00:05:09 -!- RedDak has quit (Read error: 104 (Connection reset by peer)).
00:40:00 -!- alexbobp has quit ("Leaving.").
01:13:30 -!- Sgeo has joined.
01:27:56 -!- tusho has quit.
01:28:22 -!- tusho has joined.
01:52:56 -!- tusho has quit (Read error: 113 (No route to host)).
02:18:59 -!- CakeProphet has joined.
02:25:41 -!- CakeProphet has quit (Read error: 60 (Operation timed out)).
02:32:09 -!- CakeProphet has joined.
02:58:27 -!- CakeProphet has quit (Read error: 110 (Connection timed out)).
04:44:01 -!- CakeProphet has joined.
04:52:11 -!- adu has joined.
05:26:28 -!- pikhq has quit ("leaving").
05:41:19 -!- Sgeo has quit ("Ex-Chat").
05:54:38 -!- GreaseMonkey has joined.
05:58:19 <psygnisfive> zzz
05:59:19 <adu> aaa
05:59:25 <psygnisfive> bbb
05:59:31 <adu> ccc
05:59:47 <psygnisfive> þþþ
05:59:50 <adu> lol
06:00:06 <adu> I'm surprised that showed up in my client
06:00:12 <adu> i suppose its latin-1
06:00:23 <psygnisfive> ßßß
06:00:54 <adu> 义义义
06:01:34 <psygnisfive> ʔ
06:01:46 -!- olsner has joined.
06:02:14 <psygnisfive> ƭƭƭ
06:02:49 <psygnisfive> ƥʃɗɠɦƙƥȥɖƈɓɲ
06:02:57 <psygnisfive> this is such a stupid combination method lol
06:03:45 <olsner> it does seem stupid indeed
06:04:22 <psygnisfive> well, its stupid because the things have nothing in common
06:04:53 <psygnisfive> ƥ : p !:: ʃ : s and so on
06:36:53 <adu> hm
06:48:39 -!- adu has quit.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:25:01 -!- lament has joined.
08:26:51 -!- fizzie has quit (clarke.freenode.net irc.freenode.net).
08:26:51 -!- lifthras1ir has quit (clarke.freenode.net irc.freenode.net).
08:27:41 -!- lifthrasiir has joined.
08:27:44 -!- kar8nga has joined.
08:28:55 -!- fizzie has joined.
09:05:00 -!- lament has left (?).
09:57:04 -!- Corun has joined.
10:02:26 -!- psygnisf_ has joined.
10:02:26 -!- psygnisfive has quit (Read error: 104 (Connection reset by peer)).
10:02:52 -!- Corun has quit ("This computer has gone to sleep").
10:19:54 -!- tusho has joined.
10:22:34 <tusho> ais523-log-ping:
10:22:41 <tusho> oh man you know that twiceler bot that was scouring eso-std.org
10:22:56 <tusho> and we went onto the site and it was a spidering bot for a company saying they were 'reinventing search' blah blah blah
10:23:00 <tusho> it's cuil's spider
10:23:00 <tusho> XD
10:25:00 <tusho> they were called cuill when we checked though
10:32:24 -!- GreaseMonkey has quit ("HydraIRC -> http://google.com <- Go find something better").
10:45:03 <AnMaster> tusho, link to cuil stuff?
10:45:22 <tusho> AnMaster: you don't want to use it. it is failure embodied.
10:45:27 <tusho> for example - 'cobol' gets no search results
10:45:30 <tusho> and
10:45:30 <tusho> http://xs129.xs.to/xs129/08313/picture7574.png
10:46:25 <AnMaster> tusho, who?
10:46:30 <AnMaster> never heard of that name
10:46:33 <tusho> AnMaster: eric schmidt = google CEO
10:46:36 <tusho> that picture = steve ballmer.
10:46:41 <AnMaster> oops
10:46:48 <tusho> cuil = thinks they belong together
10:47:07 <AnMaster> that *is* funny
10:47:22 <fizzie> Well, it's still not masturbating guys, like in that El Reg post: http://www.theregister.co.uk/2008/07/29/cuil_launch/
10:47:34 <tusho> hah
10:47:56 <fizzie> Maybe they think the picture embodies the concept of "Teaching".
10:48:53 <tusho> {The company says it's attempting to tag each search result with an image "that will help people visually check whether the result is something they want to click on."}
10:49:10 <tusho> "this site is GAY"
10:50:31 <AnMaster> oh god
10:50:41 <AnMaster> what happened to that powerset thingy? was all the rage some time ago
10:50:46 <tusho> AnMaster: microsoft bought them
10:50:50 <tusho> embrace, extend, extinguish
10:50:54 <AnMaster> tusho, but was it any good?
10:51:00 <fizzie> The COBOL thing is curious, since the "tab bar" there does list searches like "Cobol Programming", which do return semi-reasonable results.
10:51:30 <tusho> AnMaster: some interesting ideas, but no. not really.
10:51:40 <tusho> fizzie: Apparently 'all' means 'everything else'.
10:51:47 <AnMaster> google still wins then :)
10:52:46 <tusho> Speaking of which, omgwtfbbq pikhq is back in Agora.
11:01:56 <tusho> oh man. I just linked someone to a Jakob Nielsen article and they said they could find a link saying the exact opposite :)
11:13:14 -!- sebbu has joined.
11:17:39 <AnMaster> tusho, who is "Jakob Nielsen"?
11:25:38 -!- sebbu2 has quit (Read error: 113 (No route to host)).
11:31:50 -!- RedDak has joined.
11:31:53 <tusho> AnMaster: web usability expert
11:32:08 <AnMaster> oh ok
11:32:17 <tusho> http://www.useit.com/
11:32:28 <AnMaster> I still think cfunge's webpage is about as usable as it can get, for a geek ;P
11:32:38 <AnMaster> just a plain page
11:32:38 <tusho> not really
11:32:42 <tusho> its navigation is sorely lacking
11:32:46 <tusho> also it's inaccessible
11:32:54 <AnMaster> tusho, sure, but there isn't much to navigate to
11:33:01 <tusho> AnMaster: it's not bad, certainly
11:33:02 <tusho> anyway
11:33:02 <AnMaster> and.... google for cfunge
11:33:06 <AnMaster> first hit recently
11:33:06 <tusho> it's horrible inaccessable
11:33:15 <tusho> your links are only distinguished by colours
11:33:19 <tusho> it's a giant "fuck you" to colourblinds.
11:33:21 <AnMaster> ah ok true
11:33:23 <tusho> add an underline back in
11:33:28 <AnMaster> tusho, good point
11:33:31 <tusho> also
11:33:38 <tusho> change the visited link colour to something else
11:33:45 <tusho> because colourblind users won't be able to distinguish it either
11:33:45 <AnMaster> tusho, I just took the theme from the lighttpd directory layout
11:34:09 <tusho> not sure how to do visited links nicely for colourblind people actually
11:34:12 <AnMaster> tusho, but then wikipedia has the same problem
11:34:14 <tusho> purple should be pretty distinguishable though
11:34:16 <AnMaster> no underlined links
11:34:17 <tusho> AnMaster: yes it does
11:34:23 <tusho> doesn't make it acceptable
11:34:27 <AnMaster> why don't they fix it?
11:34:34 <tusho> i don't know. I haven't aske
11:34:35 <tusho> d
11:35:19 <AnMaster> tusho, underlines added
11:35:26 <tusho> you can turn it on in user preferences though
11:35:28 <AnMaster> as for visited links, it is not something I use myself anyway
11:35:48 <AnMaster> rather I should make them keep the same color
11:35:50 <tusho> http://www.useit.com/alertbox/20040503.html
11:35:52 <tusho> :p
11:35:55 <AnMaster> to piss of non-colorblinds too
11:36:00 <AnMaster> ;P
11:36:21 <tusho> the eso site is just going to use underline to distinguish a link
11:36:29 <AnMaster> no colour?
11:36:30 <tusho> so that we can use them to an insane degree without becoming a rainbow
11:36:33 <AnMaster> that will confuse me at least
11:36:40 <tusho> AnMaster: not really
11:36:49 <tusho> _ESO 1: Brainfuck_ has just been released
11:36:52 <tusho> where _ = underline
11:37:05 <tusho> plus _Comments_ :P
11:37:47 <AnMaster> tusho, IMO underline is uggly, you won't see underline in any printed newspaper or such
11:37:56 <AnMaster> you use italics and bold sure
11:37:58 <tusho> underline, nowadays, means a link
11:38:01 <AnMaster> yes
11:38:07 <tusho> of course, underline in type is ridiculous
11:38:12 <tusho> underline = italics
11:38:17 <AnMaster> tusho, but underline is horrible typographically
11:38:20 <tusho> yes
11:38:21 <tusho> of course
11:38:35 <AnMaster> tusho, remember I like LaTeX ;P
11:38:45 <tusho> the point is everyone knows a bit of underlined text on the web is something you can click and get some related info
11:39:04 <AnMaster> tusho, as for navigation on cfunge web page, wtf, it is hardly neeeded, on my screen it doesn't even scroll
11:39:18 <AnMaster> actually it does for the XHTML image at the bottom
11:39:23 <AnMaster> but no scrolling apart from that
11:39:42 <tusho> AnMaster: I suggest a direct link to the source and binaries at the top
11:39:56 <tusho> sourceforge's interface is terrible
11:39:59 <tusho> i hate downloading software from it
11:40:04 <tusho> link to, e.g. http://downloads.sourceforge.net/cfunge/cfunge-0.2.1-pre2.tar.bz2?modtime=1214848823&big_mirror=0
11:40:20 <AnMaster> I don't do binaries any longer
11:40:27 <AnMaster> because of no uncommon deps
11:40:37 <AnMaster> no longer depend on Boehm-GC
11:40:38 <tusho> AnMaster: link to the source directly then at least
11:40:43 <AnMaster> well hm good idea
11:40:47 <tusho> like this:
11:40:58 <tusho> This is cfunge, a fast, small and standard conforming Befunge98 interpreter in C.
11:41:09 <tusho> [Download the latest version, posix_0.348923476823423472384234324.]
11:41:21 <AnMaster> posix_ for version lol :P
11:41:33 <AnMaster> last would currently be 0.2.1-pre2
11:41:41 <AnMaster> anyway I plan to do pre3 today
11:41:47 <AnMaster> so will add that when I update for pre3 anyway
11:41:54 <tusho> AnMaster: you should name your releases after your crazy optimizations
11:42:00 <tusho> cfunge:  posix_fadvise edition
11:42:13 <AnMaster> 2008.2?
11:42:19 <AnMaster> some distros does it that way
11:42:20 <AnMaster> ;P
11:42:31 <AnMaster> (yes that may work for a distro, but not for cfunge)
11:42:45 <tusho> personally I like this version numbering scheme:
11:42:54 <AnMaster> tusho, actually wait, I'll have to fix that darn threading bug before
11:43:01 <tusho> major.month_number
11:43:09 <tusho> if you make multiple releases, add .day_number
11:43:12 <tusho> although hmm
11:43:16 <tusho> if you take a while to add to major...
11:43:26 <AnMaster> it may take a while
11:43:32 <tusho> major.years_of_existence.month_number[.day_number]
11:43:44 <tusho> 2.1.6
11:43:49 <tusho> ...
11:43:53 <tusho> 2.2.1
11:43:59 <tusho> 3.2.2
11:44:00 <tusho> etc
11:44:16 <AnMaster> 0.2008.7 would be this month?
11:44:45 <tusho> AnMaster: no
11:44:46 <AnMaster> anyway I really like both ick and CLC's version numbering schemes
11:44:49 <tusho> how many years has cfunge existed
11:44:52 <tusho> 0
11:44:59 <tusho> it isn't a year old yet
11:45:02 <AnMaster> maybe I should base mine on modular arithmetics
11:45:05 <tusho> so it'd be 0.0.7
11:45:08 <AnMaster> so I do all mod something
11:45:15 <tusho> if you make a stable release
11:45:15 <AnMaster> and then confuse the hell out of everyone
11:45:17 <tusho> then 1.0.7
11:45:55 <AnMaster> tusho, idea: use a HUGE number that is incremented, take this mod a smaller number
11:46:00 <tusho> heh
11:46:10 <AnMaster> actually lets just do RSA while we are at it!
11:46:11 <tusho> AnMaster: and each major release increase the smaller number
11:46:17 <AnMaster> tusho, :D
11:46:40 <AnMaster> tusho, and the minor ones increases the large number?
11:47:07 <AnMaster> and then not tell anyone what these numbers are of course, just give them the result hehe
11:47:13 <tusho> heh
11:47:13 <AnMaster> and the algorithm
11:48:05 <AnMaster> or maybe just sha512 or whatever the currently preferred hash function is
12:04:00 -!- tusho has set topic: FORTH LOVE IF HONK THEN http://vjn.cc/x.
12:07:16 -!- shachaf has quit (Read error: 104 (Connection reset by peer)).
12:07:20 -!- shachaf has joined.
12:19:43 -!- sebbu has quit ("@+").
12:24:44 -!- kar8nga has left (?).
12:24:57 -!- kar8nga has joined.
12:35:58 -!- sebbu has joined.
13:12:42 -!- ihope has quit (Read error: 110 (Connection timed out)).
13:17:35 -!- sebbu has quit ("brb").
13:31:47 -!- oerjan has joined.
14:19:52 -!- kar8nga has quit ("Leaving.").
14:34:34 -!- ihope has joined.
14:34:57 <ihope> I had a dream where stuff like '.r' was valid Python to output 'r'.
14:37:34 <tusho> heh.
14:39:29 <oerjan> unPython
14:41:06 <ihope> And there was a computer game (that my brain equated with Allegiance) that had spaceship doors that had the strangest tendency to remove your space suit, then open.
14:41:37 <oerjan> i really hate those
14:42:24 <oerjan> _could_ be a premonition of today's Buck Godot comic
14:45:21 <ihope> And there were robots that were apparently inspired by the movie WALL-E. I remember seeing about five of them working on something, then gesturing for them to stop, and green exclamation marks appeared on their screens.
14:48:22 <oerjan> hm, hasn't come here yet i think
14:50:31 <tusho> we should invent esolangs in dreams
14:54:02 <ihope> So, I was experimenting with the doors, and they suddenly opened, and I saw my space suit break apart. I tried to remember what the Allegiance controls on a keyboard were and use them to travel to the female/engineering ship, but being a human without a vehicle of any kind and not knowing the controls, I failed.
14:54:48 <ihope> The other ship was the male/flight ship. Adults went to whichever ship they wanted; children went with their father if they were male and their mother if they were female.
14:56:07 <ihope> Presumably, it was expected that most males would go to the flight ship and most females would go to the female ship.
14:56:19 <oklopol> o
14:56:32 <oklopol> i'm an orange
14:56:34 <ihope> ...the engineering ship, that is.
14:56:44 * ihope pats oklopol
14:56:55 * oklopol purrs
14:57:13 <oerjan> orange u glad to see us?
14:57:54 <ihope> They were generation ships, and everyone was expected to be on them; the people on the flight ship would learn to fly really well (and invent flight-related things), and the people on the engineering ship would learn engineering and design useful things.
14:58:10 <oklopol> oerjan: i'm a cockroach
14:58:16 * tusho had a weird internet-related dream 2 days ago
14:58:19 <tusho> or was it 3?
14:58:19 <tusho> 4?
14:58:21 <tusho> meh
15:00:12 * oerjan thinks _nothing_ puns with cockroach
15:01:22 -!- Tritonio_ has joined.
15:11:30 <Tritonio_> http://inshame.blogspot.com/2008/02/efficient-brainfuck-tables.html
15:12:05 <Tritonio_> oops
15:12:07 <Tritonio_> mistake
15:12:27 <Tritonio_> http://inshame.blogspot.com/2008/07/big-table-implementation-in-brainfuck.html
15:14:24 -!- RedDak has quit (Read error: 104 (Connection reset by peer)).
15:17:25 <tusho> Tritonio_: AAAGH! MARQUEES!
15:21:56 <Tritonio_> you hate them?
15:22:03 <tusho> Tritonio_: Who doesn't?
15:22:39 <Tritonio_> why? they love you.
15:23:01 <tusho> no
15:23:02 <tusho> they left me
15:23:03 <tusho> :(
15:24:04 <Tritonio_> hmm strange...
15:28:09 -!- Sgeo has joined.
15:41:40 <ihope> Just how optimizing are BF optimizing compilers these days?
15:42:12 <tusho> ihope: Very. Try that haskell one.
15:42:17 <tusho> It does crazy shit.
15:42:22 <tusho> As they say.
15:45:12 -!- Tritonio_ has quit ("Leaving").
15:46:06 <tusho> lost the game
15:49:18 -!- oerjan has quit ("Ayeeh! Heat!").
15:50:56 -!- MikeRiley has joined.
15:52:22 <MikeRiley> Deewiant: last night i found the underlying problem with wrapping in Rc/Funge-98...
15:55:01 <Deewiant> so what was it?
15:55:35 <MikeRiley> in Check_Wrap, it was comparing against the limits as greater or equal....needed to be just greater
15:55:46 <Deewiant> :-)
15:55:50 <MikeRiley> the bug that was there would have affected wrapping in all 3 dimensions...
15:55:54 <MikeRiley> so fixed it now...
15:55:58 <Deewiant> yeah, I would have expected it to
15:56:07 <Deewiant> but I only fixed x because that's all I needed :-)
15:56:11 <MikeRiley> also replaced the dynamic memory manager to a new one,,,MUCH MUCH faster than before...
15:56:17 <MikeRiley> figured that!! eheheheeheh
15:56:29 <MikeRiley> here is the run times for Mycology now:
15:56:39 <MikeRiley>   Instructions Executed: 205866 in 220435 cycles
15:56:39 <MikeRiley>         Execution time : 0.23   seconds
15:56:39 <MikeRiley> Instructions per second: 895069.56
15:56:39 <MikeRiley> Exiting with return code = 15
15:56:47 <MikeRiley> real    0m0.299s
15:56:47 <MikeRiley> user    0m0.196s
15:56:47 <MikeRiley> sys     0m0.096s
15:57:17 <MikeRiley> still passes 100%...
15:57:37 <Deewiant> I'd compare with you but the times on Windows are ridiculous since the PERL test takes a year
15:57:56 <Deewiant> spawning a process alone takes around 0.3 seconds...
15:58:04 <MikeRiley> ouch!!!
15:58:10 <Deewiant> or something like that
15:58:18 <Deewiant> anyway, you can see two visible pauses as PERL is tested :-P
15:58:38 <MikeRiley> windows was never very good for spawning tasks...
15:58:42 <Deewiant> yep
15:59:45 <MikeRiley> next will fix TRDS to work correctly,,,then will release v1.11...
16:00:01 <Deewiant> ah, so the next release will be a while from now ;-)
16:00:11 <MikeRiley> nope....do not think so...
16:00:23 <MikeRiley> already figured out a much simpler strategy for travelling to the past...
16:00:38 <MikeRiley> travel to the future and space jumps already work correctly...
16:00:43 <Deewiant> do tell, if it's good enough I might use it :-)
16:01:01 <MikeRiley> essentially my new strategy is to make everything future jumps!!
16:01:09 <MikeRiley> and treat them as such...
16:01:27 <MikeRiley> when going backwards, the interpreter will be reset to where the program started...
16:01:46 <MikeRiley> time travelling IPs that should be there at the time of arrival will then be added to the IP list,,,with runtimes in the future...
16:01:57 <Deewiant> hmm
16:01:57 <MikeRiley> making them act like future jumps from time 0...
16:02:19 <MikeRiley> in theory,,,should work,,,
16:02:23 <Deewiant> sure
16:02:30 <MikeRiley> and much easier to implement than my previous strategy...
16:02:36 <Deewiant> what was your previous strategy then
16:02:47 <MikeRiley> craziness!!!! eheheheheeheheheh
16:02:50 <Deewiant> >_<
16:03:11 <MikeRiley> before i as it ran it would compare to the IPs to look for ones that would be there and then spawn them in...
16:03:27 <MikeRiley> but that method was problematic when it came to matched uids...
16:03:32 <Deewiant> hmm
16:03:39 <Deewiant> I'm not sure what I do, actually :-)
16:03:48 <Deewiant> but I don't think either is very complicated
16:03:51 <Deewiant> or different in complexity
16:03:55 <MikeRiley> new method should be much cleaner,,,and less impact on the main interpreter loop...
16:04:31 <MikeRiley> current method requires TRDS functions to be dealt with until the time of jump of the last jumped IP,,,,
16:04:39 <MikeRiley> new method removes that check alltogether...
16:05:08 <MikeRiley> by using the future jumps from 0, the check to see if CycleCount > RunTime is enough to allow the IPs to arrive at the correct times...
16:06:19 <MikeRiley> and no seperate list is needed to check for when other time travellers should appear...
16:06:48 <Deewiant> yeah, I guess that works
16:06:57 <Deewiant> but the main complexity isn't managing the list, I think
16:07:22 <Deewiant> you're just moving the complex stuff from knowing how to manage the list to knowing what RunTime to give the IPs :-)
16:07:26 <MikeRiley> i think this method will actually remove the complexity...
16:07:40 <MikeRiley> the RunTime is already straightforward...
16:07:40 <Deewiant> if so, great and I'll probably do that too :-)
16:07:50 <Deewiant> just let me know if you can pass mycotrds
16:08:24 <MikeRiley> well,,,once i have it working the way i think that it should...then will try mycotrds,,,and hopefully your vision of how time travel is supposed to work meshes with mine!! ehehehehehe
16:08:51 <Deewiant> (note that all GOODs isn't enough, it's possible to have all GOODs but incorrect output)
16:08:52 <oklopol> "There were already some table implementations which used at least 2*n Brainfuck cells for storing an n sized table. My implementation needed only n+4."
16:09:03 <oklopol> was this tritonio's link or whose, dunno
16:09:08 <oklopol> anyway wtf :D
16:09:10 <Deewiant> aye, his
16:09:11 <MikeRiley> understood...i do agree,,,correct output is required...
16:09:16 <oklopol> 2^n cells?
16:09:24 <Deewiant> 2*n
16:09:32 <oklopol> xD
16:09:33 <oklopol> right
16:09:35 <oklopol> thanks
16:09:40 <Deewiant> :-)
16:10:04 <oklopol> "Also, all known implementations had a limit of 256 cells (Or equal to the size of a cell on the Brainfuck array. For the rest of the post I will assume that the cell size is one byte)." what??
16:10:23 <oklopol> is there a misread character there too?
16:10:36 <Deewiant> I don't know, what do you think it says :-)
16:10:39 <oklopol> or... err... are all known bf implementations 256-celled?
16:11:33 <Deewiant> no, they have 30000 cells usually
16:11:42 <Deewiant> but like he said, the limit was "equal to the size of a cell on the Brainfuck array"
16:11:42 <oklopol> 30000 or infinite
16:11:58 <oklopol> 256 is the usual size limit
16:12:11 <oklopol> 256^(1, 2 or 4) or infinite
16:12:24 <oklopol> hmm
16:12:31 <oklopol> oh
16:12:33 <oklopol> hey
16:12:38 <oklopol> right, implementations of the table :D
16:12:43 <oklopol> god i'm dumb
16:12:44 <oklopol> really
16:12:47 <oklopol> someone smack me
16:12:48 <oklopol> hard
16:12:51 <Deewiant> :-)
16:13:13 * tusho smacks oklopol hard
16:16:18 <MikeRiley> also,,,with the new memory manager, -c is not irrelevant....the dynamic memory manager allocates memory as needed now...
16:17:54 <MikeRiley> and considering the speed of the new manager, also will make the dynamic model the default instead of static...
16:18:57 <MikeRiley> -c is NOW irrelevant,,,
16:19:12 <MikeRiley> as is -r
17:08:41 <MikeRiley> you still about Deewiant??
17:09:08 <Deewiant> aye
17:09:48 <MikeRiley> question about why mycotrds is doing:
17:09:57 <Deewiant> oh noes ^_^
17:09:59 <MikeRiley> BAD: J doesn't jump through time properly
17:10:11 <MikeRiley> i assume at this point you are only testing single ip time travel?
17:10:15 <MikeRiley> and is it to the future or past?
17:10:24 <Deewiant> the past probably
17:10:33 <Deewiant> my guess is
17:10:38 <MikeRiley> and what are you expecting to find when you get there?
17:10:48 <Deewiant> that it's doing a modification to Funge-Space at some point using p
17:10:50 <MikeRiley> i just wrote a small program to test single ip travel to the past and it works fine...
17:10:52 <Deewiant> and then at some point checking it with g
17:10:57 <Deewiant> and not finding what it expects
17:11:11 <Deewiant> I can't remember the details but I remember that it was really basic and RC/Funge-98 didn't do it :-P
17:11:12 <MikeRiley> was p written before or after the jump?
17:11:14 <Deewiant> hmm
17:11:20 <Deewiant> upon reflection, I think I might remember the details
17:11:23 <Deewiant> so
17:11:24 <MikeRiley> i know it does not!! eheheheeheh trying to find out what it actually is...
17:11:27 <Deewiant> what it does is, before the jump
17:11:31 <Deewiant> it does, say, 500p
17:11:37 <Deewiant> and then it jumps to the past
17:11:44 <Deewiant> and makes sure that 00g does /not/ give 5
17:11:50 <MikeRiley> got it...ok....
17:12:02 <Deewiant> i.e. modifications to space in the future don't affect space in the past
17:12:09 <MikeRiley> correct,,,they should not...
17:12:18 <MikeRiley> funge-space has to appear as it was in the past...
17:12:29 <MikeRiley> now that i know what to check for, can continue working on this...
17:12:58 <MikeRiley> goal is to have TRDS working by the end of the day....
17:14:09 <Deewiant> :-D
17:14:24 <MikeRiley> doubt it will happen!!! eheheheheheeh but that is my goal....
17:18:17 <MikeRiley> yikes!!!!!!!! got that problem fixed and now mycotrds gives lots and LOTS of output!!!! eheheheheheeheheh
17:18:22 <MikeRiley> now to find the next problem...
18:03:37 <AnMaster> hm?
18:04:30 <MikeRiley> ????
18:05:03 <AnMaster> nothing
18:05:07 <AnMaster> (wrong tab)
18:06:02 <MikeRiley> no problem
18:23:24 <MikeRiley> still here Deewiant???
18:23:39 <Deewiant> o/
18:24:33 <MikeRiley> got a question about how you think time travel to the future shoud work....
18:25:00 <MikeRiley> everything is good up until mycotrds tries to to IJ to return back to the previous time,,,,it then is in a stuck loop....
18:25:11 <MikeRiley> how are you handing future travel??
18:25:46 <Deewiant> MikeRiley: could it be that when you have only one IP traveling to the future, it doesn't work because nobody's executing instructions and thus the time never comes?
18:26:13 <MikeRiley> possible,,,but doubt it, my main loop would still be running and ticking time....
18:26:16 <Deewiant> I just put the traveling IP to sleep
18:26:19 <MikeRiley> but let me check that for certain...
18:26:30 <MikeRiley> that is what i do, put the future travelling IP to sleep...
18:27:05 <MikeRiley> no IPs need to be running for CycleCount to increment,,,,so that is not the problem...
18:28:29 <MikeRiley> got another idea of what it could be tho.....
18:29:36 <MikeRiley> nope....IJ works on my test program....hmmmmmmmm
18:34:51 <MikeRiley> Deewiant, when you execute IJ, what instruciton are you jumping back to???? normally the jump would leave you at the J, but the J should not be executed again....it should be the instruction following...
18:36:15 <Deewiant> beats me... if it's the J at (68,40) it should go to (68,39)
18:36:36 <Deewiant> if you end up at (69,40) instead I can see that being an infinite loop
18:36:48 <Deewiant> or actually I can't
18:36:52 <Deewiant> but anyhoo :-P
18:37:48 <MikeRiley> ok,,,let me see if i can see where it is coming back to...
18:39:38 <MikeRiley> weird,,,,ran it in the debugger and it worked,,,,,run it without the debugger and it hangs......strange.......
18:39:49 <Deewiant> heh
18:40:27 <ihope> Is that called a Schroedinbug, or a Heisenbug?
18:40:50 <ihope> Whichever it is, I hear that's often caused by failing to initialize your variables.
18:41:02 <MikeRiley> i suppose that is a possibility.....
18:41:04 <Deewiant> heisen
18:42:00 <Deewiant> schrödinbugs are when you read the source and notice that something never could have worked, at which point it stops working
18:43:07 <ihope> tusho: isn't it now three days ago that you said "maybe tomorrow" you'd start work on rootnomic?
18:43:08 <MikeRiley> does this take into account that the same ip cannot make the same jump twice????
18:43:28 <tusho> ihope: Yes. I've been doing a lot. :P
18:43:51 <Deewiant> MikeRiley: of course, any TRDS program would loop forever if the interpreter allowed that :-)
18:44:00 <MikeRiley> more precisely, an IP with the same uid cannot make the same jump twice...
18:44:14 <MikeRiley> exactly,,,,which is why it is not allowed...
18:44:24 <Deewiant> hmm, same uid
18:44:33 <MikeRiley> which confuses me about your output....
18:44:36 <Deewiant> what if a time traveler makes the jump
18:44:41 <MikeRiley> looks like the same jumps are being made more than once???
18:44:51 <MikeRiley> a time traveller can make the same jump...
18:44:55 <MikeRiley> only once tho
18:44:56 <ihope> MikeRiley: I notice you use ,,, a lot.
18:45:18 <MikeRiley> yeah,,,i buy my punctuation symbols in buld,,,so tend to use lots of them!!! eheheheeheh
18:45:24 <MikeRiley> s/buld/bulk/
18:46:12 -!- Corun has joined.
18:47:14 <MikeRiley> the way that time jumps to the past work, is that a single entity cannot make the jump twice,,,when a past-travelling ip encounteres the same jump, it should be terminated...only the back-jumped ip would remain...
18:47:55 <Deewiant> yeah, but what if the back-jumped IP does the jump :-)
18:47:59 <MikeRiley> actually the past-travelling one is not the terminated one, it is the one that was there when the IP arrives, or created after the IP arrives...
18:48:06 <MikeRiley> here is the logic:
18:48:21 <MikeRiley> IP0 jumps at 1000 to 500 becoming IP0t
18:48:31 <MikeRiley> IP0 will still be created as normal
18:48:40 <MikeRiley> and at 500, there could be IP0 and IP0t
18:48:56 <MikeRiley> if IP0 reaches 1000 and makes the same jump, ti should be terminated...not jump again
18:49:00 <Deewiant> yep
18:49:10 <Deewiant> but what I'm wondering now is, what if IP0t makes the same jump
18:49:20 <MikeRiley> if IP0t reaches 1000 (or any cycle) at the same point to make the same jump, it jumps and becomes IP0ta
18:49:26 <Deewiant> yeah, exactly
18:49:28 <Deewiant> so
18:49:32 <Deewiant> 2008-07-31 20:44:00 ( MikeRiley) more precisely, an IP with the same uid cannot make the same jump twice...
18:49:34 <MikeRiley> since IP0t would also exist, since jumped at the same point...
18:49:36 <Deewiant> is false
18:49:39 <Deewiant> because they both have UID 0
18:49:41 <MikeRiley> pretty much...
18:50:07 <MikeRiley> IP0t can make the jump that IP0 made...
18:50:19 <Deewiant> yep
18:50:20 <MikeRiley> but if IP0t reaches that jump again,,,then it must terminate
18:50:22 <Deewiant> but they have the same id
18:50:28 <MikeRiley> they can....
18:50:35 <MikeRiley> i use a serial number to differentiate them...
18:50:36 <Deewiant> well, they do, because they're both 0
18:50:39 <Deewiant> certailny
18:50:42 <Deewiant> s/ln/nl/
18:50:51 <Deewiant> of course you have to differentiate them somehow :-P
18:51:06 -!- jix has joined.
18:51:07 <MikeRiley> and then a jump table keeps track of uid/serial number and the jump being made...
18:51:13 <Deewiant> but as far as the ID of the concurrent Funge-98 spec is concerned, they have the same ID
18:51:35 <MikeRiley> if an IP making a jump is found in the jump table, it is terminated...if not,,,then the jump is inserted and then the jump is made...incrementing the serial number...
18:51:43 <MikeRiley> they have the same uid....
18:51:46 <MikeRiley> different serial numbers
18:51:49 <Deewiant> yep
18:52:10 <MikeRiley> for all practical purposes they are the same entity and should have the same uid....
18:52:25 <MikeRiley> the serial number allows me to see if any particular time traveller is duplicating a jump....
18:52:25 <Deewiant> so, again, two IPs with the same IDs /can/ make the same jump twice, as long as they're actually different IPs :-)
18:52:37 <MikeRiley> as long as they are really different IPs...
18:53:02 <MikeRiley> so IP0 jumping to become IP0t which can cump to become IP0ta which can jump to become IP0tb
18:53:16 <MikeRiley> but IP0ta could not make the jump again that made IP0tb
18:53:17 <Deewiant> yep!
18:54:36 <MikeRiley> so,,,,in mycotrds, when it looks in the output like the same jump is being made, is it really the time travlled ones making those jumps????
18:54:46 <MikeRiley> looking at your output....not my run...
18:55:56 <Deewiant> I honestly can't remember the details of what goes on in mycotrds :-/
18:58:15 <MikeRiley> wished i was around when you created it...
18:58:33 <Deewiant> you can look at comments in CCBI's source, in ccbi.d and trds.d
18:58:58 <Deewiant> there are some that are a couple of paragraphs long, my long rants about TRDS being confusing :-P
18:59:19 <Deewiant> if there are cases there that you disagree with then mycotrds is probably in error somewhere
18:59:47 <MikeRiley> let me grab your source and take a look at the notes...
19:00:20 <MikeRiley> i agree that the spec for TRDS was highly confusing,,,did not really expect anybody to implement it besides me!!! eheheheehehehhehehe
19:00:22 <Deewiant> around line 450 in ccbi.d
19:01:31 <tusho> 6++
19:01:33 <tusho> -+6856+6+89+656
19:01:35 <tusho> 25
19:01:44 <MikeRiley> ok,,,,looking at source now....
19:02:27 -!- sebbu has joined.
19:05:44 <MikeRiley> a bit confused about your comment in ccbi.d....not quite sure that i follow your scenario there.....
19:06:07 <Deewiant> like it says, neither do I. :-P
19:06:32 <Deewiant> alright, we're at time 300
19:06:39 <Deewiant> IP 1 jumps to 200
19:06:48 <Deewiant> then jumps back to 300
19:06:57 <MikeRiley> jumps back to 300 when??
19:07:08 <Deewiant> somewhere between 200 and 299
19:07:11 <MikeRiley> ok,,,
19:07:22 <Deewiant> we continue to time 400
19:07:30 <Deewiant> IP 2 jumps to 100
19:07:55 <Deewiant> and some time before 300, it jumps back to 400.
19:08:43 <Deewiant> now, at time 300, IP 1 has to jump back to 200 again, since IP 2 may have changed things back at time 100.
19:09:02 <Deewiant> which I guess is the core point of this comment.
19:09:15 <Deewiant> i.e. we /do/ actually need to have the same IP do the same jump twice.
19:09:20 <Deewiant> I think.
19:09:32 <MikeRiley> IP 1 can only jump back again, if it was the time travellor, and not another IP1 that got created again sometime in the past...
19:09:44 <Deewiant> and why so
19:09:45 <MikeRiley> same ip is not jumping back tho....
19:09:55 <Deewiant> it is the same one, isn't it?
19:10:04 <MikeRiley> the ip jumping back the second time is the time travellor and not the original IP 1
19:10:10 <MikeRiley> yes and no....
19:10:12 <Deewiant> no, it's not
19:10:16 <Deewiant> it's the original IP 1
19:10:24 <MikeRiley> the oringal IP 1 cannot jump back again....
19:10:30 <Deewiant> and why not
19:10:39 <Deewiant> do you see why it has to, there
19:10:44 <MikeRiley> since it was the ip that jumped back to become the time travellor
19:11:00 <MikeRiley> the time traveller can jump back tho...
19:11:03 <Deewiant> no no no no no
19:11:06 <Deewiant> let's do this again
19:11:09 <Deewiant> we're at 300
19:11:11 <Deewiant> IP 1 jumps to 200
19:11:17 -!- pikhq has joined.
19:11:18 <Deewiant> which means we rerun time from 0 to 200
19:11:22 <Deewiant> and then place IP 1.
19:11:34 <Deewiant> IP 1 then jumps back to 300.
19:11:36 <MikeRiley> which becomes IP1.0
19:11:52 <Deewiant> or IP 1t or whatever, yes.
19:11:53 <MikeRiley> IP1.0 jumps back to 300
19:12:28 <Deewiant> yes, and then we continue.
19:12:36 <Deewiant> nothing special so far, just a normal time jump. :-P
19:12:40 <MikeRiley> maybe,,,hold on....
19:12:41 <Deewiant> okay, time 400.
19:12:48 <Deewiant> ...
19:12:50 <MikeRiley> there might be a simpler rule here....
19:12:51 <Deewiant> holding...
19:13:19 <MikeRiley> if the IP making the jump, occurs to the point where the time travelled version of itself already exists....then the jump cannot be repeated...
19:13:38 <MikeRiley> because it is the time-traveller already...
19:13:53 <MikeRiley> are you working on a premese that IP 2 changed history???
19:13:56 <Deewiant> that sounds a bit more correct
19:14:00 <MikeRiley> and therefore IP1 is not necesarily the same???
19:14:29 <Deewiant> MikeRiley: that's the idea. IP 2 changes history so we need to redo the jump of IP 1, because things might happen differently. (The reason being that IP 2 jumped back further than IP 1 did.)
19:14:52 <MikeRiley> that can be done,,,,in theory, as long as IP 1 is not still around in the past....
19:15:02 <MikeRiley> the future is rewritten anytime an ip jumps to the past....
19:15:06 <Deewiant> so can RC/Funge-98 handle it? :-)
19:15:18 <MikeRiley> only those ips that would be in existance at the time of the jump need to be checked for the jump back...
19:15:26 <MikeRiley> new version can,,,,old one cannot....
19:16:05 <MikeRiley> new version checks the following: uid, serial number, source vector, destination vector, source time, destination time
19:16:17 <MikeRiley> if all of those match, then a return to the past is not possible...
19:16:21 <Deewiant> In summary: 1: 300 -> 200. 2: 400 -> 100. 1: 300 -> 200 again, because 2 jumped to 100 which is less than 200. But after this, 300 -> 200 and 400 -> 100 don't happen any more.
19:16:25 <MikeRiley> from the original IP...
19:16:38 <MikeRiley> when was IP 1 born?
19:16:48 <MikeRiley> from 0??? or later??
19:17:03 <Deewiant> assume both were born before 100 to make this simpler :-)
19:17:16 <MikeRiley> actually,,,probably does not even matter...
19:17:32 <Deewiant> the idea is that if it's not mentioned you're not supposed to think about it ;-)
19:17:42 <MikeRiley> as long as the futures were different...then the original IP1 can jump back,,,if there is no timetravlled IP 1 at the destination time....
19:18:13 <MikeRiley> if there would be a time travelled IP1 at the same time,, then it can only jump back if the jump is not the same...
19:18:31 <MikeRiley> dealing with time travel paradoxes here,,,,have to make some definitions.....
19:18:38 <Deewiant> thing is that there kind of is an IP 1
19:18:45 <Deewiant> in that the jump has been done before
19:18:54 <Deewiant> but this new jump is that same exact jump
19:18:56 <Deewiant> and should overwrite it
19:19:02 <MikeRiley> but if the future was changed enough,,,,IP 1 may reach the jump oint at a different time....
19:19:02 <Deewiant> kinda sorta maybe.
19:19:25 <Deewiant> MikeRiley: well that's fine, then it's even easier to know that it's a different jump. :-)
19:19:33 <MikeRiley> exaclty...
19:19:41 <tusho> surely a totally consistent time travel model would just involve loads of forking
19:19:44 <tusho> and therefore be boring
19:19:55 <tusho> you'll HAVE to have some paradox errors
19:20:16 <MikeRiley> i work on the premise that the future does not exist when a back travelling ip arrives at its destinaiton....
19:20:41 <MikeRiley> therefore removing entries from the table for future events not related to the current time travellers...
19:21:22 <MikeRiley> there is no way to prevent paradox errors when dealing with time travel,  so best to define what can be done,,,and everything else is considered non defined...
19:21:58 <MikeRiley> if the rules for time travel to the past are well defined,,,,then anybody crazy enough to implement TRDS should have compatable versions with each other....
19:22:13 <Deewiant> I guess we agree on that comment in ccbi.d so if you think that scenario works in RC/Funge-98 then you can move on to trds.d :-)
19:22:34 <MikeRiley> the original spec for TRDS says that if an IP makes the same jump twice, then the second time the IP is eliminated instead of jumping....
19:22:39 <MikeRiley> emphasis on the "same jump"
19:23:07 <MikeRiley> scenarios sounds good...so long as IP 1 is not making the same jump....if IP 1 is already in the past...
19:23:27 <Deewiant> the thing is that it /is/ the same jump
19:23:36 <Deewiant> there's just the possibility that history has been changed in the meanwhile
19:23:40 <MikeRiley> spec says the same jump cannot be retaken...
19:23:54 <MikeRiley> understood about the rewritten history....
19:24:00 -!- lament has joined.
19:24:29 <MikeRiley> but if IP1 was not affected by the fact that history has changed,,,then it is in theory making the same jump as before...
19:24:41 <olsner> hmm, I think object identity as a concept doesn't go well with time travel... maybe you should think about equivalent jumps (and if history has changed, maybe it wasn't equivalent after all)
19:24:43 <MikeRiley> if IP1 is affected by the change in history, then it should jump back
19:24:54 <Deewiant> MikeRiley: it may have been affected in some way, for instance it now printed "foo" instead of "bar"
19:25:10 <Deewiant> if IP2 changed "rab" somewhere to read "oof" using p
19:25:19 <MikeRiley> i think that is more of a side effect rather than  a direct effect on the IP...
19:25:34 <Deewiant> exactly, so it doesn't change the jump itself
19:25:44 <Deewiant> but I do think the possibility of that should mean that the jump should be retaken
19:25:46 <MikeRiley> yes, does not change the jump itself..
19:25:58 <MikeRiley> but no way to know if something like that happened....
19:26:16 <Deewiant> MikeRiley: you know it only from the fact that IP2 jumped to a time earlier than the time which IP1 jumped to.
19:26:22 <Deewiant> you can't be /sure/, but there's the possibility.
19:26:24 <MikeRiley> now if IP1 is born after the arrival of IP2, then it will jump back everytime....
19:26:52 <Deewiant> MikeRiley: which is why that comment is there, and that questionable "else if" statement following it.
19:27:10 <MikeRiley> ok,,,, i can see that as a possiblity,,,erase all backjumps from the no jump table for ips that arrive after the history chaning IP did...
19:27:16 <Deewiant> so that we make sure that it IP1 jumps back after the first time when IP2 jumped back.
19:27:20 <Deewiant> but not after the second time.
19:28:12 <MikeRiley> my criteria for a repeated back jump is that the IP had to have changed phase from the original jump,,,mainly, jumps at a different cycle....
19:28:29 <Deewiant> that's strictly less useful though
19:28:37 <MikeRiley> possibly....
19:28:43 <Deewiant> and would require me to rewrite mycotrds so I refuse to change CCBI ;-P
19:29:48 <MikeRiley> well,,,,can always fall back on the original spec:
19:29:50 <MikeRiley> When the newly born ip #2 reaches the time it jumped (time point 200) it will cease to exist, it will not perform another jump into the past, the original ip #2 however can still jump.
19:30:18 <Deewiant> the original spec leaves this as undefined, IMO :-)
19:30:22 <MikeRiley> which says the IP cannot jump back again, even if history was changed, if the jump is the same...
19:30:37 <Deewiant> and even if not, like said, I don't want to have to rethink both CCBI and mycotrds...
19:30:40 <MikeRiley> i disagree...it is quite defined...
19:30:46 <Deewiant> especially since I do think that that's a somewhat pointless restriction
19:30:48 <MikeRiley> do not blame you there!!!!
19:30:57 <Deewiant> and doesn't make sense to me
19:31:07 <MikeRiley> but you have to have a way to prevent time loops....
19:31:13 <Deewiant> since the reason, as I understand it, for that restriction to originally exist, is to prevent looping forever
19:31:19 <MikeRiley> exactly
19:31:27 <Deewiant> removing the restriction does not mean that all programs will loop forever. :-P
19:31:33 <MikeRiley> if you can prevent time loops....then personally,,,i think it is ok to jump back,,,
19:31:41 <Deewiant> it just means it's harder to implement TRDS.
19:31:47 <MikeRiley> NO programs should loop forever because of a time loop...
19:32:11 <Deewiant> MikeRiley: the loop in this case is preventable without that restriction
19:32:30 <MikeRiley> if the loop is preventable,,,,then it is not necesary to have the restriction....
19:32:30 <Deewiant> since, again, we have: IP1 300 -> 200, IP2 400 -> 100, IP1 300 -> 200
19:32:41 <Deewiant> the first two cases are ordinary
19:32:56 <Deewiant> in the third case, we jump back again because IP2 jumped back further
19:33:06 <MikeRiley> actually IP2 400-100 rewrites history,,,,in which case the first IP1 300->200 does not exist!!!
19:33:17 <MikeRiley> and so the 2nd IP1 300-200 can be taken...
19:33:20 <Deewiant> MikeRiley: yeah, essentially, but I'm talking in practice. :-)
19:34:03 <Deewiant> well hmm
19:34:07 <Deewiant> you can see it like that if you want
19:34:07 <MikeRiley> but,,,what if,,,,,,ip2 changes history so ip1 nees to jump back,,,ip1 makes a change that makes ip2 jump back,,,and it makes the same change as before???? now you are in a time loop
19:34:21 <Deewiant> that can't happen
19:34:24 <Deewiant> because
19:34:37 <Deewiant> IP2 can change the history of IP1 only by jumping further back than IP1
19:34:46 <MikeRiley> true......
19:34:47 <Deewiant> and you can't have them both jumping forever further back than each other :-)
19:34:54 <MikeRiley> so you are right,,,ip1 cannot effect ip2
19:34:59 <Deewiant> or you can, but then that's not a time loop :-)
19:35:05 <Deewiant> that's just a plain old infinite loop :-P
19:36:05 <MikeRiley> ok,,,,solution would be,,,,when an ip jumps back, remove all entries from the jump table where the arrival time is after the newly arrived ip....
19:36:30 <MikeRiley> tha would have erase ip1's first jump,,,,allowing it to jump again....
19:36:40 <Deewiant> implement it however, as long as I don't have to do anything ;-)
19:36:51 <MikeRiley> eheheheheeheh
19:37:18 <MikeRiley> well....in that case i may have to not consider mycotrds....depending on what it expects....
19:37:38 <Deewiant> it expects what CCBI does
19:37:38 <MikeRiley> and use my own time travel tests instead...
19:37:40 <MikeRiley> but will see....
19:38:00 <MikeRiley> i will implement my vision of how time travel should work,,,,and go from there....
19:38:01 <Deewiant> and so far, I've managed to convince you that what CCBI does is right ;-)
19:38:20 <MikeRiley> so far,,,,seems reasonable....but have to implement first....
19:38:24 <MikeRiley> to be convinced....
19:38:30 <Deewiant> certainly
19:38:49 <MikeRiley> time for lunch,,,,be back in a bit.....
19:39:02 <Deewiant> I might be gone soon
19:39:19 <MikeRiley> ok,,,in that case,,,will talk with you tomorrow,,,hopefully with some form of working TRDS...
19:39:53 <MikeRiley> since you and i are the only implementers,,,i have no problem with changing specs on it,,,,in order to reduce both of our work to get something useable.....btw......
19:39:55 <MikeRiley> bye for now...
19:40:07 -!- MikeRiley has quit ("Leaving").
19:40:08 * SimonRC always wanted a few bucketfuls of commas.
19:40:10 <SimonRC> :-)
19:40:39 <SimonRC> ah, I just thought of a TRDS question
19:41:36 <SimonRC> If an IP a jumps back to a certain point, then later IP b also jumps back to that point, will b and a both arrive at once?
19:41:46 <SimonRC> I am not certain the question even makes sense
19:42:21 <Deewiant> well yes, they should both exist at the same time
19:44:02 <ihope> Oh, I forgot, I was going to take the inverse Fourier transform of a rectangular function.
19:48:13 <ihope> Say the Fourier transform goes to radians per second; then the inverse Fourier transform is, in this case, the complex conjugate of the Fourier transform divided by 2pi.
19:48:45 <SimonRC> tusho: HONK
19:48:51 <tusho> SimonRC: :)
19:49:27 -!- tusho has set topic: joy love [ honk ] [ ] if "http://vjn.cc/x" logs-at.
19:49:35 * ihope looks it up on Wikipedia instead of actually doing it
19:51:16 <SimonRC> oh puleezze
19:51:30 <tusho> SimonRC: what
19:52:12 -!- SimonRC has set topic: factor love [ honk ] when  "http://vjn.cc/x" logs >>url.
19:52:18 <SimonRC> :-)
19:52:46 <SimonRC> factor is like joy-like, but trying more to be "real-world"
19:52:56 <lament> i'm gonna try using git for some stuff in home directory
19:53:04 <SimonRC> it has lots of libraries, a nice IDE written in it, ect
19:53:08 <lament> config files, some text stuff, programming projects
19:54:44 <lament> git seems nice!
19:55:13 <lament> (but everybody's watching anxiously to see what will GHC guys pick)
19:55:27 <ihope> So, it's a sinc function of a certain amplitude.
19:57:56 -!- moozilla has quit (Read error: 104 (Connection reset by peer)).
19:58:23 -!- moozilla has joined.
19:58:56 <tusho> lament: git is gaining an awful lot of traction
19:59:05 <tusho> and seems to be the most widely-rangable one (from small to big projects)
19:59:09 <tusho> still, it's too early to decide a winner
19:59:12 <tusho> i certainly like git though
20:01:25 <SimonRC> what's git's "model"
20:01:26 <SimonRC> ?
20:02:13 <psygnisf_> you're alive! :D
20:02:38 <SimonRC> yeah
20:02:48 <SimonRC> I was alive a couple of days ago too
20:03:02 <tusho> SimonRC: it's a distributed, fast, small, versioned filesystem in userspace.
20:03:07 <tusho> (pretty much.)
20:04:03 <SimonRC> tusho: do you know darcs?
20:04:07 <tusho> yes
20:04:17 <tusho> and hg, and ...
20:04:18 <tusho> etc
20:04:31 <tusho> pretty much the only ones I don't know are bzr, monotone and arch ;)
20:04:44 -!- CakeProphet_ has joined.
20:04:52 <SimonRC> consider the difference between SVN, CVS, and darcs; how does git differ along that axis?
20:05:11 <SimonRC> is it patch-based, revision-based, or what?
20:13:14 <GregorR> git, like darcs and hg, is distributed, so close to patch-based than revision-based. It offers a lot of advanced features regarding branching and merging.
20:16:19 <SimonRC> ah, ok
20:18:02 <tusho> it is also darn fast and has tiny repositorise
20:18:09 <tusho> and scales down (for small projects) and up (for, well, linux)
20:18:59 <tusho> bye for 30mins
20:19:00 <tusho> +
20:19:44 -!- CakeProphet has quit (Read error: 110 (Connection timed out)).
20:20:26 <GregorR> It also scales sideways (for porn collections) and backwards (for bible-based learning software)
20:21:32 -!- atrapado has joined.
20:21:33 * olsner envisions the world's collective collection of porn collected in a git repository
20:22:38 <lament> scaling down is not exactly hard.
20:22:46 <lament> very few things can't do it (eg Java)
20:27:19 <ihope> Oh, it's not hard at all to violate the superposition principle of sound waves in a sonic computer.
20:38:09 <ihope> There, I've made a spec of ye analog signal processing programming language.
20:42:33 -!- MikeRiley has joined.
20:44:59 * ihope /pɹɒs/es
20:45:21 <olsner> ihope: huh, what was that?
20:45:30 <ihope> IPA.
20:45:48 <ihope> Yay, a tongue of programmation: http://esoteric.voxelperfect.net/wiki/Proce
20:46:01 -!- moozilla has quit (Read error: 60 (Operation timed out)).
20:46:03 -!- megatron has joined.
21:00:33 -!- pikhq has left (?).
21:03:51 <SimonRC> oh no, the ghc team are terrorists!   http://www.cuil.com/search?q=glasgow
21:04:54 <ihope> Racist. :-P
21:07:01 <SimonRC> I am sure the occasional ridiculous images are helping that company
21:10:38 -!- lament has quit ("leaving").
21:26:35 <ihope> I wonder if zombies often get spongiform encephalopathy.
21:26:54 -!- moozilla has joined.
21:32:18 * SimonRC is a godlike programmer...
21:32:50 <SimonRC> I wrote a copy routine that kept dropping bytes off the end of the memory being copied.
21:33:01 <SimonRC> Solution: add some random crap onto the end before copying.
21:33:43 <SimonRC> My custom scheduling algorithm make threads much too slow if they are idle for a while.
21:33:58 <SimonRC> Solution: make sure each thread keeps getting random bits of work to do.
21:35:01 <SimonRC> The program leaks memory like a sieve.
21:35:47 <SimonRC> Solution: make the program spawn a child (seperate memory map) every so often, copy the important data over to it, and die.
21:35:52 -!- megatron has quit (Read error: 110 (Connection timed out)).
21:35:53 <SimonRC> :-P
21:37:34 -!- MikeRiley has quit ("Leaving").
21:39:15 <oklopol> sin = i!(1 - i!sin) what?
21:39:38 <oklopol> SimonRC: that's awesome :D
21:43:04 -!- oerjan has joined.
21:45:45 * oerjan _can_ sort of imagine God programming that way.  Isn't random crap or something at the end how DNA chromosomes work, anyhow?
21:46:22 <oklopol> yeah except from you know dark angel the girl had the perfect code like
21:46:28 <oklopol> she was the one
21:46:38 -!- lament has joined.
21:47:03 <oerjan> oh and DNA copying _does_ drop things at the end
21:47:37 <SimonRC> all three of those things are God-inspired
21:47:46 <oerjan> and spawning a child, copying the important data - that's life in a nutshell
21:47:52 <SimonRC> Telomeres, muscle wasting, and reproduction
21:48:42 * oerjan things he maybe should work on his own muscle wasting.  some time.
21:48:45 <oerjan> *thinks
21:49:10 <SimonRC> figiting might help
21:49:10 <ihope> oklopol: the sine is the integral of one minus the integral of the sine?
21:50:06 <SimonRC> sin x = int (cos x) = int (int (0 - sin x))
21:50:46 <ihope> The integral of -sin x is cos x - 1, yes?
21:51:05 <SimonRC> um...
21:51:11 <SimonRC> I don't think so
21:51:16 <oerjan> you get a 1 for the constant in the middle
21:51:48 <ihope> Well, in Proce, if you integrate -sin x, you get cos x - 1.
21:51:52 <SimonRC> int (- sin x) = - int (sin x) = - cos x + K
21:51:54 <oerjan> ihope: cos x + C, C arbitrary
21:52:11 <ihope> In Proce, the integral is defined to be 0 at 0.
21:52:18 <SimonRC> but to make the double-integration work, the constants must all be 0, I think
21:52:21 <ihope> It's the definite integral from 0, rather.
21:52:29 <oklopol> ihope: does that define the sine function?
21:52:34 <ihope> Yes, it does.
21:52:43 <oerjan> SimonRC: i think you got the sign wrong
21:52:54 <SimonRC> um, yeah, oops
21:53:37 <ihope> Integrate sin x and you get -cos x + 1. Subtract that from 1 and you get cos x. Integrate that and you get sin x.
21:53:42 <oklopol> ihope: right, i thought there might be a trivial function that makes that fail, but yeah it's plausible that's a definition for sine
21:54:17 <ihope> It is a definition of the sine.
21:54:31 <oerjan> it's a property of sine, whether it is the only solution remains to be explained
21:54:37 <SimonRC> possibly if one expands it infinitely, one can show that one gets the taylor series for sin
21:54:41 <oklopol> oerjan: well that's what i thought
21:54:46 <oklopol> but i'm assuming ihope isn't lying
21:54:54 -!- CakeProphet_ has quit (Client Quit).
21:55:15 <oklopol> he's too clever to make a mistake, so it would be a dirty lie.
21:55:34 <ihope> The probability that I made a mistake is much higher than the probability that I lied.
21:55:38 <oerjan> well it _does_ make it a solution of f'' = -f
21:55:56 <SimonRC> had you defined it in terms of the differentiation, infinite expansion would be easier, but I am not so sure about the way constants accumulate for integration
21:56:02 <oerjan> and the solutions to that are just A sin x + B cos x iirc
21:56:04 <ihope> The solutions to f'' = -f differ only in constants of integration.
21:56:37 <ihope> f(x) = something to do with the integral of f(x) always has exactly one solution, I believe.
21:56:47 <oerjan> the B term must be 0 because that's the value at 0
21:57:10 <oerjan> and the A term probably falls out as 1 too
21:57:14 <ihope> Barring wacky cases, that is. I'm pretty sure it never has more than one solution.
21:57:20 <oklopol> ihope: yeah f(x) = anything + 0*i!f(x), so definitely
21:58:03 <oklopol> but yeah i guess you didn't exactly go for a mathematical truth there :D
21:58:28 <ihope> Just visualize what happens when you define a function in terms of its integral. :-)
21:59:13 <SimonRC> ah, yes, that will usually fix the constant won't it
21:59:58 <oklopol> ihope: i don't think i can
22:00:19 <oklopol> at least i don't see how that leads to just one solution
22:00:19 <ihope> It's the same as defining a function's derivative in terms of the function.
22:00:43 <ihope> And then giving the function an initial condition, just as you'd give the first one a constant of integration.
22:00:48 <oerjan> iirc differential equations usually have unique solutions when they should if not wacky.  although their domain of definition may be shrunk by infinities and stuff
22:01:03 <oklopol> ihope: oh in terms of *just* the integral
22:01:18 <oklopol> well yeah, that's kinda like an automaton
22:01:25 <ihope> Yeah.
22:01:33 * oklopol visualized \o/
22:01:41 <oerjan> f'(x) = g(x,f(x)) type equations, that is
22:01:45 <SimonRC> not "visually", as such
22:02:01 <oklopol> SimonRC: well yeah visually
22:02:20 -!- pikhq has joined.
22:02:27 <SimonRC> I was always shit at solving differential equations and the like
22:02:43 <SimonRC> I suck at ritualised, poorly-understood, methods of solving
22:02:50 <oklopol> i've never solved a differential equation i think
22:02:59 <oklopol> except for ones i've had to do in exams
22:03:24 <ihope> I think I solved a differential equation defining a low-pass filter.
22:03:24 <oklopol> SimonRC: same here
22:04:07 <oklopol> i don't like following algorithms or deriving truths, i like seeing solutions
22:04:10 <oerjan> SimonRC: note that for this you may just need the existence and uniqueness theorems, which are much less ritual than actually finding solutions
22:04:33 <SimonRC> yeah
22:04:41 <ihope> The exact differential equation on the Proce page, in fact, except it multiplied by a constant.
22:09:58 <oerjan> hm i now see that Proce uses distributions, so maybe a different existence & uniqueness theorem is needed (but probably well-known)
22:10:50 <tusho> re: cuil
22:10:51 <tusho> http://xs129.xs.to/xs129/08313/picture7574.png
22:12:16 * oerjan wonders why people trust a company that obviously has Lex Luthor as CEO...
22:13:43 <ihope> Distributions are just limits of functions.
22:13:46 -!- Judofyr has quit (Read error: 104 (Connection reset by peer)).
22:13:48 <ihope> I think.
22:14:05 <SimonRC> Lex luthor was properly bald surely?  And less fat
22:14:14 -!- Judofyr has joined.
22:14:17 <RodgerTheGreat> tusho: cuil is very broken on many levels
22:14:17 <SimonRC> and more charismatic?
22:14:27 <tusho> RodgerTheGreat: totally.
22:14:31 <tusho> but I like that screenshot
22:14:36 <tusho> it's just so ... perfect
22:14:57 <ihope> I'm sure there are no other distributions satisfying the differential equation that I defined the sine with.
22:15:12 <tusho> p.s. if anyone didn't get it (maybe oerjan?)
22:15:15 <tusho> that's steve ballmer in the pic
22:15:48 <SimonRC> a haskeller also 'shotted this:  http://cale.yi.org/autoshare/Screenshot-glasgow.png
22:16:03 <SimonRC> after I pointed it out on #haskell
22:16:08 <RodgerTheGreat> man, what the hell
22:16:10 <oerjan> hm, Eric Schmidt actually wrote Lex - more evidence.
22:16:29 <tusho> SimonRC: someone has a grudge
22:16:30 <tusho> :)
22:16:43 <lament> interesting
22:16:50 <lament> if you search cuil for glasgow now, GHC has a lambda
22:16:54 <SimonRC> I typed in that term after some commenter on the Reg pointed it out
22:17:00 <lament> ...however The Glasgow Centre for the Child & Society still has Bin Laden
22:17:05 <tusho> lament: they have irc bots to weed out that stuff, duh ;)
22:17:07 -!- sebbu2 has joined.
22:17:07 <oerjan> tusho: oh
22:17:14 <SimonRC> but there was a different article in second place next to ObL
22:17:17 <SimonRC> weird
22:17:18 <oerjan> I'VE BEEN MISLED
22:17:25 <tusho> oerjan: that's the funny bit!
22:17:39 <tusho> that's why i screenshotted it...
22:17:45 <tusho> he's involved with google and apple, and ballmer looks ready to shout DEVELOPERS
22:17:49 <pikhq> oerjan: Just thought you might like to know: I'm suggesting that Agora become a sovereign entity.
22:17:55 <tusho> pikhq: oh hi you're here
22:17:58 <tusho> and you're BACK IN AGORA :O
22:18:00 <pikhq> HAI.
22:18:02 <pikhq> :)
22:18:48 -!- atrapado has quit ("Abandonando").
22:20:42 <oerjan> hm steve ballmer being lex luthor somehow sounds much less surprising
22:20:59 <SimonRC> than?
22:21:15 <oerjan> that google guy
22:21:24 <SimonRC> ok
22:22:38 <oerjan> that picture of Simon Peyton-Jones seems familiar - i think i must have seen him on the news somewhere
22:25:18 <tusho> http://upload.wikimedia.org/wikipedia/commons/d/d0/Simon_Peyton_Jones_01.jpg
22:25:20 <tusho> evil spj
22:25:31 <tusho> USE HASKELL OR I WILL EAT YOUR SOUL
22:25:35 -!- olsner has quit ("Leaving").
22:27:00 <oerjan> he definitely looks like someone who could introduce horror movies...
22:27:12 <RodgerTheGreat> diggin' the bowtie though
22:27:17 <SimonRC> indeed
22:27:35 <tusho> his eyeballs appear to be lazily evaluating the scene
22:27:38 <tusho> *rimshot*
22:28:00 <oerjan> possibly even parallel processing
22:28:16 <oerjan> (also *rimshot*)
22:28:49 <SimonRC> wtf is up with that pic?
22:29:00 <SimonRC> it has really bizarre artefacts
22:29:42 <oerjan> don't mind the repaired gunshot holes in the wall
22:29:55 <tusho> SimonRC: its made by a low-quality digital camera
22:29:57 <tusho> maybe a mobile
22:30:01 <tusho> and it's focused on something other than him
22:30:11 <oerjan> those were just from a heated purity discussion
22:30:49 <SimonRC> I thought all Haskellers agreed on the purity thing?
22:31:35 <oerjan> it may have been an Ocaml driveby shooting
22:31:51 <tusho> aww, ephemera.org is taken
22:31:52 <tusho> :(
22:32:27 <SimonRC> oerjan: ah, ok
22:32:27 <oerjan> then you're probably right, i guess it may have been over some extended type system subtlety
22:32:41 <tusho> oerjan: actually no
22:32:49 <tusho> oleg was annoyed, and he looked at the wall
22:32:54 <tusho> you can imagine how long the repairs took...
22:33:02 <oerjan> oh...
22:33:09 <tusho> damn dependent types
22:33:25 <oerjan> well that goes under extended type system subtlety in any case
22:33:50 <tusho> oerjan: not very subtle
22:35:08 -!- sebbu has quit (Connection timed out).
22:35:18 * oerjan notes that trackpads are not well suited for sweaty fingers
22:35:32 -!- sebbu2 has changed nick to sebbu.
22:36:42 <tusho> I WANT EPHEMERA.ORG.
22:36:43 <tusho> >:|
22:36:48 <SimonRC> who has it?
22:37:25 <oerjan> someone hard to find, obviously :D
22:37:45 <tusho> some stupid photographer
22:37:50 <oerjan> hm wait
22:37:50 <tusho> who hasn't updated since jan 07
22:37:58 <tusho> "Derek Powazek"
22:37:59 <tusho> stupid name.
22:38:02 <tusho> he should give me that domain
22:38:49 <oerjan> i'm afraid that domain will never be released.  it will stay that way, forever, just to be ironic
22:39:04 <tusho> heh
23:00:01 <tusho> hrmph.
23:01:06 <SimonRC> I think I made a bad impression elsenet
23:01:19 <SimonRC> ^first^
23:01:33 <tusho> SimonRC: ???
23:02:18 <SimonRC> I asked a joke question that is not only an old joke there, but one that (AFAICT) some people are touchy about.
23:02:24 <SimonRC> *sigh*
23:02:59 <tusho> SimonRC: why are they touchy?
23:03:02 <tusho> touchiness is generally petty.
23:04:43 <oklopol> i'm not touchy about anything try me
23:05:06 <SimonRC> I dunno, it might be all in my head, but I read the only answer given as a cold response that indicates the person has suffered from the actual answer before.
23:05:58 <tusho> SimonRC: what exactly is this question? I'm having trouble imagining this situation
23:05:59 <tusho> :p
23:05:59 <oklopol> well what was the question, you know we're all dying to know
23:06:13 <tusho> 'HAVE YOU BEEN RAPED?'
23:06:18 <tusho> 'Yes.'
23:06:20 * SimonRC wonders where would be a good place to put the logs
23:06:23 <SimonRC> tusho: not that
23:06:26 <tusho> SimonRC: rafb.net/paste
23:06:34 <SimonRC> tusho: less public?
23:06:39 <tusho> it isn't public
23:06:39 <oklopol> pb.vjn.fi
23:06:41 <tusho> there's no "recent" list
23:06:44 <SimonRC> tusho: ok
23:06:56 <tusho> ah wait
23:06:57 <tusho> yes there is
23:07:00 <tusho> pb.vjn.fi then
23:07:04 <tusho> oklopol's
23:07:07 <oklopol> pb.vjn.fi is guaranteed not to be used by anyone except me and a few guys i know :d
23:07:17 <oklopol> also it's not indexed anywhere
23:07:37 <oklopol> logs are not encrypted though, so i may see them if you're unlucky.
23:08:01 <oklopol> err
23:08:03 <oklopol> *pastes
23:10:12 <SimonRC> http://www.vjn.fi/pb/p664553441.txt
23:10:37 <SimonRC> maybe I am being too negative
23:10:47 -!- lament has quit ("leaving").
23:11:02 <SimonRC> #afd is a derivated of the newsgroup alt.fan.dragons, BTW
23:11:04 <tusho> SimonRC: what's the "actual answer"
23:11:12 <SimonRC> "people. that's all that matters"
23:11:25 <SimonRC> I took that as them not liking me asking
23:11:34 <tusho> is kyreeth one of those crazy otherkin or something
23:11:34 <tusho> :)
23:11:42 -!- lament has joined.
23:11:45 <SimonRC> I don't know, 'cause they didn't say
23:11:59 <SimonRC> but it is quite possible
23:12:21 <tusho> well, furries generally tend to have the thinnest skins around, i'm pretty sure otherkin are just flesh and bones
23:12:54 <SimonRC> tusho: not when I usually encounter them
23:13:16 <tusho> well; either that or they're too barmy to know what the hell you're talking about
23:13:23 <tusho> woof woof
23:13:32 <SimonRC> plus, since alt.fan.dragons is a newsgroup, everyone has their vital (and indeed non-vital) stats in their sigs
23:13:41 <SimonRC> tusho: heh
23:13:56 <SimonRC> I have my own mythology of the future, y'know...
23:14:36 <SimonRC> In 2043, furries will protest for the right to have genetically-engineered children, and it will be legalised by early 2045
23:15:18 <tusho> SimonRC: in 2046, the world is a clusterfuck
23:15:19 <SimonRC> of course miscellaneous non-obvious boosting of people will have been happening since the mid 30s, especilly in China
23:15:26 <SimonRC> tusho: um, no
23:15:33 <pikhq> In 2008, the world is an orgy of clusterfucks.
23:15:39 <SimonRC> lol
23:16:05 <tusho> SimonRC: loads of typical furries having a genetically tailored kids?
23:16:08 <tusho> yeah. clusterfuck in a year.
23:16:56 <SimonRC> um, the kids will be only 1-2 years old at that point
23:17:26 <SimonRC> the technology could come sooner than the 2040s...
23:17:40 <tusho> SimonRC: they'll be genetically engineered to grow up 10 times as fast
23:17:41 <tusho> :D
23:18:05 <SimonRC> humans still have much of the genetic gear around
23:18:08 <SimonRC> a few people in India are born with monkey-like tails...
23:18:14 <oklopol> SimonRC: imo the response was unnecessarily hostile
23:18:23 <oklopol> and i could easily have made that same joke
23:18:39 <SimonRC> one minor german aristocratic family had "normal" hair rather than almost-invisible hair all over...
23:19:31 <tusho> tails would be fun.
23:19:34 <SimonRC> a tiny number of people in brazil are born with chimp-like skulls (due to brain not growing)
23:19:49 <SimonRC> tusho: could be awkward
23:19:55 <tusho> SimonRC: most likely
23:20:01 <tusho> = detachable tails
23:20:02 <tusho> :D
23:20:18 <SimonRC> but they would wobble too much, due to not being attache to the spine
23:20:41 <SimonRC> I have seen sites of people who built such things
23:20:59 <tusho> SimonRC: they would be attached to the spine via magic
23:21:24 <SimonRC> I have for a while been considering getting a woolen hat and cutting earholes in it, that would match pointy ears, not mine
23:21:38 <SimonRC> tusho: bah, no magic, science
23:21:47 <tusho> SimonRC: scientific magic
23:22:06 <SimonRC> just for the hell of it and the conversations it would start with people who jumped to a reasonable conclusion
23:24:54 <SimonRC> maybe a t-shirt with huge slits down the back (buttoned up at the bottom) to match it
23:44:26 <SimonRC> (the idea being that they would be great had you wings)
23:46:19 <tusho> command {/(?!^|\s)([^+\s]+)(\+\+|--)(?!$|\s)/ => [:nickname, :direction]} => "Increment or decrement someone's karma." do ... end
23:46:23 <tusho> i am insane
23:47:13 * SimonRC reads
23:47:34 * SimonRC doesn't know ruby's (? ... ) syntax
23:48:07 <tusho> SimonRC: it's just a regexp - (? ...) is for special blocks
23:48:17 <tusho> specifically, ?! is a group that doesn't cause a group in the match output
23:48:19 <tusho> python has that too
23:48:21 <SimonRC> ok
23:48:25 <tusho> with exactly the same syntax, I believe
23:48:31 <tusho> or was it ?:. whatever.
23:48:34 <SimonRC> why not:
23:48:51 <tusho> why not [^\s]? because I want ^ to mean the start of the text
23:48:56 <tusho> not the literal char &
23:48:57 <tusho> *^
23:48:58 <tusho> in fact
23:49:00 <tusho> [\^\s]
23:49:07 <tusho> [^\s] would be anything but \s. but yeah.
23:49:32 <SimonRC> {/(^|\s)([^+\s]+)(\+\+|--)($|\s)/ => [:dummy1, :nickname, :direction, :dummy2]}
23:49:42 <SimonRC> or something like that
23:49:54 <tusho> SimonRC: because although I could do that, it'd be polluting my namespace for little reason
23:50:00 <tusho> when two extra regular chars in the regexp using a widely-used extension could be done intsead :)
23:50:24 <tusho> http://rafb.net/p/IkdJgu73.html full code
23:50:36 <SimonRC> tusho: true
23:50:52 <SimonRC> what is the scope of the variables whose names are given by the symbols?
23:51:03 <tusho> SimonRC: the do ... end block attached to it
23:51:09 <tusho> I'd use do |arg, arg, arg| ... end
23:51:11 <tusho> but I have special things like
23:51:19 <tusho> command "link uri *desc" do ... end
23:51:22 <tusho> which would accept.
23:51:26 <tusho> .link foo This is cool.
23:51:32 <tusho> and give me uri="foo", desc="This is cool."
23:51:44 <tusho> so I just inject the vars into the namespace
23:52:20 <tusho> of course, things like that regexp line are pretty rare...
23:52:29 <tusho> hmm
23:52:32 <tusho> it's broken too
23:52:38 <tusho> it'd only handle the first ++ or -- in a message
23:53:27 <tusho> DAMNIT
23:53:30 <tusho> ontological.org is taken too
23:55:42 <oerjan> it's a simple matter of the domain already existing
23:56:07 <SimonRC> I could give you a subdomain of kigdatsi.org ;-)