection).
19:33:00 <oerjan> I TAKE THAT EXPLANATION
19:33:06 <elliott> ais523: ever played Valentine?
19:33:07 <oklopol> well that's why i'm here right now
19:33:42 <ais523> and this is probably a bad moment
19:34:02 <oklopol> irc makes me incredibly stupid, i just go completely numb from the neck up
19:34:32 <elliott> Valentine is like Clock Patience, but worse
19:34:33 <oklopol> maybe i should sleep at some point, and not come here tomorrow
19:34:38 <elliott> a game takes about two hours, afaik it's impossible to lose
19:35:09 <elliott> after playing it, your head is all funny for ages because it's just so numbed
19:35:34 <oklopol> today, i walked in a circle for a bit over an hour in this mall kinda thing
19:35:49 <oklopol> was waiting for someone, and realized i enjoy walking more than sitting
19:35:56 <ais523> oklopol: I've done that
19:36:21 <oklopol> i don't think i've done it in a mall, but otherwise i always walk in a circle
19:36:48 <ais523> 8 makes you less dizzy
19:37:07 <ais523> I tend not to use circles, but rather long meandering patterns that go to each of the boundaries of the area in which people might be expected to meat me
19:37:11 <oklopol> well it doesn't make me feel like i need to unwind the rotations
19:37:32 <oklopol> ais523: i need something highly repetitive or i can't think
19:37:52 <oklopol> well, of course occasionally i play the observation game called life and look at what people are doing
19:38:19 <oklopol> in which case i use more random pattenrs
19:38:20 <ais523> I do much of my thinking at bus shelters, often
19:38:26 <ais523> when I'm walking back and forth, mostly
19:38:36 <ais523> and people sitting in the bus shelter wonder why I'm walking back and forth in the rain
19:39:13 <oklopol> but i get lots of stuff done in the bus too
19:39:13 <ais523> it depends on the type of rain, really, some is nice than others
19:39:24 <oklopol> actually i just proved a theorem in the bus last week
19:39:31 <oklopol> i haven't used one this week except today
19:39:49 <oklopol> (it was a neat theorem, too)
19:40:05 <oklopol> i like all rain in theory, but i don't always want to get wet
19:40:18 <oklopol> my feet smell bad enough as is
19:40:20 <ais523> last night I was trying to sleep and thought about random stuff to try to get to sleep
19:40:26 <oklopol> quintopia 2WDFA = 2WNFA = RLG
19:40:26 <ais523> and realised I pronounced f and ph differently
19:40:28 <quintopia> and whose umbrella do you steal when you don't want to get wet
19:40:33 <ais523> although only with a very small difference
19:41:07 <oklopol> quintopia: it's obvious of course, the reason it's neat is that it fits really nicely in my framework, and is a converse to a theorem in HoFL
19:41:40 <oklopol> RLG is a class of picture languages, let L be a regular language over \Delta, and for all d \in \Delta, let L_d be a regular language
19:42:21 <oklopol> (L, (L_d)_{d \in \Delta}) defines a picture language by taking words from L, and substituting words from the languages corresponding to the letters, and making them the columns of the picture
19:42:27 <oklopol> picture = matrix over finite alphabet
19:42:49 <oklopol> also the theorem is "CDFA = CNFA = RLG"
19:43:23 <oklopol> CXFA is a class that just restricts automata of type X to only walk in a comb-like pattern on the picture, so that spikes go down
19:43:42 <oklopol> so yeah, it's pretty obvious if you know 2-way AFA give regular languages in the word case
19:44:19 <quintopia> ...this is much further than i went in automata theory actually
19:44:32 <oklopol> well there are like 7 researches in picture language theory in the world
19:45:02 <Gregor> <stan_man_can> can you return objects in JS?
19:45:06 * Gregor bashes his head into a wall.
19:45:10 <oklopol> that's not very deep, that's the languages of words such that there exists an alternating finite automata that can move left and right, that accepts exactly those words
19:45:11 <quintopia> maybe. i got DFA=NFA=regular languages. don't think i ever saw 2-way AFA
19:45:39 <oklopol> but they are more intuitive to program in
19:45:48 <oklopol> you can do function calls for instance
19:46:38 <oklopol> with DFA and NFA, you can't really do shit
19:46:51 <oklopol> when running on a more complicated graph
19:47:02 <oklopol> for instance tree-walking NFA are strictly weaker than regular tree languages
19:47:05 <elliott> dfa/nfa are for expert regex programmers
19:47:16 <oklopol> and NFA on pictures are incredibly stupid
19:47:50 <elliott> oklopol: regexps are nfas/dfas, i think with backrefs they are ndfas? or something
19:47:52 <oklopol> btw "tree-walking NFA are strictly weaker than regular tree languages" is a very new result
19:48:02 <oklopol> tree walking automata were not really studied at all until 2005
19:48:13 <oklopol> because really basic questions about them are really fucking hard to solve!
19:48:32 <j-invariant> hahah the hologram in star trek got "human" rights for holograms: And he said "it feels like a hollow victory"
19:48:47 <elliott> j-invariant: not Voyager i hope
19:50:33 <oklopol> elliott: what's a backreference?
19:50:59 <ais523> oklopol: it's an extension to the regexps used in programming that makes them not regular expressions
19:51:03 <oklopol> well pushdown automata obviously can't do that
19:51:11 <elliott> they're a different type of automaton
19:51:14 <oklopol> and it's obviously context-sensitive
19:51:14 <elliott> that has exponential worst-case behaviour
19:51:15 <Sgeo> I should probably eat
19:51:19 <oklopol> so something between those
19:51:23 <ais523> indeed, but regexp+backreference is less powerful than PDA, I think
19:51:29 <elliott> (even for non-backreferencing regexps)
19:51:35 <ais523> or at least, neither is strictly more powerful than the otehr
19:51:37 <elliott> (ofc you can use the more efficient kind if you know it has no backrefs)
19:51:40 <oklopol> you can do {ww : w is a word}
19:51:49 <ais523> PDAs can match brackets, backrefs can't
19:51:59 <elliott> for a very literal definition of word
19:52:03 <ais523> I mean, a PDA can match a^nb^n, but backreferences can't
19:52:24 <ais523> whereas backreference can match a^xb^ya^xb^y and PDAs can't
19:52:32 <oklopol> so mutually incompatible with CFG
19:53:00 <oklopol> and strictly smaller than CSG
19:53:18 <oklopol> (at least i think so, just make the thing you'll refer to later)
19:53:48 <oklopol> (and then when it's referenced, just compare the thing matched earlier, then continue)
19:54:14 <oklopol> using the fact CSG = linear space
19:57:16 <oklopol> of course would need an exact definition of backrefs
19:57:30 <oklopol> like in a star expression, is it clear what they do
19:59:00 <elliott> 16:56:33 <oerjan> and that type magic means it has varied whether you could use it with $
19:59:59 <oklopol> elliott: also wanna hear an even sexier characterization of the regular languages: L is regular iff there is an eq relation ~ on \Sigma^* such that 1) x ~ y => xa ~ ya for all words x, y and symbols a, 2) x ~ y => (x \in L iff y \in L)
20:00:11 <oerjan> elliott: referring to runST i believe?
20:00:55 <oklopol> more like runST*D* hahahahahahahahahahahahahaha
20:00:58 <oklopol> hahahahahahahahahahahahahahaha
20:01:01 <oklopol> hahahahahahahahahahahahahahahahahah
20:01:32 <oerjan> elliott: for syntactic niceness you might want to write something like runST $ do ... rather than runST (do ...)
20:01:57 <elliott> oerjan: yeah why would that not work?
20:02:01 <elliott> does haskell do stupid things w/ forall
20:02:03 <oerjan> but without being very clever, the first cannot be typed
20:02:53 <oerjan> elliott: the thing is when you use the (do ...) directly as an argument for runST it is easier for the type system to notice it needs to pass through the forall part of the type
20:03:22 <oerjan> but with $ it has to sort of chain through the type of $ itself, which knows nothing about higher rank stuff
20:03:30 <oerjan> or at least that's my impression
20:04:02 <oerjan> and it's apparently technically tricky to make it work properly while still being decidable
20:04:32 <oerjan> :t runST $ do return "hi"
20:04:54 <oerjan> :t flip ($) runST (do return "hi")
20:04:55 <lambdabot> Cannot match a monotype with `(forall s. ST s a) -> a'
20:04:55 <lambdabot> Probable cause: `runST' is applied to too few arguments
20:04:55 <lambdabot> In the second argument of `flip', namely `runST'
20:05:21 <oerjan> the workaround is a cheat that depends on the _order_ of arguments
20:05:45 <oerjan> :t flip ($) (do return "hi") runST
20:05:46 <lambdabot> Couldn't match expected type `forall s. ST s a'
20:05:46 <lambdabot> Expected type: (forall s. ST s a) -> a
20:05:48 <oklopol> why not just make $ syntactic sugar, it's not like anyone actually understands what it does anyway
20:06:19 <oklopol> haskell is like religion, people learn to preach about it but there's really no actual content
20:06:26 <oerjan> so it only works when runST is the _first_ argument of the function that the type needs to pass through
20:06:26 <oklopol> let's talk about regular languages instead
20:07:00 <oerjan> this is of course very ugly, so iirc they nearly removed it again
20:07:24 <Phantom__Hoover> <oklopol> why not just make $ syntactic sugar, it's not like anyone actually understands what it does anyway ← what's so hard to understand?
20:08:07 <oklopol> "<Phantom__Hoover> <oklopol> why not just make $ syntactic sugar, it's not like anyone actually understands what it does anyway ← what's so hard to understand?" <<< you can't understand something you can't touch with your fingers
20:08:27 <oklopol> you can try, sure, but you're living a lie.
20:08:54 <oklopol> elliott: yeah and what if it's really big? try understanding fucking $ oisjhgfoirewjfiorewjgfoiwrgijwreiguherhge
20:09:06 <elliott> becoming an ultrafinistit brb
20:11:46 <dbc> "Fuck your family" sounds like a more interesting game than "type the same thing repeatedly".
20:12:05 <elliott> what we're saying is, incest vs. spam, today on #esoteric
20:12:10 <elliott> dbc: are you just talkin' cuz i pinged you
20:12:39 <dbc> Maybe I should have swapped the pronouns, because I didn't mean MY family :)
20:13:18 <dbc> (Yes I am just talking because you mentioned my name earlier)
20:14:27 <dbc> It would be sneakier if it were not a standard feature of xchat.
20:15:56 <fizzie> The good old errno == E2BIG.
20:16:36 <elliott> dbc: xchat? you seem too old-school for xchat
20:16:40 <elliott> do we have any bitchx users here
20:17:04 <fizzie> I was one in one period of my IRC career.
20:17:27 <fizzie> Also I at some point had ircII script called "schlong", which was the most ridiculous thing ever.
20:17:47 <dbc> I'm not old-school Linux. I used Mac for most of my life. I was using Mac OS 9 for years after almost anyone else, because I thought if I was going to switch over to a Unix-based system I might as well switch over to Linux, and finally I did.
20:17:49 <oklopol> did it have buttons for drawing ascii penises
20:18:10 <fizzie> http://packetstorm.linuxsecurity.com/irc/scripts/schlong.irc
20:18:10 <elliott> "I ported my ircII script to Bitch-X. So one could say that I inserted my schlong into that Bitch-X."
20:18:15 -!- cheater99 has quit (Ping timeout: 240 seconds).
20:18:24 <elliott> i'm going to hang myself now -->
20:18:37 <fizzie> "oh and please dont run schlong with BitchX.. damn thats ugly."
20:19:04 <elliott> fizzie: oh man i love ircers, they're so irritating, look at dat ascii art splash screen
20:19:18 <fizzie> Yes, it was the very "gangsta".
20:19:20 <elliott> assign bogus_patterns *fuck* *shit* *suck* *dick* *penis* *cunt* *haha* *fake* *split* *ass* *hehe* *bogus* *yawn* *leet*
20:19:31 -!- cheater99 has joined.
20:19:33 <elliott> wow those quoats... i don't even
20:19:41 <fizzie> See the "random showers" part.
20:20:27 <elliott> wat @ those INPUT_PROMPT lines
20:20:37 <fizzie> It has some synflood-sort of "hacking" commands too.
20:20:40 <elliott> echo | /ls, /ls2, /ls3 - lists current directory
20:22:10 <fizzie> Also a "zmodem-bomb flood". And whatnot.
20:24:30 <elliott> fizzie: needs a command to write the entire contents of the latest phrack to irc
20:24:30 <fizzie> Also all kinds of attacks against different other scripts.
20:25:04 <fizzie> It was a bit more chaotic era. We used to hang on some ircnet channel that was the "command channel" of a backdoored mIRC script that was somewhat popular.
20:25:23 <fizzie> And then randomly autofix their infestations every now and then.
20:25:39 <fizzie> Until the channel was split-takeovered by (presumably) the script authors.
20:25:58 <elliott> fizzie: But... you were on the good side in a really, really gangster way, right?
20:26:28 <fizzie> Sure, I mean, I think there might have been some snooping around.
20:27:01 <fizzie> Anyway back then if you wanted to play with other people's machines you could just IP-scan a random ISP's dialup ranges for Back Orifice, there were always about five per a /24.
20:27:54 -!- impomatic has left (?).
20:28:06 <fizzie> For example there was the Quake Incident.
20:28:14 <fizzie> But maybe I shouldn't go into any incriminating details.
20:29:46 <fizzie> Let's just say that hypothetically some people might have used BO to modify a person's quake 1 configs (to set the rainbow-colorful poly-test mode on, rearrange all keys, and make most of then send weird messages to the chat channel) and then pop up a MessageBox saying "I'm Quake 1: play me, I'm better than ever".
20:30:06 <fizzie> (But there was a backup of the original config too, hypothetically.)
20:31:26 <elliott> ircnet is descendent of original IRC, right?
20:31:28 <elliott> it's the one that kept eric
20:31:55 <elliott> did eris get dropped from everyone, then?
20:32:11 <oklopol> elliott: assuming fizzie'll know that is racist
20:32:30 <fizzie> And then that thing where a guy was ICQing to an away-being friend of him, and the hypothetical people were reading his messages via BO's keylogger and replying with MessageBoxes. He was very curious to know how we were sending those "pull-up messages"; also the hypothetical people tried to convince him that they were the "root administrators of the Internet", but he caught that ruse because they were speaking Finnish.
20:33:08 <elliott> Gregor: Meanwhile, in the realm of Ridiculous Ideas: "The Zaphod addon integrates Narcissus into Firefox."
20:33:10 -!- Adiemus|2 has joined.
20:34:12 <fizzie> Yes, IRCnet split out of EFnet post the whole eris thing.
20:34:25 <fizzie> It's a bit arguable whether EFnet or IRCnet is "the" descendant.
20:34:35 <fizzie> I see EFnet's Wikipedia article claims they are.
20:34:44 <elliott> fizzie: So eris got completely abandoned?
20:34:48 <elliott> Eris Free Net is a bit of a strange name, then.
20:35:04 <fizzie> They had some sort of a short-lived network there too.
20:35:38 <elliott> Were you around before the split? This is, like, meeting someone who was alive when the dinosaurs were around! :p
20:36:22 <elliott> Right, so it went IRC ---> EFnet + A-net ---> EFnet ---> EFnet + Undernet ---> EFnet + Undernet + IRCnet ---> explosion
20:36:47 <fizzie> Not before EFnet, no. I had visited IRC before the EFnet/IRCnet split (via Freenet or whatever), but I wasn't very active back then.
20:36:53 <jix> Gregor: there is only so much time
20:37:02 <elliott> wrong, there is infinite time
20:37:03 <fizzie> I think I didn't seriously start doing the IRC thing until about '97 or so.
20:37:24 <elliott> fizzie: Erm, define Freenet.
20:37:27 <oklopol> lol, elliott wasn't even born
20:37:27 <elliott> Obviously not /that/ Freenet.
20:37:38 <fizzie> Not that; it was some sort of a modem-accessible thing.
20:37:50 -!- asiekierka has quit (Ping timeout: 240 seconds).
20:38:13 <fizzie> Free-Net Finland - Finland's K-12 school (whole nation)
20:38:17 <elliott> "Later in 2009, some major IRC servers were delinked: irc.vel.net, irc.dks.ca, irc.pte.hu, EFnet's only UK server efnet.demon.co.uk, and EFnet's only UK hub hub.uk, which were sponsored by Demon Internet."
20:38:17 <fizzie> Sort of a Finnish BBS thing.
20:38:21 <fizzie> Never all that popular.
20:38:26 <elliott> Does anyone actually /like/ EFnet?
20:39:37 <fizzie> EFnetters themselves are fond of their network, I think.
20:39:38 <elliott> Personally I wish Freenode would die and OFTC to take its place.
20:39:38 <elliott> Ho ho, us and our modern, stately, warezless, enlightened channels.
20:39:40 <fizzie> I'm not sure if I've spent any real time there; I was on a few Dalnet channels some time back, and had mostly idle-only connections to some others, including Quakenet.
20:39:50 <fizzie> I think there was a competing "fizzie" in Quakenet at some point.
20:39:59 <fizzie> Or it might've been a "fizzle".
20:40:02 <elliott> you should visit vjn, they're fun crazies
20:40:44 <fizzie> Aw, freenet.hut.fi no longer exists; I guess it's dead too.
20:40:46 <oklopol> * Spoli (~Samuvain@Spoli.users.quakenet.org) Quit (Ping timeout)
20:41:02 <elliott> oklopol: if this were like a year ago i could have pasted five pages of ooing
20:41:48 <elliott> quintopia: you're not european enough
20:41:48 <quintopia> what is the appropriate gateway server to use
20:41:53 <oklopol> we are actually thinking about doing some stuff soon
20:42:02 <elliott> hey you know what we should do
20:42:05 <elliott> blatantly break freenode policy
20:42:10 <oklopol> so it might be there's also more ooing coming up
20:42:20 <elliott> leaving freenode empty and barren
20:42:36 -!- Tritonio has joined.
20:43:00 <oklopol> i wonder how long freenode would last if we left
20:43:14 <elliott> after we, their main source of activity, disappear
20:43:15 <fizzie> irc.stealth.net used to be the main US mostly-open-for-everyone IRCnet server.
20:43:15 <fizzie> It splits somewhat often, though.
20:43:15 <elliott> (bad taste ahead) we could call it LFnet, lilo free net
20:43:16 <elliott> fizzie: "mostly-open-for-everyone"?
20:43:44 <fizzie> elliott: Yeah, they have pretty flexible I:lines. Generally the .fi servers serve a rather limited region; usually ISP-run servers tend to serve just their own clients and so on.
20:43:46 <elliott> but man, esonet, we could all have oper status, apart from vorpal
20:44:03 <elliott> but we'd just k-line him whenever he threatened to use them
20:44:13 <oklopol> and we could even all have OUR OWN CHANNELS!
20:44:13 <elliott> that way we could claim to not be discriminatoring
20:44:15 <fizzie> And many others have those "restricted" i:lines where you can't get +o on a channel.
20:44:26 <elliott> oklopol: vjn could move in! to double the activity!
20:44:41 <fizzie> They're to encourage people to use their local server, of course.
20:44:49 <fizzie> We're not made out of bandwidth!
20:45:11 <pikhq> fizzie: By US standards, you are.
20:45:23 <tswett> So, guys, it's currently colder at this spot in Michigan than it is in Helsinki.
20:45:31 <elliott> oklopol: we could have non-# channels
20:45:35 <elliott> what are the fun channel types again
20:45:38 <tswett> <elliott> I didn't know Stanford was in Michigan again.
20:45:50 <elliott> again, yeah, it used to be
20:46:05 <fizzie> What I'm on, irc.nebula.fi, seems to have I lines mostly for their own customers.
20:46:09 <fizzie> The list is horribly long though.
20:46:40 <elliott> fizzie: lol in finland your isps run irc servers...strange world
20:46:45 <elliott> here it's like, no, we don't do usenet, no
20:46:52 <elliott> but we have 200mb of free web space do you want that
20:46:59 <fizzie> ISPs and universities.
20:47:01 -!- augur has quit (Remote host closed the connection).
20:47:01 <elliott> with bad webmail, this is a feature for you?
20:49:27 <fizzie> Most ISPs here tend to have a bad webmail too.
20:50:36 <fizzie> This Internet here comes with 600 MB of webspace (of which I use none) and 5 email boxes, that's I think pretty standard.
20:51:39 <fizzie> Oh, and they often bundle a subscription to some sort of antivirii/firewall/"end-user security" product at the same or low extra (+3e/month?) price, that's a common feature too.
20:52:29 <fizzie> It also seems to have become a standard that most ADSL/cable modem links come with a system that gives you (via DHCP) 5 simultaneous public IPs, but no more; if you want more devices than that, it's going to need to be NATted.
20:52:57 <elliott> fizzie: Here we get 1 public IP.
20:53:24 <fizzie> (Some of them even try in the contract to say that "you can only connect up to 5 computers to the pipe", but I'm not sure if that's even theoretically enforceable; in practice it isn't.)
20:54:37 <fizzie> Pretty much all of them also have a very vague "you must not connect a server into our pipe" condition, that's widely ignored.
20:54:55 <ais523> fizzie: in the UK, they actually try to enforce no-servers conditions sometimes
20:55:02 <ais523> normally by blocking a whole load of ports inbound
20:55:17 <fizzie> Well, maybe that's sometimes done.
20:55:30 <fizzie> And most of places block outbound SMTP except to their own relay, but that's just common sense.
20:55:42 <fizzie> I guess some (don't know how many) block inbound SMTP too.
20:55:58 <fizzie> At least it cuts down the number of badly configured open-relay spamservers.
20:56:10 <elliott> <fizzie> And most of places block outbound SMTP except to their own relay, but that's just common sense.
20:56:33 <elliott> john gilmore's toad.com is an open proxy
20:56:33 <fizzie> Well, maybe not "common sense", but it doesn't really inconvenience anyone.
20:56:37 <elliott> he gets shit from his isp all the time for it
20:56:49 <ais523> sending email directly rather than via the ISP relay is almost never deliberate
20:56:54 <elliott> "He runs the mail server at toad.com as an open mail relay. In October 2002, Gilmore's ISP, Verio, cut off his Internet access for running an open relay, a violation of Verio's terms of service. Many people contend that open relays make it too easy to send spam. Gilmore protests that his mail server was programmed to be essentially useless to spammers and other senders of mass email and he argues that Verio's actions constitute censorship. He als
20:56:54 <elliott> o notes that his configuration makes it easier for friends who travel to send email, although his critics counter that there are other mechanisms to accommodate people wanting to send email while traveling."
20:57:02 <elliott> ais523: eh, mailing directly is fun
20:57:10 <fizzie> I mean, you can still run a completely open mail relay if you want, it's just that you have to configure it to use the ISP's box as a relay.
20:57:13 <elliott> ais523: to be fair, i basically oppose any kind of isp filtering of the internet whatsoever
20:57:16 <fizzie> Presumably they can then notice spamming better.
20:59:05 <fizzie> A reasonable number I think also blocks inbound netbios-over-TCP and naked-CIFS-at-445 ports too.
21:00:00 <fizzie> I wouldn't mind that sort of thing being an opt-out thing you could disable if you liked, but I doubt many ISPs bother.
21:00:42 <fizzie> Somewhere I think I've seen a web-form driven system where you could configure your own port filterings.
21:00:53 <fizzie> Maybe it was at Dystopia or something, those guys were pretty informal.
21:01:21 <fizzie> ("Configure" as in "select one out of three options".)
21:02:47 <elliott> ais523: "Charles Walker (CWW) has invited you to join Empire Avenue, the social stock market game! To accept the invitation and sign up – it's totally free! – click on this link:"
21:02:54 <fizzie> Got merged into Sci.fi, which got merged into Saunalahti, which currently exists as a brand of Elisa, I think.
21:02:59 <ais523> elliott: it's a pretty common name
21:03:10 <elliott> ais523: yes, but most people don't have my email
21:03:13 <ais523> c_walker (or c-walker) was an Agoran for a while, I think
21:03:16 <elliott> i only know of one cwalker
21:06:14 <ais523> elliott: send him back similar spam, but which actually goes up to the signup page for an esolang
21:06:21 <ais523> err, if any esolangs have one of those
21:06:23 <elliott> ais523: signup page for an... esolang?
21:06:31 <elliott> what were you trying to say :D
21:06:59 <ais523> elliott: well, one proto-esolang idea I've had is that it's just like a normal language (or maybe BF, or whatever), except that the spec says that you have to send me money every time you run a program with it
21:07:02 <oklopol> why couldn't you have a signup page for an esolang
21:07:03 <ais523> otherwise, it's not a conforming impl
21:07:13 <elliott> ais523: i'd be the first to run a program! prolly not
21:07:16 <ais523> the spending money is the only interesting feature
21:07:45 <fizzie> Found an ISP review article in a Finnish computer rag from '98; it goes: "even though IRC addicts curse the new HPY calling costs and cut their calls every half an hour --"; I so remember doing that.
21:08:10 <ais523> wow, I think this paper was written on a typewriter with underlines and multiline square brackets drawn by hand
21:08:20 <ais523> also, all the pages are slightly different sizes
21:08:47 <fizzie> (HPY is the Helsinki area phone service; they used to have a per-call price for the "weekdays 17-08 + weekends" time range; but then they changed it to per-call price + per-minute addition for minutes extending 30, so that for asymptotical optimum it was best to hang up every 36-37 minutes or so.)
21:10:06 <pikhq> fizzie: Per-call pricing for phone in general?
21:10:22 <fizzie> Yeah, it was before widespread use of computars, really.
21:10:31 <pikhq> I'm pretty sure that here, at least "local" calls have been flat rate per-month.
21:10:53 <pikhq> With a per-minute charge for "long distance" which has become completely stupid.
21:11:47 <pikhq> Where the definitions of either are slightly arbitrary, but *usually* the town you're in and the surrounding area will all be local.
21:11:58 <fizzie> Right; well, it was something like that except a (really rather small) per-call cost (and pretty small monthly payments), before they had to change it because people started making 12-hour "calls".
21:12:25 <pikhq> The phone companies here generally didn't give a rip about 12-hour calls.
21:12:41 <fizzie> I'm guessing they just saw a chance to make more money from Internet addicts.
21:12:47 <pikhq> Well, at least in my experience.
21:13:20 <fizzie> Anyway, when the phone companies went to per-call-time pricing for local calls too, the ISPs by contrast started offering flat-rate monthly prices.
21:13:49 <pikhq> I do know that Internet addicts here would often get a second phone line installed...
21:13:58 <Phantom__Hoover> <pikhq> With a per-minute charge for "long distance" which has become completely stupid. ← I really don't get that.
21:14:22 <elliott> should not be displayed directly, but are expected to be found in ordinary text files (such as backspace and tab).
21:14:25 <elliott> text files have backspaces in them?
21:15:01 <oklopol> well they can start with backspaces
21:15:34 <fizzie> Phantom__Hoover: At least around here "long-distance" used to be "goes into the area of another regional telephone service provider", and then it sort-of makes sense, since the other company's going to charge from the use of their network.
21:15:35 <pikhq> Ah, apparently the US domestic phone services offer not-very-expensive unlimited long distance now.
21:16:07 <fizzie> "Landlines" are pretty much dying now anyway, so the pricing model is a bit moot.
21:16:23 <pikhq> Erm, s/domestic/landline/
21:16:28 <pikhq> Phantom__Hoover: Once upon a time, long-distance calls entailed a lot of pricey infrastructure.
21:16:45 <pikhq> Now, they entail having Internet access.
21:17:31 <pikhq> (up until like 30 years ago, the entire phone system was analog...)
21:18:09 <fizzie> Some statistics say that in 2009 in Finland <30% of homes had landlines, while in Sweden the same was >55%.
21:18:47 <pikhq> elliott: Nowadays, your phone service gets digitised at the other end of the last mile.
21:19:12 <pikhq> Well, in some cases it actually gets digitised *during* that.
21:19:13 <Phantom__Hoover> fizzie, yeah, but you guys see internet access as a human right.
21:19:25 <pikhq> Say, at a box on your block, or even right at your POP...
21:19:30 <fizzie> With ISDN links it's digital all the way; those got some small amount of popularity in Finland (and more in Germany), but never really much.
21:19:48 <fizzie> I understand in most places it never caught on even to the degree it did here.
21:19:55 <pikhq> After that, it's all fairly low-fidelity PCM.
21:20:20 <pikhq> (this low-fidelity PCM is actually the *cause* of 56 kbps being the upper bound on modems)
21:20:26 <ais523> gah, the University doesn't seem to have rights to a copy of a particular paper I'm looking for, and it isn't very important
21:20:35 <fizzie> 8 kHz, 8 bits; 64 kbps.
21:20:41 <ais523> so I don't want to ask them to pay to obtain them
21:20:58 <elliott> ais523: that is... almost Sgeo-esque
21:21:06 <elliott> except with Sgeo it's feeling responsible for things
21:21:56 -!- variable has joined.
21:22:09 <pikhq> Yay, mu-law or A-law PCM.
21:22:39 <ais523> the paper isn't found by Google at all; looking in more specialised search engines I managed to discover its existence, but nothing more than that
21:22:44 <Gregor> Did fizzie or nooga talk while I was gone?
21:23:19 <elliott> nooga is only medium-active :P
21:23:28 <Gregor> Oh, in fact, fizzie was talking four minutes ago :P
21:23:42 <elliott> i thought you wanted to be old
21:23:52 <Gregor> I don't want to be oldEST.
21:23:55 <quintopia> i don't understand how we went from a telegraph network to an analog network and couldn't go back to the old digital stuff...
21:23:59 <fizzie> Oh well, the landlines still make better-quality calls than most mobile phone networks. (I guess with 3G's AMR-WB that's a bit arguable.)
21:24:03 <ais523> also, sorting directories by last-modified time is a surprisingly useful sort order
21:24:07 <elliott> Gregor: you're the real-world-oldest! ok not even close
21:24:08 <ais523> although dotfiles and my IRC logs normally win
21:24:17 <Gregor> elliott: My age is a mystery X-P
21:24:19 <quintopia> seems like they'd keep that stuff around. replace the electromechanical switchs with electronic repeaters...
21:24:20 <elliott> i think impomatic is the oldest one we know
21:24:25 <elliott> Gregor: 27 and five quarters
21:24:43 <ais523> I don't think of impomatic as being particularly real-world old
21:24:54 <elliott> definitely >=40 (if i'm wrong: sry impo)
21:25:02 <ais523> that doesn't surprise me, although it's not what I expected
21:25:04 <elliott> but like 90% sure it's >=40
21:25:08 <elliott> oerjan is 38 or maybe 39 now
21:25:24 <ais523> and will never catch up
21:25:26 <elliott> I think Ruxglo is probably pretty old since he goes on about computers being so bloated these days
21:25:32 <elliott> but i'm less old-farty about it
21:25:32 <quintopia> Gregor: because you haven't C-sectioned the inner child yet :P
21:25:58 <pikhq> quintopia: The telegraph network is still running.
21:26:07 <Gregor> quintopia, elliott: Hint: my age is between 17 and 71
21:26:11 <pikhq> quintopia: It, too, runs over the Internet.
21:27:00 <quintopia> let's just assume he's 44 with a childlike spirit from here out
21:27:01 <fizzie> elliott: If you had some sort of an IRC clippy, it'd pop up now and say "I see you're doing a binary search. Would you like any help with that?"
21:27:24 * oerjan thinks his dad had ISDN at one time
21:27:29 <elliott> He said he was 14, I think, 1999-2001
21:27:34 -!- augur has joined.
21:27:37 <elliott> So he's 24 or 25, basically :P
21:27:50 <elliott> quintopia: well i knew that
21:27:54 <pikhq> quintopia: For instance, the recent Wikileaks stuff? Those are *leaked telegrams*.
21:28:00 <elliott> trollaximafyagratorybitchatok
21:28:12 <Gregor> FFFFFFFFFFFFFFFFFUUUUUUUUUUUUUUUUUUUUU
21:28:21 <quintopia> pikhq: oh right. thankfully, we don't have to write STOP at the end of sentences now
21:28:22 <fizzie> oerjan: It was called the "I Still Don't Need it" network, too. Or the "I Smell Dollars Now" network.
21:29:53 <oerjan> <elliott> oerjan is 38 or maybe 39 now <-- 40
21:30:26 <fizzie> Isn't the Wikileaks stuff from SIPRNet? I don't think that's exactly "the telegraph network", since it's a closed non-civilian thing.
21:30:50 -!- cheater99 has quit (Ping timeout: 240 seconds).
21:30:59 -!- cheater00 has joined.
21:31:09 <pikhq> fizzie: Okay, it's not so much a single network as it is a bunch of networks with similar, archaic standards.
21:31:16 <elliott> The Secret Internet Protocol Router Network (SIPRNet) is "a system of interconnected computer networks used by the United States Department of Defense and the U.S. Department of State to transmit classified information (up to and including information classified SECRET) by packet switching over the TCP/IP protocols in a 'completely secure' environment".[1] It also provides services such as hypertext document access and electronic mail. As such, S
21:31:16 <elliott> IPRNet is the DoD’s classified version of the civilian Internet.
21:31:23 <pikhq> fizzie: And the diplomatic cable leaks used one such archaic standard.
21:31:30 <pikhq> elliott: They tunnel Telex over it.
21:33:07 <fizzie> I'm surprised they don't use X.400 messages instead.
21:34:21 <pikhq> *Anyways*, it's pretty funny how all of that stuff is still in use just because it takes time to stop using it.
21:35:48 <fizzie> "UUCP was in use over special-purposes high cost links (e.g., marine satellite links) long after its disappearance elsewhere[4], and still remains in legacy use." Yes, it certainly seems that way.
21:36:23 <pikhq> And it'll probably be quite a while for UUCP to actually *stop*. Especially as it's easy to tunnel over IP.
21:36:43 <elliott> but it should have been done with the normal cp command >:D
21:37:34 <pikhq> An email address that needs to go through UUCP would be pretty funny to have.
21:37:37 <fizzie> "Unix-to-Unix Copy Program," said PDP-1. "You will never find a more wretched hive of bugs and flamers. We must be cautious." [From DEC WARS.]
21:38:10 <pikhq> "Oh, yeah, my email address is pikhq@uucp-proxy.com!foo!bar!baz"
21:39:28 <fizzie> Back in the '90s I think the emails sent from a Finnish insurance company went through a X.400-to-email-and-back gateway; this made for some pretty horrible-looking email addresses.
21:39:38 <fizzie> Maybe not as silly as bang paths, but long anyway.
21:40:01 -!- cheater00 has quit (Ping timeout: 240 seconds).
21:40:14 <pikhq> Ah, sorry "foo!bar!baz!pikhq@uucp-proxy.com"
21:40:18 <coppro> rumour has it the smtp daemon here at the university will forward mail with no envelop addresses
21:40:48 <ais523> bah, even formal papers can be really vague
21:41:14 <ais523> this paper says "alternating", and ababababab is obviously allowed, but I'm not sure if abababababa is
21:41:18 <ais523> and it's not obvious from context either
21:41:32 <Phantom__Hoover> <coppro> rumour has it the smtp daemon here at the university will forward mail with no envelop addresses ← surely that's easy to test?
21:42:20 <coppro> Phantom__Hoover: that requires me to figure out where it is
21:42:22 -!- cheater00 has joined.
21:42:26 <pikhq> elliott: Well who is actually *on* UUCP any more for email?
21:42:29 <coppro> and that sounds like effort
21:42:58 <pikhq> elliott: Okay, okay, you can just go to "foo!bar!baz!pikhq"
21:43:21 <elliott> are there any public uucp nodes still around :P
21:43:30 <fizzie> pikhq: Isn't the usual way to do {foo,fii,faa}!bar!baz!pikhq so that the user has easier time finding the starting point? :p
21:44:15 <ais523> any ideas here as to whether this paper intended (in a C-like syntax) {auto semaphore s; grab(s);} to be legal?
21:44:22 <ais523> (i.e. the semaphore is grabbed and then immediately goes out of scope)
21:44:43 <ais523> you could make a case either way
21:44:49 <elliott> ais523: that sounds "unsound"
21:44:53 <elliott> so i'd expect not, if it's like
21:44:57 <elliott> trying to make type-safe semaphores
21:45:05 <ais523> it's not trying to make type-safe semaphores
21:45:11 <elliott> no idea without more context, then
21:45:16 <ais523> it's just trying to express them in a strongly typed language
21:45:20 <ais523> and I don't have enough context to tell either
21:45:42 <fizzie> The X.400-derived addresses were something like "G=Ernie;S=Egbert;O=Company;OU=Sales;PRMD=strngabbrv;ADMD=strngabbrv;C=fi@x400.wht.evr.isp.fi", possibly with some of the more special pre-@ characters escaped in some way.
21:45:51 <ais523> the paper says that when a semaphore goes into scope, there's an alternating sequence of grabs and releases, then it goes out of scope
21:45:55 <ais523> that's the semantics of a semaphore
21:46:29 <ais523> and "alternating" is such a tricky word there
21:46:58 <elliott> ais523: I'd say it's invalid, then
21:47:10 <elliott> ais523: I would imagine that they grabs and releases must be balanced
21:47:15 <elliott> go on that assumption unless you find a contradiction later
21:47:21 <fizzie> Is this paper somewhere?
21:47:29 <Sgeo> Well, Atomo is not multiple-dispatch, despite claims to the contrary
21:47:30 <elliott> fizzie: I doubt ais523 would BREAK THE LAW like that.
21:47:31 <ais523> well, the paper does allow semaphores to be initialised in a grabbed state, which makes me think they needn't be
21:47:49 <ais523> actually, this one is: http://www.cs.bham.ac.uk/~drg/papers/scc-tcs.pdf
21:48:14 <elliott> <ais523> well, the paper does allow semaphores to be initialised in a grabbed state, which makes me think they needn't be
21:48:19 <elliott> don't you know that Ghica guy?
21:48:21 <Sgeo> http://raynes.me/logs/irc.freenode.net/atomo/today.txt
21:48:25 <ais523> yes, he's my supervisor
21:48:31 <elliott> ais523: ask him, you doofus :P
21:48:49 <ais523> and besides, I'm trying to figure out what the paper says, not what he means
21:48:56 <ais523> it's the sort of definition that can easily be changed on a whim
21:49:03 <elliott> [15:46:35] BrianRice-work: pattern-matching *is* multiple dispatch plus destructuring and some other conditionalized flow-control
21:49:05 <ais523> as long as you own up to it
21:49:07 <elliott> i hate that characterisation
21:49:20 <elliott> i guess it works for OOP languages
21:49:23 <ais523> elliott: it isn't incorrect, just useless
21:49:29 <elliott> ais523: it's incorrect in Haskell!
21:49:37 <elliott> and every language where pattern matching is common, pretty much
21:49:44 <ais523> because multiple dispatch doesn't really have meaning in Haskell
21:49:53 <ais523> but if it /did/ have meaning, it would be one way to define pattern matching
21:59:50 -!- variable has quit (Read error: Connection timed out).
22:00:25 <j-invariant> speaking of minecraft, http://www.youtube.com/watch?v=x2mCDkqXki0
22:01:31 -!- variable has joined.
22:05:25 -!- variable has quit (Max SendQ exceeded).
22:05:41 <ais523> j-invariant: wrong channel...
22:06:34 -!- variable has joined.
22:07:30 <ais523> hey, now I get to enforce topicality here again!
22:07:37 <ais523> and I thought we'd all but given up on that years ago
22:07:49 <elliott> ais523: no you don't, only for minecraft
22:07:50 <ais523> that said, we do launch into ontopic conversation whenever something new in the world of esoprogramming comes up
22:07:52 <elliott> other offtopic things still belond here
22:07:55 <ais523> elliott: well, sometimes
22:08:14 <elliott> admittedly, I don't move offtopic, non-esoteric conversations in -minecraft here, due to that being just silly
22:08:28 <ais523> my research is sufficiently weird that I can almost claim it's ontopic here
22:08:41 <ais523> I mean, who uses lazy imperative languages anyway?
22:08:43 <elliott> oh, I'd say it's on-topic, in that it's relevant to all our interests :P
22:08:55 <coppro> probably someone at this university
22:08:57 <ais523> I just need to break through the mathsese to explain it here
22:09:02 <elliott> admittedly, I can conceive of an esolanger who wouldn't be interested in your research coming in here, but I doubt they'd be well-liked
22:09:07 <coppro> ais523: I mean waterloo
22:09:12 <ais523> well, not for programming, just as an intermediate step
22:09:17 <elliott> ais523: hey there are plenty of people in this channel fluent in mathsese :P
22:09:25 <oklopol> regular languages also have to following characterization, methinks: let M be a finite monoid, and let h be a monoid homomorphism from \Sigma^* to M, then the preimage of every subset of M is regular, and all regular languages are obtained this way
22:09:31 <ais523> yes, but it's a particular style of mathsese
22:09:40 <ais523> I struggle with oklopol's, for instance, just because I'm not used to it
22:09:44 <ais523> and probably vice versa
22:09:57 <elliott> no oklopol is just a genius who understands everything
22:11:20 <oklopol> i'm too tired to see if that characterization is true
22:11:54 <ais523> hmm, this website that gives the information needed for citing papers just puts it in a text file called "science"
22:13:23 <ais523> hmm, this report cites six different papers by my supervisor
22:13:30 <oklopol> okay so at least it's true for unary languages
22:13:35 <ais523> somehow I'm not surprised he took me on as a PhD student, now
22:13:39 <ais523> he must have known it'd get him more citations
22:13:55 <ais523> (it's quite a small field, there's something like six or eight people in it)
22:14:16 <coppro> lazy imperative languages?
22:14:22 <elliott> ais523: it's quite a strange field, i don't understand how anyone can get a phd without being converted to the functional order ;D
22:14:36 <oklopol> that's cool because mine is too
22:14:45 <elliott> oklopol: but do you play football on it
22:14:45 <ais523> elliott: oh, they're functional as well
22:14:58 <ais523> we compile imperative languages into lazy imperative languages into call-by-name functional languages into hardware
22:15:08 <ais523> but there are higher-order functions even in the original imperative languages
22:15:10 <elliott> side-effects, who cares about the world, i don't get it, the world sucks
22:15:11 <coppro> ok, that's pretty narrow
22:15:26 <coppro> oklopol: what's your field?
22:15:27 <ais523> because computer scientists are of the opinion that a language isn't really a language at all if it doesn't do higher-order functions
22:15:30 <elliott> ais523: higher-order functions are a requirement for being called a language :)
22:15:36 <oklopol> coppro: currently, it's picture languages
22:15:39 <elliott> ais523: /functional/ is when a language /just/ has functions
22:15:46 <elliott> or, well, ADTs are also acceptable
22:15:48 <oklopol> it's not really a "field" :P
22:16:00 -!- cheater00 has quit (Ping timeout: 240 seconds).
22:16:01 <elliott> but yes, languages without higher-order functions aren't
22:16:12 <ais523> elliott: I'm trying to figure out if our functional languages that go straight into hardware have side effects or not
22:16:31 <ais523> I think they cheat by effectively using an IO monad, and hiding it better than Haskell does
22:16:39 -!- Wamanuz has quit (Remote host closed the connection).
22:16:54 <elliott> the haskell conspiracy denies these awful allegations of the IO monad's impurity!
22:16:58 -!- cheater00 has joined.
22:17:08 <elliott> (that they are true is besides the point)
22:17:17 <ais523> well, ICA has a type "com" which is basically just an IO action
22:17:23 <ais523> and all the other types are IO actions too
22:17:30 <elliott> you also have com -> exp -> exp
22:17:48 <elliott> what's the type of the read-char function in ICA
22:18:05 <elliott> it has to be an exp, right?
22:18:14 <elliott> your type system is more like an effect system
22:18:17 <ais523> or, well, exp x exp x exp x exp x exp x exp x exp x exp
22:18:21 <elliott> as it doesn't type the actual /values/
22:18:21 <ais523> because expressions are booleans
22:18:54 <ais523> my whole thesis is basically "you can use type systems for anything, even hardware"
22:18:55 <elliott> ais523: obviously, your only type should be .5 bits
22:19:13 <ais523> so we use type systems to represent anything that sensibly needs representing
22:19:24 <elliott> that way, it has 1.5 values
22:19:37 <elliott> and if you stick two of them together, you get a trit that you can just ignore one value of
22:19:38 <ais523> err, that doesn't map into hardware well
22:19:44 <oklopol> so anyway regular unary = eventually period binary sequences (where s_i = 1 iff word of length i is in the language);so take the monoid with one generator a, and a^m = a^k for m < k, so if you start from 1 and start multiplying by a, you get 1, a, a^2, ..., a^k, ..., a^{m-1}, a^k and the loop starts repeating, now it's obvious how to get eventually periodic sequences as preimages, just take homomorphism h(b) = a where your language is
22:19:51 <ais523> also, two 1.5-valued numbers gives you a 2.25-valued number
22:20:00 <oklopol> you get 1, a, a^2, ..., a^k, ..., a^{m-1}, a^k and the loop starts repeating, now it's obvious how to get eventually periodic sequences as preimages, just take homomorphism h(b) = a where your language is unary over b, and take preimages of 1, a, a^2, ..., a^k to get the nonperiodic part, and preimages of a^k, ..., a^{m-1} to get the periodic part
22:20:06 <oklopol> i should get an irc client
22:20:22 <ais523> oklopol: mIRC is an IRC client, sot of
22:20:29 <elliott> ais523: er, what's half of a trit then... oh wait, that's easy
22:20:31 <oklopol> or did you mean it didn't come through as in, nonsense filter caught it in your brain
22:20:44 <ais523> oklopol: the end of the line was cut off the first time
22:21:03 <oklopol> yeah well luckily you have it now
22:21:07 <elliott> ais523: ok wait, a trit is 1.5 bits, right?
22:21:16 <ais523> no, it's log 3 / log 2 bits
22:21:16 <oerjan> <oklopol> i'm too tired to see if that characterization is true <-- looks plausible to me, for one direction let M = S^S where S is your DFA states, for the other let S = M. i think.
22:21:30 <ais523> (the nice thing about that formula is that it doesn't matter what base you take the logarithms to)
22:21:37 <elliott> ais523: right, so how many bits is half a trit, i'm getting myself really confused
22:21:46 <ais523> elliott: log 3 / log 2 / 2
22:22:05 <ais523> and log 3 / log 2 is not log 1.5, log 3 - log 2 is log 1.5
22:22:22 <ais523> apparently log 3 / log 2 / 2 is 0.79248125
22:22:25 <elliott> ais523: I conclude, then, that exp should hold log(3)/2log(2) bits.
22:22:27 <oklopol> so S = M, then when you read a symbol, take its image, and multiply what you have by that
22:22:45 <elliott> ais523: to make a boolean, pack two of them together, and treat 2 as file-not-found
22:22:57 <ais523> elliott: unfortunately the semantics we're using requires all storage to have an integral number of states
22:23:01 <ais523> as we're storing things in regular expressions
22:23:07 <elliott> ais523: that's a shame, that you have to redo your semantics
22:23:17 <ais523> (the actual regular expressions, not regexps)
22:23:34 <oklopol> oerjan: yeah those look correct to me
22:23:37 <elliott> ais523: this is making me think of turkey bomb
22:23:57 <oklopol> not that you weren't sure already, prolly
22:24:53 <elliott> ais523: hmm, I want to make that log(3)/2log(2) esolang now
22:25:07 <elliott> ais523: obviously, you'd have to have an even number of variables at any one time (if you have an odd one, an extra one is added)
22:25:15 <elliott> ais523: then, the variable storage is just an array of trits
22:25:23 <elliott> ais523: but how does one extract the lower half-trit out of a trit?
22:25:36 <oklopol> oerjan: for S^S, the language is preimages of functions that map some initial to a final?
22:25:39 <elliott> i.e., if this was two bits, then it'd be (x&1) for 1 bit
22:25:43 <elliott> what's the equivalent for a trit, and half a trit?
22:26:12 <oklopol> yeah surely that's what it is
22:26:24 <ais523> elliott: make sure you add my nondeterministic lazy maximum function
22:26:24 <elliott> oerjan: fail: http://esolangs.org/wiki/User:Yasser
22:26:31 <elliott> ais523: that's not an answer to my question :D
22:26:38 <elliott> how do you extract the lower half-trit from a trit
22:26:58 <ais523> (it takes two of its three arguments in the order they're given, and works out the maximum value of them, not evaluating the second if the first was already at the maximum possible value)
22:27:07 <ais523> elliott: you don't, the operation doesn't make sense
22:27:31 <elliott> ais523: a trit is composed out of two half-trits, ergo there is some way to extract them
22:28:09 <elliott> ais523: and, obviously, the result is 0 <= result < 3^(log(2)^2 / 2)
22:28:14 <ais523> haha, not only do I have six papers from my supervisor cited, but they're dated 2006, 2007, 2008, 2009, 2010, 2011
22:28:24 <elliott> excuse me ais523, we're on the verge of a breakthrough here!
22:28:26 <ais523> (I coauthored the 2010 and 2011 papers)
22:29:46 -!- Wamanuz has joined.
22:29:57 <oklopol> the esolang page links to their blog, and the blog maps to the esolang page
22:29:59 <elliott> oklopol: some random crap ->bf language that the creators are treating all enterprisey
22:30:08 <elliott> they made a blog post announcing they made a page on esolang
22:30:16 <elliott> which makes the evil man in me want to have it deleted :D
22:30:24 <elliott> it says nothing about the language
22:30:30 <elliott> surely that's criteria for deletion? ais523?
22:30:37 <oklopol> well yeah that was very funny and all, but i'd like to know what the language is like, because that would be much funnier
22:30:38 <elliott> it's basically marketing material with not a _single_ bit of info about the actual language
22:30:44 <ais523> elliott: I haven't looked at it yet
22:30:51 <ais523> but that's par for the course for non-eso langs, isn't it?
22:31:00 <elliott> ais523: there is /no/ info about it /anywhere/
22:31:03 <oerjan> <oklopol> oerjan: for S^S, the language is preimages of functions that map some initial to a final? <-- yeah
22:31:27 <coppro> oklopol: is piet considered the leading research in your field?
22:31:34 <elliott> ais523: the _entire_ info about the lang that can be gleaned from their esowiki page and blog: The syntax is "basically" inspired by Tbf.
22:31:34 <oklopol> oerjan: i like how i couldn't even see S = M :D
22:31:38 <elliott> the only info about the implementation:
22:31:45 <elliott> ClearBF is based on a C implementation using the Flex library to generate the lexical analyser for the ClearBF language. We have implemented the majority of algorithms used when translating the ClearBF code to the BF. However we've used some famous Brainfuck algorithms.
22:31:45 <oklopol> maybe i should consider sleeping...
22:31:49 <elliott> ais523: the rest is a development team list
22:32:05 <elliott> and the blog is filled with crap like this:
22:32:06 <elliott> [[The objective of “Compilation Project” at ENSIAS – Rabat is to apply the content of the Compilation course on the development of a compiler, which is a relatively large application. The project is therefore both a “compilation” project and a “software engineering” project. The main points are: compliance with requirements, software design, validation and verification techniques, quality process.
22:32:06 <elliott> The software, which is to be developed in C programming language with Flex (The Fast Lexical Analyzer), is a compiler for a simple language. In our case, we’ve created the ClearBF language. This theme has been chosen firstly because the requirements can be understood quickly, and secondly because it allows a deeper understanding of the underlying concepts of programming languages.
22:32:07 <elliott> The projects was developed with supervised and non-supervised lab work sessions.]]
22:32:20 <elliott> now this is /partly/ because i'm evil, but I do think it's a useless page even by esowiki's standards
22:32:34 <oklopol> coppro: robozzle was pretty close to what i'm doing research on, actually
22:32:35 <elliott> and I think new articles about languages that are clearly already designed should definitely include _some_ info about the language
22:32:41 <elliott> especially if the language creator is writing it
22:33:03 <ais523> they don't actually say it, but I get the feeling that ClearBF is actually an implementation of TBF, which is a trivial BF derivative
22:33:12 <ais523> the link to a page describing TBF wouldn't make sense otherwise
22:33:20 <elliott> ais523: they say the syntax is inspired
22:33:28 <quintopia> mark it for deletion so as to force them to expand the description
22:33:37 <elliott> also, tbf is a to-brainfuck compiler, not a brainfuck derivative
22:33:41 <ais523> I think a user talk note is probably simplest
22:33:48 <ais523> elliott: well, its input is a BF derivative
22:33:51 <elliott> ais523: probably, but they sound irritating :D
22:33:58 <elliott> and fail at wikis: http://esolangs.org/wiki/User:Yasser
22:34:55 <oklopol> The ClearBF is based in a C implementation using the Flex library to generate the lexical analyser for ClearBF language.
22:35:18 <Sgeo> Sounds enterprisey
22:35:23 <elliott> ais523: if you don't get a meaningful response i say delete
22:35:33 <oklopol> sounds like something fungot would say
22:35:35 <Sgeo> It's clearly meant for big businesses
22:35:36 <elliott> We hope to be able to describe the ClearBF language at a later date blah blah blah blah blah blah blah! ~~~~
22:35:46 <elliott> ais523: in which case I say delete until they are going to tell us anything ...
22:36:00 <oklopol> big businesses run by bots written in befunge
22:36:05 <elliott> ok, so really i just want their self-congratulatory blogpost announcing their innovative creation of a page on the wiki to be invalidated
22:36:33 <quintopia> elliott: see the clearbf talk page
22:36:46 <j-invariant> elliott: it already is, just by its nature
22:36:50 <Sgeo> Is it someone's school project?
22:36:58 <elliott> Sgeo: many people's, it seems
22:37:04 <elliott> This project was a scholar project at ENSIAS - Morocco to build a compiler in C/Flex. The team is supervised by Dr. Karim Baîna and Mrs. Mounia ABIK. Members of the team are:
22:37:04 <elliott> Mohammed ASSOUKTI (Team’s leader)
22:37:15 <elliott> requires a phd and a team leader to write a to-bf compiler!
22:37:31 <elliott> here's my predicted grammar for the language
22:38:03 <ais523> we have some information from the use of flex but not bison
22:38:12 <elliott> <prog> ::= 'forward' | 'bacward' | 'add1' | 'dec1' | 'loopstart' | 'loopend;' | 'read_char' | 'print_char;'
22:38:13 <ais523> but I'm not entirely sure what it means
22:38:15 <j-invariant> "I have no plans to learn UML before I die. I'm sure they'll be teaching it in hell."
22:38:18 <elliott> ^^^ ClearBF Enterprise Langauge
22:38:37 <elliott> every program consists of one statement
22:38:46 <elliott> putting the parser in a loop proved impractical!
22:38:54 <Phantom__Hoover> Even ignoring that, it completely fails to balance loopss.
22:38:55 <elliott> but it has the same commands as brainfuck
22:38:58 <elliott> so obviously it's turing complete
22:39:05 <elliott> Phantom__Hoover: i'm joking if it's not clear
22:39:11 <elliott> <elliott> here's my predicted grammar for the language
22:39:20 <oklopol> Phantom__Hoover is just joking
22:39:36 * Sgeo splashes acid on Phantom__Hoover
22:39:39 <oklopol> i'm actually joking, of course i got that.
22:39:44 -!- Phantom__Hoover has changed nick to potential_of_Hyd.
22:39:52 -!- BeholdMyGlory has quit (Remote host closed the connection).
22:39:59 -!- potential_of_Hyd has changed nick to Phantom_Hoover.
22:41:48 <ais523> should I apologise to Evince for making it open eight PDFs at once?
22:42:03 <ais523> somehow I doubt Adobe Reader could manage that without slowing the computer down to an unusable pace
22:42:08 <elliott> ais523: is it struggling with that? :P
22:42:46 <ais523> I didn't even notice I had so many PDFs open, except via the crowded taskbar
22:42:57 <elliott> ais523: btw, get ready for gnome to become terrible: http://gnome3.org/
22:43:00 <ais523> also, a Word document, using openoffice in its pretend-to-be-a-pdf-reader mode
22:43:06 <oklopol> ais523: Tbf is not exactly trivial from what i can tell
22:43:08 <elliott> ais523: their FAQ strongly suggests gnome-panel is going to become unmaintained, but I plan to maintain it
22:43:14 <oklopol> not very complicated either
22:43:21 <elliott> ais523: this of course includes their stupid gnome-shell crap...
22:43:27 <ais523> anyway, Ubuntu apparently invented Unity just because they didn't like gnome-shell
22:43:28 <elliott> ais523: ubuntu are switching to Unity, though, which is unusable in entirely different ways
22:43:36 <elliott> ais523: and Ubuntu Desktop is going to switch to it
22:43:44 <ais523> also, what's with all the retheming?
22:43:54 <elliott> ais523: but it's all right, I plan to make sure gnome-panel keeps working until hell freezes over
22:44:19 <ais523> and why would they integrate IM into the shell?
22:44:21 <elliott> i hope it doesn't need too many bugfixes :D
22:44:32 <elliott> ais523: but in @, everything is integrated into the shell!
22:44:37 <ais523> that's the sort of thing that it makes sense to have a separate program for
22:44:39 <ais523> elliott: then it isn't a shell
22:44:46 -!- macrohauler has joined.
22:44:51 <elliott> ais523: everything is integrated into the kernel too :D
22:44:57 <oerjan> <elliott> Sgeo: that cures cancer <-- i thought from yesterday that it was _alkaline_ water which cured cancer?
22:45:11 <elliott> oerjan: no no, sgeo's alluded-to girlfriend(?) thinks vitamin c cures cancer
22:45:15 <ais523> "Our system settings have been completely redesigned for GNOME 3, making them easier to use than ever before. There are some great new features in our system settings too, such as the ability to use Flickr images as desktop backgrounds."
22:45:16 <elliott> and that chemotherapy basically does the same things as vitamin c
22:45:25 <elliott> ais523: (you could make an argument that @ actually has almost no kernel)
22:45:30 <elliott> ais523: (although it isn't really a microkernel)
22:45:45 <ais523> that line I quoted really contradicts itself in style
22:45:48 <macrohauler> good day everyone *trying not to interrupt
22:45:55 <elliott> ais523: I'm not going to start maintaining all of GNOME 2, though, so everyone's going to have to deal with that... just with a panel
22:45:59 <ais523> don't worry about interrupting, we jump topics really quickly
22:46:04 <ais523> and can carry on multiple situations simultaneously
22:46:09 <ais523> elliott: can you make it work in xmonad?
22:46:19 <ais523> *multiple conversations simultaneously
22:46:26 <elliott> also, /me doesn't think of ais523 as the kind of person to use xmonad
22:46:38 <coppro> ais523: the answer is always yes
22:46:39 <ais523> elliott: xmonad tries to treat it like any other window
22:46:40 <macrohauler> im known as fr34k at the forums, so i was wondering, could i create a new page using my current nick and then redirect from the fr34k page?
22:46:42 <coppro> everything can be made to work in xmonad
22:46:50 <elliott> fr34k, haven't you been in here before?
22:46:55 <fizzie> ais523: That's not very impressive evidence of doing multiple convesations simultaneously, getting the words all giraffe like that.
22:47:01 <elliott> macrohauler: certainly, but i think sysops can rename accounts
22:47:03 <ais523> macrohauler: oh, usernames?
22:47:08 <ais523> elliott: no we can't, I don't have the perms for that
22:47:11 <oerjan> elliott: i was referring to kurzweil i believe, not SATG
22:47:16 <elliott> or is that just beaeurjocrats?
22:47:17 <ais523> mere sysops can't rewrite history
22:47:22 <elliott> macrohauler: just redirect then
22:47:27 <ais523> it's crat-only on wikipedia too
22:47:47 <elliott> macrohauler: stick around here, we're nice! well sometimes.
22:47:49 <ais523> elliott: we can distort history to take credit for other people's edits, but not attribute them to people other than ourselves
22:48:08 <ais523> except by merging two edits and attributing the resulting merge to the second user
22:48:17 <elliott> do that, for every one of macrohauler's edits
22:48:18 <macrohauler> elliott: i was actually thinking of doing it, have started using irc much more lately
22:48:21 <ais523> note that this probably isn't intentional, and there are rules saying "don't do that"
22:48:42 <nooga> elliott: is the mc server up?
22:48:50 <elliott> nooga: um yes but you don't have the address afaik
22:48:52 <macrohauler> so wait, should i make a redirect or is my username getting changed x)
22:48:59 <ais523> macrohauler: put a note on your userpage saying that you and Fr34k are the same person; also, make sure you have both accounts registered to stop someone else registering one
22:49:03 <elliott> only graue is a bear-o-crat and he's never online
22:49:11 <elliott> macrohauler: don't expect too much on-topicness in here btw :p
22:49:15 <ais523> not that it's likely to happen, but we had an account impersonation troll once
22:49:20 <Sgeo> Would I make a better beu..I can't spell
22:49:39 <ais523> elliott: crats can't rename users on Esolang either, the extension to do so isn't installed; it has to be done directly on the database
22:50:40 <oklopol> Sgeo: you would make an awesome beu
22:51:04 <oerjan> <elliott> oerjan: SATG? <-- i'm disappointed.
22:51:17 <ais523> "Spreading the word about GNOME 3 is a great way to help out. A positive tweet, dent, status update or post is always appreciated. You can also join our Facebook group or become a friend of GNOME. Additionally, you can help out with one of the many varied tasks involved in producing GNOME. Just check out our contribution page for more details."
22:52:00 <ais523> Mathnerd314: what did you have before?
22:52:08 <ais523> I need to figure out the proper way to flame and/or troll you
22:52:17 <elliott> ais523: he's using arch, that's enough to flame and/or troll anyone
22:52:19 <ais523> not that I actually would, but it's an interesting intellectual exercise
22:52:23 <elliott> if you need more information than that you're bad at trolling
22:52:46 <ais523> elliott: but you need to figure out whether to go for Linux distro flamewar, Linux vs. Windows flamewar, or Linux vs. Mac flamewar
22:52:51 <elliott> Mathnerd314: oh, that's easy
22:52:59 <ais523> and don't you dislike Arch just because Vorpal used it?
22:53:02 <Mathnerd314> elliott: and I haven't switched yet, just decided too
22:53:07 <elliott> no, I used arch for a length of time
22:53:12 <elliott> it was an exercise in pain and suck
22:53:18 <elliott> and the community is worse than ubuntu's
22:53:27 <Mathnerd314> elliott: thus why I'm telling you, to see if there's a better option
22:53:27 <elliott> because the average IQ is the same, but they think they're 10x as knowledgable
22:53:33 <elliott> because they use tiling wms and translucent terminals :P
22:53:37 <elliott> Mathnerd314: uh i usually use debian
22:54:01 <ais523> I use Ubuntu atm because I actually need to get on with my life
22:54:07 <elliott> ais523: hey debian lets you do that
22:54:10 <ais523> and it's what was on here before, and I didn't want to go to the trouble of jumping distro
22:54:11 <elliott> just without so many gradients
22:54:15 <oerjan> <ais523> not that I actually would, but it's an interesting intellectual exercise <-- ais523 secretly has plans to murder us all, not that he would ever _do_ it...
22:54:17 <pikhq> elliott: So, they're basically Gentoo a few years ago.
22:54:37 <elliott> Mathnerd314: but debian isn't
22:54:42 <pikhq> (as far as I'm aware, *most* of the idiots completely and utterly bolted from Gentoo... Probably helps that Gentoo no longer has stage1)
22:54:45 <elliott> pikhq: gentoo users are smarter than arch users, unbeliveably
22:55:15 <ais523> oerjan: note that it's mathematically impossible for me to confirm that statement without lying, because you stated "secretly", so either I don't have secret plans and would lie if I confirmed it, or I do and they wouldn't be secret if I confirmed it
22:55:59 <ais523> elliott: I have decided to stick with Ubuntu Lucid for the time being, though, which is strange as I normally upgrade pretty quickly
22:56:14 <elliott> ais523: 10.10 is pretty good, IMO
22:56:21 <elliott> upgrading ubuntu is "fun" though
22:56:57 * pikhq is quite happy with Debian.
22:57:02 <ais523> it's been 50:50 for me, but every time it's screwed up I've managed to fix it
22:57:18 <macrohauler> now that we all seem to enjoy irc, i'd like some feedback on terpir from anyone interested :3
22:57:39 <elliott> i think i looked at that for one second and then moved along :D
22:57:51 -!- Tritonio has quit (Quit: Leaving).
22:57:51 <elliott> macrohauler: i don't quite see why it's embedded in irc
22:57:56 <elliott> [00:01:02] <~jrunyon> define newfunc with output ohai thar [newline] alongside output how are you? without alongside newfunc
22:58:02 <elliott> IMO, the timestamp and <> part of that is just fluff
22:58:16 <elliott> seems like a language in search of a gimmick, frankly :P but then I think that about most esolangs...
22:58:30 <elliott> "[00:00:01] would be executed 2 seconds before [00:00:05] - half the time between messages"
22:58:36 <macrohauler> is thought of to be somewhat like real-time programming in an irc client
22:58:46 <ais523> I like the way it has its own wiki
22:59:03 <elliott> ais523: yes, only the most robust, enterprise esolangs have wikis!
22:59:05 <ais523> it could make for a great competitive programming game, though
22:59:09 <ais523> try to hack each other's programs in realtime
22:59:14 <ais523> by appending bits to the end of your own
22:59:23 <elliott> ais523: hmm, appending, or spawning new threads
22:59:24 <ais523> I'd hate to actually play it, though
22:59:33 <elliott> really? that sounds awesome
22:59:35 <ais523> btw, I just emailed myself to backup some files: opinions on whether this is a good idea?
22:59:40 <elliott> you'd have to have a super-concise language
22:59:41 <ais523> elliott: I wouldn't have the concentration
23:00:00 <elliott> ais523: I email myself notes, like passwords that have to obey certain rules, that I'd normally forget
23:00:06 <elliott> because gmail's search is really good :D
23:00:14 <elliott> it's basically a ghetto note-taking app
23:00:24 <ais523> Evolution's isn't bad either, except when it permanently hides all the emails in an entire folder
23:00:30 <elliott> and I can do "from:me in:inbox" to see everything i've sent myself
23:00:36 <ais523> (they're still /there/, just there's no way to tell except screwing around with the FS directly)
23:00:52 <ais523> elliott: oh, I'd just type ais523 into the subject-or-sender box
23:01:11 <oklopol> speaking of awesome languages, we should have some clue-related ring-jerking with elliott some time soon
23:01:18 <elliott> ais523: that makes me see all my to-email-list messages
23:01:30 <elliott> oklopol: only if you've added function refs
23:01:41 <oklopol> oh shit the ball is in my court
23:01:46 <ais523> elliott: oh, I have a separate inbox for those
23:02:01 <elliott> oklopol: ...to my latest version.
23:02:15 <oklopol> well you know maybe some day
23:02:20 <ais523> oklopol: thanks for indirectly giving Esolang its first disambiguation page
23:02:29 <oklopol> actually it was pretty direct
23:02:53 <ais523> except you didn't create it yourself, or did you?
23:03:10 <oklopol> but i did directly ask someone to make it for me
23:03:25 <elliott> grr, i hate the required edit summary for anons
23:03:33 <elliott> if you ever see an anonymous edit with an edit summary like "ASflzkfdsgfdlkjx'd5" that's me :D
23:03:35 <ais523> elliott: it's an anti-spambot measure
23:03:39 <ais523> also, the spambots do that too
23:03:40 <elliott> i know, but it irritates me
23:03:45 <elliott> ais523: i know, that's why it's so great :D
23:03:46 <ais523> elliott: then get an account?
23:03:54 <elliott> ais523: i have one, but logging in is tedious
23:03:59 <ais523> set it to remember the login
23:04:02 <elliott> so i only do it to like, post talk page messages
23:04:10 <elliott> ais523: except i switch OSes too much for that and stuff :D
23:04:11 <ais523> unless you have an aversion to long-term logins
23:04:20 <ais523> elliott: hmm, there should be a way to share cookies between browsers
23:04:30 <ais523> that would really screw with the minds of certain sites, I imagine
23:05:14 <oerjan> is there any reason why you couldn't be logged in from several browsers at once?
23:05:36 <elliott> so it'd need to be cross-partition
23:05:49 <elliott> where one OS can't read one of the FSes
23:05:57 <ais523> you could use a cookie daemon
23:06:01 <ais523> on an entirely different system
23:06:09 <elliott> ais523: on my modem-router!
23:06:14 <ais523> also, ext4 sounds like the sort of system I'd expect OS X to have eventually
23:06:22 <elliott> OS X was going to use ZFS but then didn't
23:06:26 <elliott> I think HFS+ is here to stay :P
23:06:26 <pikhq> Good God, *still* no IANA allocation. What's holding them up?
23:06:27 <ais523> admittedly, I just suggested that because I like the name "cookie daemon"
23:06:31 <fizzie> There are cookie-sync systems that use external servers.
23:06:41 <ais523> pikhq: the IANA don't want to admit it's the end of the world
23:06:49 <elliott> IANA look like this right now: ;__;
23:06:51 <ais523> where do you check to see where the allocation's happened, anyway?
23:06:57 <ais523> *if the allocation's happened
23:07:02 <ais523> elliott: yep, but I mean websitewise
23:07:04 <elliott> ilari probably checks it every 20 seconds!
23:07:10 <ais523> that's one of the things I'd expect the web to be good for
23:07:11 <elliott> ergo we are the most up-to-date source
23:07:28 <elliott> ais523: there's that chart
23:07:30 -!- augur has quit (Remote host closed the connection).
23:07:33 <oerjan> the chinese government are holding APNIC hostage to prevent the inevitable rioting in the streets
23:07:43 <pikhq> http://www.iana.org/assignments/ipv4-address-space/ipv4-address-space.xhtml Here's the IANA address space registry.
23:07:46 <oerjan> it's the only possible explanation
23:07:46 <elliott> the depletion things have been predicting the current day for the past N days, btw :P
23:07:48 <ais523> APNIC = asia? or APNIC = africa? I keep forgetting
23:07:51 -!- zzo38 has joined.
23:07:59 <pikhq> APNIC is "Asia & Pacific".
23:08:06 <ais523> elliott: that's becaues it's long since depleted but nobody wants to admit it
23:08:10 <elliott> Asiapacific? be more spacific
23:08:14 <oerjan> ais523: asia. africa is still trivial in comparison.
23:08:25 <elliott> ais523: no, it's because the models think it should have happened already
23:08:25 <ais523> how did level 3 end up with two /8s?
23:08:28 <elliott> ais523: it only hasn't because IANA is stalling
23:08:29 <pikhq> AfriNIC is probably going to be the *last* RIR to be depleted.
23:08:35 <ais523> (and do they actually use them?)
23:08:37 <elliott> ais523: (previously APNIC were stalling but they have surely submitted by now)
23:09:21 <ais523> anyway, what will really cause fireworks is when 10/8 gets allocated
23:09:32 <pikhq> elliott: It's easy to tell when they started realising that IPv4 space was precious. :)
23:09:36 <ais523> it'll be like allocating 1/8 but on a much larger scale
23:09:45 <elliott> ais523: will it ever be allocated?
23:09:48 <pikhq> ais523: 10/8 cannot be allocated.
23:09:55 <ais523> I know, but it'd be funny if they did
23:09:59 <ais523> and shock people into using IPv6
23:10:05 <fizzie> Prior to CIDR (pre-'93) if you needed more than 64k addresses, /8 would have been it.
23:10:05 <elliott> (is that the correct notation?)
23:10:13 <pikhq> ais523: They'd have to actually replace every router.
23:10:13 <elliott> fizzie: this is from 1994, 1995
23:10:17 <elliott> fizzie: the allocations i quoted
23:10:31 <ais523> and 14/8 was reclaimed from a previously reserved use, after all
23:10:38 <elliott> they're probably preparing the biggest party ever
23:10:48 <pikhq> elliott: Everything but 192.168/16 is publically allocated.
23:10:50 <elliott> for some definition of probably
23:10:55 <elliott> pikhq: allocate 192.168/16!
23:11:02 <elliott> and whatever prefix of 127 is reserved!
23:11:03 <pikhq> elliott: Replace all routers!
23:11:22 <fizzie> Whole 127/8 is for loopback.
23:11:26 <pikhq> 127/8 is all loopback.
23:11:42 <fizzie> Allocate 240/4 instead, that's even bigger block.
23:11:53 <elliott> wait, no, it's /0 isn't it
23:12:01 <fizzie> (And also problematical when it comes to software.)
23:12:02 <ais523> so currently, 39/8, 102/8, 103/8, 104/8, 106/8, 179/8, 185/8 are free
23:12:07 <elliott> /0 is the entire space, right?
23:12:15 <elliott> and 127/1 is half the space?
23:12:25 <ais523> 127/1 is the same half as 0/1
23:12:27 <ais523> 128/1 is the other half
23:12:31 <pikhq> fizzie: 240/4, eh? Ouch.
23:12:34 <elliott> erm, can ips start with 0.?
23:12:37 <elliott> if not, then surely it's 1/1 and 129/1
23:12:53 <fizzie> They can, but 0/8 is reserved.
23:13:05 <ais523> they can, it's used for "local identification"
23:13:14 <fizzie> It's 0/1 and 128/1 if you want to divide into two halves.
23:13:33 <pikhq> elliott: 0/8 is link-local.
23:13:37 <elliott> because you could visit it as 0
23:13:39 <fizzie> I used to have static routes for 0/1 and 128/1 somewhere for an explicit "override default route except not really" somewhere.
23:13:52 <ais523> I saw an interesting article on why 240/4 wasn't used ages ago (I'm not sure where); apparently it was because routers didn't understand it, and upgrading them to understand it would be just as difficult as just upgrading to IPv6 so why don't we do that instead
23:13:57 <fizzie> Or maybe it was for 0/127 and 8000::/1.
23:14:00 <elliott> come on, wouldn't you put 0:6667 into your client
23:14:05 <elliott> would you not go WOOO THIS IS AWESOME
23:14:31 <ais523> I still think it was awesome when ARIN put a box on 1.2.3.4 that did nothing but listen for traffic and respond to pings
23:14:36 <ais523> to see just how badly it was being DDOSed
23:14:46 <elliott> 1.2.3.4 can't be addressed as "0"
23:14:50 <elliott> it is, therefore, inferior to 0.0.0.0
23:14:51 <ais523> and the answer was, unusably so, they had to take it down after a while
23:14:56 <ais523> anyway, time to go home
23:14:58 -!- ais523 has quit (Remote host closed the connection).
23:15:05 <elliott> PING 0.0.0.0 (127.0.0.1) 56(84) bytes of data.
23:15:07 <elliott> forgot it was used for that :
23:15:21 <elliott> "connect: invalid argument" nice :D
23:17:57 <zzo38> Is zero a valid IP address?
23:17:59 <macrohauler> ais523: thanks for the tip, terpir is now evolving into an IRC RPG using live programming ;'D
23:18:42 <fizzie> For some values of "valid".
23:18:47 <fizzie> It's not valid for global unicast use.
23:18:50 <fizzie> 0.0.0.0/8 - Addresses in this block refer to source hosts on "this"
23:18:50 <fizzie> network. Address 0.0.0.0/32 may be used as a source address for this
23:18:51 <fizzie> host on this network; other addresses within 0.0.0.0/8 may be used to
23:18:51 <fizzie> refer to specified hosts on this network ([RFC1122], Section
23:19:03 <pikhq> zzo38: RFC 5735 defines it as the source address for link-local packets.
23:19:33 <fizzie> Because INADDR_ANY == 0, it also has some trouble when it comes to the sockets interface.
23:21:12 <elliott> macrohauler: bah, not as good as MY version of it
23:21:28 <elliott> <bot> st1: 349 349 3498 293 48 28 47
23:21:36 <elliott> <bot> st2: 9 9 9 9 9 9 9 9 >9< 93 5 6
23:21:38 <elliott> <p2> OAU*@*$YH(@UOKE()RJF*G£&Y%
23:21:56 <elliott> ofc you'd have to cap the execution speed way down :D
23:22:45 <zzo38> Do you have a server which I can make remote version control system with?
23:23:41 <elliott> zzo38: you're making your own VCS too?
23:24:40 <zzo38> elliott: Actually that isn't what I was asking about. It was suggested I can use VCS for storing TeXnicard and other things, and it should probably be remote VCS so that we have two copies.
23:24:59 <zzo38> (I could make my own VCS too, but that isn't what I was asking about.)
23:25:05 <elliott> zzo38: You can sign up to http://github.com/ to get free git hosting.
23:25:22 <elliott> Or http://repo.or.cz/ if you're old-school.
23:25:41 <elliott> Or http://patch-tag.com/ or http://darcsden.com/ if you're a darcs person...
23:26:28 <fizzie> Or that thing, what is it, bitbucket, for mercurial?
23:26:31 <fizzie> I think it's free too.
23:26:46 <elliott> fizzie: But nobody likes Mercurial.
23:26:57 <elliott> (Gregor doesn't count as a person.)
23:28:38 <olsner> everyone likes git, except mercurial fans
23:28:52 <olsner> haskell people like darcs in theory, but git in practice
23:28:54 <coppro> even monotone fans like git
23:29:07 <elliott> olsner: i like darcs in practice
23:29:18 <elliott> olsner: well scapegoat is better but :)
23:29:25 <elliott> there's nothing wrong with darcs... although it has a stupid merge algorithm
23:29:31 <elliott> and "patch theory" is bullshit
23:29:36 <elliott> but the cherry-picking is unmatched
23:29:43 <elliott> olsner: admittedly for ghc-scale projects even darcs 2 is unusable
23:29:51 <elliott> but for most projects darcs 2 is very usable
23:30:41 <olsner> I haven't used darcs 2...
23:30:51 <zzo38> I did in fact think of a simple protocol which could be used: send the ASCII SOH, the username, the commit ID, the list of files and sizes (and text/binary mode), and the data of the files. And then it check for changes, and make list, and then you can use another command for receiving the data.
23:31:03 <olsner> but darcs 1 instantly went into the "take forever, do nothing" state as soon as I tried "branching" ny having repos with different sets of patches
23:31:11 <elliott> olsner: yeah, there were edge-cases
23:31:15 <elliott> one really big exponential one
23:31:21 <elliott> olsner: but darcs 2 focused on performance
23:31:38 <elliott> Darcs 2 introduces format changes designed to improve performance and safety. The changes are the hashed repository format and the darcs-2 patch format. The hashed repository format can be used in a manner that is interchangeable with older darcs -- darcs 1 cannot read the hashed format, but darcs 2 allows you to exchange patches between repositories in new and old formats. The darcs-2 patch format requires a repository conversion that is not bac
23:31:38 <elliott> kwards-compatible, so projects switching to darcs-2 patch format will have require that all their users upgrade to darcs 2.
23:31:41 <olsner> I think I got the "nothing fucking works" edge-case
23:31:44 <elliott> Identical primitive changes no longer conflict. See below for a discussion of these implications.
23:31:54 <elliott> olsner: I'm not saying darcs 1 is any good, just that darcs 2 is :P
23:32:08 <elliott> such important projects as C-INTERCAL were maintained in darcs until esr decided to be a jerkass!
23:32:23 <olsner> haha, darcs had its chance and blew it, I'm ignoring it completely now since git is good enough
23:32:45 <j-invariant> that's what happens when you write a program in haskell
23:32:53 <elliott> olsner: git's review of the repository is awful, though
23:32:53 <olsner> (just for the heck of it: I have had better luck using CVS than darcs)
23:33:01 <elliott> olsner: oh stop trolling :)
23:33:08 <elliott> <olsner> haha, darcs had its chance and blew it, I'm ignoring it completely now since git is good enough
23:33:12 <zzo38> I don't need all the services usable by github and repo.or.cz and so on, and I don't want to do it by email or webpages.
23:33:14 <elliott> i realise you're not serious but this kind of thinking stinks
23:33:16 <olsner> well, it's not even trolling though
23:33:19 <elliott> like all this shit is in competition
23:33:30 <elliott> fighting for users is stupid
23:33:33 <elliott> zzo38: repo.or.cz is not based on webpages
23:33:36 <elliott> you just sign up for webpages
23:33:39 <elliott> and then use the git command-line tool
23:34:09 <zzo38> But I still need email and SSH to sign up??
23:34:19 <elliott> zzo38: git handles pushing with SSH
23:34:26 <olsner> ok, that part is not completely serious... it's more like git is sufficiently close for my needs that I'm not re-evaluating new versions of darcs
23:34:43 <olsner> but I *have* more successfully used CVS than darcs
23:35:09 <elliott> ok, that part is not completely serious... it's more like C++ is sufficiently close for my needs that I'm not re-evaluating new versions of GHC
23:35:16 <elliott> but I *have* more successfully used Pascal than GHC
23:37:05 <zzo38> Does that mean I can do it on Cygwin?
23:37:12 <olsner> hmm, the difference between git and darcs is not quite comparable to C++ vs GHC I think
23:37:31 <elliott> olsner: well C++ has a stupid, backwards view of the universe, like git :)
23:37:35 <elliott> git doesn't even know what a change is, or a patch
23:37:53 <elliott> zzo38: Try this: http://code.google.com/p/msysgit/
23:37:59 <elliott> zzo38: It's the best git for Windows.
23:38:15 <elliott> zzo38: http://code.google.com/p/msysgit/downloads/detail?name=Git-1.7.3.1-preview20101002.exe
23:38:23 <zzo38> Actually I just tried it in Cygwin, I have the commands ssh-keygen, ssh, git, already installed.
23:38:32 <elliott> zzo38: Yes, but Cygwin git is very slow because Cygwin is slow.
23:39:25 -!- oerjan has quit (Quit: Good night).
23:39:35 <zzo38> How do I specify which files should be in the repository?
23:39:51 <elliott> zzo38: "git add x" adds the file x to the current revision.
23:39:58 <elliott> Then "git commit -m 'My commit message.'" makes a commit.
23:40:08 <elliott> You can also say "git commit -am 'blah'" if you don't want to have to "git add" every file you have changed first.
23:40:23 <elliott> Also you can say just "git commit" and it will start an editor to let you enter a longer commit message on multiple lines.
23:41:44 <zzo38> Why is a commit message mandatory? And what if I just want a file that lists which files belong in the repository?
23:42:09 <elliott> zzo38: Because it shows the commit messages in the log.
23:42:14 <elliott> And that isn't how git works, so you cannot do that.
23:42:29 <elliott> If you just "git add" then, you can just use "git commit -a" later, to make it commit all the files in the repository.
23:43:35 <pikhq> elliott: I'd think that git on Cygwin wouldn't be *too* bad.
23:43:39 <pikhq> I mean, it's not using fork.
23:43:50 <elliott> pikhq: Worse when git was partly shell scripts ...
23:44:03 <pikhq> Ah, yeah, that would be terrible.
23:44:19 <pikhq> Such a shame Microsoft can't just add a freaking fork() call.
23:45:00 <pikhq> The NT kernel already has such a system call; Win32 just doesn't expose it.
23:45:25 <pikhq> The POSIX subsystem uses it, though.
23:45:44 <elliott> pikhq: can you write a win32 dll using the posix subsystem?
23:46:27 <pikhq> No, but you can write a POSIX program that calls out to Win32.
23:47:23 <elliott> pikhq: hmm, why can't you go the other way?
23:47:46 <pikhq> Fuck you, that's why.
23:48:42 <pikhq> Anyways, even with fork() spawning processes will probably be pricy for Cygwin.
23:48:57 <pikhq> For various reasons, Win32 makes processes expensive to spawn in general.
23:49:44 <pikhq> Unlike POSIX, where that's what it does best.
23:50:52 <elliott> pikhq: hmm, could cygwin harness the posix subsystem?
23:51:01 <elliott> cygwin's fork() is basically "restart this process, and skip to here"
23:51:09 <elliott> and very slow, since it has to restore so much stuff
23:51:22 <pikhq> Not without a lot of work.
23:51:52 <coppro> IIRC someone has written a fork version which operates reasonably quickly
23:51:55 <coppro> by skipping all the win32 overhead
23:51:56 <zzo38> Are there any simpler version control systems?
23:52:02 <coppro> the actual kernel overhead is small
23:52:13 <pikhq> coppro: Fork is *actually in the NT kernel*.
23:52:29 <coppro> pikhq: wait, directly, including memory set up and everything?
23:52:45 <coppro> my understanding was there was a process split primitive, but it was not equivalent to fork
23:52:49 <pikhq> elliott: Because the subsystems are basically completely seperate things that happen to run on the same kernel.
23:52:55 <elliott> zzo38: None that are of any use. git is very simple.
23:53:02 <pikhq> coppro: Yes, it's for the sake of POSIX.
23:53:07 <elliott> pikhq: Could Cygwin run entirely in the POSIX subsystem? :P
23:53:14 <pikhq> coppro: Windows has a full, entire POSIX subsystem.
23:53:29 <pikhq> elliott: No, but a full UNIX userspace can.
23:53:38 <elliott> pikhq: I mean [could Cygwin be ported to].
23:53:45 <pikhq> elliott: No point.
23:54:00 <pikhq> All Cygwin is is a Win32 libc and patches to make stuff work.
23:54:21 <pikhq> And said libc is also a partial kernel emulator.
23:54:49 <elliott> pikhq: erm it's also a libposix, isn't it
23:55:07 <pikhq> When you've just straight-up got libc, you might as well just install Debian.
23:56:00 <elliott> 12:37:38 <fizzie> You find the strangest things when digging through old home directories; here's a "cp" replacement that copies files by starting two processes, having the first read the input file, the second write the output file, and doing all communication between processes by using the SIGUSR1 and SIGUSR2 signals as the "dit" and "dah" symbols for morse code, and suitable pauses to distinguish words.
23:56:06 <elliott> fizzie: That is the greatest thing ever created.
23:56:25 <elliott> pikhq: ^ This is Kitten's cp.
23:56:31 <pikhq> ... Cross root-filesystem cp.
23:56:51 <elliott> 12:38:24 <fizzie> I'm not sure why it doesn't just use those two signals as "on" and "off" events (or just one signal to toggle) and timing for even the dit/dah distinguishement; but it's reasonably silly as-is.
23:57:00 <elliott> 12:45:06 <AnMaster> fizzie, so it doesn't work for non-text files?
23:57:01 <elliott> 12:45:28 <AnMaster> fizzie, as for using timing: would be unreliable on a non-realtime OS
23:57:01 <elliott> 12:45:29 <fizzie> AnMaster: Right. But be honest, how often do you need to copy non-text files anyway? Almost never!
23:57:12 <elliott> fizzie: Make it convert bits in the file to "zero" and "one" words first.
23:57:27 <elliott> So it ends up doing "zero one one zero zero one zero one one ..." in Morse code.
23:57:52 <fizzie> It's not exactly blazing fast now either.
23:58:21 <fizzie> It was for a "unix programming" course, the assignment was to use signals for something.
23:58:54 <pikhq> I'd have probably just implemented pikhq coreutils kill or something.
23:59:04 <pikhq> But that's more awesome.