00:00:16 <oerjan> btw chiqrsx9+ _may_ take a file name.
00:02:03 <GregorR> It doesn't put anything else on the command line *shrugs*
00:03:21 <EgoBot> Interpreter chiqrsx9p deleted.
00:04:26 <oerjan> !addinterp chiqrsx9p perl http://oerjan.nvg.org/esoteric/chiqrsx9+.pl
00:04:27 <EgoBot> Interpreter chiqrsx9p installed.
00:04:35 -!- okloduk has joined.
00:04:47 <pikhq> Is that HQ9+ improved?
00:05:03 <EgoBot> Deep recursion on subroutine "main::interpret" at /tmp/input.25214 line 56, <STDIN> line 1.
00:05:11 <oerjan> GregorR: addinterp needs to tell when the URL doesn't exist ;D
00:05:44 <pikhq> WTF do c, r, and s do?
00:06:15 <oerjan> c is cat, r is rot13, s is sort
00:07:45 <oerjan> works fine, although only i see most lines
00:08:08 * pikhq thinks it needs ski
00:09:00 <oerjan> oh wait the sort is line based
00:09:14 <oerjan> not too useful on irc :D
00:09:31 <oerjan> pikhq: ski would be against the spirit of it
00:09:49 <pikhq> Ski that can't take s or k as arguments?
00:09:51 <GregorR> WTF, 1l_a_mmi REFUSES to output a 0 >_<
00:10:16 <pikhq> So, khq would be valid, but skkh wouldn't. ;)
00:20:41 -!- psygnisfive has joined.
00:22:56 -!- oklopol has quit (Read error: 113 (No route to host)).
00:26:23 <GregorR> I made a "language" called 1l_butnot to help me write 1l_a code :P
00:26:41 <GregorR> It's 1D, but has only >, <, [] and {} ({} is a loop-while-0)
00:26:55 -!- puzzlet has quit (Remote closed the connection).
00:27:00 <GregorR> I think I need to restrict it a bit more though, because I'm ending up having to add in a bunch of weird code to make it do what I want :P
00:27:10 -!- puzzlet has joined.
00:28:04 <GregorR> (Oh, and it's bitwise :P )
00:38:24 -!- okloduk has quit (Read error: 110 (Connection timed out)).
01:28:41 -!- coppro has quit (Read error: 104 (Connection reset by peer)).
01:29:27 -!- coppro has joined.
01:32:53 <GregorR> Whoot, I have 1l_a outputting 010 :P
01:59:22 * GregorR is meticulously and by-hand converting his 1l_butnot code into 1l_a :P
02:01:20 <GregorR> YAY, I printed an 'H' in 1l_a 8-D
02:03:10 <GregorR> (In 1l_a, that's something to be proud of :P )
02:06:07 <oerjan> so you'll get to the d sometime next week?
02:06:50 <psygnisfive> http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=saint-louis-du-ha!+ha!&sll=37.0625,-95.677068&sspn=37.735377,54.316406&ie=UTF8&t=h&z=11&iwloc=A
02:07:36 <GregorR> Maaaaan, Canada gets the funny versions of our cities :(
02:11:14 <psygnisfive> why is it that when i do a google image search for "laughing frenchman" i get images of horses
02:11:45 <GregorR> psygnisfive: What did you expect?
02:11:54 <GregorR> (haw haw haw we Americans hate ze Franch)
02:13:00 <oerjan> http://en.wikipedia.org/wiki/Westward_Ho!
02:15:25 <pikhq> Not sure why I hit Ctrl-R, anyways.
02:16:16 <oerjan> psygnisfive is out of control
02:18:24 <oerjan> huh, i didn't know irssi was that dominant, about half the VERSION responses were from it
02:18:38 <pikhq> Well, irssi is awesome.
02:20:19 <oerjan> um, the capital R's are mirrored as well
02:21:02 <oerjan> psygnisfive: in the logs they show as mirrored (also rotated upside down)
02:21:29 <oerjan> the upper case ones, not the lower case ones
02:21:33 <psygnisfive> which i guess is mirrored along the horizontal axis
02:21:36 <oerjan> oh they're all upside down
02:21:47 <psygnisfive> the lowercase r's are vertically and horizontally mirrored
02:22:17 <oerjan> the upper case are only vertically flipped
02:26:06 <psygnisfive> the first is a decent approximation of english r's, the second of parisian and munichian r's
02:27:36 <oerjan> i was suddenly wondering if there was something about parisian and munichian ortography i wasn't aware of
02:47:27 <GregorR> psygnisfive: Which English 'r'?
02:48:07 <GregorR> psygnisfive: American English's schwer isn't the same as the consonant 'r'.
02:48:21 <psygnisfive> consonant r. schwar is a rhotic vowel, not r.
02:48:33 <GregorR> Schwer is a rhotic vowel spelled 'r' :P
02:48:41 <GregorR> (With any vowel before that)
02:48:48 <GregorR> (Well, really just 'u' or 'e')
02:51:04 <GregorR> psygnisfive: It's not simply any V+/r/ sequence, but in American English many of them are. (quoth wikipedia:) standard, dinner, Lincolnshire, editor, measure, martyr
02:51:15 -!- puzzlet has quit (Read error: 60 (Operation timed out)).
02:51:24 <GregorR> (I love that that includes every vowel :P )
02:51:52 <psygnisfive> theyre all written with different letters, but thats completely irrelevant.
02:52:00 <GregorR> Yes. And that vowel is a schwer.
02:52:02 <psygnisfive> theres only one vowel represented in those words.
02:52:30 <psygnisfive> and its the schwa+r sequence, in a single syllable, and not across syllable boundaries, which becomes schwar
02:52:38 -!- puzzlet has joined.
02:53:40 <GregorR> I have no idea what you're trying to argue, because what I'm trying to argue is that saying that something is pronounced like English 'r' is ambiguous because it could be the consonant 'r' or a schwer, and what you just said seems to agree with that.
02:54:09 <GregorR> Oh, or are you saying that you would have said "er" if you meant schwer or something...
02:54:21 <psygnisfive> english r is pronounced lik english r. schwar is not an english r. its a separate sound that results from phonological processes
02:55:13 <psygnisfive> and it only arrises in place of an underlying /@r\/. its not, in itself, a way of pronouncing /r\/, but rather a way of pronouncing /@r\/
02:56:16 <GregorR> This seems like a silly argument :P
03:21:01 -!- Sgeo has joined.
04:00:05 -!- Slereah has joined.
04:13:21 -!- WangZeDong has quit (Read error: 110 (Connection timed out)).
04:22:26 <GregorR> I'm starting to un-suspect that 1l_a is T.C.
04:22:47 <GregorR> I think that making practical loops may be impossible.
04:30:57 <GregorR> How (if possible) do you put underscores in the titles of Mediawiki pages?
04:34:07 -!- bsmntbombdood has joined.
04:49:53 <GregorR> BTW, what do people use to edit programs in 2D languages?
04:50:35 <GregorR> I'm using oocalc. It works surprisingly well, and can even save without needing any further modifications to get it in the target form if you know how to ask it right, but it's a bit of a PITA because you can't just type, you have to press an arrow after every key.
04:50:46 <GregorR> I don't suppose somebody's written an editor?
04:53:36 <oerjan> it's been a while, but i think i used vim with the virtualedit option
04:55:08 <oerjan> hm that doesn't help with writing in other directions though...
04:57:48 <oerjan> i vaguely recall i may even have tried emacs for it
05:23:16 <GregorR> Yeah, that's not bad ... I think I prefer oocalc by a liiiiiiiiiiittle bit.
05:46:46 <oerjan> GregorR: i think the emacs thing was called picture mode or something, and did allow for writing in other directions
06:13:55 -!- oerjan has quit ("Good night").
06:20:21 <GregorR> One other very nice feature of using oocalc is that, since it's a spreadsheet, I can really trivially copy 2D areas. Can presumably do that in emacs too?
06:34:45 -!- Sgeo has quit (Read error: 110 (Connection timed out)).
07:23:22 -!- olsner has joined.
07:33:32 -!- GreaseMonkey has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:14:22 -!- tombom has joined.
08:40:04 -!- lereah_ has joined.
08:48:51 <lereah_> The number of people dying is determined by the Fermi-Dirac distribution
09:22:55 -!- coppro has quit (Read error: 110 (Connection timed out)).
10:04:25 -!- WangZeDong has joined.
10:04:44 <lereah_> Madre de dios, es el Dong!
10:19:08 -!- Slereah has quit (Read error: 110 (Connection timed out)).
10:26:44 -!- kerlo has quit (wolfe.freenode.net irc.freenode.net).
10:27:18 -!- MizardX- has joined.
10:29:02 -!- MizardX has quit (Read error: 54 (Connection reset by peer)).
10:29:09 -!- kerlo has joined.
10:29:14 -!- MizardX- has changed nick to MizardX.
10:51:30 -!- GreaseMonkey has quit ("YES -> thor-ainor.it <- THIS IS *DELICIOUS*!").
10:57:40 -!- puzzlet_ has joined.
10:58:18 -!- puzzlet has quit (Remote closed the connection).
12:22:01 -!- jix has joined.
12:54:07 -!- MizardX has quit ("Proclamation of invalidity!").
13:08:46 -!- KingOfKarlsruhe has joined.
13:11:11 -!- MizardX has joined.
14:32:49 -!- olsner has quit (wolfe.freenode.net irc.freenode.net).
14:32:49 -!- Gracenotes has quit (wolfe.freenode.net irc.freenode.net).
14:32:49 -!- AnMaster has quit (wolfe.freenode.net irc.freenode.net).
14:32:54 -!- olsner has joined.
14:32:54 -!- Gracenotes has joined.
14:32:54 -!- AnMaster has joined.
14:37:11 -!- Gracenotes has quit (Remote closed the connection).
14:40:11 -!- FireFly has joined.
14:42:35 -!- MigoMipo has joined.
14:58:47 <ehird> bsmntbombdood: what did you want me for yesterday/day before?
15:15:59 -!- Sgeo has joined.
15:16:05 -!- MigoMipo has left (?).
15:22:35 <ehird> 3d realms shut down
15:22:38 <ehird> Duke Nukem Forever is over
15:23:49 <Sgeo> That reminds me, even though Free Realms looks like junk, I should try it
15:26:04 <ehird> "Poor developers got screwed over. They coded the whole thing in Arc before an implementation was released, then Paul Graham pulled a stunt and released a set of MzScheme macros instead of the auto-vectorizing native code compiler he promised them." —Slava Pestov
15:28:01 <ehird> Sgeo: on Duke Nukem Forever in response to 3D Realms shutting down
15:33:34 -!- lereah_ has quit ("Leaving").
15:35:07 -!- tombom has quit ("Peace and Protection 4.22.2").
15:38:02 -!- oerjan has joined.
15:44:58 -!- Slereah has joined.
15:49:50 <ehird> http://www.reddit.com/r/programming/comments/8ihf5/3d_realms_shuts_down_no_duke_nukem_ever_forever/c09e6or
15:49:53 <ehird> Please let that bullshit be true.
15:52:00 <ehird> http://duke.a-13.net/?080907
15:52:12 <ehird> Things done during the development of duke nukem forever
15:52:16 <ehird> and bonus: things that took shorter.
15:52:23 <ehird> Asztal: could have been registered just to post that :p
15:52:32 <ehird> i mean, by a real ex-employee, highly doubtful though
15:52:36 <ehird> but I fucking want it to be true
15:52:41 <Sgeo> Um, that site mentions a twitter user, but doesn't link to it
15:52:58 <ehird> Sgeo: 'hodapp' is a link, retardo :P
15:53:05 <ehird> it links to twitter.com
15:53:12 <pikhq> The freaking *Beatles* took less time than Duke. :)
15:53:15 <ehird> Sgeo: just above, though: Updated by Eli Hodapp - http://twitter.com/hodapp - http://a-13.net/ - Last Edit: 5/06/09
15:53:27 <ehird> pikhq: the beatles weren't especially long lived...
15:53:39 <Sgeo> But still, having @hodapp link to twitter.com is braindead
15:53:51 <ehird> pikhq: the rolling stones have lasted 47 years so far...
15:53:56 <ehird> Sgeo: or an honest mistake zomg
15:57:03 -!- WangZeDong has quit (Read error: 110 (Connection timed out)).
15:57:32 <ehird> i totally didn't give a shit about dnf until that recent 15-second-or-so gameplay video, that thing just looked amazing
15:58:16 <pikhq> ehird: True. However, the Beatles had a very innovative and, well, *filled* career.
15:58:24 <pikhq> Fine, for something more stunning.
15:58:31 <pikhq> *Project Manhattan* took less time.
15:58:33 -!- puzzlet_ has quit (Remote closed the connection).
15:58:37 -!- puzzlet has joined.
15:59:00 <ehird> It'd be hard to even get it going elsewhere; nobody would fund it or buy it
15:59:26 <oerjan> project manhattan actually needed to get finished first...
15:59:42 <ehird> Project Manhattan was the name of a duke nukem game
15:59:58 <ehird> he meant the nuclear shit
16:00:01 <ehird> but it was a duke game too
16:00:04 <ehird> http://en.wikipedia.org/wiki/Duke_Nukem:_Manhattan_Project
16:15:27 <ehird> http://developer.valvesoftware.com/wiki/Valve_Time ← hahaha
16:19:27 * ehird starts the Bailout 3D Realms Party
16:20:11 <oerjan> but is that the fake rapture or the real one?
16:24:08 <fungot> BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB ...too much output!
16:31:24 <ehird> lawl, dumb redditors responding to me with the same tired, broken anti-piracy arguments
16:31:37 <ehird> i replied to one and was going to bother with the others but then I realised I don't give a shit <:)
16:37:17 -!- ais523 has joined.
16:39:38 <ehird> http://blog.interop.com/blog/2009/05/06/the-father-of-the-crc-passes-away/
16:42:06 <ais523> wow, for ages my keyboard wasn't working for anything but MouseKeys
16:42:13 <ais523> but it's started working again, it seems
16:46:21 * ehird starts W:A in VM. Let's see if I can't pass grenade training this time.
16:48:33 <ehird> Bah, only got the first one. As usual.
16:48:38 <ehird> Sgeo: indeed, 'tis the awesome
16:48:52 <Sgeo> ehird, we agree on something? *gasp*
16:50:41 <ais523> Sgeo: do you disagree with AnMaster on anything?
16:51:07 <Sgeo> I don't remember ever agreeing or disagreeing with AnMaster on anything
16:51:54 <ehird> It's funny, W:A in this 512MB-of-RAM VM running a full virtualized Windows is as fast as on my real pc
16:54:49 <Sgeo> Ugh, I don't know if I should go to school today
16:55:28 <Sgeo> On the one hand, the only thing that the class is today is finding out if I passed or failed, which I may or may not be able to find out via email
16:55:44 <Sgeo> On the other hand, there are girls at the pizza place on the way home that I might not see again
16:56:30 <EgoBot> Supported commands: addinterp bf_txtgen daemons delinterp fyb fyb.orig help info kill userinterps 1l 2l adjust asm axo bch befunge befunge98 bf bf16 bf32 bf8 bfbignum c chiqrsx9p cintercal clcintercal cxx dimensifuck echo forth glass glypho hello kipple lambda lazyk linguine malbolge pbrain perl qbf rail rhotor rot13 sadol sceql sh slashes test trigger udage01 underload unlambda whirl yodawg
16:56:55 <ais523> why do EgoBot commands not tab-complete?
16:56:57 <ais523> it would be really useful...
16:57:03 <AnMaster> ais523, make your client do it
16:58:24 <Sgeo> !chiqrsx9p xgibberish
16:59:28 -!- puzzlet has quit (Remote closed the connection).
16:59:32 -!- puzzlet has joined.
16:59:57 <oerjan> the x command is a _tiny_ bit hard to use, i fear
17:00:08 <ehird> it doesn't even make it TC
17:00:10 <ais523> nah, 1 in 256 of the time it's trivially easy
17:00:26 <oerjan> you could rerun it until it worked
17:01:03 <EgoBot> 99 bottles of beer on the wall,
17:01:14 <Sgeo> aww, it messages me the rest
17:01:23 <Sgeo> Instead of being an evil menace to the channel
17:01:49 <ehird> it should just do like lambdabot
17:01:55 <ehird> do the first few lines, then require a command to do more
17:01:57 <Sgeo> Ok, it's going a bit slowly
17:02:00 <ehird> thus if it floods it's your fault :)
17:02:32 <GregorR> Sgeo: It does exponential backoff, and only sends 8 messages. There is no way to get the full output.
17:02:34 <Deewiant> 99 bottles of beer * 2 seconds / bottle = over 3 minutes of spam via PRIVMSG
17:02:57 <Sgeo> It's slower than 2sec/bottle
17:03:25 <Deewiant> Meh, 2 seconds / line is the acceptable rate
17:03:39 <ehird> i forgot how amazing worms' music is
17:03:43 <GregorR> <GregorR> Sgeo: It does exponential backoff, and only sends 8 messages. There is no way to get the full output.
17:04:06 <Sgeo> It sent 5 messages, including the one to the channel
17:04:20 <ehird> and it'll send a few more
17:04:51 <GregorR> Nowait, I'm sorry, five messages, the last one is /eight/ seconds.
17:05:18 <fizzie> Deewiant: Your "acceptable" rate of drinking is 2 seconds / bottle? Rather hard-core.
17:05:44 <GregorR> Any slower and you're a cop-out.
17:05:45 <Deewiant> To be pedantic, I did not claim 2 seconds / bottle was acceptable
17:06:32 <ehird> actually i'm not sure it counts as music, more ambience
17:06:46 <oerjan> !chiqrsx9p xprint "Hi!";
17:07:02 <ehird> !chiqrsx9p xprint "Hi!";
17:07:03 <ehird> !chiqrsx9p xprint "Hi!";
17:07:05 <ehird> !chiqrsx9p xprint "Hi!";
17:07:07 <ehird> !chiqrsx9p xprint "Hi!";
17:07:09 <ehird> !chiqrsx9p xprint "Hi!";
17:07:11 <ehird> !chiqrsx9p xprint "Hi!";
17:07:14 <ehird> by the birthday paradox, how many times do we have to do this>
17:07:15 -!- Hiato has joined.
17:07:15 <GregorR> ehird: Yes, very helpful of you.
17:07:28 <oerjan> birthday paradox doesn't apply
17:07:39 <ais523> nah, it's not birthday
17:07:58 <ais523> and no, oerjan, multiplying by 0.5 gives the expectation, which isn't what's wanted here
17:08:10 <ais523> for the median number of times, you have to multiply by approximately ln 2
17:08:17 -!- Hiato has quit ("Leaving.").
17:08:58 <GregorR> "-----###" looks more like a mace than a fly-swatter to me.
17:09:04 <GregorR> And a mace is a mean thing to swat somebody with.
17:09:58 <Sgeo> D&D style alignment system for nomics: Silly/Serious Lenient/Pedantic
17:10:25 <GregorR> Does "Silly/Pedantic" make sense?
17:10:41 <ais523> GregorR: yes, I've seen such nomics in the past
17:10:59 <GregorR> (I've generated 35 potential 26-letter pangrams, but they all include either "VFW" or "WV" >_<)
17:11:10 <oerjan> so is agora still at Serious/Pedantic?
17:11:21 <ehird> agora's serious/pedantic-with-handwaving
17:11:26 <ais523> in fact, it may be in a B-style crisis atm
17:11:29 <ais523> which is almost unprecedented
17:11:37 <ais523> because if it is a genuine crisis, ratification wouldn't have worked
17:11:43 * GregorR assumed that Sgeo was referring to participants, not the nomics themselves.
17:12:26 <oerjan> GregorR: i'd suggest doing the rarest letters first, since that gives fewest options
17:12:26 <GregorR> The participants can span the board and just sort of "average out" to something.
17:12:50 <GregorR> oerjan: A friend of mine suggested that, but I had already been running this for days, so now he's running that :P
17:12:51 <ehird> oerjan: the latest herald's report mentions that you're behind insane proposals
17:12:53 <ehird> will you marry me?
17:13:46 <oerjan> it was a silly proposal
17:14:01 <oerjan> insane proposals still exist?
17:14:26 <ehird> oerjan: did the map exist in your time?
17:15:13 <oerjan> there may have been a game something map-based
17:15:27 <ehird> oerjan: http://agora.qoid.us/current_flr.txt; first rule is the map
17:15:40 <ehird> it's not much of a rule, having no effect
17:17:20 <oerjan> ehird: that same silly week i also voted by phone
17:18:15 <oerjan> alas, i only hit steve's answering machine
17:19:08 <ehird> i only know steve from the spam scam
17:20:05 -!- KingOfKarlsruhe has quit (Remote closed the connection).
17:20:31 <oerjan> that map is definitely from after my time
17:20:57 <ehird> i don't actually know when you left :P
17:22:23 <ehird> H. Distributor Steve gets a spam email to the lists. He edits it to include a without-objection rule amendment
17:22:26 <ehird> and lets it go on to the lists
17:22:34 <ehird> Everybody's spam filters ignore it
17:25:05 <ais523> GregorR: I didn't know you were a nomic fan
17:25:07 <ehird> GregorR: Yeah, it is a bit of an abuse of office though
17:25:12 <ehird> ais523: does he have to be to appreciate that?
17:25:29 <ais523> not knowing X does not necessarily imply that X is true, though
17:25:47 <GregorR> I've never nomic'd because I think the whole thing is a waste of time, but that doesn't mean I don't enjoy them :P
17:26:03 <ehird> GregorR: dude you operate a site about choosing what hat you will wear
17:26:16 <ehird> and use a neural net to match colours
17:26:18 <GregorR> WARNING. WARNING. ABOVE STATEMENT WAS SARCASM
17:26:44 <GregorR> I /still/ want to find a group of people to play my card-game Nomic.
17:26:49 * ehird kills two worms in his first two turns by (1) fire punch (2) baseball bat into the sea. Yay.
17:26:56 <ehird> GregorR: make it playable online and you're on :-P
17:27:16 * ais523 vaguely advertises ##nomic
17:27:18 <GregorR> ehird: That would sort of kill it, as any change to the rules that deviates too much would require a rewrite of the engine >_>
17:27:25 * ehird vaguely disadvertises ##nomic
17:27:28 <ais523> which has been really active recently despite ehird's attempts
17:27:29 <Sgeo> ehird, that's why I set the Agora filters to not let emails from there go into spam
17:27:38 * GregorR vaguely vagueomits all over ##nomic
17:28:14 <Sgeo> Why does ehird hate ##nomic ?
17:28:20 <Sgeo> Not because I'm there, I hope
17:28:31 <ehird> one which is not on topic for #esoteric
17:28:38 <ehird> in case ais523 wants to but in with his own version of events ;-)
17:28:50 * ehird thinks of an esolang based on Worms.
17:28:59 <ehird> that would be quite hard to program in, I imagine
17:29:31 <ehird> you'd have worms which when you hit them with weapon X output a certain character, etc
17:35:27 -!- BeholdMyGlory has joined.
17:39:42 * AnMaster ponders writing an SQL database engine in befunge
17:40:22 <ehird> ais523: do you know any reason a background animation in an old game should speed up because you move your mouse?
17:40:33 <ehird> I'm trying to figure out how you do animation that leaves that oddity...
17:40:41 <ais523> ooh, I'm aware of the general principle
17:40:45 <ehird> maybe they're doing it every event?
17:40:52 <ehird> and setting a timer to trigger the event every N
17:40:52 <ais523> all sorts of things speed up if you move your mouse
17:40:56 <ehird> but your mouse moving triggers it
17:40:56 <ais523> and yes, it's generally based on event loops
17:41:13 <ais523> also to do with borken implementations of slowing down idle programs in a multitasking environment
17:41:23 <ais523> you move your mouse -> the program isn't idle -> it isn't slowed
17:41:24 <ehird> there should be a bit of hardware that just gives an event to the OS as fast as possible, evidently
17:42:43 <AnMaster> using SOCK to allow client(s) to connect.. hm
17:45:53 -!- oerjan has quit ("leaving").
17:53:08 <GregorR> There are now two programs written in 1L_a
17:54:01 <GregorR> Now, for a Boolfuck interpreter!
17:59:40 <Deewiant> Intel's phasing out the i7 940
17:59:50 <ehird> Nuthin' wrong with it.
18:00:04 <Deewiant> http://www.tgdaily.com/html_tmp/content-view-42332-135.html
18:00:08 <ehird> Well fuck Intel :-P
18:00:14 <ehird> The 940 is a nice balance.
18:00:49 <GregorR> !boolfuck ;;;+;+;;+;+;+;+;+;+;;+;;+;;;+;;+;+;;+;;;+;;+;+;;+;+;;;;+;+;;+;;;+;;+;+;+;;;;;;;+;+;;+;;;+;+;;;+;+;;;;+;+;;+;;+;+;;+;;;+;;;+;;+;+;;+;;;+;+;;+;;+;+;+;;;;+;+;;;+;+;+;
18:00:50 <ehird> doesn't matter though
18:00:53 <ehird> if you have one, you have one
18:00:59 <ehird> Deewiant: do you use ecc memory btw?
18:01:05 <ehird> the i7 doesn't support it which is weird
18:01:34 <ehird> it doesn't really seem worth it
18:02:14 <AnMaster> what exactly does static mean in the context of C++...
18:02:21 <ehird> bsmntbombdood: wut?
18:02:32 <Deewiant> AnMaster: It's overloaded even in C
18:02:38 <AnMaster> Deewiant, yes, but not as much
18:02:45 <bsmntbombdood> ehird: they say that non-ecc has 1 bit error per gigabyte per month
18:02:59 <ehird> that's funny, my computer isn't crashing randomly due to bits being misplaced.
18:03:04 <Deewiant> It's even more overloaded in D
18:03:04 <ehird> are you sure that's not marketing talk from the ecc people :)
18:03:09 <ais523> a random misplaced bit generally has no effect anyway
18:03:11 <AnMaster> Deewiant, err... static == local to file in C? Can be used for functions and variables.
18:03:38 <Deewiant> AnMaster: Yes, at global scope. At function scope, it means that a variable is global but visible only from there.
18:03:51 <AnMaster> <ehird> that's funny, my computer isn't crashing randomly due to bits being misplaced. <-- a lot of the time it will be in currently unused memory I suspect.
18:03:59 <ais523> ehird: because of all the bits that might flip in memory, only a small proportion of them generally have an effect on control flow
18:04:09 <ais523> many will be unused, or in an image or text
18:04:11 <AnMaster> Deewiant, ah right true, don't use that a lot so didn't think of that.
18:04:12 <ais523> or loop counters in delays
18:04:20 <ehird> surely it can't be such a problem
18:04:23 <ehird> if they sell non-ecc memory
18:04:24 <ais523> tbh, even a random bitflip in an executable will be unlikely to do anything worse than making it crash
18:04:25 <ehird> and people are fine with it
18:04:47 <Deewiant> In D, we have: static variables, static constructors/destructors, static assert, static if, static import
18:04:49 <ehird> bsmntbombdood: i've never used ecc memory and I've never had a problem at all
18:04:58 <ais523> bsmntbombdood: much less than once per month, most likely on a program that crashes more often than that for other reasons?
18:05:00 <ehird> I think the problems may be greatly exaggerated
18:05:11 <ais523> bsmntbombdood: what OS do you run?
18:05:12 <Deewiant> And stuff crashes more than once per month for me anyway.
18:05:15 <ehird> also, it slows down ram access a tiny bit :P
18:05:28 <Deewiant> If I get an additional crash due to not using ECC memory, I really won't notice.
18:05:35 <Deewiant> ehird: 2% is the usual figure IIRC.
18:05:47 <bsmntbombdood> ehird: write a program to md5sum 1 gb of randomly initialized memory and put it at nice 19 for a month
18:05:57 <ais523> bsmntbombdood: I write lots of programs, they crash all the time until I get them working properly
18:05:58 <ehird> bsmntbombdood: i don't do that shit :P
18:06:00 <AnMaster> Deewiant, Any reason you can't introduce new keywords instead?
18:06:18 <Deewiant> AnMaster: Walter doesn't like doing that.
18:06:21 <ehird> is ECC memory really needed
18:06:29 <ehird> I can't think of one corruption I've had of memory
18:06:30 <Deewiant> I understand that doing it too much is bad, but I think static is too overloaded :-P
18:06:32 <ais523> anyway, the vast majority of memory used by an executable is in large malloced objects
18:06:46 <ais523> and bitflips there are unlikely to be particularly disastrous
18:06:53 <ais523> probably just cause one particular result to go haywire
18:06:58 <ais523> or a segfault, if it hits a pointer
18:07:15 <ehird> ais523: that sounds pretty bad to me
18:07:21 <ehird> OTOH that's never ever ever ever EVER happened to me
18:07:30 <ehird> and i'll bet it's never happened to most other people
18:07:35 <ehird> if it does, then non-ecc memory is a sham
18:07:52 <Deewiant> http://www.reddit.com/r/programming/comments/8i77h/do_you_use_ecc_dram_why/
18:07:54 <ais523> ehird: what proportion of your memory do you normally use?
18:08:01 <ehird> Deewiant: that's where I got my question from
18:08:08 <ais523> this sort of thing would probably happen about once a decade
18:08:12 <AnMaster> Deewiant, sure yes, you can run into issues with user named identifiers in some cases. But in most that interpretation doesn't make sense
18:08:22 <AnMaster> on the other hand I guess that makes the grammar even harder to parse...
18:08:43 <ehird> Deewiant: can you buy superfast ram with ecc, even?
18:09:05 -!- tombom has joined.
18:09:08 <ehird> Deewiant: as in "excessive like that ddr ssd" sort of fast :-)
18:09:42 <ehird> I wonder if xeon has any DISadvantages vs i7
18:09:44 <AnMaster> ehird, with enough money I bet you could develop such a thing
18:13:27 <bsmntbombdood> AnMaster: http://www.ramsan.com/products/ramsan-400.htm
18:13:42 <ehird> http://www.ramsan.com/products/ramsan-440.htm
18:13:50 <ehird> also that's not RAM :-P
18:21:09 -!- Sgeo has quit (Read error: 110 (Connection timed out)).
18:24:35 -!- ais523 has quit ("getting dinner, I'll be back later").
18:40:12 <GregorR> Bleh, Boolfuck is little-endian.
18:40:24 <GregorR> Is there a *fuck that's JUST bitwise brainfuck?
18:40:54 <GregorR> That is, the only change is removing - and + switches bits? Not little-endian like Boolfuck, not combining more operators like Bitchanger?
18:41:09 <Deewiant> It should take about 2 minutes to whip up Bitfuck :-P
18:41:34 <GregorR> Yeah, but I was hoping for one that already existed and already had a trivial translation from BF so I wouldn't have to do that first :P
18:41:36 <Deewiant> I.e. no, I don't know of one, but feel free to make one
18:42:12 <GregorR> AnMaster: It does the bits backwards. It's bit-wise little-endian.
18:42:16 <GregorR> AnMaster: Which makes no kind of sense X_X
18:42:38 <AnMaster> GregorR, big endian makes just as much sense.
18:43:17 <GregorR> When was the last time you wrote a binary number in little-endian order :P
18:43:54 <AnMaster> GregorR, well, bit-endianness usually isn't visible on most platforms. But for bytes... hm... half a month ago I think.
18:44:29 <AnMaster> It was when messing with inline asm and SSE.
18:45:12 <GregorR> OK, how about I go from this angle instead: 1L_a uses "big-endian" bitwise reads/writes, so writing a Boolfuck interpreter in it would suck because I'd have to buffer.
18:45:28 <AnMaster> I don't even know what 1L_a is
18:45:42 <GregorR> http://www.esolangs.org/wiki/1L_a
18:46:29 <GregorR> I want to prove that it's T.C., although I'm not yet convinced that it is :P
18:52:28 -!- M0ny has joined.
18:57:01 -!- MizardX has quit ("I'll be back in a few minutes. I have to change location...").
18:58:18 * ehird listens to Graue's noise to attempt to come to esozen
18:59:22 <ehird> I like this. I don't think I'm supposed to.
19:01:32 <AnMaster> ehird, err... what is "esozen"
19:01:54 <ehird> http://oceanbase.org/data/files/music/Scrap%20Heap%20-%20Live%20at%20Electric%20Possible,%20Feb.%202009.mp3
19:01:59 <AnMaster> somehow I read it as "reason" on the first pass. I have no idea how that happened.
19:02:00 <ehird> recorded quietly, he sez.
19:02:08 <ehird> also, esozen is the exact opposite of reason
19:02:17 <ehird> it stands to esozen that it should be very similar in appearance
19:02:34 <ehird> AnMaster: creator and curator of esolangs.org
19:02:41 <ehird> aka catatonic purpoise
19:03:12 <ehird> It's not actively hostile, it just doesn't have melody.
19:03:17 <ehird> It's quite soothing
19:03:43 <ehird> Mostly a bit of bashing some stuff and a little not-that-whiney feedback.
19:03:47 <lifthrasiir> i'm tired of pattern matching in python and trying to write esotope-bfc in ocaml again, aww.
19:03:53 <ehird> lifthrasiir: HASKELL.
19:04:21 <ehird> Bah, but I want to write mine in haskell :P
19:04:28 <AnMaster> ehird, I only hear a few clicks?
19:04:38 <ehird> AnMaster: Turn the sound up (it's recorded quiet) and wait.
19:04:47 <ehird> I promise you it doesn't suddenly turn into white noise :P
19:04:58 <AnMaster> ehird, or suddenly turn into high volume?
19:05:02 <lifthrasiir> at least i have written some more codes in ocaml than haskell, so... :p
19:05:07 * AnMaster wgets and imports into audio editor just in case
19:05:23 <ehird> i have better things to do than find noisy music that turns int o white noise just for you AnMaster :P
19:05:35 <lifthrasiir> anyway they both are not familiar to me, i indeed have to learn functional languages seriously.
19:06:28 <AnMaster> ehird, doesn't need to be white noise to break eardrums. I have headphones so I'd prefer to be extra careful. I do want to be able to hear anything tomorrow too.
19:06:42 <ehird> goddamn, it's not loud, honestly
19:07:03 <ehird> AnMaster: 5-7 minutes?
19:07:14 <lifthrasiir> well i also have used prolog (not free software environment, btw) but i felt uncomfortable whenever i use it.
19:07:18 -!- MizardX has joined.
19:07:20 <ehird> prolog kind of sucks
19:07:28 <ehird> what i want is haskell+prolog syntax
19:07:35 <AnMaster> ehird, it sucks in a brilliant way?
19:08:25 <AnMaster> lifthrasiir, talking about slowness. Link to your optimising compiler?
19:08:33 <AnMaster> I assume it is even better now
19:08:45 <lifthrasiir> AnMaster: http://code.google.com/p/esotope-bfc/
19:09:02 <AnMaster> lifthrasiir, how good is it at lostkingdom now
19:09:41 <AnMaster> ehird, audacity says 15 minutes..
19:09:55 <ehird> i was guesstimating.
19:10:27 <lifthrasiir> i was fixing some bugs and considering several corner cases so far. the main optimizer didn't see much progress.
19:11:08 <lifthrasiir> on the other hand i'm planning array optimization, but i'm not decided how to do yet.
19:11:20 * ehird gains esolightenment
19:12:03 <AnMaster> ehird, I listened to it for a bit. Not my type of, mu^Wno^Wperformance.
19:12:13 <ehird> I never claimed it was
19:12:18 <ehird> You're the one who brought it up :P
19:12:29 <ehird> Graue has like 10 releases of that kind of stuff
19:13:54 <AnMaster> lifthrasiir, you could remove those stores to cells right
19:14:11 <AnMaster> lifthrasiir, plan to implement that?
19:14:13 <ehird> lifthrasiir: use fputs
19:14:17 <ehird> in case someone outputs %s
19:14:22 <ehird> (not puts, it gives a newline)
19:14:37 <lifthrasiir> ehird: well you're right. of course i escaped % properly ;)
19:14:43 <ehird> printf is slower! :-(
19:14:58 <Deewiant> You'd think puts(x) and fputs(x, stdout) are equivalent but no
19:15:00 <ehird> printf("Hello World!");
19:15:02 <AnMaster> lifthrasiir, well you could just turn it into fputs("Hello World!"); and drop those assignments before
19:15:02 <ehird> you don't need to write to p
19:15:04 <ehird> since they're never read
19:15:06 <ehird> trivial optimization
19:15:08 <ehird> Deewiant: yeah it pisses me off
19:15:10 <ehird> fputs("Hello World!")?
19:15:18 <Deewiant> ehird: GCC is quite clever enough to figure those out?
19:15:27 <ehird> Deewiant: GCC can figure a lot of shit out
19:15:33 <ehird> But clang also does array→variable conversion
19:15:34 <Deewiant> Assuming the resulting C code is put through an optimizing compiler I wouldn't bother
19:15:36 <ehird> So relying on just gcc is silly
19:15:37 <lifthrasiir> i also plans to do so but i think global optimizer is needed to do it efficiently
19:15:39 <ehird> Also, might as well be sure,.
19:16:00 <AnMaster> unsigned char m[30000], *p = m;
19:16:01 <Deewiant> ehird: My point being that any decent compiler should be able to see that it's not used.
19:16:10 <ehird> Deewiant: not worth relying on
19:16:14 <ehird> it's trivial to remove anyway
19:16:19 <Deewiant> Not necessarily, if it is trivial
19:16:22 <AnMaster> lifthrasiir, allows optimiser to see it won't be used outside the single file
19:16:25 <Deewiant> But if it isn't, I wouldn't bother.
19:16:34 <AnMaster> it doesn't know that it is only one file yet. The linker does.
19:16:49 <AnMaster> of course you can tell the compiler that
19:17:09 <lifthrasiir> AnMaster: that is ad-hoc too, i really have to analyze how many cells are used.
19:17:12 <AnMaster> but usually that involves finding whatever switch the compiler use.
19:17:22 <lifthrasiir> for example, static unsigned char m[5], *p = m;... etc.
19:17:23 <AnMaster> lifthrasiir, not decidable in the general case.
19:17:36 <ehird> nor are most optimizations
19:17:41 <ehird> lifthrasiir: are you gonna make an auto-growing tape?
19:17:50 <ehird> e.g. let ,[a thousand >s] work
19:18:07 <lifthrasiir> AnMaster: at least you know optimized hello world program won't use any memory cells.
19:18:22 <AnMaster> lifthrasiir, yes. It will work for trivial programs
19:18:47 <AnMaster> in fact it will work for any "balanced loops only"-program
19:20:18 <AnMaster> and possibly for some with unbalanced ones (for example, you could potentially know what cell is set in this program: >>>,<<<[>].)
19:21:12 <AnMaster> in case you optimise [>] into a memchr() or similar anyway.
19:25:02 <ehird> AnMaster: I lie -- it does get really loud all of a sudden at one point.
19:25:06 <ehird> The applause at the end :P
19:25:24 <AnMaster> ehird, applause from people or from GM?
19:26:10 <lifthrasiir> AnMaster: there are many cases such as [>>>>] (memory operation independent to p[0] here) [<<<<] in lostkng; my current concern is how to optimize them, as it is effectively an array operation.
19:26:46 <AnMaster> ehird, I don't remember instrument number, but GM and/or GM2 contains applause.
19:27:08 <lifthrasiir> i can optimize [>] into memchr(), but i cannot [>>] or [<]. nor i think using memchr is important optimization.
19:27:29 <ehird> [>>] is very common for th 1 elem 1 elem 1 elem 0 structur
19:27:51 <ehird> lifthrasiir: you can optimize [>>] to
19:28:07 <ehird> but I assume you do
19:28:50 <lifthrasiir> well of course we can use similare techniques used in memchr to [>>] or [>>>>]; but that's another story.
19:28:51 <AnMaster> lifthrasiir, you can do -=n for [<] [<<] and so on
19:29:20 <lifthrasiir> example in lostkng output: while (p[43] != 0) p -= 2;
19:29:22 <AnMaster> lifthrasiir, anyway memchr is _probably_ better than for (;*p; p++)
19:30:08 <ehird> lifthrasiir: you mean the (int*) way of reading shit?
19:30:15 <ehird> that sort of stuff
19:30:35 <AnMaster> since glibc memchr() is insanely optimised. Like strlen(), memchr() is made to search <machine word size> of bytes at once.
19:30:58 <lifthrasiir> AnMaster: SIMD instructions are good for some thousand bytes... iirc.
19:31:59 <lifthrasiir> AnMaster: am i wrong? my knowledge on x86 is quite limited. :S
19:32:14 <Deewiant> memchr just reads using using (unsigned long*)
19:32:29 <AnMaster> lifthrasiir, SSE operands and registers are 128 bits wide.
19:32:44 <AnMaster> not sure you can do byte searching with it
19:33:04 <AnMaster> memcpy() and such definitely can use SSE to copy fast.
19:33:14 <AnMaster> bsmntbombdood, really? which SSE instructions would you use for it.
19:35:29 <bsmntbombdood> i think theres a compared packed bytes -> mask instruction
19:35:48 <AnMaster> bsmntbombdood, which SSE version?
19:36:45 <AnMaster> can't find anything like it in the AMD64 reference docs.
19:36:55 <bsmntbombdood> PCMPEQB, Compare packed bytes in mm/m64 and mm for equality.
19:39:07 <lifthrasiir> AnMaster: as far as i know libc memchr() does similar thing, xoring current dword with (byte * 0x1010101) and find null byte in it.
19:39:12 <AnMaster> hm that could work. But testing if you hit any match would not be that trivial. Control flow + SSE doesn't match well.
19:39:40 <AnMaster> lifthrasiir, err 0x1010101... don't you mean 0x101010101010101
19:40:13 <AnMaster> lifthrasiir, well actually it would be a machine sized word on my computer.
19:40:42 <lifthrasiir> AnMaster: i think that's a very reason that SIMD-based memchr is slower than original memchr for smaller cases.
19:40:48 <bsmntbombdood> AnMaster: load one register full of the byte you are looking for, the other with a portion of the string, then xorps, then use the bytes -> bitmask operation, then ffs
19:42:09 <AnMaster> bsmntbombdood, you need to store the result to temporary memory then
19:42:32 <AnMaster> afaik you can't easily find first set on an sse register directly
19:43:02 <AnMaster> bsmntbombdood, since my computer only has SSE3 that feels completely irrelevant to me :P
19:43:51 <lifthrasiir> bsmntbombdood: "Packed Compare Explicit Length Strings, Return Index"? does it work like repe/repne prefix?
19:43:55 <ehird> MORE REASON WHY THE i7s/XEONS ARE TOTALLY AWESOME <_<
19:44:16 <fizzie> SSE4.2 has a CRC32 opcode, that's also pretty funky.
19:44:19 <ehird> bsmntbombdood: with your new machine you can write hyper efficient code that nobody else can run! woohoo :-)
19:44:24 -!- puzzlet has quit (Read error: 60 (Operation timed out)).
19:44:49 <Deewiant> The newer Core 2s have SSE 4.1, the older ones have SSSE3, AMD's still at SSE3
19:44:55 <ehird> AnMaster: What's that do
19:45:05 <AnMaster> ehird, it's like SSE5 or something iirc
19:45:17 <ehird> INTEL FUCKS AMD SUCKS. I can't think of anything better to rhyme with sucks
19:45:18 <GregorR> bsmntbombdood: I will sell you 1 assembly knowledge for 1 proof of 1L_a-Turing-completeness.
19:45:25 <AnMaster> ehird, actually I think ABM was intel
19:45:31 <ehird> GregorR: STOP SELLING NON-SCARCE ASSETS BITCH
19:45:35 <ehird> AnMaster: I was talking to Deewiant :-P
19:45:43 <ehird> 19:44 Deewiant: The newer Core 2s have SSE 4.1, the older ones have SSSE3, AMD's still at SSE3
19:45:59 -!- puzzlet has joined.
19:46:00 <Deewiant> AVX is Intel's and due in 2010
19:46:03 <bsmntbombdood> AnMaster: i think i remember seeing an ffs instruction on sse registers
19:46:07 <ehird> amd are adopting avx
19:46:15 <Deewiant> AnMaster: Doesn't ring a bell.
19:46:18 <ehird> Therefore avx is better qed
19:46:28 <AnMaster> Deewiant, it is mentioned in man gcc for the -msse* section
19:46:39 <Deewiant> AMD are incorporating parts of AVX, to be exact.
19:46:52 <Deewiant> Intel may or may not incorporate parts of SSE5.
19:47:00 <AnMaster> bsmntbombdood, well that is irrelevant to me too.
19:47:19 <bsmntbombdood> AnMaster: it's not my fault you have an archaic proc
19:47:30 <AnMaster> bsmntbombdood, um. archaic is way older
19:47:38 <AnMaster> this is just "not last edition"
19:47:58 <ehird> bsmntbombdood: dude, you use a P4
19:48:04 <ehird> unless your computer arrived already :P
19:48:34 <bsmntbombdood> ehird: newegg is slow, my order is "processing", whatever that means
19:48:55 <AnMaster> bsmntbombdood: Here is how it is defined: archaic = PDP/11, very old = pentium3, old = pentium 4, early 64-bit sempron = not state-of-art but still rather new
19:48:56 <fizzie> It means they're using your computer.
19:49:03 -!- Hiato has joined.
19:49:05 <fizzie> "Hey, it's processing."
19:50:39 <Hiato> what's the best primality test? Bu that I mean which uses the least/lowest complexity time for the best probability?
19:50:52 <Deewiant> Your early 64-bit Sempron is slower than some Pentium 4s.
19:51:01 <bsmntbombdood> Hiato: it depends if you want deterministic or nondeterministic
19:52:34 <AnMaster> lifthrasiir, "This pass is likely to change, since it depends on ad-hoc analysis currently. SSA-based optimizer is planned, but not yet implemented."?
19:52:40 <Hiato> bsmntbombdood: well, at this point, its pretty immaterial. I was looking at AKS, but I can't find a good implementation. The number I want to test is too massive for the standard trial division or elliptic methods
19:52:48 <ehird> PDP11 is more than archaic
19:53:01 <AnMaster> ehird, what do you call PDP11 then
19:53:08 <ehird> AnMaster: an abacus
19:53:11 <lifthrasiir> AnMaster: mainly for constant & copy propagation and dead code elimination.
19:53:14 <Deewiant> ehird: What do you call an abacus
19:53:16 <AnMaster> ehird, that is technically incorrect.
19:53:25 <ehird> Deewiant: A collection of atoms
19:53:32 <lifthrasiir> current optimizer doesn't do any optimization across basic blocks
19:53:47 <AnMaster> ehird, what do you call a collection of atoms.
19:54:06 <Hiato> heh, well, the thing is I just wanna plug (3^797161-1)/2 into something and see what I get
19:54:08 <EgoBot> Use: !fyb <program name> <program>
19:54:31 <EgoBot> Sorry, I have no help for fyb!
19:54:41 <EgoBot> Sorry, I have no help for befunge98!
19:54:45 <GregorR> http://codu.org/eso/fyb/README
19:55:00 <Deewiant> Oh right, that thing we played with yesterday :-P
19:55:38 <Hiato> bsmntbombdood: cool, will check
19:55:41 <GregorR> http://codu.org/eso/fyb/report.txt // Somebody tied with logicex-2 :P
19:55:42 <Deewiant> Hey, I beat everybody except you, w00t.
19:55:48 <AnMaster> lifthrasiir, what about polyhedra loop optimisation?
19:56:37 <Hiato> awesome, will do, thanks
19:56:56 <lifthrasiir> AnMaster: that's for distant future right now. heh. i didn't view any paper or articles on it yet.
19:57:06 <AnMaster> "One is to use two or more cells per item, where one is used for actual value, one is used to mark the boundary of array (for [>>] or variants), and others are scratch cell."
19:57:29 <AnMaster> lifthrasiir, iirc ais bf backend to gcc used 5 cells
19:57:33 <ehird> 19:53 AnMaster: ehird, what do you call a collection of atoms.
19:57:37 <ehird> a collection of quarks
19:57:48 <AnMaster> ehird, and what do you call a collection of quarks? ;P
19:57:49 <bsmntbombdood> AnMaster: are you talking about brainfuck optimization?
19:58:02 <ehird> AnMaster: a collection of quarks
19:58:07 <ehird> Discrete space, bitch.
19:58:18 <lifthrasiir> bsmntbombdood: i'm making highly-optimizing brainfuck compiler for now. :)
19:58:30 <AnMaster> bsmntbombdood, no. I'm talking about optimising intercal of course... what do you think
19:59:06 <lifthrasiir> if you are interested, please visit http://code.google.com/p/esotope-bfc/
19:59:47 <AnMaster> read the page and the linked ones.
20:00:20 <AnMaster> lifthrasiir, where is the "browse code" thing?
20:00:41 <AnMaster> lifthrasiir, no browser there, only download info.
20:00:52 <ehird> it has a web viewer
20:01:02 <AnMaster> ehird, not clickable. Too much work ;P
20:01:11 <ehird> AnMaster: hg.mearie.org/esotope/bfc/
20:01:15 <ehird> http://hg.mearie.org/esotope/bfc/
20:01:39 <AnMaster> ehird, anyway I could just select it in the browser and right click and select "open url in new tab"
20:04:17 <AnMaster> lifthrasiir, why all in one file btw?
20:05:00 <AnMaster> personally I would probably split emitter code into one file, loading code in one, and optimisation in one. Possibly several for different optimisation passes
20:06:33 <lifthrasiir> AnMaster: since it is in heavy development i won't split them before its structure once stablized.
20:07:48 <lifthrasiir> and it turned out that current intermediate representation has many deficiencies
20:07:58 <AnMaster> lifthrasiir, sure, but why does splitting into modules make that harder? You have the same interfaces between the classes anyway? Just easier to manage in the editor
20:08:15 <AnMaster> I assume your editor has no issues with having more than one file open at once of course
20:08:50 -!- jix has quit (No route to host).
20:09:09 <AnMaster> "structured into multiple private modules" does not imply "stable public API" to me...
20:09:31 <lifthrasiir> AnMaster: for that, it was just borethsome and there is no other reason.
20:10:08 <AnMaster> huh. I find long files more bothersome than multiple files. :P
20:11:52 <AnMaster> bsmntbombdood, it isn't hosted on google code
20:12:05 <lifthrasiir> i write many codes both for work and for fun, and i just don't think of quality for fun unless there are a ton of codes.
20:12:07 <AnMaster> but if you open the non-clicky link..
20:12:30 <AnMaster> lifthrasiir, meh. I always try to think of quality when coding for fun ;P
20:12:39 <AnMaster> but then again I guess that is a matter of taste.
20:14:10 <lifthrasiir> AnMaster: of course some quality is automatically achieved, but i'm not aiming at best quality ;)
20:15:05 <ehird> bsmntbombdood: yer right.
20:15:19 <ehird> fastest, most direct
20:15:22 <lifthrasiir> i totally ignored that. thank you for pointing out.
20:15:26 <ehird> lifthrasiir: listen to the man GregorR ;-)
20:16:07 <GregorR> But buffering is bad and write is portable.
20:16:19 <ehird> ah, buffering plays with lostkng
20:16:31 <ehird> if you enable long descriptions you have to input something to get the first room description
20:16:33 <ehird> if you have buffering
20:16:36 <GregorR> And makes PSOX etc difficult.
20:19:17 <lifthrasiir> okay, it emits fwrite(stdout, "Hello World!", 12); now.
20:19:20 <bsmntbombdood> how good can a c compiler be at using the sse instructions?
20:19:30 <Deewiant> As good as a C compiler can be
20:20:01 <ehird> lifthrasiir: write(1)
20:20:24 <ehird> lifthrasiir: death to buffering! death to no speed!
20:20:29 <ehird> bsmntbombdood: as I said, buffering breaks shit
20:20:29 <bsmntbombdood> do not use the system call interface if you can avoid it
20:20:51 <ehird> why not just use write(2)
20:20:56 <ehird> and get some extra microspeed
20:21:05 <Deewiant> ehird: Death to buffering == death to speed, in some cases.
20:21:21 <Deewiant> write(2) isn't Windoze-compatible.
20:21:25 <ehird> Deewiant: Yes, but you have to not buffer.
20:21:32 <ehird> For LostKng and PSOX, for two examples
20:21:37 <ehird> Also, what GregorR said
20:21:44 <Deewiant> GregorR: Cygwin doesn't count.
20:21:48 -!- ais523 has joined.
20:21:51 <GregorR> Deewiant: Windows has write(2)
20:22:28 <GregorR> Haven't you ever used MingW? :P
20:22:29 <ehird> SO WHAT I'M SAYING IS, lifthrasiir use write(2) :-P
20:23:01 <lifthrasiir> btw, is there any brainfuck compiler worth adding to Comparison page?
20:23:20 <ehird> lifthrasiir: ah, wait
20:23:24 <AnMaster> <ehird> if you enable long descriptions you have to input something to get the first room description
20:23:27 <ehird> lifthrasiir: GregorR's EgoBF suite.
20:23:53 <ehird> lifthrasiir: And the two best interps: http://swapped.cc/bf/ http://mazonka.com/brainf/index.html
20:23:53 <AnMaster> lifthrasiir, whenever you have a input command, do an fflush() first.
20:24:02 <AnMaster> then you can still allow buffering
20:24:11 <AnMaster> because it will flush before it waits for input
20:24:12 <ehird> wtf bsmntbombdood, be patient
20:24:15 <lifthrasiir> ehird: those are mentioned in the page already. ;)
20:24:18 <ehird> postage can take like a week in the uk
20:24:20 <ehird> lifthrasiir: ah, kay
20:24:21 <AnMaster> lifthrasiir, this is what cfunge does btw.
20:24:22 <ehird> lifthrasiir: even egobf?
20:24:25 <ehird> it has a compiler, jit and interpreter
20:24:45 <Deewiant> GregorR: Yes, I have, but obviously I don't use POSIX functions with MinGW. :-P
20:25:07 <ais523> it used to be a pain getting stdin/stdout working on Windows
20:25:08 <Deewiant> Anyway, I see that it is in io.h.
20:25:12 <ais523> although allegedly, it's easier nowadays
20:25:12 <ehird> bsmntbombdood: you canucks are so impatient. you are a canuck right
20:25:34 <ehird> bsmntbombdood: canadian
20:25:35 <Deewiant> So using write(2) isn't non-GCC-compliant unless you're careful to import the right thing.
20:25:40 <AnMaster> <lifthrasiir> btw, is there any brainfuck compiler worth adding to Comparison page? <-- don
20:26:03 <ehird> bsmntbombdood: err i seem to remember you were
20:26:08 <ehird> bsmntbombdood: k, USians are so impatient
20:26:09 <GregorR> Deewiant: People use compilers other than GCC? Hyuk hyuk]
20:26:10 <lifthrasiir> since it didn't have any project page i assumed that is not public
20:26:24 <Deewiant> GregorR: You've used DMC yourself. :-P
20:26:41 <GregorR> Only to see how shitty and not-GCC it is!
20:26:55 <GregorR> I've also used wcc, pcc and tcc :P
20:26:57 <ehird> GCC is pretty the crap...
20:27:11 -!- bsmntbombdood has quit (Read error: 104 (Connection reset by peer)).
20:27:13 <Deewiant> DMC will error out on #include <unistd.h> on Windows.
20:27:31 <ais523> with "can't find unistd.h"?
20:27:35 <Deewiant> #error "unistd.h is not for Windows use"
20:27:36 <ais523> or does it have a unistd.h with a #error in?
20:27:43 -!- bsmntbombdood has joined.
20:28:00 <ais523> I love the way that the old Windows documentation used to have portability boxes in it
20:28:08 <ais523> saying which of the functions were POSIX-compatible
20:28:21 <ais523> the answer is, not really any of the useful ones, unless they were in the C standard as well
20:29:38 <ehird> Hmm, Linus uses Fedora
20:29:44 <ehird> I'd have thought he would be more of an LFS man
20:30:16 <ehird> Well, I agree, but he's a make-a-kernel guy.
20:30:19 <GregorR> He probably compiles his own kernel ;)
20:30:28 <ehird> I'd expect him to, say, build his own base and then use some distro's packages on top of it
20:30:29 <Deewiant> He programs low-level stuff, sure, but he's not much of a sysadmin.
20:32:46 <ehird> Damn, I just realised that I first tried Ubuntu w/ 5.10. That's ages ago :P
20:33:09 <GregorR> ehird: If the first three digits of the year are the same, it's not ages ago.
20:33:31 <Deewiant> I've used it since our school has it, but not otherwise.
20:33:40 <ehird> GregorR: I only started existing ten years before that release was made.
20:33:42 <ehird> So it's ages ago for me.
20:34:03 <ehird> My first linux distro was a crappy thing called PCLinuxOS; I don't remember why
20:34:16 <ais523> ehird: I thought PCLinuxOS was rather popular
20:34:19 <GregorR> ehird: Well geeze, is it MY fault you couldn't get off your butt and into existence?
20:34:21 <ais523> although not as popular as Ubuntu, of course
20:34:26 <ehird> The name is really cheesy.
20:34:33 <ehird> It's an operating system for your personal computer!
20:34:37 <GregorR> It's a Mandr{ake,iva} derivative.
20:34:38 <ehird> I never stuck with any of them 'til 2006, ofc, because of winmodem issues
20:35:31 <fizzie> "PCLinuxOS, often abbreviated as PCLOS" makes it sound like a yet another Common LISP object system.
20:36:08 <ehird> Pico Common Lisp Object System
20:36:10 <ehird> It's really minimalist!
20:36:10 <GregorR> The Perl Common LISP Object System :P
20:36:12 <ais523> Deewiant: anyway Ubuntu is basically Debian, but with a nicer interface, some dubious patches, a really rigid release schedule, and a rather lax attitude towards bug reports
20:36:30 <GregorR> <ais523> Deewiant: anyway Ubuntu is basically Debian, but with a nicer interface, some dubious patches, a really rigid release schedule, and a rather lax attitude towards bug reports // an excellent description
20:36:48 <fizzie> "Ubuntu is basically Debian, but with the core values sacrificed."
20:36:54 <lifthrasiir> but that doesn't made Debian obsolete. they just have different goals.
20:37:04 <ehird> fizzie: a bad interface is a core value of Debian?
20:37:07 <ehird> That explains *a lot*.
20:37:08 <ais523> more to the point, Ubuntu isn't a fork, just a patchset
20:39:23 <ehird> I'd like an Ubuntu without the dubious patches, without the non-free stuff by default, more upstream interaction, and less tampering with other people's software.
20:39:33 -!- bsmntbombdood has quit (Read error: 104 (Connection reset by peer)).
20:39:40 <ehird> Debian gets the first two but doesn't have the nice interface and release schedule and shit
20:39:56 <ehird> And other distros are either source based, have a tiny community and thus aren't really very polished, ...
20:40:08 -!- bsmntbombdood has joined.
20:40:08 <ehird> I'd probably be all over fedora core if it didn't use rpm.
20:40:22 <ais523> it's the dubious patches that give Ubuntu its interface niceness, though, mostly
20:40:38 <ehird> No, that's the "tampering with other people's software"
20:40:46 <ehird> Most of ubuntu's niceness can be done w/ separate packages
20:40:50 <ehird> like the notification system, etc
20:42:12 <ehird> So separate polish packages for a nicer interface + nice release schedule + FOSS only by default (according to DFSG more or less) + upstream interaction + doesn't patch other's software more than necessary + good, medium-to-large sized community = distro heaven.
20:42:14 <ehird> Doesn't exist though.
20:42:22 <ehird> Oh, and + nice package manager + that means not RPM.
20:42:37 <ehird> (Because Fedora seems to fulfill most of the rest, heh.)
20:42:57 <ais523> what in particular do you have against RPM, by the way?
20:43:06 <ais523> I've never used it, so I don't know why it's so hated, and I'm interested
20:43:13 <ehird> ais523: You cannot possibly imagine.
20:43:53 <ais523> no, that's why I'm asking
20:44:01 <ais523> it's got to be better than CPAN, surely?
20:44:19 <ehird> Imagine distilled hell as a package manager, with many things that should be bugs but are design decisions. Imagine it breaks shit wildly. Imagine a weeping child. Imagine Satan typecasted to a package manager. Imagine this. Imagine this. Imagine -- you're falling, oh god it's me, *I'm* falling, where am I, who am Imagine this.
20:44:33 <ehird> That's basically what using rpm is like.
20:44:48 <GregorR> I think that's a bit over the top :P
20:45:07 <ais523> I'm after technical info, not metaphors
20:45:20 <ehird> ais523: That WAS the technical info.
20:45:25 <ehird> Are you sure you want to hear the metaphors?
20:45:38 <ais523> ehird: I don't believe you
20:45:51 <GregorR> I have no idea what's better about .deb and pals, but I know that I've always had weird packaging issues on RPM-based distros and never on .deb-based distros. Not much of a technical detail, because I've never been able to distill particular issues, it's just always that way.
20:48:18 <AnMaster> personally I prefer source based ones. IME it works better when you are mixing stable and testing packages.
20:48:24 <AnMaster> and also better in many other cases
20:48:35 <ais523> well, source-based is slower
20:48:50 <ais523> I generally use binary for most things, but swap out to source-based for things I'm testing or modifying
20:49:02 <ais523> one thing I like about debs is you can mix source and binary without much trouble
20:49:06 <ehird> AnMaster: my apparent issue with the arch linux installer is that when it tries to update the package manager definitions I get a Transient resolver failure.
20:49:10 <ehird> ais523: checkinstall++
20:49:18 <AnMaster> ais523, not really: scenario: glibc 2.6 -> 2.7 upgrade, discover valgrind package is broken and needs to be recompiled against new glibc
20:49:28 <ais523> well, for two versions of Ubuntu in a row, now, there's an error in at whenever I run the package manager
20:49:31 <ais523> it's been reported for months
20:49:34 -!- Gracenotes has joined.
20:49:35 <ais523> but they haven't fixed it yet
20:49:39 <AnMaster> a) file a bug and build a new binary package locally then install it
20:49:50 <AnMaster> b for source based, a for binary ones
20:49:58 <AnMaster> there I say source based ones are faster.
20:50:01 <ais523> AnMaster: pull updates, and find that it's been automatically rebuilt against the new glibc by the build farm
20:50:02 <ehird> AnMaster: you have clearly never used a binary package manager.
20:50:14 <ehird> AnMaster: I meant to say a good one
20:50:32 <ehird> anything other than dpkg/apt is shit, and if you don't know what you're doing with dpkg/apt they're shit too.
20:50:43 <ehird> getting cosy with dpkg/apt = package manager heaven.
20:50:55 <ais523> dpkg and apt are actually a separation of powers
20:51:02 <ais523> dpkg is a package manager, apt manages repositories
20:51:02 <GregorR> Yay slackware .tar.gz packages!
20:51:07 <ais523> dpkg doesn't care where the packages come from
20:51:08 <GregorR> tar vzxf mypkg.tar.gz LAWL
20:51:14 <AnMaster> ehird, getting cosy with any good package manager == heaven. That is true for portage and pacman at least
20:51:16 <ehird> GregorR: YOU FORGOT THE POST-INSTALL SCRIPT
20:51:19 <ais523> oh, and .debs are tar.gz too, just with a different extension
20:51:27 <ehird> AnMaster: not as good as dpkg/apt.
20:51:44 <AnMaster> ehird, only if you don't know what you are doing with portage and/or pacman :P
20:51:46 <GregorR> ais523: Actually they're three files in a .ar, one of which is data.tar.gz :P
20:51:49 <ais523> I'd say a good source-based package manager can be as good to use as a good binary-based manager
20:52:03 <ais523> and the diff file, and the dsc
20:52:09 <AnMaster> ehird, anyway you totally ignored my example above
20:52:11 <ais523> just you get them unpacked if you use apt-get source
20:52:14 <ehird> Deewiant: it's not going on to a tape!
20:52:22 <ais523> that's for legal reasons, I think
20:52:24 <ehird> AnMaster: no, because ais523 covered it, and also dpkg/apt can handle BOTH SOURCE AND BINARY PACKAGES
20:52:27 <ehird> which you don't seem to have realised
20:52:32 <AnMaster> ehird, it needs the package manager to remember to rebuild valgrind and such, IME they often forget that
20:53:47 <AnMaster> ehird, maybe debian managed that somehow. Marking "must be rebuilt when dependency foo is updated in the y bit of x.y-patchlevel" or such could work.
20:53:47 <pikhq> ehird, dpkg/apt is not a very good source-based package manager.
20:53:49 <ais523> ehird: whenever anyone on Slashdot claims that open source program (insert name here) is really hard to compile, I grab the source package and dependencies and compile the source package to prove them wrong
20:53:58 <pikhq> It *is*, however, the best damned binary package manager.
20:54:04 <ehird> pikhq: it isn't the best, but it's the best binary package manager, and it does okay at doing source stuff
20:54:12 <ehird> so you can mix them without too much pain; binary works in most cases
20:54:12 <ais523> yes, as source-based you have to compile by hand
20:54:17 <AnMaster> ais523, what about nethack. That is one hard to compile IMO.
20:54:19 <ehird> AnMaster's only example was an edge case
20:54:21 <ehird> ais523: you can write a script to do that
20:54:29 <AnMaster> ais523, at least, for multi-user installs.
20:54:32 <ehird> but yeah, AnMaster's only example was an edgecase, and when doing source stuff with apt as an edge case it works
20:54:35 <pikhq> I'd say that either Portage or Ports is the best source-based package manager...
20:54:44 <ais523> but the assumption is that, if you got the source package, you probably wanted to modify before compiling
20:54:54 <AnMaster> ehird, um. I find that kind of stuff is rather common with rolling release.
20:55:05 <ehird> AnMaster: and you attribute this to everything but rolling release :)
20:55:33 <pikhq> Portage, of course, is a poor binary package manager. Mostly, its usage for binary packages is rolling out the same configuration to a bunch of systems.
20:55:52 <GregorR> FLAAAAAAAAAAAAAAAAAAAAAAAAARGHLE
20:55:59 <GregorR> This conversation never ends.
20:56:05 <AnMaster> pikhq, I find ports a bit bad at updates to packages like perl and python, where you need to rebuild various modules afterwards. Generally ports seems to work well but need more work (reading /usr/ports/UPDATING and so on)
20:56:08 <ehird> blame that other guy
20:56:25 <GregorR> Somebody write some FYB code! :P
20:56:34 <AnMaster> ehird, non-rolling releases have lots and lots of other issues.
20:56:53 <pikhq> GregorR: I'll give you some credit here: Slackware is definitely the most *elegant* package management system. ;)
20:57:34 <GregorR> Somebody write some FYB code! D8<
20:57:34 <pikhq> There's something nice about "Well, it's just a tarball."
20:57:54 <GregorR> Somebody write some good FYB code :P
20:58:06 <AnMaster> ehird, like bumpy rides when it is time for major upgrades. A lot of work then. Personally I prefer to have the non-trivial upgrades spread out over the year, like one non-trivial package / month rather than 5 non-trivial ones at the same time.
20:58:07 <ais523> GregorR: you have to limit FYB program length
20:58:19 <ehird> click update manager, like you always do
20:58:24 <ehird> click upgrade on the "new distro!" thing
20:58:32 <ais523> because otherwise you can start with something like %?[%%%%%%%%%... 1000000 more no-ops here...%%]?
20:58:33 <AnMaster> ehird, well debian is better than ubuntu when it comes to that iirc.
20:58:44 <ehird> ubuntu's the best at easy distro upgrades
20:58:47 <GregorR> ais523: The best you can hope for is a tie with that.
20:58:58 <ais523> GregorR: you put the rest of your program at the end
20:59:03 <AnMaster> ehird, from what I heard from ais that seems not true.
20:59:09 <GregorR> ais523: Hmmmmmmmmmmmmmmmm, 'struth.
20:59:15 <ais523> the point being that the other program needs to run a million > commands to reach your pointer
20:59:19 <GregorR> ais523: Maybe I should make nops not-typable.
20:59:23 <ehird> AnMaster: ais is, I fear, one of those people who breaks everything he touches.
20:59:51 <ais523> GregorR: either you should make all [] run at least once, or limit program length, really
21:00:01 <ais523> otherwise something similar could be done
21:00:23 <ais523> maybe you could do it INTERCAL-style, where [ abstains from everything up to and including the next ]
21:00:32 <AnMaster> ehird, well, maybe I'm one of them too. Thus I prefer a very robust system. One that is easy to mend when you broke it. One with less GUI abstractions, since I will always need to edit the text configs anyway sooner or later.
21:00:33 <ais523> actually, even that's speed-of-light movemetn
21:01:07 <ehird> I need to learn to not argue with people who only use the same arguments and who I already know are wrong ;-)
21:01:10 <ais523> fwiw, just starting with one million % without the loop would have a similar effect, just not as strongly
21:01:23 <AnMaster> ehird, Gentoo is robust in my experience.
21:01:31 <ehird> AnMaster: see what I just said?
21:01:35 <ehird> Not interested, stfu.
21:01:46 <ais523> ehird: anyway, from what I've heard, Ubuntu distro upgrades work well, but Debian's work even better
21:01:58 <AnMaster> ehird, ah you say that every time you realise I won the discussion.
21:02:01 <ais523> I admit that only one out of four Ubuntu upgrades have gone smoothly for me
21:02:09 <ehird> AnMaster: no, I didn't
21:02:12 <ais523> but in at least one of those, and probably two, it was my fault
21:02:15 <ehird> stop reading a fuckin' conspiracy into everything
21:02:23 <ehird> you've said all your arguments, stop rallying your goddamn troops
21:02:31 <AnMaster> ehird, You can have a conspiracy with just a single person in it?
21:02:46 <ehird> only you could imagine such a thing, but it seems so.
21:03:14 <AnMaster> ehird, you seemed to be suggesting it. I never even considered the idea before.
21:03:27 <pikhq> AnMaster: Gentoo is actually not all that robust.
21:03:36 <pikhq> Gentoo is a distro that assumes you know what you're doing.
21:03:47 <pikhq> And if you don't, bad shit happens.
21:04:02 <AnMaster> pikhq, well yes. But it is easy to mend if bad thing *does* happen.
21:04:03 <GregorR> No, LFS is a distroy that assumes you know what you're doing :P
21:04:20 <AnMaster> not that bad things happened since 2004 on gentoo to me.
21:04:20 <pikhq> (generally next time they do a major change, like XFree86 -> X.org or 2.4 -> 2.6)
21:04:28 <pikhq> GregorR: LFS is more so.
21:05:18 <AnMaster> GregorR, LFS is "follow the manual" then "realise the system is unmaintainable due to no easy way to upgrade" then a few days later "switch back to the gentoo install"
21:05:30 <ais523> incidentally, I've once before now built a Linux system via makefile
21:05:37 <ais523> it was a makefile that was full of wget
21:05:42 <ais523> basically, linux-from-scratch
21:05:55 <AnMaster> ais523, yes that exists already. ALFS iirc
21:05:58 <ais523> it got all the bits it used from their original websites
21:06:04 <ais523> the kernel from kernel.org, for instance
21:06:11 <GregorR> AnMaster: LFS is for learning, not for being your "real" distro.
21:06:27 <AnMaster> GregorR, and I have done LFS, HLFS, CLFS
21:06:41 <ais523> GregorR: why don't they make LFS more user-friendly, then
21:06:54 <ais523> so it does work as a real distro
21:07:04 <AnMaster> wasn't it the last stable version when I did it too
21:07:06 <ehird> the author of lfs uses it
21:07:18 <AnMaster> ehird, there is no single author of it
21:07:29 * AnMaster has been on the linux from scratch irc server
21:07:32 <ehird> there's the original author.
21:07:35 <GregorR> <AnMaster> They're all married.
21:08:34 <AnMaster> GregorR, care to explain that totally non-sequitur reply...
21:08:46 <GregorR> No, I'm tired of explaining my jokes :P
21:09:02 <AnMaster> GregorR, well since you directed it at me...
21:09:04 <ehird> GregorR: that's the harsh reality of AnMaster
21:09:14 <ehird> <foo> x isn't directing, it's impersonating.
21:09:31 <GregorR> Except when you're MAKING A FUCKING JOKE
21:09:47 <GregorR> <AnMaster> ehird, there is no single author of it <AnMaster> They're all married.
21:09:48 <AnMaster> GregorR, well since you impersonated me... What was the joke.
21:09:53 <GregorR> Because they're all MARRIED
21:09:56 * ehird rubs eyes, swigs whiskey ← don't do anmaster, kids
21:10:04 <ehird> it'll ruin your life
21:10:06 <GregorR> See how it's not funny when I have to explain it X_X
21:10:10 <AnMaster> GregorR, is this a reference to something?
21:10:25 <ehird> GregorR: let's go mad together
21:10:39 <ehird> ahahahahahaahahahaha
21:11:28 <Deewiant> AnMaster: "Single" as in unmarried.
21:11:36 <AnMaster> Deewiant, ah right. that meaning.
21:11:48 <ehird> GregorR: so, about that going mad
21:11:50 <ehird> I'm done, how about you?
21:12:01 <GregorR> ehird: I'm holding on to my sanity by thinking about eating breakfast for dinner.
21:12:19 <ehird> <GregorR> Furthermore, time is cyclic.
21:12:25 -!- tombom_ has joined.
21:12:57 <AnMaster> would be interesting with a cyclic universe
21:13:51 <AnMaster> ehird, well, if you invented some way to become immortal you could have fun the second time around.
21:14:09 <ehird> if time's cyclic, you'd do the exact same thing each time
21:14:14 <ehird> and the matter would be exactly the same
21:14:19 <ehird> thus you would not have any memories
21:14:30 <ehird> effectively you'd be a new person
21:14:33 <AnMaster> ehird, depends on cyclic in what way
21:14:36 <GregorR> ehird: Unless you were a living time paradox :)
21:14:41 <ehird> AnMaster: In the cyclic way.
21:15:27 <AnMaster> ehird, what if you could survive the end and pass through to the beginning next time around. Then you could use it as a time machine
21:15:39 <ehird> Then time is not cyclic.
21:15:52 <AnMaster> ehird, that doesn't make good sf...
21:16:10 <ehird> "Time is cyclic except when it's not."
21:16:31 <AnMaster> ehird, s/when it is not/when you reverse the polarity/
21:16:42 <AnMaster> there you have a perfect concept!
21:19:09 <fizzie> Hm; that's no fun: VirtualBox's "OSE" open-source edition doesn't support the SATA controller, according to their web page; but since the configuramator GUI still has a "enable additional controller" checkbox in the hard disks tab, and lets you attach things up to 30 SATA ports, so I thought maybe they enabled that feature in the open-source version too; but then it just gives a cryptic VERR_PDM_DEVICE_NOT_FOUND error when starting.
21:19:13 <ehird> Hey, Arch installed.
21:19:19 <ehird> fizzie: it works for me.
21:19:30 <ehird> Do I have the closed source one?
21:19:46 <fizzie> How should I know what you have? It's free-for-personal-use, though, so you might well have it.
21:20:03 <AnMaster> <ehird> How surprising. <-- why?
21:20:05 <ehird> fizzie: Use qemu/virtualbox's closed source one?
21:20:15 <ehird> AnMaster: It hasn't worked in two VMs previously?
21:20:47 <ehird> So you keep saying. No, it was Arch not supporting a part of the VM configuration.
21:21:01 <ehird> Although I suppose that's heretical to the Holy Church of Distro Perfectness.
21:21:10 <AnMaster> ehird, You haven't given enough details to debug the issue
21:21:20 <fizzie> I don't mind the lack of SATA controller so much (that I'd bother doing something that's not directly aptable), it's just that I don't like how they're giving me false hopes by making it available in the config screens, then crashing them down by not booting.
21:21:27 <ehird> I didn't ask anyone to debug it, AnMaster.
21:21:41 <AnMaster> ehird, you complained loudly though
21:21:59 <ehird> If by loudly you meant I mentioned I gave up on arch due to not working in a vm as a one-line remark.
21:22:15 <AnMaster> constructive feedback and details bug reports is more useful
21:22:30 <ehird> And I don't give a damn what you want me to do regarding Arch bugs.
21:24:08 <fizzie> And anyway the open-source edition supports a couple of SCSI controllers, so I can build this "going-to-be-transferred-on-a-real-hardware-where-the-HD-will-be-/dev/sda" thing without any device-naming differences.
21:25:10 <fizzie> Or maybe not. Strange. I thought it was supposed to do the SCSI controllers.
21:25:23 <ehird> fizzie: just use qemu
21:25:27 <ehird> it's slow but acurrate
21:27:13 <fizzie> I guess I could, but since I have it in the 'box already... did qemu happen to do the .vdi thing? Not that I couldn't just convert the image.
21:27:29 <ehird> Qemu uses .raw and .qcow, I think.
21:27:40 <GregorR> qemu supports tons of formats, but not .vdi
21:28:14 <GregorR> Supported format: nbd parallels qcow2 vvfat vpc bochs dmg cloop vmdk qcow cow host_device raw
21:28:35 <fizzie> Well, it's a fixed-size .vdi, so it's pretty much .raw with an offset; and "VBoxManage converthd --format RAW" should get rid of the offset, too.
21:28:41 <AnMaster> ehird, so you recommend to not file any bugs to any projects, instead only complain in some totally unrelated irc channel and not report the issue at all
21:28:45 <fizzie> Er, clonehd, not converthd.
21:29:21 -!- tombom has quit (Read error: 110 (Connection timed out)).
21:29:22 -!- tombom_ has changed nick to tombom.
21:30:10 <AnMaster> "open source editions" lacking most of the useful features are worse than "closed source, but no-cost" software IMO
21:30:19 <AnMaster> at least the latter doesn't pretend to be open
21:30:33 <fizzie> It doesn't really lack "most of the useful features"; it's pretty usable as-is.
21:30:37 <GregorR> Actually, VirtualBox OSE excludes only some spectacularly useless featuers.
21:30:49 <AnMaster> fizzie, lacking SCSI emulation is "lacking major feature" to me
21:31:14 <AnMaster> also what about forwarding serial port and usb devices?
21:31:17 <GregorR> Lacking emulation of a technology no one uses?
21:31:29 <AnMaster> oh wait, it can't do the latter in any edition iir
21:31:36 <AnMaster> not sure about forwarding serial and usb
21:31:45 <fizzie> GregorR: Lacking the possibility of using >3 HDs.
21:32:16 <fizzie> (Assuming the open-source edition doesn't do the SCSI controllers; I'm not quite sure about that.)
21:32:30 <AnMaster> fizzie, does it do serial and usb forwarding?
21:33:08 <AnMaster> http://www.virtualbox.org/wiki/Editions <-- seems like no
21:33:10 <fizzie> I think USB-passthrough is also only in the closed-source edition. Not that I've ever wanted to have that.
21:33:22 <fizzie> Serial is supported everywhere.
21:33:52 <AnMaster> I have some silly devices that only support windows sadly. Like a GPS unit I need to update the maps on every now and then
21:34:07 <fizzie> That's a good candidate for some reverse-engineering fun.
21:34:14 <fizzie> Oh, and you can use native partitions.
21:34:31 <AnMaster> fizzie, even sillier: the gps unit actually runs linux on arm internally
21:34:40 <GregorR> I was under the impression that USB forwarding had made it into the OSI.
21:34:44 <fizzie> (You have to create a .vmdk file referring to the raw partition.)
21:35:11 <GregorR> AnMaster: VirtualBox can use VMWare files.
21:35:33 <fizzie> AnMaster: Yes, and the raw-partition support is done as a part of the vmdk support.
21:35:39 <AnMaster> well I think I prefer vmware-server then. At least it doesn't pretend to be open source.
21:35:49 <AnMaster> while stripping vital features
21:36:38 <fizzie> I don't prefer it, because last time I tried it, it made an unholy mess of things. I really don't like installers touching my files all around, anyway.
21:37:03 <AnMaster> gentoo has a package for it, that contains the mess
21:37:33 <AnMaster> fizzie, from what I can see it uses sed a lot.
21:37:36 <fizzie> Might be; I guess no Debian-enthusiast has bothered to go through all that trouble.
21:38:11 -!- Hiato has quit ("Leaving.").
21:38:24 <fizzie> There is "vmware-package" which gives a make-vmpkg tool that you can use to build .debs out of vmware workstation/player/server installer files.
21:38:37 <fizzie> I guess that would handle the mess too.
21:38:51 <GregorR> "Again,don't be angry with me if you see this message in SPAM of your mailbox; it will be because I send it with low internet connection."
21:39:00 <fizzie> It wasn't there the last time I looked.
21:39:33 <AnMaster> GregorR, um. what is a "low internet connection" supposed to be...
21:39:51 <GregorR> AnMaster: I don't know, but don't you just wurve how stupid spam is? :P
21:40:10 <AnMaster> GregorR, except I don't know what "wurve" means. *googles*
21:40:28 <AnMaster> No definitions were found for wurve.
21:40:31 <GregorR> I especially love how it then goes on to say "I am the Executive Auditor and Head Of Computing Department in my bank." And yet for some reason, he can only afford "low internet connection" :P
21:40:49 <Deewiant> "wurve" is some kind of cutesy-or-something spelling for "love"
21:40:57 <AnMaster> http://www.urbandictionary.com/define.php?term=wurve
21:41:06 <AnMaster> and google doesn't know it at all
21:41:26 <GregorR> You're trusting Urban Dictionary? :P
21:41:26 <Deewiant> That would make me smarter than Google and Urban Dictionary combined, now wouldn't it
21:41:42 <AnMaster> GregorR, it is usually useful for irc
21:42:00 <GregorR> Well, Deewiant was right :P
21:42:16 <AnMaster> GregorR, aspell doesn't like it either
21:42:23 -!- WangZeDong has joined.
21:42:26 <Deewiant> It's not a real word, it's onomatopoeic.
21:42:37 <AnMaster> nor is it in /usr/share/dict/words
21:49:48 -!- tombom has quit ("Peace and Protection 4.22.2").
21:51:33 -!- Slereah has quit (Read error: 110 (Connection timed out)).
21:51:51 -!- GregorR has set topic: Logarithms: <http://dickensurl.com/1b34/It%E2%80%99s_my_old_girl_that_advises._She_has_the_head._But_I_never_own_to_it_before_her._Discipline_must_be_maintained.> | ~fish~.
21:55:54 <fizzie> I noticed it just recently when twiddling through Linux kernel menuconfig; there's a CONFIG_CRYPTO_CRC32C_INTEL setting that makes use of it.
21:56:07 -!- GregorR has set topic: the international hub for esoteric programming language design and deployment - #esoteric is not associated with the joke language Perl, please visit www.perl.org - logs: http://tunes.org/~nef/logs/esoteric/.
21:56:16 -!- GregorR has set topic: #esoteric, the international hub for esoteric programming language design and deployment - #esoteric is not associated with the joke language Perl, please visit www.perl.org - logs: http://tunes.org/~nef/logs/esoteric/.
21:57:15 <ais523> ah, I always get a warm fuzzy nostalgic feeling when we have "the international hub for esoteric programming language design and deployment" in the topic
21:57:35 <GregorR> I decided it was time to switch it back to my original topic :P
21:57:58 <GregorR> 05.10.10:19:09:35 --- topic: set to '#esoteric, the international hub for esoteric programming language design and deployment - #esoteric is not associated with the joke language Perl, please visit www.perl.org - logs: http://tunes.org/~nef/logs/esoteric/ or http://meme.b9.com/cdates.html?channel=esoteric' by GregorR
22:00:53 <fizzie> I can't seem to find any speed benchmarks between the crc32c and crc32c-intel modules. :/
22:01:39 <GregorR> fizzie: I should hope that somebody did benchmark them, rather than just assuming the latter was faster :P
22:02:25 <Deewiant> I'd hope that it's a fairly safe assumption that it is.
22:03:42 <fizzie> I'd still hope that the original patch-writer would've tested hir implementation at least with some sort of rudimentary speed benchmark; but if that happened, it's not a very well-published result. I might be using the wrong sort of keywords, though.
22:07:07 <fizzie> I could theoretically speaking be motivated enough to try it myself, but I don't have any sort of fancy-pants CPU for that; I guess my second-newest Intel CPU would be... a Pentium-150 in a laptop. (And then I have this Atom box, but it doesn't do anything newer than ssse3 either.)
22:09:39 <Deewiant> Atom doesn't count as Intel? :-P
22:10:39 <fizzie> It does; that's why the P150 is the second-newest.
22:10:56 <Deewiant> Ah. "And then" made it sound like it was a third box.
22:11:13 <fizzie> Although I think I managed to forget a Pentium-M somewhere.
22:12:46 <bsmntbombdood> "crypto" and "crc32" should not be in the same sentence
22:13:53 <ehird> 21:28 AnMaster: ehird, so you recommend to not file any bugs to any projects, instead only complain in some totally unrelated irc channel and not report the issue at all ← not what I said.
22:13:57 <ehird> is that what I said?
22:14:19 <AnMaster> ehird, I extrapolated from your behaviour.
22:14:29 <ehird> That's some sucky extrapolation.
22:14:31 <ais523> bsmntbombdood: you could use crc32 in order to verify that a public-key had transmitted properly
22:14:40 <AnMaster> but yes it seems to match your opinion
22:14:43 <ais523> you don't want to use it for the actual crypto, though
22:14:53 <ais523> it doesn't guarantee that things haven't been tampered with
22:14:55 <ais523> just that they don't contain typos
22:15:02 <ais523> which isn't very useful for crypto
22:15:07 <fizzie> It was a bit unclear why crc32 is in the "cryptoapi" part; the help says it's used in iSCSI headers and things like that.
22:16:33 <fizzie> Yes, there's still the "library routines" crc32 thing too; but the intel-accelerated patch was for a cryptoapi digest.
22:16:42 <fizzie> I guess they just liked the cryptoapi digest-calculating API more.
22:17:12 <fizzie> It is not immediately obvious what would be a more suitable way of exporting that sort of stuff to userspace.
22:17:58 <fizzie> (Or can you use "raw" cryptoapi quasi-directly from a userspace application? I have no clue. The dm-crypto thing uses it, but that's inside the kernul.)
22:19:25 <ehird> Let's have a change. /me tries "sudo pacman -S kde".
22:19:47 <ehird> I wonder if this package manager is good enough to do "remove this and all its dependencies apart from vital shit, 'kay?" to wipe out ze kde.
22:20:11 <ehird> Or, as apt puts it: remove packages that aren't depended on.
22:20:40 <ais523> "remove X and all its dependencies" is what Windows uninstallers try to do, which is why they're so broken
22:20:48 <ais523> "remove X, then remove all unused dependencies" is so much saner
22:21:04 <ehird> I was just describing the general operation
22:21:14 <ehird> "rid me of this foul beast and all that it sides with! apart from my chums."
22:21:26 -!- puzzlet_ has joined.
22:21:44 <ehird> Coo, it's downloading gcc.
22:21:50 <ehird> ^C let's just install the build-essential equivalent.
22:21:54 <ehird> ... also, why does kde depend on gcc?
22:22:07 <ehird> Anyway, sudo pacman -S core-devel was it? or dev. or base.
22:22:11 <ehird> Deewiant: do you know?
22:23:16 <ehird> Deewiant: would there happen to be a way to streamline its output and not make it ask me whether to install it twice ;-)
22:23:44 <ehird> "INSTALL THIS SHIT?" "OKAY, HERE ARE THE DEPENDENCIES. REALLY INSTALL THIS SHIT?"
22:23:51 <ehird> ais523: "Do you want to erase your system for this package? [Yn]"
22:24:17 <Deewiant> --noconfirm do not ask for any confirmation
22:24:23 <ehird> Deewiant: That's exactly what I don't want
22:24:29 <ehird> I just don't want it to ask me the same question twice dammit :P
22:24:42 <ehird> you can run vista without uac
22:25:10 <Deewiant> ehird: It only asks you twice for groups, btw.
22:25:15 <Deewiant> Which you won't be installing often.
22:25:17 <ais523> ehird: but that's insecure!
22:25:23 <ehird> It has been my misfortune to mainly install groups ;-)
22:25:30 <ehird> ais523: just run as a non-admin?
22:25:41 <ais523> that's how the Vista computers over here are set up
22:25:54 <ais523> they're as secure as I can get Vista without much Windows knowledge
22:25:58 <ehird> you set up a windows machine willingly?
22:26:00 <ais523> and with only Norton 360 as the antivirus
22:26:15 <ehird> turn into a christian and repent
22:26:18 <ais523> with any luck they haven't been hacked into yet
22:26:23 <ehird> you can stop being a christian after that, it's ok
22:29:30 <pikhq> I think I'd rather run DOS than Vista.
22:29:58 <pikhq> (at least DOS does what you ask it and gets the fuck out of your way)
22:30:08 <ais523> nowadays there's FreeDOS
22:30:25 <ais523> whose main advantage is that it's crazy fast
22:30:27 <pikhq> Well, yeah. I of course was thinking of FreeDOS.
22:30:30 <ehird> bsmntbombdood: good god, have some patience
22:30:34 <ehird> do you really need it in 5 minutes? :P
22:30:42 <pikhq> It's a damned good DOS.
22:30:43 <ehird> I've waited 3 weeks for things to be shipped
22:30:54 <ais523> Dell actually offer it on new computers as an OS choice along with Ubuntu or Windows
22:30:58 <ehird> bsmntbombdood: >_<
22:31:14 <ehird> bsmntbombdood: so you get everything the day after you order it?
22:31:55 <ehird> ehh, bug their support team?
22:32:22 <bsmntbombdood> Status: Submitted - Your order has been successfully submitted. Your credit/debit card has not yet been charged. Please allow 1-2 business days for your order to process and ship.
22:32:46 <ehird> you americans are spoilt rotten
22:32:54 <ehird> 1-2 business days is luxury over here!
22:33:17 <Deewiant> Is 1-2 business days a long time for an online order? O_o
22:33:54 -!- puzzlet has quit (Read error: 110 (Connection timed out)).
22:34:41 <ehird> bsmntbombdood: nope
22:34:42 <Deewiant> Like ehird said, it's luxury :-P
22:34:54 <Deewiant> 1-2 weeks is my default expectation.
22:34:54 <ehird> an online order usually takes about a week from ordering to getting it
22:35:10 <pikhq> The only time I expect it to ship next day is when the web site says so...
22:35:17 <ehird> bsmntbombdood: where do you order shit where you get it a day after ordering
22:35:30 <AnMaster> <ehird> Let's have a change. /me tries "sudo pacman -S kde". <-- kdemod. Read on arch wiki
22:35:48 <ehird> AnMaster: explain in 5 words, I don't wanna cancel my download :P
22:35:59 <ehird> AnMaster: ah. tweaked.
22:36:01 <pikhq> Which generally means it's one of the few bits of inventory that they've got right next to the boxing guys, and they just need to stick a label on the box...
22:36:07 <ehird> I'm trying arch to escape from modified upstreams, y'know
22:36:36 <ehird> bsmntbombdood: wtf is up with you :D
22:36:42 <ehird> you first said about wanting a new pc a few months ago
22:36:47 <ehird> and now 1-2 business days is terrible?
22:37:05 <Deewiant> ehird: Modularized version of KDE from a third-party Arch package repo, or something like that
22:37:15 <ehird> Upstream modification, boo
22:37:26 <Deewiant> IME that means that it works better. :-P
22:37:38 <ehird> It's not like I want to use KDE in practic
22:37:40 <pikhq> Not all such modification is bad.
22:37:47 <ehird> It's just that the last 3458973489573495345 distros I've used use gnome :P
22:37:49 <ehird> bsmntbombdood: why?
22:37:59 <pikhq> Some packages need such modification to freaking *build*.
22:38:05 <ehird> pikhq: That's fine.
22:38:08 <Deewiant> Most KDE programs I've tried still crash for me, but with kdemod I at least managed to run some things. :-P
22:38:11 <ehird> I just don't like changes beyond "make it work properly"
22:38:16 -!- BeholdMyGlory has quit (Remote closed the connection).
22:38:34 <pikhq> Ah. Well, then, you'd like Gentoo...
22:38:42 <ehird> pikhq: Noooooooooooo way.
22:38:51 <pikhq> The only package that does more than "make it work properly" is gentoo-sources.
22:38:53 <ehird> Fuck source distros, fuck Gentoo's community, fuck Gentoo's relation with upstreams. :P
22:39:03 <AnMaster> <ais523> nowadays there's FreeDOS <ais523> whose main advantage is that it's crazy fast <-- stop making me want to port cfunge... AUGH
22:39:15 <ehird> AnMaster: no posix :-)
22:39:20 <ehird> OTOH, it'd work on Windows then
22:39:22 <GregorR> ehird: There is (nearly) with DJGPP
22:39:22 <pikhq> Gentoo has an excellent relationship with most of its upstreams; exceptions being, well, dicks.
22:39:35 <ehird> GregorR: Not exactly nearly.
22:39:40 <GregorR> DJGPP supports a bunch of POSIXy stuff. No, not mmap.
22:39:41 <ehird> pikhq: I haven't seen that.
22:39:41 <Deewiant> AnMaster: 2009-05-08 00:36:11 ( AnMaster) I'm going to bed
22:39:47 <ehird> Deewiant: he does that all the time
22:39:48 <GregorR> DJGPP supports what can be supported without killing yourself ;)
22:39:50 <ehird> he takes a few hours
22:39:54 <pikhq> And Gentoo's community is mostly smart people with a few dumbasses.
22:39:57 <ehird> GregorR: AnMaster uses those things.
22:40:11 <GregorR> I mean supported /on DOS/ without killing yourself, of course.
22:40:15 <ehird> pikhq: like hell it is, I've never seen more people think they're cool because they can run a compiler and turn on optimization flags
22:40:18 <pikhq> (you can tell whether they're smart depending on their opinion on funroll-loops.info)
22:40:34 <ehird> GregorR: Just saying that it wouldn't work for cfunge.
22:40:38 <AnMaster> ehird, not representative. I have met lots of gentoo users who aren't like that.
22:40:52 <AnMaster> in fact I met very few gentoo users like that
22:40:53 <ehird> Do they hang out in #the-secret-gentoo-channel-wihtout-all-the-idiots?
22:41:03 <ehird> Because I've never, ever seen them.
22:41:13 <ehird> 'part from a few in here.
22:41:15 <pikhq> And you're saying "fuck source distros" while using arch.
22:41:17 <AnMaster> ehird, oh main #gentoo? Well the sub-channels tend to be saner
22:41:24 <ehird> pikhq: Arch... is a binary distro...
22:41:29 <ais523> ehird: settled on a distro yet?
22:41:37 <ehird> ais523: nope; thank god for cheap VM creation
22:41:37 <AnMaster> though it has makepkg to build binary packages
22:42:00 <ehird> ais523: probably Ubuntu is what I'd use atm, for the "goddamn, just work" value.
22:42:06 <GregorR> On the twelfth day, the LORD created cheap VM creation. And it was good.
22:42:24 <pikhq> ehird: That's what I use Debian for, myself...
22:42:50 <pikhq> Or maybe Slackware.
22:42:50 <ehird> pikhq: Debian's default gnome isn't quite as polished.
22:43:07 <ehird> Also, installing and using Slackware is not much of a "hey, I just clicked this and it did it all for me, life is good" experience.
22:43:42 <pikhq> Maybe I was just insane, but when I used Slackware, it kinda... Just Worked (tm).
22:43:59 <pikhq> (probably just insane; Slackware was my first distro)
22:44:45 <pikhq> Actually, Slackware 8.
22:44:52 <ehird> Hehe, slackware started because SLS switched to ELF...
22:45:07 -!- M0ny has quit ("When you get sad stop being sad and be awesome instead.").
22:45:18 <ehird> bsmntbombdood: not imo
22:45:24 <pikhq> Funny, I thought Slackware started because SLS *wouldn't* switch to ELF...
22:45:28 <ehird> it's a bit of a pain to configure but after that it's pretty smooth sailing
22:45:37 <ehird> kde is awful indeed
22:45:42 <ehird> but I like a bit of glue between my components
22:46:45 <ehird> bsmntbombdood: out of curiosity, what wm do you use?
22:47:14 <ehird> i consider tiling wms unergonomical :)
22:48:41 <fizzie> Slackware 3.2 "broken edition" was my first distro.
22:49:13 <pikhq> bsmntbombdood: Why, hello, fellow Tiling WM user.
22:49:22 <ehird> fizzie: my friend's got you all beat
22:49:30 <pikhq> (shame you use one by a dick, though)
22:49:57 <fizzie> I do Awesome3 here; it's a tiling wm too.
22:50:22 <pikhq> bsmntbombdood, he also is distributing Ion3 illegally. :p
22:51:01 <ehird> bsmntbombdood: licensing issues
22:51:07 <pikhq> (at least, he is if there is a single bit of code from Ion3 before the non-free change not written by Tuomo)
22:51:14 <ehird> tuomov's opinion on them seems to be "I CAN DO WHAT THE FUCK I WANT, YEAAAAH!"
22:51:57 <ehird> does the latest ion distro even have a LICENSE file?
22:52:24 <ehird> it's extended LGPL with bullshit terms
22:52:29 <ehird> - Versions not based on the copyright holder's latest release (on
22:52:29 <ehird> the corresponding "branch", such as Ion3(tm)), must within 28 days
22:52:30 <ehird> of this release, be prominently marked as (potentially) obsolete
22:52:34 <ehird> HAHA he actually added that
22:52:38 <ehird> due to debian having a few month old version
22:52:44 <ehird> - Significantly altered versions may be provided only if the user
22:52:44 <ehird> explicitly requests for those modifications to be applied, and
22:52:45 <ehird> is prominently notified that the software is no longer considered
22:52:48 <ehird> the standard version, and is not supported by the copyright holder.
22:52:49 <ehird> The version string displayed by the program must describe these
22:52:51 <ehird> modifications and the "support void" status.
22:52:53 <ehird> - A version that does not significantly differ from one of the
22:52:55 <ehird> copyright holder's releases, must be provided by default.
22:52:57 <ehird> i just quoted all 3 clauses out of order
22:53:00 <ehird> because they're all absolute gold
22:53:10 <ais523> I would so love it if Debian somehow put on a script which took the latest version and patched it back to an old version
22:53:25 <ehird> - A version that does not significantly differ from one of the
22:53:25 <ehird> copyright holder's releases, must be provided by default.
22:53:33 <ehird> tuomov is such a retard
22:53:48 <ais523> the third term there is a weird wording of one DFSG-approved term
22:53:58 <fizzie> That "significantly altered" refers (among other things, I guess) the incident where Arch's non-official "user repository" had a script that fetched Ion and installed the xft "blurry-fonts patch", yet called the result "ion".
22:54:01 <ais523> which is "all modified versions must be provided in the form of a diff from the original version"
22:54:02 <GregorR> When did Slackware switch to ELF? :P
22:54:10 <ehird> ais523: provided to the user
22:54:11 <ais523> Debian don't like that rule, but don't think it makes software nonfree
22:54:19 <ehird> apt-get install ion
22:54:22 <ehird> they must get a stock one
22:54:27 <ehird> apt-get install ion-megadebian
22:54:28 <pikhq> GregorR: Probably about when they switched to libc 6.
22:54:34 <ehird> and which then must display prominently
22:54:39 <ehird> WE SUCK DICKS AND YOU CANNOT GET SUPPORT FOR THIS VERSION <3
22:54:57 <pikhq> The thing I really "like" about all this is that he thinks that he can repeal the previous LGPL versions.
22:55:04 <ehird> basically, tuomov has no fucking idea what foss is and demands that packagers work like crack addicts to keep up to date with his shitty updates :-D
22:55:29 <ehird> "Significant change: Bug fixes are insignificant as additions."
22:55:44 <ehird> Debian should consider their changes to Ion to be a bugfix for tuomov's idiocy ;-)
22:56:05 <pikhq> I'm surprising Ion's not been forked by now.
22:56:07 <ais523> it wouldn't be the first time...
22:56:18 <ehird> ais523: well, you couldn't call it eion
22:56:22 <ehird> you'd have to completely change the name
22:56:30 -!- MizardX has quit (Read error: 54 (Connection reset by peer)).
22:56:32 <pikhq> eglibc itself isn't the first time.
22:56:34 <ehird> although debian could redirect ion3 to opm
22:56:36 -!- MizardX has joined.
22:57:07 <pikhq> bsmntbombdood: Ratpoison.
22:57:15 <pikhq> Mean to switch to StumpWM one of these days.
22:57:28 <fizzie> Slackware went from libc5 into glibc when they jumped versions from 4 -> 7 because redhat and friends were already at number 6. I don't remember the ELF thing; I think there was some sort of ELF support in 3.x already.
22:57:47 <ehird> I once drafted up in my head what would be needed to make a tiling wm usable (mainly manipulated w/ mouse, fitts law compliant etc) and it basically turned into a shit floating wm
23:00:20 <ehird> the clue's in the name
23:00:52 <pikhq> I use the mouse for Flash.
23:02:13 <ehird> (It's not completely unheard of, though - GNU emacs used to greet Symbolics users with the message "In doing business with Symbolics, you are rewarding a wrong.")
23:02:23 <ehird> Poor rms and his Symbolics hate.
23:02:32 <ehird> Rewarding a wrong!
23:02:49 -!- bsmntbombdood has quit (Read error: 104 (Connection reset by peer)).
23:02:53 <pikhq> At least Emacs didn't have an invariant clause. :p
23:02:53 <ais523> <perl6 doc> Conjecture: other such patterns may be recognized in the future, depending on which unrealistic benchmarks we want to run faster.
23:03:11 <ehird> Okay, arch installed kde
23:03:22 <ais523> the Emacs manual used to have a lengthy rant about why it didn't run on Windows
23:03:26 -!- bsmntbombdood has joined.
23:03:28 <ais523> although that's been fixed now too
23:04:07 <ehird> http://edward.oconnor.cx/2005/04/rms
23:04:42 <ais523> wow, assignment in Perl6 is as lazy as possible without breaking imperative semantics
23:04:59 <ais523> you can write @array = 0..*;
23:05:01 <ehird> Perl 6 is just shit.
23:05:06 <ais523> and get an infinitely long array
23:05:13 <ais523> Haskell doing that makes sense, seeing it in Perl is pretty mindblowing though
23:06:26 <ais523> it's also the only language I've seen with todo operators
23:06:40 <ais523> most people just put todos in comments...
23:06:59 <ehird> OK, KDE installed on Arch and KDM set.
23:07:16 <ehird> Time to reboot and check it out.
23:07:28 <ehird> At least, nothing so far has happened that's any worse than Debian, so.
23:07:38 * ehird sudo shutdown -r now
23:07:41 <ais523> are you planning to uninstall KDE again as a test?
23:08:09 -!- oerjan has joined.
23:08:18 -!- ais523 has quit (Remote closed the connection).
23:08:26 <ehird> It's asking me for a login.
23:08:56 <oerjan> !perl print "What do you mean, not associated?";
23:08:57 <EgoBot> What do you mean, not associated?
23:14:58 <ehird> Deewiant: I have KDE, I just don't have X11. Wtf.
23:14:58 <oerjan> <Deewiant> If I get an additional crash due to not using ECC memory, I really won't notice.
23:15:46 <oerjan> i do recall reading about a Java exploit that could use a large fraction of memory bit errors to break security
23:18:59 <oerjan> by arranging for a reference to point to an object of the wrong class
23:19:46 <oerjan> this could be an issue for any language based on strong static typing, i guess
23:20:42 <oerjan> well, probably dynamic too
23:25:48 <ehird> holy unantialiased text and no mouse batman
23:25:59 <ehird> holy no keyboard batman
23:26:14 <ehird> holy Deewiant how do you uninstall a package and remove unused dependencies batman
23:37:23 <ehird> I'll bet fizzie has
23:39:54 <fizzie> I might have tried it, but not really used.
23:40:15 <ehird> fizzie: isn't awesome based on dwm
23:40:26 <fizzie> It seems a bit like evilwm, except tiling. (Evilwm also does the whole "no title bar, no configuration files, not much of anything" thing.)
23:40:39 <ehird> I'm sure awesome is based on dwm
23:40:42 <fizzie> I don't know the genealogy.
23:40:52 <ehird> fizzie: awesoem does floating windows, right?
23:40:54 <ehird> are they a pain to use?
23:41:08 <pikhq> ehird: Well, you might just have libX11. Technically, you don't need an X server. ;p
23:41:14 <ehird> pikhq: indeed that was the case
23:41:21 <fizzie> I'm not quite sure what you mean by "pain". They float, that's about it.
23:41:28 <ehird> bsmntbombdood: are you actually running into troubles with your current one or just impatience?
23:41:35 <oerjan> 11:57:37 <ehird> a collection of quarks
23:41:38 <ehird> fizzie: i mean, what if I basically just wanted to use floating windows
23:41:43 <ehird> would that mesh with awesome? :p
23:41:49 <oerjan> what did the poor leptons do to you since you ignore them?
23:41:58 <ehird> oerjan: sry leptons!
23:42:20 <ehird> bsmntbombdood: Making noise? Hate to break it to ya but that i7 rig isn't going to be _quiet_...
23:42:33 <fizzie> I... guess you could. One of the possible window layouts is "everything floating". I'm not sure how nice it is to use in a window-managmenty sense; my needs are so very simple.
23:43:38 <ehird> bsmntbombdood: anyway, calm down or you'll forget t he spacers :-P
23:43:39 <oerjan> most people don't want their fans to die
23:44:03 <ehird> Please tell me you know what I meant.
23:44:39 <ehird> bsmntbombdood: the spacers are the screws that come with your case that you attach so your mobo doesn't touch the case
23:45:04 <ehird> i mean when you get it :
23:45:10 <ehird> it was just a joke since i've known one or two cases where people have fried mobos by not putting the spacers on
23:45:15 <ehird> one guy did it _thrice_
23:47:13 <ehird> bsmntbombdood: well that'd explain why it took "so long"
23:47:16 <ehird> what's their problem
23:49:35 <bsmntbombdood> now it prompts me to change it but won't let me...
23:49:52 <ehird> bsmntbombdood: wat?
23:55:23 <ehird> pikhq: what do you think tuomov would do if I took his code and blatantly violated the license :-)