00:01:54 -!- jix has quit ("This computer has gone to sleep").
00:07:12 -!- calamari has quit (Read error: 110 (Connection timed out)).
00:29:52 -!- tusho has quit.
01:59:58 -!- Sgeo has quit (Connection timed out).
03:03:48 -!- Corun has quit ("This computer has gone to sleep").
03:45:19 -!- comex has quit (Read error: 113 (No route to host)).
03:48:11 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | it can..
04:24:38 -!- CO2Games has joined.
04:34:39 -!- GreaseMonkey has joined.
04:36:58 -!- CO2Games has quit ("And I invented doors, no joke!").
04:39:39 -!- ihope has quit (Read error: 110 (Connection timed out)).
04:48:11 -!- oklofok has joined.
07:11:53 -!- oklopol has joined.
07:12:07 -!- oklofok has quit (Read error: 113 (No route to host)).
07:13:15 <oklopol> i somewhat proved my original idea wouldn't work, but devised another scheme that seems more promising
07:23:38 <oklopol> http://www.vjn.fi/pb/p466532445.txt
07:24:11 <oklopol> the dictionary of references is just the normal references an object has to other objects
07:24:40 <oklopol> backrefs are just the other directions of references, because we need to be able to follow them backwards
07:24:41 -!- kar8nga has joined.
07:25:07 <oklopol> rootrefcount is what (i think) makes this work
07:25:33 <oklopol> root refs are references directly from a variable on stack
07:26:14 <oklopol> only when they reach zero will we have to do any checking for whether there are paths to the object
07:26:49 <oklopol> and the idea is, we mostly reference objects directly from local variables, so we don't need to do these searches that often
07:27:07 <oklopol> then again, i might be wrong, but i think this one would at least work.
07:43:41 <fizzie> Just intuitively speaking, "search through backrefs to find a path to root" sounds like it would take a while, given that it can (apparently?) be an arbitrary, cyclic graph.
07:45:36 -!- KingOfKarlsruhe has joined.
07:46:48 <oklopol> fizzie: it could definitely take a while
07:47:31 -!- KingOfKarlsruhe has quit (Remote closed the connection).
07:50:24 <oklopol> for simple cases it will only run the search a few times, and that's all i care about :P
07:50:37 <oklopol> i have more useful ideas regarding typed structures
07:55:11 <fizzie> Not right now, too early for thinking. I guess it sort-of sounds like it would not leak.
07:55:37 <oklopol> yes, that's very probably, it's just currently it's a bit useless.
07:59:39 <oklopol> i have a demonstration in 15 minutes
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:03:24 -!- kar8nga has left (?).
09:48:11 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | it's pretty obscure as these things go..
09:50:43 -!- oerjan has joined.
11:14:19 -!- oerjan has quit ("leaving").
11:23:47 -!- GreaseMonkey has quit ("So, how much do you love noodles?").
11:38:03 -!- kar8nga has joined.
11:44:48 -!- ihope has joined.
12:41:36 -!- tusho has joined.
13:04:46 -!- Slereah has quit (Read error: 113 (No route to host)).
14:00:03 -!- comex has joined.
14:41:57 -!- Slereah has joined.
14:45:15 -!- Sgeo has joined.
14:48:47 -!- Slereah has quit (Read error: 113 (No route to host)).
14:48:54 -!- Slereah has joined.
15:04:00 -!- Corun has joined.
15:22:09 <GregorR> If I had tons of money I would air a commercial that was just a black screen with the white text "You just lost the game."
15:29:17 <Slereah> Can Mathematica use the +/- operator?
15:29:34 <Slereah> It's in the character set, but it never tries to solve the problem I submit with it
15:32:42 <Slereah> Plusminus is apparently in the "symbol without built-in meaning" category.
15:33:56 <tusho> ±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±±
15:37:02 <Slereah> I'm looking at a shitload of cases here.
15:37:22 <tusho> Slereah: implement ± yourself?
15:37:24 <GregorR> Just solve in 2^(number of instances of ±) cases?
15:37:24 <Slereah> Error computation mean that I'll have to search for the worst case scenario.
15:37:43 <Slereah> Yeah, that's what I wanted to do
15:37:48 <Slereah> But first, I have to learn Mathematica's syntax
15:37:54 <Slereah> Usually I just use it for problem solving
15:39:19 <Slereah> I have that so far by hand :
15:39:20 <Slereah> a^3 \[PlusMinus] 3 a^2 \[CapitalDelta]a +
15:39:20 <Slereah> 3 a \[CapitalDelta]a^2 \[PlusMinus] \[CapitalDelta]a^3 = (
15:39:20 <Slereah> P^2 \[PlusMinus] 2 P \[CapitalDelta]P + \[CapitalDelta]P^2)/(
15:39:20 <Slereah> 2 \[Pi] \[CapitalDelta]\[Pi] + \[CapitalDelta]\[Pi]^2)) *(G \
15:39:22 <Slereah> \[PlusMinus] \[CapitalDelta]G) ((M \[PlusMinus] \[CapitalDelta]M) + \
15:39:24 <Slereah> (m \[PlusMinus] \[CapitalDelta]m))
15:39:46 <Slereah> Of course, Mathematica translates poorly on IRC
15:48:12 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | yeah.. not quite sure what I was thinking with that.
16:00:22 -!- comex has quit (Remote closed the connection).
16:08:10 <optbot> AnMaster: it does pass on true linux and true freebsd I know
16:08:51 <AnMaster> wonder what the context of that was
16:14:35 -!- Slereah has quit (Read error: 113 (No route to host)).
16:25:34 -!- Slereah has joined.
16:57:31 -!- olsner has joined.
17:01:43 -!- tusho has quit (Remote closed the connection).
17:02:00 -!- tusho has joined.
17:03:38 -!- Hiato has joined.
17:20:07 -!- Slereah has quit (Read error: 54 (Connection reset by peer)).
17:26:47 -!- Slereah has joined.
17:29:45 -!- ais523 has joined.
17:34:48 -!- Slereah has quit (Read error: 113 (No route to host)).
17:38:04 -!- Hiato has quit ("Leaving.").
17:44:13 -!- Mony has joined.
17:48:02 -!- Slereah has joined.
17:48:17 <Slereah> I surrender to you, mighty moon men!
17:48:46 * GregorR sucks out Slereah's brain with a straw.
17:51:02 -!- ais523_ has joined.
17:51:08 <Slereah> Mathematica has way too much stuff
17:51:13 <Slereah> It be hard to find something
17:51:41 <ais523_> actually, half the time it isn't there
17:51:42 <GregorR> Tragically, Slereah spent so long looking for a function in Mathematica that in the process he forgot how to speak English.
17:51:45 <ais523_> one big problem with Mathematica is that it seems to like a huge number of single-purpose functions
17:51:47 <ais523_> which can be chained into one big inefficient function
17:52:10 <ais523_> but really, if you want to do something the Wolfram people didn't think of, the code ends up monstrous and a couple of computational orders slower than it should be
17:52:21 -!- ais523 has quit (Nick collision from services.).
17:52:30 -!- ais523_ has changed nick to ais523.
17:53:14 <Slereah> But I'm not looking at something hard
17:53:33 <Slereah> Like it has no defined PlusMinus function.
17:53:53 <GregorR> programHalts(program, input)
17:53:54 <Slereah> And the definition proposed in the help file isn't very helpful
17:55:46 <Slereah> Right now, I'm hoping for something useful to plot orbits.
17:56:15 <Slereah> Like something where you can make evolve points with t, instead of just the whole orbit
17:56:31 <ais523> Slereah: unfortunately Wolfram seem to think half their functions are general when they aren't
17:56:45 -!- kar8nga has left (?).
17:58:15 <tusho> who wants to explain to me why i'd want a pgp key for anything that isn't email
17:59:19 <Slereah> It's weird. Most articles don't contain the time dependance of orbits.
17:59:27 <Slereah> They just give you the whole orbit.
17:59:32 <Slereah> Fuck it, I'll just compute it.
18:01:22 <Slereah> It's been a while since orbital mechanics.
18:12:06 <ais523> tusho: other cases where you need to confirm your identity?
18:12:12 <ais523> what if you were a DNS server, for instance?
18:12:42 <tusho> although that just confirms i own tusho.net
18:12:45 <tusho> and is also http bound
18:13:03 <tusho> microid, possibly?
18:13:21 <tusho> pgp has the advantage of being older ofc
18:14:06 <ais523> nowadays people use OpenPGP instead to avoid paying licensing
18:14:11 <ais523> and OpenPGP isn't all that old
18:14:42 <tusho> yes, but the actual key system...
18:34:06 -!- Corun_ has joined.
18:34:42 -!- Corun has quit (Read error: 104 (Connection reset by peer)).
18:34:50 -!- Corun_ has changed nick to Corun.
18:48:49 -!- oerjan has joined.
18:55:31 -!- Corun has quit (Remote closed the connection).
18:55:43 -!- Corun has joined.
19:21:22 -!- Slereah has quit (Read error: 60 (Operation timed out)).
19:23:46 -!- Corun_ has joined.
19:25:51 -!- Corun has quit (Read error: 60 (Operation timed out)).
19:31:00 <AnMaster> as for mathematica, I'm sure maxima can't do as much, but maxima is 1) free 2) works fairly well for a lot of problems, and have decent speed
19:31:36 -!- kar8nga has joined.
19:31:46 <AnMaster> tusho, you could use gpg to sign tarballs you release
19:31:54 <tusho> AnMaster: like that ever happens :D
19:31:55 <AnMaster> that is about the only reason apart from email I can think of
19:32:07 <AnMaster> tusho, well if you don't do that then I guess that is not a reason for you then
19:32:17 <tusho> i was joking about my not ever doing anything.
19:32:20 <tusho> oh wait, that's not a joke.
19:32:33 <AnMaster> 11:29:20 up 16 days, 23:38, 3 users, load average: 429.31, 219.79, 596.82 <-- hm... I don't like that on a server...
19:32:47 <tusho> 18:32:41 up 26 days, 21:49, 1 user, load average: 0.15, 0.03, 0.01
19:32:55 <tusho> wonder why we rebooted last?
19:32:56 -!- Corun__ has joined.
19:32:58 <AnMaster> (so about nothing apart from reporting it that I can do)
19:33:32 <ais523> not surprising as I turn it off when I'm not using it
19:33:40 <AnMaster> 20:33:32 up 31 days, 7:50, 32 users, load average: 0.37, 0.35, 0.22
19:33:51 <AnMaster> but I was talking about the horrible load average
19:34:12 <AnMaster> tell me if http://supertux.lethargik.org/wiki/Main_Page loads at all, and if it does how long does it take
19:34:32 <AnMaster> (that is the server with the horrible load average, and it is shared hosting so I can't do anything about it really)
19:34:59 <AnMaster> 11:34:45 up 16 days, 23:43, 3 users, load average: 651.20, 482.98, 247.35 <-- ugh
19:35:19 <tusho> 7:35pm up 9 days 20:54, 0 users, load average: 0.15, 0.14, 0.16
19:35:29 <AnMaster> it takes about 10 seconds from hitting enter to the output currently
19:35:30 <tusho> i rebooted for an efi upgrade, iirc
19:35:43 <tusho> i bet optbot is still hogging memory
19:35:43 <optbot> tusho: what i have is basically a haskell-like syntax, from the standpoint of minimal keywords and you just say foo x y = .. to define a function, and give it a haskell like type signature
19:36:13 <ais523> tusho: is that a reply to optbot, or something up on rutian?
19:36:14 <tusho> PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
19:36:14 <optbot> ais523: !bf +++[>++++++++++++<-]>.
19:36:14 <tusho> 15752 root 16 0 14588 1296 1180 S 9999 0.5 0:00.00 ApplicationPool
19:36:14 <tusho> 15753 www-data 16 0 345m 4508 1748 S 9999 1.7 0:00.00 apache2
19:36:14 <tusho> 15755 www-data 24 0 345m 4560 1776 S 9999 1.7 0:00.00 apache2
19:36:15 <tusho> 1919 mysql 16 0 222m 7044 3008 S 9999 2.7 0:04.41 mysqld
19:36:15 <tusho> 2450 mysql 16 0 222m 7044 3008 S 516 2.7 0:17.56 mysqld
19:36:16 <tusho> 1933 mysql 16 0 222m 7044 3008 S 493 2.7 0:16.79 mysqld
19:36:18 <tusho> 14922 mysql 16 0 222m 7044 3008 S 355 2.7 0:12.09 mysqld
19:36:20 <tusho> 14535 mysql 15 0 222m 7044 3008 S 310 2.7 0:10.56 mysqld
19:36:21 <oerjan> AnMaster: with that domain name what do you expect? :D
19:36:22 <tusho> 1930 mysql 16 0 222m 7044 3008 S 153 2.7 0:05.22 mysqld
19:36:24 <tusho> 1928 mysql 16 0 222m 7044 3008 S 50 2.7 0:01.69 mysqld
19:36:26 <tusho> 1927 mysql 16 0 222m 7044 3008 S 44 2.7 0:01.51 mysqld
19:36:29 <tusho> it sorts itself out after a second though...
19:36:32 <tusho> so i guess a top glitch
19:36:50 <AnMaster> oerjan, true. Though I didn't pay for it, another developer on the supertux project pays for hosting
19:37:07 <AnMaster> but better than sf.net or berlios.de at least
19:37:20 <tusho> dreamhost are pretty good as far as shared hosting go
19:37:31 <tusho> they used to oversell but they seemt o have stopped that now
19:37:36 <AnMaster> tusho, hm... as far as I can tell it is swap trashing
19:37:45 <tusho> someone misbehaving perhaps
19:37:49 <AnMaster> over 1 GB swap filled now, was just 500 MB a while ago
19:38:00 <AnMaster> tusho, can't say, ps aux only show own processes
19:38:04 <tusho> 9480 tusho 16 0 172m 30m 980 S 0 11.8 0:00.86 ruby
19:38:05 <tusho> 15880 www-data 16 0 134m 16m 5668 S 0 6.5 1:02.10 php-cgi
19:38:05 <tusho> 15817 www-data 17 0 106m 9620 5004 S 0 3.7 1:08.50 php-cgi
19:38:05 <tusho> 1919 mysql 16 0 222m 7044 3008 S 0 2.7 0:04.41 mysqld
19:38:11 <tusho> followed by tons of mysqls at 2.7%
19:38:17 <tusho> the 11.8% ram usage for ruby is optbot
19:38:17 <optbot> tusho: of course to really fry your brain you would want the set of dimensions itself to be an infinite-dimensional space
19:38:20 <tusho> as it loads all the logs into memory
19:38:24 <AnMaster> tusho, they blocked ps aux so you can only see yourself
19:38:41 <AnMaster> Linux millhouse 2.4.32-grsec+f6b+gr217+nfs+a32+fuse23+tg+++opt+c8+gr2b-v6.194 #1 SMP Tue Jun 6 15:52:09 PDT 2006 i686 GNU/Linux
19:39:34 <tusho> Millhouse is not a server.
19:39:35 <AnMaster> tusho, it is swap trashing and cpu trashing, since the swap trash ratio is rather moderate but the system is very very slow
19:39:43 <AnMaster> tusho, it is the one I'm sshed to
19:39:51 <tusho> Millhouse is NOT a server.
19:40:38 <AnMaster> oh god.... please please stop trolling, no one think you are funny
19:41:35 <tusho> I like to imagine AnMaster begging on his knees. "Oh god... please, please, PLEASE stop trolling..."
19:41:47 <AnMaster> tusho, no it wasn't begging on my knees
19:41:50 -!- Corun_ has quit (Connection timed out).
19:42:06 <AnMaster> it was a "sigh, soon time to ignore again"
19:43:44 <AnMaster> btw how many of you know of the ssh abort sequence?
19:44:07 <ais523> now everyone paying attention does
19:44:20 <AnMaster> ais523, certainly and they can claim they knew it all along
19:53:52 -!- Corun__ has changed nick to Corun.
20:03:36 -!- oklofok has joined.
20:04:24 -!- oklopol has quit (Read error: 113 (No route to host)).
20:10:37 -!- Slereah has joined.
20:13:14 <Slereah> http://membres.lycos.fr/bewulf/Russell/balls.jpg
20:13:20 <Slereah> What the fuck is this giant ball
20:13:32 <Slereah> Isn't that shit a triple of numbers?
20:17:29 <oerjan> i don't know but maybe it's a syntax problem? try liberally adding parentheses, including around the whole triple
20:17:43 * oerjan doesn't know mathematica though
20:18:04 <Slereah> Here's the expression : Sphere[{xcharyb[\[Epsilon], \[Alpha], t],
20:18:04 <Slereah> ycharyb[\[Epsilon], \[Alpha], t], 0}, .5]
20:18:30 <Slereah> I know it can work, I stole it from somewhere else
20:18:41 <Slereah> I just tweaked some stuff to make it fit the problem
20:19:25 <oerjan> maybe something is not being calculated all the way to a number?
20:19:48 <Slereah> That's my main problem with Mathematica
20:20:03 <Slereah> Once in a while, it simply refuses to compute a value while plotting something
20:20:10 <Slereah> And I never have any idea why
20:20:26 <oerjan> so it works when you calculate the coordinates separately?
20:22:17 -!- psygnisfive has joined.
20:22:29 <Slereah> Ah, I think I found the problem
20:22:53 <Slereah> xcharyb is defined as x[\[Epsilon], aua[80, Sdef], \[Alpha], t]
20:23:03 <Slereah> And when I compute that with some values
20:23:15 <Slereah> I get fucking 1.57557*10^-7 a[80, Sdef]
20:23:32 <Slereah> It can't compute the function a?
20:24:13 <Slereah> I apparently forgot some function-validating
20:25:20 <Slereah> Well, the plotting still doesn't work, but we're getting somewhere
20:26:51 <Slereah> How come Mathematica can display its character set, but not the errors?
20:26:58 <psygnisfive> we've been building turing machines in my philosophy and computers class
20:27:13 <psygnisfive> i built a machine that does remainder division. :T
20:27:13 <Slereah> psygnisfive : I was doing it before it was cool
20:27:35 <oerjan> that's it. someone hand me the mouthwash.
20:27:35 <Slereah> Because it is the Love Machine 9000.
20:28:10 <psygnisfive> im doing it all with quintuples and a single tape machine
20:29:00 <psygnisfive> and i want to just built a universal machine so i dont have to come up with any more specific machines
20:29:25 <Slereah> psygnisfive : There's one in the original Turing article
20:29:28 <psygnisfive> i mean, i want to just like.. do a MIPS-esque type architecture
20:29:34 <psygnisfive> because thats easy to conceptualize for me
20:30:01 <psygnisfive> if im not mistaken, he discusses a machine that takes another TM specification and simulates it
20:30:14 <Slereah> Well, when you say "specification"
20:30:14 <psygnisfive> but that still requires that i design other turing machines
20:30:21 <Slereah> It's actually a terrible unary encoding
20:30:45 <psygnisfive> i want to just get away from having to design task specific machines and abstract out to something general
20:31:02 <Slereah> psygnisfive : Use my Love Machine 9000!
20:31:05 <psygnisfive> not that TM specifications arent general for this machine
20:31:08 <Slereah> It's a general Turing machine.
20:31:31 <Slereah> And I'm pretty sure it works. I haven't tried in a while.
20:31:36 <Slereah> http://esolangs.org/wiki/NTCM
20:31:48 <oerjan> psygnisfive: maybe write a compiler...
20:32:16 <psygnisfive> oerjan: im just gonna try and build a mips-like machine.
20:32:29 <psygnisfive> we're using VisualTuring to build out machines
20:32:50 <psygnisfive> but it doesnt have like.. proper "states" right
20:33:22 <psygnisfive> so that each state has a particular action necessarily associated with it
20:33:28 <psygnisfive> either do nothing, move right, move left, or write a symbol
20:33:45 <psygnisfive> so the machines end up with these absurdly hard to follow designs
20:34:06 <psygnisfive> and the states themselves dont seem to be able to REPRESENT anything
20:34:26 <psygnisfive> tho luckily you can use the do-nothing states as proper states
20:34:48 <psygnisfive> it also doesnt show state transitions, so you cant follow the behavior of the machine
20:35:32 <psygnisfive> you have to drag out these action-state things
20:35:46 <psygnisfive> and then you have to CHANGE TOOLS to connect it to another state
20:36:13 <psygnisfive> then you have to change tools AGAIN to specify the conditions of the transition
20:36:25 <psygnisfive> and you have to change tools AGAIN to delete anything
20:37:03 <psygnisfive> and when you edit the tape, you have to click the tape, and then navigate it, tape-cell by tape-cell, using your arrow keys
20:37:11 <psygnisfive> you cant just click on a tape cell and edit it.
20:38:43 <oerjan> so basically, the only positive thing about this is that you can sue your school for RSI damages afterwards?
20:39:28 <oerjan> repetitive strain injury
20:39:39 <Slereah> Can I sue Wolfram for that?
20:40:04 <oerjan> doubtful, software always contains a heap of disclaimers doesn't it
20:40:53 <psygnisfive> we should build an awesome turing machine simulator
20:41:54 <oerjan> psygnisfive: now you're really hurting Slereah ;(
20:43:46 <Slereah> It totally simulates a Turing machine dude
20:44:39 <oerjan> Slereah: pay attention to the grammar
20:45:36 * oerjan wonders if slereah's's would be legal english
20:46:43 <Slereah> I think that two elisioned sesses are combined in one, technically
20:46:46 <psygnisfive> i doubt you'd pronounce it like that tho. it'd probably just be pronounced /sliri@z @z/
20:46:53 <psygnisfive> which is actually i guess how you'd pronounct slereah's's
20:47:35 <psygnisfive> you can only combine them with plural and possessive
20:48:51 <psygnisfive> cmon guys, lets design a universal machine using 5-tuples
20:49:45 <psygnisfive> maybe not mips like but simple and properly CPU-ish
20:50:09 <psygnisfive> so what are some fundamental operations we need?
20:50:25 <Slereah> Operation 1 : Getting the Love Machine 9000 :o
20:50:53 <psygnisfive> if you want me to have sex with you just say so.
20:51:42 <Slereah> The Love Machine 9000 is the official name of NTCM.
20:52:05 <Slereah> I never really had to use it officially
20:59:33 <Slereah> Finally, that piece of shit Mathematica is working
20:59:41 <Slereah> It's not very good at error warnings
21:00:41 <psygnisfive> anyone have a good reference for a simple but not-completely incomprehensible RISC design?
21:00:58 <oerjan> i'd say "love machine" is a perverse name for a Nested Tropical Cyclone Model
21:01:28 <oerjan> psygnisfive: what's wrong with subleq? :)
21:02:09 <Slereah> oerjan : Sum tropical love
21:02:28 <Slereah> Also, define "instruction"
21:04:09 <Slereah> How many instructions would you consider a RISC?
21:04:28 <tusho> psygnisfive: I meant mips
21:04:32 <oklofok> it's not about the quantity, it's about having reduced some original quantity
21:04:49 <psygnisfive> im considering mips. i know some stuff about it already.
21:04:51 <Slereah> oklofok : Wouldn't any qualify?
21:05:02 <tusho> psygnisfive: jsmips
21:06:09 <oerjan> optbot: Which CPU architecture do you prefer?
21:06:10 <optbot> oerjan: Here's a hint for writing Glass code: make every line a 0-stack-change element. The code is less efficient, but wildly more understandable.
21:06:14 <tusho> GregorR: http://github.com/kobs/js-mips/
21:06:24 <psygnisfive> tusho: im actually going to try and implement it on a 5-tuple turing machine. :P
21:06:33 <tusho> that might be hard
21:06:51 <tusho> psygnisfive: subleq
21:07:05 <tusho> psygnisfive: tough shit
21:07:34 <oklofok> didn't GregorR make quite an extensive set of basic operations for subleg
21:09:35 <tusho> i see what you did thar
21:10:46 <ais523> oerjan: unfortunately sublet is a real word
21:10:50 <ais523> not a common one though
21:11:26 <Slereah> I subleted your mom last night.
21:11:36 <oerjan> i submit that we sublet that meaning
21:11:41 <oklofok> i heard your mom last night
21:12:08 <Slereah> Well I heard you like mudkips
21:15:02 -!- chrisdb has joined.
21:15:50 <Slereah> So I herd u want to make ur own programming language
21:16:35 <chrisdb> Well, I was thinking about it for a hobby.
21:16:59 * oerjan subletty hides the mudkips
21:17:04 <Slereah> That's what this is all about.
21:17:27 <oklofok> chrisdb: tell us all about it
21:18:11 <Slereah> We're aroused by such concepts
21:18:15 <tusho> chrisdb: ignore Slereah
21:18:20 <tusho> also, you need to sacrifice some goats
21:18:21 <oerjan> Slereah: well some of you
21:18:36 <ais523> well, this channel is about making new and strange languages half the time
21:18:47 <ais523> the rest of the time it's about random crap because we're rubbish at enforcing topicality
21:19:12 <ais523> oerjan: well I tend to enforce topicality more than most people, so I see the channel as topical more than the rest of you do, I expect
21:19:22 <tusho> this channel is effectively an anarchy
21:19:32 <tusho> lament is never on, and when he is he never uses his op powers
21:19:49 * ais523 expects lament to randomly join and kick tusho at that moment
21:20:12 <oerjan> ais523: tusho is not spamming _now_ ...
21:20:22 <chrisdb> There isn't much to tell. I haven't done anything yet except reread the books gathering dust on my shelves about compiler writing (although I was only planning on a slow demo interpreter). I like programming languages which let you customise pretty much everything - Smalltalk and Ruby are examples of languages which allow you to override stuff which is hard coded in other languages. But...
21:20:22 <ais523> I mean, just to make the point
21:20:24 <chrisdb> ...most, like Ruby, have some limits, or have annoyances like the lack of support for proper operator precedence (Smalltalk).
21:20:32 <chrisdb> Of course, if I ever got it working it will be slow as hell.
21:20:44 <ais523> don't worry, most of the languages we come up with are slow as hell
21:20:44 <tusho> smalltalk's operator precedence is fine imo :P
21:20:48 <ais523> we don't bother about impractical ideas
21:20:59 <tusho> ais523: imagine an AnMaster written feather interp...
21:21:02 -!- kar8nga has quit (Read error: 145 (Connection timed out)).
21:21:09 <ais523> in fact a decent proportion of the languages we come up with are theoretically impossible to implement, and that's been proved in some cases
21:21:26 <chrisdb> tusho: it just works left to right!
21:21:31 <tusho> "I added some _posix and now it runs Hello World in 328472384723487234 years instead of 328472384723487239"
21:21:39 <tusho> operator precedence is confusing anyway
21:21:41 <tusho> AnMaster: a newbie
21:21:53 <AnMaster> anyway what language were you talking about?
21:22:08 <psygnisfive> which languages are impossible to implement? :o
21:22:14 <tusho> AnMaster: the time it took you to type that > the time it'd take to look up
21:22:38 <chrisdb> Left to right is confusing. I mean, you just don't expect 5 + 3 * 2 to evaluate as 16.
21:22:40 * oerjan gets ready to poison the competition
21:23:01 <psygnisfive> but only because you have knowledge of mathematical operator precedence. :P
21:23:06 <ais523> here 5 + 3 * 2 would probably evaluate as 10023 for no particularly good reason, some of the time
21:23:08 <AnMaster> chrisdb, I expect it to evaluate according to math rules ;P
21:23:12 <ais523> Java2K does that, IIRC
21:23:30 <chrisdb> Clearly, operator precedence should exist but be configurable. :p I want a programming language where I can increase the precedence of + or decrease it as I want.
21:23:35 <ais523> no, INTERCAL doesn't have the usual arithmetic operators at all
21:23:39 <tusho> also designed by mathematicians
21:23:49 <tusho> which leads me to believe that you're all a bunch of precedence-obsessed crazies
21:23:56 <chrisdb> tusho: Pure functional languages are annoying.
21:23:57 <ais523> haskell's a real language, but a very elegant one
21:24:03 <ais523> and would be almost eso if it hadn't caught on
21:24:04 <AnMaster> chrisdb, actually I get 5 + 3 * 2 to a stack of 15,2 with 2 at the top
21:24:06 <tusho> chrisdb: mathematics is a purely functional language.
21:24:12 <Slereah> Haskell is annoying, chrisdb
21:24:16 <tusho> either you want mathematics or you don't, chrisdb
21:24:18 <AnMaster> chrisdb, that is assuming rpn and that empty stack is 0 ;P
21:24:33 <GregorR> Haskell's a real language, but also an unstable equilibrium :)
21:24:57 <fizzie> I think Prolog (at least the swi-prolog implementation) lets you manipulate operator precedence. At least you can specify the precedence for user-defined operators.
21:24:59 <chrisdb> Slereah: Scheme isn't truly pure functional though. I like the ability to express mathematics concepts easily.. I just want to be able to store state in a non-painful way.
21:25:10 <chrisdb> For some programming problems, the easiest solution is persistent state.
21:25:28 <chrisdb> Haskell makes persistent state feel like having teeth extracted.
21:25:30 <AnMaster> chrisdb, sure, start another process that calls itself tail recursively (the erlang way)
21:25:57 <AnMaster> (erlang is purely functional, and concurrent)
21:26:11 <tusho> also, ignore AnMaster, he's still pleased to himself to no end that he managed to grasp basic erlang
21:26:20 <psygnisfive> state is really just a state object that continuously gets replaced as it gets passed from function to function ;D
21:26:20 <tusho> anyway, state in haskell is trivial
21:26:22 <AnMaster> also ignore tusho, he hates me
21:26:41 <tusho> AnMaster: If by "no reason" you mean "close to everything you say is wrong or misguided"...
21:26:44 <oerjan> always ignore tusho and AnMaster when talking to each other
21:26:47 <chrisdb> fizzie: This is true. I think the upcoming (for about 50 years) Perl 6 will have user defined operators won't it?
21:26:58 <tusho> chrisdb: Also, this channel is never this fast, ever.
21:27:00 <psygnisfive> hes at that tender young age where a boy is supposed to hate the entire world
21:27:01 <ais523> both tusho and AnMaster are good to talk with individually
21:27:17 <ais523> but never allow them to talk to each other, it always ends in tears
21:27:24 <fizzie> chrisdb: At least Perl 6 will have a whole lot of operators; at some point there were some non-ascii ones, too.
21:27:29 <oerjan> tusho: critical mass, i said
21:27:30 <tusho> shut up psygnisfive
21:27:38 <AnMaster> well no one can blame him, he will grow up at some point
21:27:43 <tusho> psygnisfive: shut up.
21:27:50 <tusho> psygnisfive: shut up.
21:27:55 <Slereah> psygnisfive is a dirty pedo
21:28:17 <Slereah> Yes, but my shirt is awesome
21:28:20 <ais523> ah, I still have psygnisfive on ignore
21:28:25 <tusho> also, someone tell AnMaster to stop fucking acting like "SIGH, well he will GROW UP at one point", I can't because I'm too busy telling psygnisfive to shut up.
21:28:28 <ais523> I was wondering why I only saw half the conversation
21:28:37 <ais523> and based on what I saw, probably I'll leave the ignore there a bit longer
21:28:39 <chrisdb> tusho: about the only easy way to get persistent state in Haskell is to pass your variables as arguments through all your functions. Which is damn annoying. I'm not a member of the 'global variables are always evil' brigade.
21:28:40 <AnMaster> ais523, why did you have him on ignore?
21:28:46 <tusho> chrisdb: StateT, man.
21:28:50 <psygnisfive> tusho: if you'd stop freaking out about your age i wouldn't enjoy poking fun at you so much. :D
21:28:54 <chrisdb> Although I am a member of the 'goto is evil' brigade.
21:28:56 <ais523> AnMaster: saying offtopic stuff I didn't particularly want to listen to
21:29:06 <tusho> psygnisfive: I haven't fucking freaked out about my age for months.
21:29:07 <ais523> chrisdb: do you consider COME FROM evil?
21:29:15 <psygnisfive> and yet i mention it once and you freak out
21:29:19 <ais523> well, I use ? in C but not goto
21:29:24 <psygnisfive> and i mention it in a purely humorous context!
21:29:24 <ais523> except in generated code
21:29:25 <tusho> psygnisfive: because it's fucking old, and pisses me off to no end
21:29:30 <ais523> my generated code contains lots of gotos
21:29:37 <psygnisfive> which is a clear sign that you're immature.
21:29:38 <tusho> esp. because it's the perfect setup for AnMaster whining about how "OHHH HE'LL GROW UP"
21:29:39 <chrisdb> ais523: Come from would be evil if it existed in a mainstream language. As it is, it's just funny.
21:29:41 <AnMaster> ais523, there are some valid reasons for goto in C
21:29:50 <ais523> AnMaster: yes, but they've never happened to me
21:30:01 <psygnisfive> unfortunately, most programmers are idiots. :P
21:30:01 <ais523> I know what they are, though
21:30:12 <tusho> "I am annoyed by things. Ergo, I am immature." Somebody give psygnisfive a Logic Medal.
21:30:26 <psygnisfive> no, you're pissed off by the mere mention of your age
21:30:39 <AnMaster> ais523, some error handling, you need to free a lot of stuff at the end of all error paths, and it can error out in many places, so a goto error; then error: <cleanup code> at the end of the function after the normal return
21:30:46 <AnMaster> ais523, I used that in FILE and SOCK in cfunge
21:30:50 <psygnisfive> a mature person would realize that it was said in a humorous context and would GO WITH THE HUMOR
21:31:03 <ais523> AnMaster: normally I instead put the relevant section into a function and use return
21:31:05 <psygnisfive> because they realize that it's not an insult but merely good fun
21:31:13 <tusho> psygnisfive: yes they would - if they found it funny. But your "joke" contained little other content than "Ha ha, tusho's age, ha"
21:31:15 <AnMaster> that and generated code are the only reasons to me
21:31:23 <tusho> which isn't amusing, it's just annoying, as it's been played out for how many months now?
21:31:26 <ais523> the other famous one is jumping out of nested loops
21:31:32 <AnMaster> tusho, I found psygnisfive funny
21:31:36 <ais523> but there's a nice rule of thumb: never use goto to jump backwards
21:31:38 <psygnisfive> you'd notice i was half making fun of anmaster.
21:31:48 <ais523> because all the legit cases involve jumping forwards
21:31:56 <psygnisfive> but you were too caught up in your OMG STOP SAYING IM 13 GUYS UGH shit to see it
21:31:58 <ais523> ofc there are illegit cases involving jumping forwards too
21:31:59 <tusho> psygnisfive: so i should find it funny because i dislike AnMaster and you were making fun of him.
21:32:30 <AnMaster> ais523, in generated code you could jump backwards with it though
21:32:55 <psygnisfive> well whatever tusho. now im just going to use your age against as often as possible since you're such an immature twat
21:33:03 <AnMaster> ais523, oh and I avoid nested loops where I have to jump out and try some other design if possible
21:33:06 <ais523> hmm... actually, IIRC C-INTERCAL's generated code never does a backwards goto
21:33:11 <tusho> psygnisfive: i'd say that doing intentionally to annoy me would be a sign of immaturity.
21:33:16 <ais523> although it does do backwards longjmps on occasion
21:33:16 <tusho> do you want to stoop to my level now?
21:33:20 <ais523> they can't really go forwards
21:33:42 <AnMaster> ais523, doing forwards longjmp wouldn't be valid
21:33:52 <ais523> well, not forwards in time
21:34:03 <ais523> I suppose you could jump forwards with longjmp inside a loop
21:34:13 <ais523> by getting the destination on the first iteration and jumping on the second
21:34:15 <AnMaster> ais523, I'm not sure that is allowed
21:34:19 <ais523> but I don't see why that's useful
21:34:22 <ais523> and yes, it is allowed
21:34:48 <chrisdb> Also, currying seems limited. What happens if I want to create a new function by specifying the second argument of another function rather than the first argument? I know there are ways around it, but it becomes longer. I want a short, elegant syntax for deriving a function which takes n-1 arguments from a function which takes n arguments no matter which argument I want to eliminate.
21:35:06 <tusho> it's in an srfi somewhere
21:35:22 <ais523> the restrictions are: setjmp must have been called before longjmp, the function containing setjmp mustn't have returned, and the call to setjmp itself must be on its own, the control expression of a switch, or compared to a constant in the control expression of an if
21:35:30 <ais523> also, auto variables + longjmp = craziness
21:36:06 -!- bsmntbombdood has quit (Read error: 113 (No route to host)).
21:36:17 <chrisdb> psygnisfive: just in general.
21:36:29 <psygnisfive> if you have the ability to define higher-order functions
21:36:48 <chrisdb> tusho: I really, really, like a lot of the abilities of lisp. It's just the billions of brackets which put me off.
21:37:01 <tusho> chrisdb: lol, are you serious
21:37:14 <psygnisfive> chrisdb: the brackets aren't that much of an issue if you have an editor and you do proper indentation.
21:37:16 <tusho> that's what java programmers say about lisp, not people who have actually done research on it and tried it :)
21:37:46 <psygnisfive> furthermore, any C-like language that's programmed functionally will use the same number of parens.
21:38:03 <chrisdb> And it also seems to lack infix operators. Which makes the bracketing even worse.
21:38:12 <oklofok> there's craziness, people mocking each other, and people actually having a conversation about coding
21:38:28 <psygnisfive> but the operators that replace them are not strictly binary
21:38:28 <chrisdb> Everything has to be a list surrounded by brackets.
21:38:49 <psygnisfive> chrisdb: the same is true of any situation where you have precedence issues, or where you're using functions.
21:39:22 <AnMaster> chrisdb, hm I agree to some extent that S-Expressions are ugly. though I quite like it's syntax anyway
21:39:35 <psygnisfive> get the fuck over it chris, the downsides of the parens are minimal. :P
21:39:41 -!- bsmntbombdood has joined.
21:39:46 <psygnisfive> or, look into some of the alternative versions of lisp that have infix operators.
21:40:11 <ais523> some sort of preprocessed Lisp might be both faster to write and annoy the hell out of sexp fans
21:40:24 <ais523> which is probably enough reason to write the preprocessor by itself
21:40:30 <chrisdb> psygnisfive: I don't really like many C-like languages either. I use them for work sometimes, but counting damn {s and }s is annoying. I prefer languages where block delineators are a bit bigger and easy to count than a little squiggle.
21:40:37 <AnMaster> psygnisfive, now I think you annoyed tusho more than ever
21:40:45 <psygnisfive> chrisdb: why the fuck are you counting {}s?
21:40:47 <ais523> chrisdb: what about Python indentation?
21:41:00 <ais523> I don't like that because indentation gets mangled too easily, with the sort of things I do
21:41:08 <AnMaster> I dislike indention based blocks
21:41:10 <tusho> ais523: then your editor sucks
21:41:19 <ais523> but C isn't too bad with a decent editor because you can reconstruct the indentation from the { and }
21:41:23 <psygnisfive> chrisdb, i think you've got some fucking stupid editing habits there
21:41:25 <ais523> tusho: no, my editors are fine for editing Python
21:41:29 <ais523> it's not editing I'm worried about
21:41:36 <ais523> consider pasting into IRC, for instance
21:41:41 <chrisdb> psygnisfive: you forget one, get a cryptic unhelpful error message from the compiler, and then spend ages trying to figure out where you missed off a }.
21:41:56 <tusho> chrisdb: so use an editor that autobalances them
21:41:58 <ais523> when I programmed bsmnt_bot to do Brainfuck, the entire program was in an eval with \n followed by varying numbers of spaces
21:42:02 <ais523> which was basically impossible to read
21:42:08 -!- oerjan has quit ("ZZZZ").
21:42:10 <psygnisfive> chris: or you could just get an editor that does automatic balancing and never worry...
21:42:35 <AnMaster> chrisdb, try emacs, it is good
21:42:49 <AnMaster> chrisdb, or kate, if you want something easier to learn
21:42:55 <ais523> emacs is a bit hard to learn but excellent once you've learnt it
21:42:59 <ais523> and Kate is pretty good all-round
21:43:03 <chrisdb> I have heard emacs is a good OS... I mean text editor.
21:43:08 <psygnisfive> chrisdb: well, i dont know what editors there are for non-macs but find one with paren balancing and highlighting
21:43:09 <AnMaster> ais523, I still use kate a lot
21:43:35 <tusho> AnMaster: i believe you should now be yelling at psygnisfive for being an apple fanboy like you do me (", i dont know what editors there are for non-macs")
21:43:36 <AnMaster> tusho, he said text editor, not headache
21:44:16 <AnMaster> tusho, why? as he acts nicely otherwise I don't see any reason to do that
21:44:25 <psygnisfive> so why should i know of non-mac text editors?
21:44:58 <psygnisfive> maybe when you're older you'll know how to take pot shots at people with more skill.
21:45:28 <tusho> psygnisfive: i don't think you are
21:45:42 <tusho> but AnMaster slings accusations of 'mac fanboy' to me all the time for similar things
21:45:42 <psygnisfive> lets see if we can make tusho commit suicide! :D
21:45:57 <tusho> and I don't see how liking someone less or more makes them more or less of a mac fanboy, do you?
21:46:05 <psygnisfive> again, tusho, you care too much about what other people think.
21:46:38 <tusho> i don't, just pointing out the obvious and glaring hypocrisy :D
21:47:33 <ais523> tusho: actually, everyone I hate is not, as far as I know ,a mac fanboy
21:47:36 <ais523> but then I don't hate many people
21:47:40 <ais523> so that's just coincidence
21:47:48 <chrisdb> I don't understand why someone would take a decent underlying OS and build a UI that looks like it's made of duplo blocks on top of it. :(
21:48:04 <chrisdb> And have a mouse with only one button, but god knows two or three is confusing obviously.
21:48:12 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | variables are "bound" simply because that's a very easy and explicit way to represent structure.
21:48:15 <AnMaster> chrisdb, that last thing they changed
21:48:23 <AnMaster> iirc modern macs got more than one button
21:48:33 <ais523> fungot: please break up the argument, by any means possible
21:48:33 <fungot> ais523: by the compiler
21:48:36 <psygnisfive> yeah, and its not like you can't use multi-button ice
21:48:38 <AnMaster> chrisdb, don't think their laptops do yet, but their mice have
21:48:45 <ais523> fungot: how will I break up the argument with the compiler?
21:48:45 <fungot> ais523: i'm checking out the sources of scheme48... can you 8d i
21:48:46 <tusho> nobody ever used one button mice, really
21:48:49 <chrisdb> psygnisfive: I mean macs are obviously aimed at the 4 - 8 yo market segment.
21:49:10 <tusho> chrisdb: 'lisp has WAY TOO MANY PARENTHESES' 'macs are for stupid users and only have one button'
21:49:20 <tusho> could you fill any more stereotypical, uninformed blanket criticisms pls
21:49:37 <ais523> tusho: well, you could combine the insults, and have an iLisp which used iParentheses
21:49:49 <tusho> ais523: and is marketed to the 4-8 year old market segment!
21:49:51 <psygnisfive> with its BLUE AND GREEN theme that looks like a tonka toy
21:49:53 <tusho> and only has one parentheses
21:49:59 <psygnisfive> yes, Mac OS X is definitely aimed at children.
21:50:11 <chrisdb> psygnisfive: Windows XP is as bad. Because they were trying to copy the Mac look... and almost succeeded.
21:50:14 <ais523> tusho: but then you'd end up with Unlambda Junior
21:50:27 <psygnisfive> they made something that looks NOTHING like OS X
21:50:43 <tusho> I like the part where chrisdb says something without basis to 'prove' his stereotypical, uninformed insult!
21:50:48 <psygnisfive> the only very colorful thing about OS X is the default window button colors
21:51:19 <chrisdb> Windows is crap, but there's some software that won't run on anything else.
21:51:26 <AnMaster> psygnisfive, what about the "aqua" look in 10.0 or so?
21:51:28 <tusho> How do I shot virtualbox
21:51:39 <psygnisfive> the aqua look? you mean the horizontal bars?
21:51:47 <AnMaster> psygnisfive, blue buttons in 10.0
21:51:49 <tusho> ais523: 'how do I shot X' meme
21:51:57 <tusho> originally spiderman saying 'how do i shot web'
21:52:02 <AnMaster> psygnisfive, haven't you even seen OS X 10.0
21:52:13 <tusho> os x 10.0 is hilariously awful
21:52:19 <tusho> but 10.3 onwards is fine...
21:52:23 <psygnisfive> i thought the window controls were always red yellow and onward
21:52:30 <ais523> isn't OS X 10.0 redundant?
21:52:36 <ais523> surely it should just be OS 10.0?
21:52:45 <tusho> ais523: not really
21:52:56 <tusho> OS X is a brand...
21:52:59 <tusho> and so is the 10.x versions
21:53:13 <tusho> Mac OS X v10.major_number_that_actually_changes
21:53:15 <tusho> quite silly, really
21:53:21 <AnMaster> ais523, ah no, I think the X is like NT, meaning "we changed the whole thing and made it not crash when one program crashes"
21:53:26 * ais523 understands computers, but doesn't really understand computer marketing people, no matter what it is they're marketing
21:53:41 <tusho> psygnisfive: http://upload.wikimedia.org/wikipedia/en/7/78/Macosxpb.png
21:53:49 <tusho> warning: will make you puke with hilarity
21:53:59 <tusho> the public beta had the APPLE IN THE MIDDLE OF THE MENU BAR
21:54:01 <psygnisfive> tusho: ok.. what about it? the only annoying thing is the horizontal bars
21:54:05 <tusho> that makes no sense at all
21:54:22 <AnMaster> tusho, yeah was that menu clickable?
21:54:34 <ais523> tusho: it lets you open two programs at once, and use the menu for both of them
21:54:37 <psygnisfive> raskin wasn't working there anymore, what do you expect :P
21:54:54 <ais523> it's a capability that Windows has and Mac OS X doesn't have, mostly for the reason that it isn't actually useful
21:54:55 <AnMaster> tusho, then it must have been uggly when menus didn't fit in
21:55:05 <tusho> AnMaster: that was in the public beta...
21:55:08 <tusho> before retail release
21:55:13 <psygnisfive> ais: windows most certainly DOESNT have that ability
21:55:19 <tusho> "mac os x programs" was a funny joke
21:55:21 <AnMaster> tusho, indeed, and OS 9 had the apple in the same place they have now
21:56:09 <AnMaster> <ais523> it's a capability that Windows has and Mac OS X doesn't have, mostly for the reason that it isn't actually useful <-- yes so does linux, depending on settings
21:56:22 <tusho> 'linux' has no gui, it's meaningless
21:56:27 <ais523> AnMaster: yes, especially in KDE you can choose how the menus work
21:56:29 <tusho> you mean 'yes so does KDE' or 'yes so does Gnome'
21:56:35 <ais523> and no, 'Linux' doesn't by itself have a UI
21:56:47 <AnMaster> psygnisfive, he means menus local to the window instead of at the top of the screen
21:56:55 <ais523> but people talking about a "Linux GUI" generally are referring to either KDE or Gnome
21:56:58 <GregorR> Linux by itself doesn't have a TUI :P
21:56:59 <AnMaster> chrisdb, I couldn't agree more
21:57:02 <psygnisfive> but you cant click on two menu bars at once in ANY situation
21:57:05 <ais523> because people who use something else are unlikely to make that mistake
21:57:09 <tusho> chrisdb: how is os x more childish than KDE :|
21:57:16 -!- Corun has quit (Remote closed the connection).
21:57:20 <pikhq> Of course KDE > Gnome.
21:57:24 <AnMaster> tusho, depends on what version you compare
21:57:24 <pikhq> After all, K > G. :p
21:57:25 <tusho> chrisdb: plz say how http://upload.wikimedia.org/wikipedia/en/c/c0/Leopard_Desktop.png is more childish than kde
21:57:29 -!- Corun has joined.
21:57:33 <ais523> you can connect to a serial port and send magic SysRq over the serial connection
21:57:35 <AnMaster> tusho, http://upload.wikimedia.org/wikipedia/en/7/78/Macosxpb.png is more childish than anything
21:57:46 <tusho> AnMaster: yes, but that's irrelevant nowadays
21:57:47 <tusho> http://upload.wikimedia.org/wikipedia/en/c/c0/Leopard_Desktop.png
21:58:02 <chrisdb> I ought to try 4.1. They say that they've ironed out a lot of the problems with Plasma.
21:58:09 <chrisdb> I want to know if it is any good or not.
21:58:11 <tusho> chrisdb: justify how http://upload.wikimedia.org/wikipedia/en/c/c0/Leopard_Desktop.png is more childish than kde
21:58:21 <AnMaster> tusho, 3D look and transparent menu bar, I hope you can turn those off?
21:58:27 <tusho> AnMaster: 3d look of the dock: yes
21:58:29 <tusho> transparent menu bar: no
21:58:31 <GregorR> OS X does its best to act like a toy in every regard. Useless 3D effects, things bouncing and "genie"ing about, nice and flashy for a two-year-old but I want a fucking OPERATING SYSTEM, not a toy.
21:58:39 <psygnisfive> turning off the transparent menu bar is trivial
21:58:50 <tusho> GregorR: useless 3d effects, genieing about?
21:58:53 <tusho> have you ever used an os x system
21:59:05 <pikhq> GregorR: OS X does two different things: it tries to be easily usable and it tries to be a toy.
21:59:05 <GregorR> tusho: You can disable /many/ effects.
21:59:07 <ais523> not all 3D effects are useless
21:59:08 <chrisdb> tusho: pretty HUGE icons which pop up where you hover over them, special effects all over the place, wobbling windows...
21:59:09 <ais523> some of them are, though
21:59:14 <pikhq> Unfortunately, the two don't necessarily go hand in hand.
21:59:19 <chrisdb> it's all eye candy for five year olds.
21:59:25 <ais523> also, I like the wobbly windows in Compiz, although my brother doesn't
21:59:50 <psygnisfive> so chrisdb, do you have any arguments that AREN'T strawmen?
21:59:50 <AnMaster> http://gentoo-wiki.com/images/d/df/AnMaster_KDE20080112.png
21:59:54 <ais523> I use Gnome because it gives me a nice simple UI for when I don't want to think, and I'm more used to doing complicated stuff over the command line so don't care what the desktop environment is then
22:00:03 <GregorR> I'll totally destroy my case by saying that I would /love/ a UI for X11 that resembled Windows 3.11.
22:00:06 <pikhq> GregorR: Compiz has a ton of stuff that's rather useless, for example.
22:00:12 -!- Slereah has quit (No route to host).
22:00:12 <tusho> GregorR: twm is close
22:00:16 <pikhq> Also, I wouldn't want that.
22:00:23 <GregorR> tusho: Yeah, but it has annoying focus properties.
22:00:26 <ais523> GregorR: also I've used CDE, that's pretty like Windows 3.11
22:00:28 <tusho> GregorR: Patch it.
22:00:30 <psygnisfive> anmaster: mac os doesnt have lots of 3D stuff. :P
22:00:33 <ais523> but it loads to shell by default, IIRC
22:00:35 <pikhq> My preferred UI tends to range from KDE to Ratpoison, depending upon my mood.
22:00:49 <GregorR> TOO MANY PEOPLE TALKING CAN'T KEEP UP *gives up*
22:00:49 <psygnisfive> well, unless you count time machine but thats useless anyway.
22:00:53 <chrisdb> I like KDE because it offers options and configurability instead of catering to the lowest common denominator.
22:01:00 <AnMaster> psygnisfive, windows that are like a genie in a bottle?
22:01:07 <AnMaster> psygnisfive, I used OS X, I seen it
22:01:13 <chrisdb> And because KDE media frameworks etc tend to work well.
22:01:16 <tusho> AnMaster: that's just the minimize effect
22:01:24 <GregorR> psygnisfive: You can change it from genie to scale, but not /off/.
22:01:24 <tusho> and i rarely ever minimize
22:01:26 <chrisdb> Whereas the Gnome ones are generally just about adequate.
22:01:26 <ais523> chrisdb: I like Gnome because if I'm going to do complicated options and configurability I'm unlikely to use my desktop environment for it anyway
22:01:29 <GregorR> psygnisfive: The effect is unremovable.
22:01:36 <AnMaster> psygnisfive, oh and you can't easily change font size with built in settings, sure you could edit some *.plist file
22:01:43 <AnMaster> or use some freeware tool that allows it
22:01:53 <AnMaster> doesn't allow a lot of settings
22:01:53 <GregorR> psygnisfive: That option must be pretty damn well hidden because I've never managed to turn off all the stupid effects X_X
22:01:59 <AnMaster> in fact gnome allows *more* settings
22:02:12 <chrisdb> but no-one's ever going to write software that takes advantage of its libraries.
22:02:17 <pikhq> And Gnome gets criticised for not *having* settings.
22:02:28 <AnMaster> pikhq, I say OS X is worse than Gnome
22:02:33 <pikhq> ais523: Enlightenment v. 17.
22:02:37 <pikhq> AnMaster: And I agree.
22:02:46 <psygnisfive> actually, no, gregor, you're right, you cant turn it off.
22:02:52 <ais523> anyway, Gnome vs. KDE is one of those arguments that never goes away
22:02:55 <ais523> and I'm glad both exist
22:03:03 <psygnisfive> thats what i get for never actually minimizing windows ;D
22:03:05 <pikhq> Well, actually, OS X *looks* a bit better out of the box...
22:03:13 <ais523> (n.b. although I use Gnome as the window manager and GDM, I tend to use KDE applications)
22:03:15 <AnMaster> pikhq, agreed, if you like eye candy
22:03:18 <GregorR> AXIOM A: Linus prefers KDE to gnome. AXIOM B: Linus Torvalds is incapable of wrong. CONCLUSION: KDE is better than Gnome.
22:03:19 <tusho> i kind of hate how people give 10 points about os x, 9 of them are debunked, and then even with only 1 point they consider that to be evidence that os x sucks
22:03:21 <pikhq> Which is good, since it's hard to change how it looks. :p
22:03:25 <AnMaster> pikhq, http://gentoo-wiki.com/images/d/df/AnMaster_KDE20080112.png
22:03:37 <pikhq> AnMaster: That's a decent KDE setup.
22:03:40 <ais523> tusho: actually all operating systems suck
22:03:47 <AnMaster> pikhq, thanks, and no eye candy
22:03:56 <pikhq> I've got mine somewhat similar...
22:03:56 <ais523> you pick the one whose type of suckiness interferes least with what you want to do
22:03:57 <tusho> but it's only relevant to talk about relative suckitude
22:04:03 <pikhq> Except I use Plastik and the default icons...
22:04:11 <chrisdb> One thing people don't use to advocate Linux enough: installing software is truly easy. Point your tool of choice at an online repository, tell it what you want, and let it sort out all the dependencies for you automatically.
22:04:14 <pikhq> And full-screen windows.
22:04:15 <AnMaster> pikhq, can't stand that, I use kdeclassic
22:04:18 <ais523> so in my case I mostly don't care about Linux being bad at running Windows applications and Gnome being hard to configure
22:04:21 <chrisdb> It's hard if you don't have fast internet access..
22:04:33 <chrisdb> but otherwise it's easier than other systems, not harder.
22:04:37 <pikhq> I prefer Plastik, of course, but KDEclassic is a fine theme.
22:04:38 <ais523> and I solved the hardware compatibility problems by buying a laptop designed to work with Linux
22:04:42 <AnMaster> <chrisdb> One thing people don't use to advocate Linux enough: installing software is truly easy. Point your tool of choice at an online repository, tell it what you want, and let it sort out all the dependencies for you automatically. <-- yep
22:04:46 <GregorR> http://www.kde-look.org/CONTENT/content-pre2/54708-2.png // my KDE look
22:04:56 <psygnisfive> nevermind that knowing all this crap about repositories is NOT intuitive
22:04:59 <AnMaster> psygnisfive, it works very well on both Gentoo and Arch
22:05:13 <tusho> PROPOSING A TRUCE: We all shut the fuck up about operating systems because we have never convinced one person and all we do is fling shit at other people without achieving anything.
22:05:16 <ais523> given that the Ubuntu repos contain two different INTERCAL implementations, I assume they're pretty complete
22:05:19 <AnMaster> psygnisfive, never used ubuntu
22:05:21 <ais523> AnMaster: on repos being important
22:05:23 <chrisdb> psygnisfive: yes it is. What's hard about the idea of telling a tool where to look for software and what the name is of the software you want?
22:05:24 <pikhq> Now, my *Ratpoison* setup is quite different from everything else seen here.
22:05:30 <AnMaster> psygnisfive, how do you uninstall a *.pkg on OS X
22:05:32 <pikhq> Full-screen terminal.
22:05:34 <tusho> PROPOSING A TRUCE: We all shut the fuck up about operating systems because we have never convinced one person and all we do is fling shit at other people without achieving anything.
22:05:36 <pikhq> And, well, not much else.
22:05:39 <GregorR> psygnisfive: http://oblisk.codu.org/
22:05:45 <AnMaster> psygnisfive, not just delete the app, since it installed stuff in /System/Library/
22:06:04 <psygnisfive> well, thats a problem of the app being ridiculous then
22:06:10 <chrisdb> psygnisfive: how is rpm any different from a .msi on windows?
22:06:17 <AnMaster> psygnisfive, I'm sure it is ridiculous yeah
22:06:22 <tusho> I like the part where everyone decided that they prefer flinging shit to actually talking about relevant stuff!
22:06:24 <chrisdb> Apart from the dependencies thing?
22:06:29 <pikhq> GregorR: You have an LCARS theme? :)
22:06:32 <AnMaster> psygnisfive, it was apple's own xcode.
22:06:38 <chrisdb> You just basically have files and locations to automatically put them.
22:06:45 <ais523> tusho: well, talk about something relevant then
22:06:52 <ais523> for instance, Keymaker's come up with a new language
22:06:56 <ais523> which is very tarpitty
22:06:58 <psygnisfive> AppZapper, anmaster. it should be built into Mac OS but its not
22:06:59 <tusho> ais523: I would if it weren't for the GIANT FLOOD OF PEOPLE YELLING ABOUT OPERATING SYTSEMS
22:07:00 <AnMaster> psygnisfive, even windows got better uninstall than OS X
22:07:09 <tusho> which makes it POINTLESS
22:07:18 <ais523> http://esolangs.org/wiki/Figurehead
22:07:28 <tusho> EVERYBODY SHUT UP K THX BAI
22:07:30 <psygnisfive> i never really uninstall stuff tho so im not familiar with how to do it :p
22:07:32 <chrisdb> The windows registry sucks. Uninstalling a program should be as simple as deleting the directory containing it in most cases.
22:07:38 <tusho> fffffffffffffffffffffffffffffffffffffffffffffffFFFFFFFFFFFFFFFFFFFFFFFFFFFF
22:07:42 <AnMaster> psygnisfive, while I never had linux package manager fail to uninstall stuff
22:07:45 <chrisdb> Settings stored centrally is a crappy idea.
22:07:53 <tusho> hi, some of us would like to TALK ABOUT ESOLANGS
22:07:57 <tusho> I am reading figurehead now, seems interesting
22:08:05 <ais523> AnMaster: I did, once, but that's due to deliberately running the version of the Ubuntu repos which break at random now and again
22:08:07 <psygnisfive> anmaster: i ahree. apple needs to fix its uninstall utilities.
22:08:12 <ais523> and I figured out what had gone wrong and fixed it by hand
22:08:18 <ais523> tusho: yes, figurehead is interesting AFAICT
22:08:21 <AnMaster> ais523, well I used many OS but never ubuntu
22:08:29 <ais523> but I'm not entirely sure how different it is from a Minsky machine
22:08:39 <tusho> ais523: i'd discuss how interesting it is ... if it weren't for everyone getting nowhere with a stupid conversation about OSs! :D
22:08:40 <ais523> AnMaster: it's just Debian with different defaults
22:08:53 <AnMaster> ais523, hardly used debian very much either
22:09:32 <AnMaster> ais523, and it was ages ago I used something *.rpm, Red Hat-pre-fedora, maybe 7.0 or so
22:09:54 <AnMaster> psygnisfive, well on linux you get it with the OS :P
22:10:09 <ais523> AnMaster: so you don't really use binary packages much at all?
22:10:13 <psygnisfive> you get the uninstaller free with the OS, yes, i know
22:10:19 <chrisdb> AnMaster: On Linux you get *everything* with the OS. Including the kitchen sink.
22:10:21 <ais523> which format are those in?
22:10:30 <psygnisfive> apple has a lot of problems with how it does things, unfortunately.
22:10:38 -!- Slereah has joined.
22:10:41 <chrisdb> I like it, but... you do get a shitload of stuff you never use if you select a default install for most distros.
22:10:47 <ais523> $ apt-cache search kitchen sink
22:10:47 <ais523> xemacs21-basesupport - Editor and kitchen sink -- compiled elisp support files
22:10:52 <tusho> dsfhkudsklfabhjkafkdjsdafhksjlfehkuwlhiuafwhioadfshiuafhiufwfwqdwjdhjdjhbhjbxakhjsadkjqwrjkhqwkhjqwehjkqewjkhhjkqerwhjkl qwehrjkeqwjkrkehjajahkrwhjkewr
22:10:55 <AnMaster> ais523, PKGBUILD to construct packages, actual packages are *i686.pkg.tgz iirc
22:11:38 <AnMaster> ais523, and the PKGBUILD files are bash on turbo basically
22:11:39 <ais523> anyway, I think I'll just stick with my .pax.{lzma|bz2|gz} format I'm using for C-INTERCAL
22:12:02 <ais523> because it confuses people, yet works anyway
22:12:32 <AnMaster> ais523, sure that is source code format
22:12:50 <ais523> by the way, how many people here use cpio?
22:13:00 <ais523> tar v. cpio is supposed to be one of the big Holy Wars in Unixen
22:13:05 <ais523> but I seem to have missed it completely
22:13:11 <AnMaster> ais523, I did use it once to extract the files from a *.rpm iirc
22:13:21 <ais523> maybe it ended before I got used to Linux?
22:13:27 <ais523> or maybe it's just raging somewhere I don't know?
22:13:48 <AnMaster> thought cpio was simply an old format
22:13:58 <ais523> apparently pax was invented in an attempt to end the war
22:14:57 <tusho> Meanwhile, http://tusho.net/ is the first XHTML5+RDFa page ever, especially because that hasn't actually ben specced yet.
22:15:44 <chrisdb> Does Lua have operator overloading?
22:16:22 <ais523> tusho: well, it works in Konq atm
22:16:28 <tusho> ais523: of course, it should
22:16:38 <tusho> view the source, it's just xhtml5+extra properties and shit that define RDF stuff
22:16:42 <tusho> AnMaster: html5 serialised as XML, basically
22:16:49 <tusho> html5 is defined in the abstract
22:16:54 <tusho> then html and xml serializations are defined
22:16:57 <ais523> tusho: hmm... there should actually be a page that's written in Turing-complete XSLT
22:17:05 <tusho> i use xhtml5 so that i can import fancy namespaces
22:17:31 <ais523> actually, I was aware that there was an XML version of HTML5
22:17:37 <ais523> don't think I've ever seen anyone use it before, though
22:17:51 <tusho> intertwingly.net, iirc
22:17:53 <AnMaster> tusho, don't you need a <!DOCTYPE html>?
22:18:02 <ais523> AnMaster: that's for the HTML serialisation
22:18:06 <tusho> that doctype is only to force standards-compliant mode for text/html
22:18:15 <tusho> whereas xml mime-types are treated as strict by all browsrs today
22:18:23 <tusho> psygnisfive: XHTML5+RDFa, actually.
22:18:31 <AnMaster> tusho, except IE offers a download dialog iirc
22:18:41 <tusho> and people are starting to drop support for IE6
22:18:43 <ais523> hmm... are there any XSLT interps in XSLT lying around, I wonder?
22:19:01 <ais523> also, /me tries to guess who wrote the topic
22:19:33 <ais523> I've actually seen some people who were still using Windows 2000
22:19:55 <tusho> ais523: it was oklofok
22:20:08 <tusho> anyway - XHTML1+RDFa is specified, and XHTML5 is specified, and there's not anything in XHTML5 that would change the way you apply +RDFa
22:20:09 <ais523> wow, I think I'm beating 50% at this now
22:20:10 <tusho> so I just did that
22:20:16 <ais523> not bad given how many people are here in the channel
22:20:20 <tusho> essentially I wrote an XHTML1+RDFa document but with HTML5 tags
22:20:23 <tusho> which is, er, what it'd be
22:22:58 <ais523> presumably a standard XML validator would work
22:23:08 <tusho> AnMaster: http://validator.nu/ will validate the XHTML5 part
22:23:12 <tusho> but give errors for the RDFa part, obviously
22:23:21 <tusho> http://www.w3.org/RDF/Validator/ might validate the rdf parts, not sure
22:23:27 <ais523> XML is pretty trivial to validate, that's one of its goals in existance I think
22:23:52 <tusho> nah, http://www.w3.org/RDF/Validator/ doesn't validate it because you don't use actual rdf elements directly
22:24:03 <tusho> <foaf:name>foo</foaf:name> becomes something like <h1 property="foaf:name">foo</h1>
22:24:12 <tusho> which is reasonable - it's for xhtml documents decorated with rdf semanticity
22:24:21 <chrisdb> Bah humbug. Good old HTML 4.01 Transitional was good enough in 1999, and it's good enough now.
22:24:42 <tusho> chrisdb: i'm interested in ai, and the semantic web is an interesting step to things like that
22:24:51 <tusho> plus i've always been a fan of metadata
22:25:12 <ais523> some day I must get round to writing that program that extracts command-line options for ick from a file's last-modified date
22:25:54 <chrisdb> I had an interest in AI, but the more I read the more depressed I got about the limited accomplishments in that area. That's not to say that putting the intelligence into AI is impossible, of course, but I'm not sure I know how to do it.
22:26:26 <ais523> AnMaster: I proposed it in a.l.i, they said I was mad
22:26:38 <tusho> chrisdb: well, I'm also interested in sort-of-AI
22:26:40 <ais523> and when a.l.i say you're mad you know you're onto a good idea
22:26:58 <tusho> the semantic web is a huge web of relationships in a mesh of human-and-machine readable info & data
22:27:16 <tusho> timbl supports it heavily, and i think it has quite some potential
22:27:23 <tusho> also it's not a fad, started in like 2000 and has been going slowly but steadily since
22:27:33 <tusho> i'm fine having my humble little homepage be a tiny, tiny part of it
22:27:36 <AnMaster> ais523, got copies of the thread?
22:27:46 <tusho> AnMaster: google groups yo
22:27:53 <ais523> AnMaster: it's still on Google Groups IIRC
22:28:21 <chrisdb> tusho: I think it does have some potential. The main problem with AI reasoning systems (apart from their poor handling of ambiguity and uncertainty sometimes) if their lack of a massive amount of basic knowledge that human beings take for granted.
22:28:21 <ais523> http://groups.google.com/group/alt.lang.intercal/browse_thread/thread/5a0696843eeeb5b6#
22:28:36 <tusho> chrisdb: yes, i like to think about stuff like that but it's not why i like the semantic web
22:28:45 <tusho> which is founded on more practical concerns, really
22:29:20 <AnMaster> ais523, "Perl's method of doing this, in particular, is delightfully twisted, so INTERCAL has to come up with something weirder. " <-- hm?
22:29:26 <ais523> <ais523> (and posting here to ask whether it's a good idea and whether I should do it, which in INTERCAL aren't necessarily correlated)
22:29:39 <ais523> AnMaster: try writing #!/bin/cat and running it as Perl some time
22:30:33 <ais523> if Perl sees a #! line mentioning an application other than Perl, it execs that application rather than running normally
22:30:44 <AnMaster> ais523, that is strange, the OS should do that
22:30:48 <ais523> thus #!/bin/cat causes Perl to exec cat, thus outputting a copy of the source
22:30:56 <ais523> AnMaster: yes, the OS does do that if the file's marked executable
22:30:57 <chrisdb> The nastiest thing about Perl is its lack of support for proper named function arguments . You have to pop them off @_ don't you?
22:31:09 <ais523> but Perl does that if the file's run deliberately with Perl
22:31:21 <ais523> chrisdb: yes, but there's an idiom which pretty much makes it like K&R C
22:31:35 <tusho> ais523: because K&R C is totally awesome
22:31:39 <AnMaster> I think prototypes are a good thing(TM)
22:32:03 <ais523> #perl (or is it ##perl?) tells people not to use them because their semantics are confusing
22:32:03 <AnMaster> ais523, oh? why don't ppl use it then?
22:32:18 <ais523> AnMaster: partly because they don't give you named arguments
22:32:23 <ais523> they change the syntax of the language instead
22:33:00 <chrisdb> Does Perl let you return multiple values from a function? Like an inverse @_?
22:33:11 <GregorR> You can return an array *shrugs*
22:33:31 <chrisdb> I meant without the packing. Something like Matlab's multiple return values.
22:33:50 <chrisdb> [A B] = SomeFunction(C,D,E)
22:34:04 <ais523> sub test {1;}; test / 4; # /;
22:34:12 <ais523> sub test ($) {1;}; test / 4; # /;
22:34:20 <ais523> two legal Perl programs
22:34:23 <ais523> the first one returns 1
22:34:45 <tusho> the second returns 1/4
22:34:46 <ais523> the first one returns 1, and the second returns 1/4
22:35:07 <ais523> also, the second one ends in a comment but the first one doesn't
22:35:21 <ais523> (this is the point where people normally go wtf)
22:35:47 <chrisdb> Meh, we've all seen perl line noise programs, it takes more than that to shock.
22:35:53 <tusho> / ... / is treated as a regexp argument in the first one
22:35:55 <ais523> chrisdb: this isn't about line noise
22:36:01 <ais523> that program is pretty readable
22:36:03 <tusho> the second program needs 'sub test()"
22:36:17 <tusho> hahahahahahahah i fixed ais523's crazy perl's error
22:36:18 <ais523> sub test {1;}; test / 4; # /;
22:36:21 <ais523> sub test () {1;}; test / 4; # /;
22:36:29 <AnMaster> <chrisdb> [A B] = SomeFunction(C,D,E) <-- you mean a tuple?
22:36:40 <chrisdb> First place, Best 'The Perl Journal', 1 st Annual Obfuscated Perl Contest: Joe Futrelle.
22:36:48 <chrisdb> ;while(<SZzBN3>){/\s\((.*p\))&/
22:36:49 <chrisdb> &&(@S2zBN3=unpack$age,$1)}foreach
22:36:54 <chrisdb> 30){$_=<SZz8N3>}/^(.)/|print $1
22:36:56 <chrisdb> ;$S2z8M3=0}s/.*//|print}sub w1HC{$age=c17
22:36:57 <chrisdb> ;socket(SZz8N3,PF_INET,SOCK_STREAM,getprotobyname('tcp'))&&
22:36:59 <chrisdb> connect(SZz8N3,sockaddr_in(023,"\022\x17\x\cv"))
22:37:01 <tusho> chrisdb: Oh, you flood too!
22:37:12 <AnMaster> chrisdb, please use a pastebin for that long pastes in the future
22:37:14 * ais523 wonders why the S2zBN3
22:37:16 <tusho> Also, that is totally not obfuscated.
22:37:18 <ais523> probably it's really significant
22:37:27 <tusho> ais523: because silly names make things reallllllllly complex...
22:37:34 <tusho> most obfuscation is based on trivial stuff liek that
22:37:45 <ais523> and tusho: no, but it would cause utter chaos if someone messed around with the S2zBN3 typeglob, as all the variables in the program have the same name...
22:38:14 <ais523> also, typeglobs are sufficiently esoteric that i have trouble just thinking about htem
22:38:53 <ais523> AnMaster: basically a typeglob is a sort of metavariable that holds all the variables with a given name
22:39:22 <ais523> so *S2zBN3 refers to a scalar, array, hash, filehandle, and typeglob, each of which is called S2zBN3
22:39:36 <ais523> normally you ignore them, but you can cause utter chaos by assinging to them
22:39:39 <tusho> (i.e. @S2zBN3, $S2zBN3)
22:39:45 <chrisdb> AnMaster: Matlab has special syntax for returning multiple arguments (the [] syntax). Apart from in this use, sticking multiple variables between square brackets will usually attempt to cat them to form a single matrix (which would fail unless their dimensions were compatible).
22:39:50 <ais523> e.g. IIRC if you do *foo = *bar, then $foo will actually mean $bar
22:39:55 <AnMaster> ais523, you mean several variables can have same name, as long as their types are different?
22:40:12 <AnMaster> chrisdb, but isn't that really like tuples?
22:40:13 -!- kar8nga has joined.
22:40:15 <ais523> actually quite a lot of sigil-using languages do taht
22:40:25 <tusho> the variable is called $foo, not foo
22:40:28 <tusho> that's the sane way to think about it
22:40:58 <chrisdb> AnMaster: I guess so, except it's limited to returning arguments and not found elsewhere.
22:41:57 -!- kar8nga has left (?).
22:43:17 <ais523> tusho: no, the sane way to think about it is "scalar called foo"
22:43:32 <ais523> for instance, the array called foo can be referred to either as $foo or @foo depending on context
22:43:52 <ais523> both refer to the array foo
22:44:01 <ais523> the sigil doesn't depend on the type of the variable
22:44:08 <chrisdb> Why isn't it @foo[1]. Does that do something else?
22:44:10 <ais523> but on the type of the expression you're extracting from the variable
22:44:14 <chrisdb> Or is it just a pointless alternation?
22:44:28 <ais523> chrisdb: yes, it gives you a 1-element array that's a slice of the array foo on its first element
22:44:36 <ais523> an expression that starts with @ always returns an array
22:44:46 <ais523> anyway, apparently they're 'fixing' this weirdness for Perl6
22:45:04 <tusho> the fix is more confusing
22:45:12 <tusho> it removes all the consistency
22:45:22 <ais523> well at least it gives variables consistent names
22:45:31 <ais523> but makes contexts even harder to figure out in one's head
22:46:32 <chrisdb> A couple of years ago I thought Perl 6 sounded fairly interesting.. a slightly saner version of Perl.
22:46:50 <chrisdb> Now I've gotten tired of waiting for it and all my interest has drained away.
22:47:09 <ais523> imagine C where every { } was in fact a lambda that took no arguments
22:47:20 <ais523> oh and functions were first-class
22:47:56 <pikhq> That part sounds like Plof.
22:48:24 <ais523> also, Perl6 regexen are powerful enough to write parsers with
22:48:33 <ais523> which automatically deduce a lexer from the syntax of the language
22:48:35 <chrisdb> ais523: I meant saner in the syntax and presentation. It has proper named arguments to functions and other basic things which will make the code look a bit cleaner.
22:50:00 <AnMaster> perl6 is like GNU Hurd, we can go on and wait for as long as we like
22:50:13 <chrisdb> And the {} thing doesn't sound that insane. Reducing a language to as few basic principles as possible can sometimes be helpful.
22:50:18 <chrisdb> Isn't that why people like Lisp?
22:50:35 <chrisdb> Underneath all the complexity is a few basic principles and a lot of Lisp code.
22:50:37 <tusho> AnMaster: i disagree, perl 6 has a kind of better schedule
22:50:44 <tusho> it IS being worked on, slowly, but progress IS being mae
22:50:48 <tusho> whereas the hurd...is just dead
22:51:04 <chrisdb> Is the Hurd dead? Or are a few people still working on it?
22:51:09 <AnMaster> tusho, ah but hurd back around the time linux was first invented
22:51:10 <tusho> i believe duke nukem forever will be released sometime in my lifetime...
22:51:18 <tusho> AnMaster: perl 6 has been going for ages, though
22:51:21 <tusho> and died for a bit
22:51:26 <tusho> and then never got back up again
22:51:28 <chrisdb> I could never imagine the design being fast though.
22:51:28 <tusho> because nobody cared
22:51:50 <tusho> Observation: A bright purple colour that is not pink is a hard thing.
22:52:05 <ais523> tusho: make it a bluish magenta
22:52:08 <ais523> but as bright as possible
22:52:22 <ais523> let's think... #CC00FF
22:52:23 <chrisdb> Poor Richard Stallman, with his insistence on calling it GNU/Linux or whatever.
22:52:40 <ais523> chrisdb: I've used Linux systems which had no GNU software on
22:52:59 <pikhq> ais523: Such systems, of course, shouldn't be referred to as GNU/Linux.
22:53:24 * pikhq wonders how many people use Busybox/ucLibc systems, though.
22:53:33 <ais523> mine was Busybox/uclibc/Linux
22:53:35 <tusho> ais523: that is pink :P
22:53:57 <tusho> i mean bright as in light
22:54:03 <tusho> as in ... er, more white
22:54:17 <ais523> tusho: that is so not pink
22:54:20 <tusho> and turning up the saturation on #CC00FF totally gives you pink
22:54:34 <ais523> tusho: and yes, that's a maximally saturated colour in its colour group
22:54:40 <ais523> as it has both a 00 and an FF in
22:54:48 <chrisdb> I read the webpage of a guy who'd stripped down a Linux system so that all it did was load emacs after it booted once.
22:54:59 <ais523> so you can't saturate it more without making it a different colour
22:55:39 <tusho> ais523: so i stand by what i said - a whiteish (you could put black text on it) purple is very hard
22:56:22 <AnMaster> ais523, you could, in a wider colour space
22:56:36 <ais523> AnMaster: yes, but you'd need a better screen
22:56:56 <ais523> anyway, I rather like what CC00FF does on an LCD screen
22:57:01 <ais523> from below it's bright purple
22:57:11 <ais523> looking from slightly above it becomes a reddish purple
22:57:16 <AnMaster> ais523, mine got very wide "viewing angle"
22:57:24 <ais523> then as I move up it becomes a bluish purple
22:57:28 <ais523> and then a greyish orange
22:57:53 <oklofok> you can look at my laptop's screen from any direction
22:57:53 <tusho> my lcd seems to have an infinite "viewing angle"
22:57:54 <ais523> tusho: that isn't pink
22:58:00 <ais523> and yes, I tested on a web-browser
22:58:02 <tusho> but looking at this one from any angle works fine
22:58:07 <ais523> using a Javascript: URL to fill the screen with colour
22:58:12 <oklofok> yarr my so called "viewing angle" is infiniteous too
22:58:13 <ais523> as I couldn't remember how to do it with data:
22:58:21 <ais523> this screen's viewing angle stuff is rubbish
22:58:31 <tusho> ais523: data:text/html,<html bgcolor=#...>
22:58:48 <ais523> except html body bgcolor=
22:58:50 <AnMaster> <tusho> but looking at this one from any angle works fine <-- does it have blank screen?
22:59:03 <ais523> tusho: I couldn't remember how you specified the encoding
22:59:06 <ais523> but it looked like you didn't
22:59:09 <tusho> maybe that has something to do with it
22:59:10 <AnMaster> tusho, those with blank seem to have a huge viewing angle
22:59:17 <tusho> ais523: that's just data:text/html;base64,...
22:59:18 <tusho> AnMaster: define blank
22:59:19 -!- chrisdb has quit ("ChatZilla 0.9.83 [Firefox 3.0.2/2008091620]").
22:59:44 <tusho> had a tiny viewing range
22:59:50 <AnMaster> tusho, err blank is wrong English word I bet
22:59:51 <tusho> so uh, i'd say that mattes have larger range :|
23:00:05 <tusho> but yes, I previously had a glossy LCD: tiny range
23:00:10 <tusho> this matte one: infinite range, it seems
23:02:04 <tusho> ais523: anyway, it's not pink
23:02:09 <tusho> but it's not just a lighter version of a dark purple
23:02:35 <ais523> tusho: colour bands are skewed between computer perception and human perception
23:02:50 <ais523> so it's possible some colours expressible in English don't even exist in RL
23:03:04 <tusho> which is unfortunate
23:03:07 <tusho> as i can picture it in my head
23:05:08 * tusho goes all out to make his XHTML5+RDFa page look flashy as it doesn't really matter if it's form over function because it's just my little web identity card and it's only one entry page
23:05:24 <ais523> anyway, I imagine the best way to do it would be to take a violet or indigo wavelength
23:05:34 <ais523> and just make it really really bright by shining a lot of photons at once
23:05:40 <ais523> to create the RL colour
23:05:53 <ais523> although that would hurt your eyes and you couldn't get a computer to do that
23:06:09 <tusho> CONCLUSION: Fuck computers.
23:06:30 <ais523> tusho: unfortunately Japanese sex robots aren't yet sufficiently advanced to make that feasible
23:07:10 <ais523> but it'll probably be all over the news when it happens
23:13:26 <AnMaster> ais523, I love your "just-in-case compiler" in http://groups.google.com/group/alt.lang.intercal/browse_thread/thread/9e9ad7d6875e9582#
23:14:05 <ais523> yes, it's an interesting strategy, CLC-INTERCAL had a just-too-late compiler so I needed to think up a snazzy name for mine
23:14:37 <AnMaster> ais523, "just-too-late" means?
23:15:04 <ais523> CLC-INTERCAL works by not compiling the program originally and just running it and getting an error. Then it compiles the bit of the code that caused the error and tries running again
23:15:11 <ais523> until eventually the whole program is compiled
23:15:12 -!- atrapado has joined.
23:15:20 <ais523> it's a bit like JIT compilation, just stupider
23:16:06 <AnMaster> ais523, didn't you say it compiled to some byte code that was then interpreted by perl?
23:16:08 <ais523> still, INTERCAL needs /some/ way around the fact that syntax errors are legal in a program
23:16:15 <ais523> AnMaster: yes, that's CLC-INTERCAL
23:16:27 <ais523> the bytecode for the compiler is mixed in with the bytecode for the program as far as I can tell
23:16:31 <ais523> and the program compiles itself as it's run
23:16:37 <ais523> from inside the bytecode
23:16:54 <ais523> I think it has to be done like that as the compiler can be dynamically modified from inside the program
23:19:07 <tusho> is an intercal interp possible
23:20:59 <ais523> probably easier than in intercal compiler, tbh
23:21:19 <ais523> as long as you can easily keep track of where statements begin and end, which is not trivial especially if it changes
23:21:28 <ais523> (it would be easy except that READOUT contains the substring "DO")
23:21:34 <tusho> COMPILER language with no pronouncable acronym
23:21:58 <ais523> well, I think all known implementations of it are compilers
23:22:05 <AnMaster> ais523, once you finish gcc-bf what about gcc-intercal?
23:22:11 <ais523> CLC-INTERCAL is sort of, C-INTERCAL definitely is, and J-INTERCAL compiles into Java bytecode
23:22:17 <tusho> AnMaster: gee, that's only the 5th time you've asked him that
23:22:30 <ais523> AnMaster: I've been thinking about it, especially as I was horrified to realise that ABSTAIN is expressible to some extent in gcc
23:22:34 <ais523> tusho: no, that's the first time
23:22:42 <ais523> presumably you get so used to the questions you don't actually read them...
23:22:53 <tusho> i'm sure he's asked something similar, at least
23:23:04 <AnMaster> ais523, how can you do ABSTAIN in gcc?
23:23:17 <ais523> AnMaster: it has a feature called conditional execution
23:23:22 <ais523> at the backend, not frontend
23:23:37 <ais523> which basically tells gcc that the target CPU has an instruction that causes it to ignore instructions for a while
23:23:58 <ais523> if is a goto forwards based on some condition
23:24:05 <ais523> whereas conditional execution blanks out commands
23:24:09 <ais523> the timing rules are completely different
23:24:15 <AnMaster> ais523, jump relative program counter?
23:24:36 <ais523> for instance on the PIC16F84, a bit-test instruction temporarily causes the next instruction to be a NOP if it fails
23:24:49 <ais523> the timing characteristics of that are very different from those of jumping forwards two instructions
23:24:54 <AnMaster> I think I remember that from PIC12F*
23:25:03 <tusho> #2C3300 is a nice colour
23:25:15 <ais523> let me try to imagine it first
23:25:20 <ais523> then I'll load it up and look
23:25:33 <ais523> let's see... a dark greenish-yellow?
23:25:54 <tusho> AnMaster: data:text/html,<body bgcolor=#2C3300>
23:26:18 <AnMaster> tusho, one of the shades of military green I'd say
23:26:26 <tusho> hmm, yes, possibly
23:26:36 <tusho> I just like the kind of grimey sort of feeling
23:26:55 <AnMaster> "No definitions were found for grimey."
23:27:18 <ais523> data:text/html,<html><body bgcolor='#2c3300'>
23:27:18 <ais523> and tusho: I like it when seen from above
23:27:18 -!- ais523 has quit (Remote closed the connection).
23:27:27 <tusho> that's some impressive lag.
23:27:33 -!- ais523 has joined.
23:27:48 <AnMaster> tusho: Did you mean: define:glamalicious
23:28:04 <tusho> ais523: what is it like seen from above?
23:28:11 <AnMaster> No definitions were found for glamalicious.
23:28:15 <ais523> well, my connection dropped
23:28:19 <ais523> AnMaster: wtf was the original question, to get /that/ reply?
23:28:23 <ais523> from below it's a sort of dull reddish uck
23:28:30 <tusho> ais523: ah, your screen is weird
23:28:31 <AnMaster> <AnMaster> "No definitions were found for grimey."
23:28:45 <AnMaster> now do anyone care to tell what the word means?
23:28:49 <tusho> it is meant to be uck
23:28:58 <ais523> but it isn't uck from above
23:29:14 <ais523> AnMaster: it isn't a real word, ordinary Googling from it returns a few pages which are using it as a made-up word
23:29:46 <tusho> glamalicious = glam
23:29:50 <ais523> it's glamalicious I was talking about
23:30:18 -!- CO2Games has joined.
23:30:45 <ais523> well, grimy is a real word
23:30:55 <ais523> I don't see why grimey would be considered anything other than a misspelling of it
23:31:01 <ais523> as the pronounciation would be the same
23:31:33 <AnMaster> ais523, there are several words with different spelling, same pronunciation and different meanings
23:31:40 <AnMaster> in both Swedish and English I think
23:31:48 <AnMaster> can't think of one right now however
23:32:05 <ais523> poor and pour is one common one in English
23:32:09 <ais523> tusho: that's same spelling
23:32:14 <tusho> ais523: no it isn't
23:32:17 <tusho> one is duck and the other is duck!
23:32:23 <tusho> also, poor is not pronounced the same as pour
23:32:30 <ais523> tusho: it is in Birmingham
23:32:42 <tusho> yes and that is because birmingham is the official home of satan.
23:32:48 <ais523> and everywhere else too I think
23:32:49 <tusho> ha, what do you say to THAT
23:33:13 <ais523> see, I counteract your bold and stupid assertion with a bold and uncited counter-assertion
23:33:34 <ais523> tusho: you are left-handed, sad and have 6 eyes?
23:33:46 <ais523> hmm... that doesn't make sense, my smiley-fu is fading
23:33:58 <tusho> also, birmingham is the official home of satan
23:34:11 <ais523> tusho: no way will I believe that from a sad left-hander with 6 eyes
23:34:39 * tusho eats ais523 with his gigantic (50meters) feeth for questioning him
23:34:55 * ais523 tries hard not to wonder what a footh is
23:35:11 <tusho> ais523: don't think about a pink elephant
23:35:23 <ais523> I actually managed that for about 2 seconds
23:35:31 <ais523> and am managing it occasionally even now
23:35:42 <ais523> the problem being that a pink elephant just isn't something I can easily bring to mind
23:35:56 <ais523> whoops, I'd better go home
23:36:00 <tusho> ais523: don't think about a train
23:36:01 <ais523> almost missed my bus...
23:36:02 -!- ais523 has quit (Remote closed the connection).
23:36:05 <tusho> DON'T THINK ABOUT A BUSS
23:36:20 <tusho> AnMaster: that hoax?
23:38:01 <tusho> ijust crashed safari
23:38:28 <AnMaster> tusho, nice, send a bug report
23:38:53 <tusho> i think i caused an infinite loop of hover&display/unhover¬ display
23:38:59 <tusho> with some crazy css
23:39:30 <tusho> i think with firefox it'd cause an endless flicker of the element appearing/disappearing
23:40:08 <AnMaster> tusho, still possible to abort by moving the mouse elsewhere?