00:00:01 -!- bsmntbombdood has quit (Read error: 110 (Connection timed out)).
00:01:53 -!- oklowob has quit ("PJIRC @ http://webirk.dy.fi").
00:02:37 -!- bsmntbombdood has joined.
00:03:41 <ehird> I hope this ubuntu in a vm craps out soon
00:03:47 <ehird> I don't want to like it too much
00:04:00 <FireFly> Does anybody know how the sample data in wav files is created?
00:05:05 <fizzie> You can stick different kinds of data in a .wav, though.
00:05:21 <FireFly> I should've googled that.. I knew that I was using data type 1, eg. PCM >_<
00:05:47 <FireFly> Well, I'm going to generate the data, so I don't mind other types
00:05:51 <fizzie> Uncompressed .wav files are pretty simple, anyway.
00:05:55 <ehird> WHAT THE FUCK, I like gnome
00:05:57 <ehird> what's happening to me
00:06:23 <fizzie> ehird: You forgot to wear your velostat hat, now you're being abducted.
00:06:45 <ehird> http://hovika.ytmnd.com/
00:07:14 <ehird> (not the author of that site)
00:07:46 <FireFly> ...it feels like generating wavs isn't really the primary use of JS, though
00:09:00 <FireFly> I though a JS sound editor would be cool
00:09:02 <fizzie> What is strange is: when BitsPerSample==8, wav file samples are unsigned bytes; but when BitsPerSample==16, the format seems to be signed little-endian two-byte integers.
00:09:20 <ehird> FireFly: how do you embed it?
00:09:31 -!- KingOfKarlsruhe has quit (Remote closed the connection).
00:09:46 <FireFly> But I havn't actually played it in the browser yet :D
00:10:04 <FireFly> I base64'd it and played it with mplayer, seem to work
00:10:32 <FireFly> Some random chippy noise, but it does get the correct length/metadata
00:12:26 -!- AnMaster_ has joined.
00:12:32 <fizzie> The wigi says IE<=7 doesn't support data: URIs at all, and IE8 has a length restriction of 32k. That's very limitsy.
00:13:12 <FireFly> Don't destroy my visions :(
00:13:20 <fizzie> The wikipedia article also has the following gem, in "advantages": "When fine tuning trafic volue: an embedded Base64 encoded image < 666 Byes of original data results in less then 866 Bytes Base64 data. This become more efficent then transfering this image + HTTP overhead (666 + 200 = 866 bytes)."
00:15:14 -!- olsner has quit ("Leaving").
00:16:01 <fizzie> Someone's been -- http://sk89q.therisenrealm.com/playground/jswav/ -- doing that sort of stuff too. I wonder if any JavaScript demoscene prods have opted for something like that instead of just using music-in-a-file; perhaps not.
00:17:12 -!- calamari has quit ("Leaving").
00:18:51 <oerjan> those pesky time travelers
00:18:58 <fizzie> You can do better. It's not a sound editor in there, after all.
00:21:15 <ehird> This Ubuntu-in-a-vm is working a little too well.
00:23:55 <ehird> For a vm with only 512mb it's sure hobbling along
00:24:09 -!- AnMaster has quit (Connection timed out).
00:31:05 -!- FireFly has quit ("Later").
01:20:01 <kerlo> The English word "if" generally corresponds to conditional probability.
01:23:32 <ehird> kerlo: why are you obsessed with bayesian probability logic vs natural language
01:28:50 -!- Dewio has changed nick to Dewi.
01:55:53 <andreou> ehird: you should see windows xp in a vm
01:56:02 <ehird> i did that a few days ago
01:56:07 <ehird> it was fast when i stripped it down
01:56:12 <andreou> the speed advantage is noticeable, to say the least
02:14:47 -!- neldoreth has quit (Read error: 104 (Connection reset by peer)).
02:14:51 -!- neldoret1 has joined.
03:01:37 <kerlo> ehird: because AI and language happen to be interests of mine.
04:06:41 -!- neldoret1 has quit (Read error: 113 (No route to host)).
04:08:16 -!- oerjan has quit ("Good night").
04:45:32 -!- Sgeo has joined.
04:54:39 -!- GreaseMonkey has joined.
05:27:47 -!- Sgeo has quit ("Leaving").
07:01:19 -!- psygnisfive has joined.
07:14:09 -!- tombom has joined.
07:24:56 -!- olsner has joined.
07:45:54 -!- rodgort has quit ("Coyote finally caught me").
07:46:07 -!- rodgort has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
10:06:40 -!- GreaseMonkey has quit ("X-Chat -> http://xchat.org <- At least when I quit I don't look like a lamer").
10:26:36 -!- lament has quit (Read error: 60 (Operation timed out)).
10:29:53 -!- AnMaster_ has changed nick to AnMaster.
10:46:42 -!- neldoreth has joined.
10:54:55 -!- swistakm_ has joined.
10:54:55 -!- swistakm has quit (Read error: 104 (Connection reset by peer)).
13:01:05 -!- KingOfKarlsruhe has joined.
13:11:45 -!- KingOfKarlsruhe has quit (Remote closed the connection).
13:14:43 -!- KingOfKarlsruhe has joined.
13:42:21 -!- neldoreth has quit ("leaving").
13:43:01 -!- oerjan has joined.
13:43:40 -!- neldoreth has joined.
13:43:57 -!- neldoreth has quit (Client Quit).
13:44:24 -!- neldoreth has joined.
13:46:11 -!- neldoreth has quit (Client Quit).
13:46:49 -!- neldoreth has joined.
13:46:51 -!- neldoreth has quit (Client Quit).
13:47:27 -!- neldoreth has joined.
13:47:27 <oerjan> AnMaster: aww, poor cthulhu
14:00:40 -!- Judofyr has joined.
14:13:03 -!- Slereah_ has joined.
14:26:38 -!- Slereah has quit (Read error: 110 (Connection timed out)).
14:41:43 -!- KingOfKarlsruhe has quit (Remote closed the connection).
14:44:19 -!- Judofyr has quit (Remote closed the connection).
14:49:26 -!- oerjan has quit ("leaving").
15:12:11 -!- Judofyr has joined.
15:34:55 <AnMaster> Deewiant, fizzie: There? SUBR: What should happen when stack cell count for the call or return is equal to zero?
15:35:07 <AnMaster> and does mycology test it for zero
15:35:28 -!- BeholdMyGlory has joined.
15:54:27 -!- FireFly has joined.
15:57:11 -!- ehird has left (?).
15:58:42 -!- ehird has joined.
15:59:46 <ehird> andreou: catseye has been resurrected since like 2006
16:07:09 <andreou> it must've been my resurrection that took longer then
16:08:17 <andreou> nice to see it still keeps that abominable background image
16:09:36 <andreou> i had a dream the other night, about a nice form on a website that presented you, after a series of questions, the open source license of your unconscious choice
16:15:35 <ehird> andreou: there's tons of those
16:15:48 <ehird> also, which abominable background
16:18:11 <andreou> the optical illusion, http://catseye.tc/images/backgrounds/sinewhite.png
16:19:40 <ehird> andreou: err, what is it meant to illuse
16:19:46 <ehird> it looks just like a regular background to m
16:21:13 <andreou> may be the color or the curving, yet it always made me dizzy
16:36:37 <AnMaster> warning: assuming signed overflow does not occur when changing X +- C1 cmp C2 to X cmp C1 +- C2 <-- I wonder how to interpret this one... for while(n--) where I know n is >= 0
16:40:37 -!- FireFly has quit ("Later").
16:40:59 -!- FireFly has joined.
16:55:28 <lifthrasiir> i don't know why... perhaps because n = <someexpr>; in the loop merges into condition check (just guessing)?
16:58:14 <AnMaster> stack_push(ip->stack, stack_pop(tmpstack));
16:58:30 <AnMaster> that causes the warning on the line with "while (n--)"
16:59:18 <AnMaster> lifthrasiir, it seems to work anyway, just a bit confusing
17:00:45 <AnMaster> X +- C1 cmp C2 to X cmp C1 +- C2 that would mean in this case (n--) != 0 into "n != 1-0"?
17:03:39 <lifthrasiir> if (n-- != 0) transforms into (n -= 1, (n + 1) != 0) then it can be the case
17:04:02 <lifthrasiir> i don't know gcc's internal well but that's only possible interpretation
17:06:32 <AnMaster> well ok that transformation you suggested seem harmless in this case
17:15:51 -!- Hiato has joined.
17:24:17 -!- swistakm_ has changed nick to swistakm.
17:30:31 <AnMaster> __builtin_ia32_movntps(((float*)(void*)&cfun_static_space) + i*4,
17:30:31 <AnMaster> *((const v4sf*)(const void*)&fspace_vector_init));
17:31:00 <AnMaster> cast to void needed to shut up warnings about strict aliasing rules (they do not matter in this case)
17:32:28 <lifthrasiir> movntps is for copying aligned bytes? (i'm not good at x86 assembly)
17:33:23 <AnMaster> lifthrasiir, WC is write combining too, and doesn't fill cache. Since I'm just space filling the static array for the commonly used part of funge space around 0,0 here...
17:34:09 <AnMaster> Move Non-Temporal Packed Single-Precision Floating-Point
17:35:18 <AnMaster> lifthrasiir, I don't know if I told you, but I use a static array for the most commonly used area around 0,0 in funge space, then a hash table for storing stuff outside that
17:36:28 <AnMaster> lifthrasiir, and basically, storing 2 MB of spaces in the binary wouldn't be sane, so I fill it at startup. -ftree-vectorise vectorised the loop, but it still caused lots of L1 and L2 cache misses. Using a WC-style streaming store solved that.
17:36:48 <lifthrasiir> good, i'm using just hash table (that's how the dict works in python) and i feel i have to add memory block for efficient storage...
17:37:08 -!- neldoreth has quit ("Lost terminal").
17:38:17 <lifthrasiir> in current pyfunge space operation accounts for 50% of time or so, which is a huge bottleneck
17:39:00 <AnMaster> lifthrasiir, hum? I don't know how much this would actually help in python... Since some of this stuff is pretty low level.
17:39:04 <lifthrasiir> (other bottleneck being command dispatch, but it is plenty of slow codes)
17:39:36 <AnMaster> I can reason about C code and often figure out what is most effective, because C is a rather thin abstraction on top of asm really. While for a high level language that is much harder.
17:39:48 <AnMaster> Of course you should still use profilers in both cases :)
17:39:59 <AnMaster> (since you can't always reason about either C or asm)
17:40:28 <AnMaster> lifthrasiir, hm gcc compiles my command dispatching into a big jump table basically.
17:40:38 <AnMaster> Which is quite ok, but could be faster
17:40:42 <lifthrasiir> AnMaster: i'm just trying to revise storage mechanism, since pyfunge once had big bottleneck related to sysinfo's boundary calculation
17:42:41 <lifthrasiir> with respect to this current design is very inefficient, since it uses direct coordinate as key (e.g. {(0,0): 64, ...})
17:45:41 <lifthrasiir> now pyfunge use caching and shortcut for sysinfo command, but it still takes too much time
17:46:25 <AnMaster> lifthrasiir, oh boundary? I store it as two pairs of x.y
17:46:43 <AnMaster> lifthrasiir, used for wrapping too
17:47:16 <lifthrasiir> how to calculate exact boundary (as required for sysinfo)?
17:48:06 <ehird> *Main> let Cast a = Cast 3
17:48:06 <ehird> <interactive>:1:4:
17:48:07 <ehird> My brain just exploded.
17:48:09 <AnMaster> lifthrasiir, well every time I write to funge space I have a couple of if that checks if bounds would grow. There is a known issue with shrinking bounds currently. But ccbi has that too
17:48:09 <ehird> I can't handle pattern bindings for existentially-quantified constructors.
17:48:11 <ehird> (real error message)
17:48:28 <AnMaster> lifthrasiir, and thus mycology doesn't test it. in cfunge and ccbi bounds currently only grow
17:48:52 <AnMaster> lifthrasiir, those if compiles down to a couple of cmov btw :)
17:49:07 <lifthrasiir> AnMaster: oh, it seems pyfunge is correct for that case ;)
17:49:30 <lifthrasiir> pyfunge also uses inexact (only growing) boundary for wrapping, only updates to exact boundary if requested
17:50:03 <AnMaster> lifthrasiir, well a slow y would be a mess for HRTI test.
17:50:16 <lifthrasiir> and that update takes too much, that's why i consider new algorithm
17:51:00 <AnMaster> I guess if I had to scan I would scan inwards from the bounds in sysinfo to find the real ones, and while I was at it also update the wrap bounds to not have to scan as much next time
17:51:16 <AnMaster> + cache positions of non-space so I could just check next time if they were still non-space
17:51:31 <lifthrasiir> AnMaster: i know, and i ended up with doing update only if some put command can shrink the space
17:52:32 <AnMaster> didn't you say scan only in sysinfo not put?
17:52:59 <AnMaster> oh you mean invalidating cached inner bounds if a space is put in some edge cell?
17:53:30 <AnMaster> lifthrasiir, I assume you basically scan around the edge in a spiral inwards?
17:53:31 <lifthrasiir> (and sorry for slow answer, i cannot write so fast in english...)
17:53:53 <AnMaster> lifthrasiir, heh I can write faster in English than my native language (Swedish). Too much IRC I guess...
17:54:13 <lifthrasiir> i think i have to do so but not yet implemented, i'm testing with possible solutions right now
17:54:18 -!- Sgeo has joined.
17:54:53 <ehird> AnMaster: swedish and english is trivial compared to korean and english I imagine
17:55:01 <ehird> er, I assume .kr is korean
17:55:11 <AnMaster> ehird, oh I didn't know where he was from
17:55:22 <AnMaster> and I wasn't implying anything like that
17:55:23 <ehird> AnMaster: whois magick
17:56:03 <lifthrasiir> but i don't think that shouldn't be hard besides from nuance and idiom differences...
17:56:19 <ehird> well, korean and english seem really far apart to me
17:56:25 <ehird> but then I only know the latter
17:57:41 <AnMaster> lifthrasiir, cache bounds, invalidate only when writing a space in an edge cell (remember fingerprint writes too!) sounds like the best way. Also update wrap bounds when you calculate new cached bounds if they grow inwards, will mean you won't have to go over so many spaces when wrapping possibly, thus reducing unneeded fetching of spaces when you are about to wrap after you have done an y
17:58:29 <AnMaster> anyway such a thing may be in the next cfunge release too, but not this one that I'm about to make tomorrow
17:59:05 <lifthrasiir> AnMaster: i should deploy such solution someday but i have one homework due to tomorrow, two homeworks due to 04-02 etc. ;)
17:59:54 <lifthrasiir> i shouldn't see mycology before 04-02... pyfunge thing took up my time too much (see hg log! ;)
18:07:44 * AnMaster fixes an internal icc header heh
18:30:18 -!- Hiato has quit (Read error: 104 (Connection reset by peer)).
18:37:41 -!- neldoreth has joined.
18:39:00 -!- KingOfKarlsruhe has joined.
18:43:53 <ehird> wow, mark pilgrim has an entire post, including video, about some 2006 supertux release.
18:44:09 <ehird> that's like post-ironic blogmocking without the irony, post, or mocking
18:59:26 <ehird> lol, guy in #haskell arguing visualworks smalltalk, which lets you view the source if you sign an NDA, is open source
19:11:42 -!- comex has quit (Remote closed the connection).
19:21:08 -!- BeholdMyGlory has quit (Remote closed the connection).
19:21:52 <ehird> I just bound tons of unicode shit to keys ☺
19:22:55 <ehird> can't do subscripts
19:23:09 <ehird> shift ones don't work
19:23:17 <ehird> so no blackie either → ☻
19:23:58 <ehird> The key combos are ⌃⌥(key)
19:24:07 <ehird> So ⌃⌥0-9 give superscript numbers
19:24:31 <ehird> ⌃⌥- gives ☺, ⌃⌥= gives ☹ (paren keys)
19:25:42 <ehird> shift-num is meant to be subscript
19:25:57 <ehird> shift - is meant to be ol' unicode blackie → ☻
19:26:10 <ehird> fizzie would be proud
19:27:47 <ehird> For unicode haskellry
19:31:43 <fizzie> I had interrobang shortcut key too, but I spent all the alphabetics to get most of the greek characters there are. Don't have that keymap in use any longer, though.
19:32:07 <fizzie> Now I just pick from gucharmap when I need it, which is very uncomfortable and slow.
19:32:20 <ehird> compose :: (a -> b) -> (c -> a) -> c -> b
19:32:20 <ehird> compose f g = \x -> f (g x)
19:32:24 <ehird> compose ∷ (α → β) → (γ → α) → γ → β
19:32:26 <ehird> compose α β = λγ → α (β γ)
19:32:36 <ehird> fizzie: What did you need all Greek for, apart from unicode fauxskell?
19:32:59 <ehird> Wonder why my shift-y shortcuts aren't workin
19:33:34 <fizzie> I just thought that since I had λ there, I might as well have the others too. They would be more useful in equation-writing except that there it's just simple to use LaTeX \alpha and so on.
19:34:15 <fizzie> Does that sort of stuff work with a real house-cell implementation?
19:34:32 <ehird> Oh, you mean jewnicode haskell?
19:34:41 <fizzie> ∷ and → in place of :: and ->, yes.
19:34:46 <ehird> I think everything there works with GHC if you enable unicode apart from the funky one-character ∷
19:34:52 <ehird> Well, I think the arrows work.
19:35:01 <ehird> The greek stuff and the lambda does for certain.
19:36:10 <ehird> http://aralbalkan.com/2067 LOL WAT
19:36:14 <fizzie> MzScheme didn't even do λ without a macro.
19:36:26 <ehird> fizzie: Tch. Haskell is the only language for enterprise uni-coding.
19:38:00 <ehird> http://pastie.org/431713.txt ← My ~/Library/KeyBindings/DefaultKeyBinding.dict.
19:38:18 <ehird> ^ = ⌃ (control), ~ = ⌥ (option/alt), $ = ⇧ (shift).
19:38:24 <ehird> Patterns with $ in don’t work for some reason.
19:38:29 <ehird> If anyone wants to fix it feel free :P
19:38:47 -!- KingOfKarlsruhe has quit (Remote closed the connection).
19:41:44 <ehird> I am addicted to ↑ already.
19:43:23 <fizzie> That's funny, this keymap does altgr-yui as ←↓→ by default. But I don't see up-arrow anywhere.
19:44:46 <fizzie> Ah, altgr-shift-u is ↑.
19:45:19 <ehird> I kind of have "further away = more obscure arrows"
19:45:24 <ehird> , and . are because of < and >
19:45:29 <ehird> k and l are "above"
19:45:31 <fizzie> Don't seem to have diagonals in this.
19:45:38 <ehird> then to the left a bit we have a nice 4x4 block yuhj
19:46:12 <fizzie> Yes, I first thought you had hjkl for the "obvious" arrow-directions, but seems that there were those 2x2 blocks.
19:46:27 <ehird> Unicode needs more smiliez
19:46:56 <fizzie> "This" keymap being XkbRules xorg, XkbModel pc105, XkbLayout fi, XkbVariant nodeadkeys, XkbOptions altwin:hyper_win,lv3:ralt_switch_multikey.
19:47:35 <ehird> US Extended + that keybinding here.
19:47:48 <fizzie> ⚇ is almost a smiley. Or ⚍.
19:47:55 <ehird> UK has shift-3 as £ and option-3 as #, which is back-asswards. US fixes that.
19:48:07 <fizzie> And ♾ is a guy with glasses on.
19:48:23 <fizzie> Even though it's "officially" U+267e permanent paper sign.
19:48:45 <ehird> ⍣ ← Butthole + Eyes
19:49:50 <ehird> ⁇ ← That’s actually a separate character.
19:50:22 <fizzie> I sort of toyed about making a ^3d6 command to fungot which would return something like "⚄ + ⚁ + ⚂ = 10" but there are only normal six-sided die faces in there.
19:50:26 <FireFly> Hm, I only knew about the double-bang
19:51:11 <FireFly> The there's the religious dingbats, hazardous signs, chess pieces
19:51:28 <ehird> is there a uncidoe for this?
19:51:39 <ehird> like ERROR CORRECTION ASTERISK
19:51:59 <fizzie> You can use "★foo" if you want to be "different".
19:52:14 <fizzie> Or even ☄foo for maximum sillitude.
19:52:21 <fizzie> (That's BLACK STAR and COMET.)
19:52:39 <ehird> fizzie: But that's not ^Isematically correct^I
19:52:52 <ehird> Maybe TIP-EX CORPORATE LOGO.
19:53:46 <ehird> DF’s logo is star-cut-out-of-circle.
19:53:51 <fizzie> "†foo" is sometimes used for corrections, but for best results you'd have to be able to stick that † also to the place you're correcting.
19:53:54 -!- swistakm has quit ("Lost terminal").
19:54:00 <fizzie> It looks a bit t-like, too.
19:54:02 <ehird> We should make our own Unicode, with blackjack and hooker.
19:54:19 <ehird> fizzie: ooh, maybe something like an overline, but an overstrike
19:54:25 <ehird> it goes way above the baseline
19:54:30 <ehird> so it goes through the middle of the previous line
19:54:41 <ehird> so just pop some spaces in, put them in, and put your correction in
19:55:16 <ehird> You could script s// to do that
19:57:16 <ehird> Anti-knuth↓↓anti-arrows
19:57:54 <fizzie> You can also use * for small corrections, ⁑ for really bad typos, and ⁂ for just completely hopeless cases.
19:58:08 <lifthrasiir> how about ideograph description sequence (IDS) to get really complex and useless ideographs?
19:58:14 <ehird> heleo i am fcugking drnku
19:58:22 <ehird> ⁂Greetings, I am here to inform you that I am currently intoxicated.
19:58:42 <ehird> ⁑Erm. Also, Fornication.
19:59:04 <fizzie> It's a bit sad that ⁂ has a funky name (asterism) but ⁑ has to get by with just "two asterisks aligned vertically".
19:59:21 <ehird> It’s the funkay new religion.
19:59:34 <fizzie> ⁂! I just had an asterism!
19:59:58 <FireFly> ..I read "Greek Extended" as "Geek Extended"
20:00:03 <ehird> “I’ll asterism your genitals. If you know what I mean.”
20:00:57 <ehird> those are bound by default. Option-[ and option-].
20:01:18 <fizzie> x⃣ ← the combining keycaps are also funky, but combining-character-rendering is always so iffy.
20:01:20 <FireFly> Too bad the alt gr mapping got all mixed up when I switched to Swedish Dvorak :(
20:01:20 <ehird> ] is single quotes.
20:01:24 <ehird> fizzie: Woah. Awesome.
20:01:30 <Sgeo> !glass {M[m(_m)M!(_m)m.?]}
20:01:52 <Sgeo> Well, what would the code do? Crash?
20:02:11 <ehird> Sgeo: Er, what does it do?
20:02:16 <ehird> Call itself a lot?
20:03:45 -!- tombom has quit ("Peace and Protection 4.22.2").
20:04:49 <ehird> And so on, and so forth, and they all lived happily ever after.
20:05:19 -!- Judofyr_ has joined.
20:07:10 <fizzie> FireFly: That's what it is here too; I mentioned that.
20:08:09 <ehird> Lunix enterprise operation system
20:10:10 -!- BeholdMyGlory has joined.
20:15:10 -!- BeholdMyGlory has quit (Remote closed the connection).
20:15:33 -!- BeholdMyGlory has joined.
20:15:41 -!- mtve has quit (calvino.freenode.net irc.freenode.net).
20:15:43 -!- FireFly has quit (calvino.freenode.net irc.freenode.net).
20:15:43 -!- Asztal has quit (calvino.freenode.net irc.freenode.net).
20:17:24 -!- FireFly has joined.
20:17:24 -!- Asztal has joined.
20:17:24 -!- mtve has joined.
20:20:50 -!- Sgeo has quit ("Leaving").
20:23:08 -!- Judofyr has quit (Read error: 110 (Connection timed out)).
20:23:53 -!- Hiato has joined.
20:27:08 -!- comex has joined.
20:29:37 -!- kar8nga has joined.
20:38:59 -!- kar8nga has quit (Remote closed the connection).
20:39:16 -!- neldoret1 has joined.
20:42:19 -!- AnMaster_ has joined.
20:47:09 -!- neldoreth has quit (No route to host).
20:49:28 -!- neldoret1 has quit (Client Quit).
20:49:39 -!- neldoreth has joined.
20:49:41 -!- neldoreth has left (?).
20:54:27 -!- AnMaster has quit (Connection timed out).
20:58:54 -!- Hiato has quit ("Leaving.").
21:03:11 -!- tombom has joined.
21:50:37 -!- nooga has joined.
21:51:08 <nooga> developing a kernel under windows is so much pain in the ass
21:55:00 -!- neldoreth has joined.
21:59:34 -!- tombom has quit ("Peace and Protection 4.22.2").
22:07:36 <Deewiant> AnMaster_: (Re: SUBR) If a subroutine takes zero parameters then of course none are pushed onto/popped from the stack...
22:08:06 -!- AnMaster_ has changed nick to AnMaster.
22:11:28 <AnMaster> freenode seems unstable today. I didn't loose connection to any network, so it is not my bouncer as ehird likes to pretend usually
22:12:06 <ehird> Unicode so needs proper symbols for correction.
22:13:12 <Deewiant> ㊣ - CIRCLED IDEOGRAPH CORRECT?
22:13:31 <ehird> Why doesn’t Unicode have useful things‽
22:17:55 <fizzie> You can use ⎀ if your correction happens to be a missing "a", but you have to know in-advance where you'll be needing it, since you need to put it on the previous line.
22:18:26 <fizzie> Maybe that's a bit limited.
22:18:27 <AnMaster> <nooga> developing a kernel under windows is so much pain in the ass <-- why do it then?
22:18:42 <Deewiant> Because installing an OS is also a pain in the ass?
22:19:23 <ehird> Deewiant: Er, it is?
22:19:27 <ehird> Fooled me you could have.
22:19:48 <AnMaster> ehird, I have to agree with your sarcasm there
22:20:06 <ehird> I installed Ubuntu fully in ~20 minutes. It gets harder if you move data across, but not much: stick them on USB sticks or something.
22:20:06 <AnMaster> I never found OS installing (apart from windows) to be painful
22:20:14 <ehird> Or, y'know, have data on a different partition.
22:20:23 <AnMaster> yes windows is painful, a lot thanks to microsoft minutes being used
22:20:31 <ehird> <3 microsoft minutes
22:20:43 <ehird> AnMaster: Well I found installing xp in a vm quite easy.
22:20:51 <ehird> took about 40 mins
22:20:53 <Deewiant> You have to repartition a hard drive, which possibly involves getting another from somewhere, and I don't consider a freshly-booted OS fully installed yet.
22:21:02 <AnMaster> ehird, a pain with all the registering and activation thingies
22:21:05 <ehird> repartitioning a harddrive is trivial
22:21:07 <ehird> AnMaster: oh well yeah
22:21:29 <AnMaster> ehird, but windows/microsoft update is way more painful I agree than the actual install
22:21:42 <fizzie> U+23E5 FLATNESS: ⏥. Certainly, that is the concept of flatness, compressed into a single symbol.
22:21:54 <ehird> fizzie: That’s not… well… flat.
22:22:01 <ehird> It’s poking upwards.
22:22:11 <fizzie> You probably have to just look at it in the right way.
22:22:13 <AnMaster> ehird, I only reboot for kernel upgrade. I even do glibc upgrades in runlevel <whatever it is gentoo use for normal operation, 3 I think>
22:22:21 <fizzie> Become one with the flatness, you know.
22:22:21 <ehird> AnMaster: Remember unpatched XP? Anyone on the internet could cause a dialog to pop up on your computer if they knew your IP.
22:22:24 <ehird> You got one every 10 seconds or so.
22:22:35 <ehird> All ads for shady antiviruses and whatnot.
22:22:54 <ehird> AnMaster: there's a command meant to do that over a LAN
22:22:59 <ehird> messaging service or something
22:23:05 <ehird> it was left open to any networkable address
22:23:17 <fizzie> That net-messaging thing was in wfw3.11 already, although I think you had to run a program to make it work.
22:23:42 <ehird> AnMaster: loads of shady sites sold tools that basically mined for ip addresses then spammed them with that
22:23:47 <Deewiant> AnMaster: Made a .bazaar/bazaar.conf which allowed me to get cfunge: it runs through Mycology in 00.01elapsed, CCBI takes 00.10
22:23:49 <ehird> marketed as "new age marketing tool", that sort of shite
22:24:15 <ehird> Deewiant: time varies wildly
22:24:19 <Deewiant> 0.00user 0.00system 0:00.01elapsed 90%CPU (0avgtext+0avgdata 0maxresident)k
22:24:19 <Deewiant> 0inputs+0outputs (0major+2108minor)pagefaults 0swaps
22:24:20 <ehird> (Lunchtime doubly s——
22:24:32 <AnMaster> Deewiant, I never seen time output that
22:24:37 <ehird> Two dashes = I was just shot.
22:24:53 <ehird> 0.00 real 0.00 user 0.00 sys
22:24:58 <ehird> No, I am sparctacus!
22:25:24 <fizzie> That almost looks like the tcsh "time".
22:25:28 <AnMaster> the time I use is the bash builtin
22:25:29 <ehird> AnMaster: /bin/time, maybe?
22:25:36 <ehird> Everyone has an external time.
22:25:44 <ehird> I believe it's required.
22:25:52 * ehird 's time is BSD time
22:25:58 <ehird> It uses BSD minutes
22:26:21 <fizzie> 0.00u 0.00s 0:00.00 0.0% +k(k,^:k) +io +sock pf+w sig cs is what tcsh's time outputs.
22:26:24 <Deewiant> It's http://www.gnu.org/software/time/time.html
22:26:55 <Deewiant> It gives me the count of page faults and other pointless stats so I find it fun
22:27:14 <fizzie> Solaris' external time (/usr/bin/time) is "\nreal 0.0\nuser 0.0\nsys 0.0\n" with a bit more whitespace.
22:27:31 <fizzie> usage: time [-p] utility [argument...]
22:27:36 <ehird> AnMaster: ) echo --help
22:27:55 <ehird> (Yes, "echo" is external, in Fish)
22:28:07 <fizzie> This GNU time has: Usage: /usr/bin/time [-apvV] [-f format] [-o file] [--append] [--verbose] [--portability] [--format=format] [--output=file] [--version] [--quiet] [--help] command [arg...]
22:28:12 <nooga> http://img12.imageshack.us/img12/8484/dupkax.jpg how does it look like?
22:28:12 <Deewiant> GNU time also has a -v option which makes it spout over 20 lines of data
22:28:26 <ehird> AnMaster: my prompt is (compressed pwd) in green, where compressed = ~/foo/bar/baz becomes ~/f/b/baz
22:28:33 <AnMaster> Deewiant, I prefer callgrind, cachegrind, gprof and gcov when I need such things
22:28:42 <ehird> nooga: what typeface is that?
22:28:55 <ehird> The whole thing looks nice apart from the colours of "rad".
22:28:58 <ehird> They're kind of clashy.
22:29:07 <ehird> Also, the bottom of the d's tail looks odd.
22:29:10 <ehird> Apart from that I really like it.
22:29:35 <AnMaster> I don't like the shape of the r really
22:29:44 <nooga> ehird: circumsized Myriad pro
22:30:12 <ehird> nooga: This is the first time I’ve ever heard “circumsized” applied to a typeface. This is a historic moment.
22:30:24 <ehird> Actually my manly self protection reflexes are kicking in round about now.
22:30:37 <ehird> Don’t do that. :-P
22:30:38 <nooga> probably an artefact
22:30:50 <AnMaster> Usage: /usr/bin/time [-apvV] [-f format] [-o file] [--append] [--verbose]
22:30:50 <AnMaster> [--portability] [--format=format] [--output=file] [--version]
22:30:57 <ehird> nooga: it looks like someone opened paint, selected a large block on green, and put it slightly off from the bottom of the d
22:32:21 <nooga> when i edit it in vector form it's okay
22:32:41 <ehird> nooga: just fix up the colours of that there rad. Also use png or svg dammit
22:32:42 <AnMaster> Deewiant, the bash builtin time is more accurate
22:33:09 <AnMaster> Deewiant, for cfunge it matters. it is so quick
22:33:24 <Deewiant> It doesn't matter, precisely because it is so quick
22:33:37 <Deewiant> If you get to the point that the wall clock time is 0.00s... you're done
22:33:48 <ehird> Deewiant: you’re done well before that
22:33:48 <Deewiant> All you have to do then is get a slower computer :-P
22:33:58 <ehird> btw, is anyone appreciating my “”‘’ quotes?
22:34:02 <AnMaster> Deewiant, cpufreq-set -g powersave ;P
22:34:03 <Deewiant> ehird: Well no, you're not 'done', you're just 'good enough'
22:34:03 <ehird> Because it’s quite a lot of work.
22:34:19 <ehird> Deewiant: This is a shit definition of “done”
22:34:26 <AnMaster> ehird, I didn't notice those quotes even...
22:34:36 <Deewiant> By 'done' here I mean there's nothing conceivable left to do
22:34:39 <ehird> AnMaster: Crank up the font size :-P
22:34:58 <ehird> …are you saying the quotes are fugly?
22:35:03 <ehird> Since when did good typography become *fugly*?
22:35:12 <ehird> Good. Now I don't have to kill you.
22:35:23 <ehird> AnMaster: You know, 16px on screen = 11pt in print.
22:35:35 <ehird> You should probably not strain your eyes.
22:35:43 <AnMaster> ehird, it is easy to read on this screen
22:35:51 <ehird> Only because you’re used to it.
22:35:59 <ehird> Ah well, let me know when you go blind ey.
22:36:05 <Deewiant> That's actually what I'm using too; I used to have it at 8 but I raised it to 9
22:36:33 <AnMaster> it was 10 before I got new glasses last time
22:36:52 <Deewiant> urxvt says pixelsize=12, maybe that's 9
22:37:07 <ehird> AnMaster: I’m going to have to demand you use the new industry standard: ↑
22:37:12 <ehird> Deewiant: Oh I thought we were talking pixels.
22:37:21 <Deewiant> ehird: I'm fairly sure it's pt.
22:37:26 <Deewiant> The 9, that is, that Vim uses.
22:37:33 <AnMaster> ehird, which open and free standard that is FSF supported and freely available are you referring to?
22:37:36 <ehird> 12px is… well, it’s still bad but I can make it out at least. ☺
22:37:53 <ehird> AnMaster: My butt’s newly decreed “IRC Unicode usage application”
22:38:12 <AnMaster> ehird, haven't even passed through ISO
22:38:27 <ehird> AnMaster: Dude, ISO accepted Microsoft’s OOXML.
22:38:35 <ehird> And you still listen to them‽
22:38:43 <ehird> (See that? That’s an interrobang.)
22:38:50 <fizzie> "URxvt.font: xft:DejaVu Sans Mono:size=8" nowadays here.
22:38:55 <ehird> Bit obvious isn’t it when you think about it
22:38:56 <AnMaster> ehird, I used a ‽ on irc recently
22:39:19 <ehird> http://site.ringce.com/products/slammer/slammer.html ← Oh fuck I need this a few months ago, pronto. No more baseline.png.
22:40:15 <ehird> AnMaster: a png of (baseline size in ems converted to pixels (the size when viewed on my display))px height and 1px width with one gray pixel at the bottom.
22:40:26 <ehird> So I can design to a baseline, see.
22:41:44 <AnMaster> what are they for? baseline product I heard of but I never heard of it in what I guess (since it is you) is typography
22:42:01 <ehird> AnMaster: The baseline is the line height, essentially.
22:42:08 <ehird> So: vertical size of font + line spacing.
22:42:27 <AnMaster> ehird, I let LaTeX worry about that :D
22:42:45 <ehird> LaTeX doesn’t really have enough line height by default IME
22:42:53 <ehird> But then I'm a readability stickler to the max
22:43:04 <AnMaster> ehird, I'd suggest 2-3 lines per page
22:43:05 <ehird> AnMaster: that’s not letting latex worry about it
22:43:30 <ehird> also, 1.75em line height works best IME
22:43:39 <ehird> so quite close to double spacing
22:44:38 <AnMaster> I like two columns per page layouts
22:44:52 <ehird> AnMaster: those are awful on screen, your eyes blit about so much
22:44:55 <ehird> in a book it works
22:45:08 <AnMaster> ehird, hm works well for me in either case.
22:45:52 <AnMaster> I fail to translate blitting as in GPUs to eyes
22:45:54 <ehird> Moving quickly. You can gleek the meaning from context if you’re a native speaker.
22:46:17 <AnMaster> and I'm no native speaker you know..
22:47:08 <nooga> but probably my english is even worse than AnMaster's ;]
22:47:43 <ehird> nooga: yes, well, I’ve tried to instill got vs has and suchlike in him so you can’t blame me
22:47:45 <AnMaster> ehird, yes it is. See *my* butt's newly decreed "IRC Accessibility and Internationalisation Guidelines"
22:47:51 <ehird> the english lobe of his brain is immutable
22:48:09 <ehird> (1) the symbol for “bottom” in computer sicence (really)
22:48:24 <ehird> (2) An ASCII art picture of a butt, if you put parens around it
22:48:29 <ehird> (3) A middle finger sticking up
22:48:35 <ehird> In short, 3 facets of relevance.
22:48:35 <fizzie> "This gubblick contains many nonsklarkish English flutzpahs, but the overall pluggandisp can be glorked [sic] from context"
22:49:25 <AnMaster> fizzie, it is saying the line can be read even though it is confusing
22:50:19 <fizzie> ehird: Given the recent jew-knee proliferation, that really should've been ⊥.
22:51:33 <AnMaster> I wonder why wikipedia in general looks so bad in konq, most other sites work well-ish in it
22:51:47 <ehird> fizzie: I have a patent on the “jew” nomenclature for referring to Unicode thankyouverymuch.
22:51:47 -!- ehird has left (?).
22:51:47 -!- ehird has joined.
22:51:47 -!- ehird has left (?).
22:51:50 -!- ehird has joined.
22:51:54 <fizzie> I wonder how a CURLY LOGICAL AND (⋏) differs from your garden-variety logical and. You know, besides the fact that it's curly.
22:52:33 <Asztal_> It makes a nicer pattern: ⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏⋏
22:52:35 <AnMaster> fizzie, I would have assumed that curlyness was a function of font
22:53:45 <ehird> fizzie: I have a patent on the “jew” nomenclature for referring to Unicode thankyouverymuch.
22:53:49 <ehird> Also fuck my irc client
22:53:58 <fizzie> I've seen ≺ (U+227A PRECEDES) used, and that's pretty much a curly less-than. I've forgotten the context; it was some sort of ranking that wasn't really an ordering.
22:55:33 <Asztal_> I'm confused as to the need for both ≸ and ≹.
22:55:35 <ehird> CAN YOU HEAR ME AM HAVE INTERNET PROBLEM
22:55:55 <fizzie> ehird: Yes, I saw that patent thing twice.
22:56:36 <fizzie> Oh. Well, I don't think there's a curly caret.
22:56:49 <fizzie> AnMaster: But there's ≻ of course.
22:57:30 <fizzie> There's also an extra-curly <, the ⊰.
22:58:53 <fizzie> AnMaster: "precedes under relation". And similarly ⊱ for s/precedes/succeeds/.
22:59:05 <AnMaster> fizzie, what about downwards curly arrow?
22:59:35 <AnMaster> ehird, we haven't noticed anything...
22:59:40 <ehird> That would be the bouncer.
22:59:57 <ehird> Alt-shift-; produces №
23:00:02 <ehird> Which then, given a number, does that
23:00:05 <fizzie> LaTeX has \prec and \succ for the normally curly versions, but I'm not seeing the extra-crispy, I mean, curly ones, not even in the AMS set of binary operations.
23:00:05 <ehird> Except when it doesn’t.
23:01:44 <AnMaster> ehird, AlgGr-shift-; == ˛ here
23:02:14 <ehird> AnMaster: This is a “dead key”, sort of.
23:02:30 <fizzie> My ˛ is not dead, but that's probably because of the nodeadkeys option.
23:03:09 <ehird> Like how Option-e gives ´, but with an orange background, and pressing “e” after gets you é. But you can just left-arrow away from it.
23:03:09 <ehird> And it deactivates.
23:03:43 -!- FireFly has quit ("Later").
23:04:05 -!- FireFly has joined.
23:04:48 -!- nooga has quit (Client Quit).
23:06:13 <ehird> http://pckeyboards.stores.yahoo.net/adcolkey.html APL keyboard
23:06:19 <ehird> AnMaster: Now back from my pilgrimage of disconnection,
23:08:56 <ehird> AnMaster: http://pastie.org/431713.txt
23:08:58 <ehird> Damn disconnection.
23:10:33 <AnMaster> ehird, all your bouncers fault btw
23:11:54 <ehird> it’s my connection since adium is fucking up and my bouncer doesn’t punish me for freenode’s faults
23:15:48 -!- FireFly has quit ("zzzz").
23:19:56 <ehird> http://packages.debian.org/unstable/admin/vrms
23:21:29 -!- BeholdMyGlory has quit (Read error: 104 (Connection reset by peer)).
23:21:38 -!- oerjan has joined.
23:21:53 <fizzie> It's funny to get a monthly vrms report bitching about the fact that I have autoconf-doc and gdb-doc installed.
23:22:18 <ehird> http://packages.debian.org/unstable/admin/vrms
23:22:23 <ehird> fizzie: why did you install it
23:22:31 <ehird> I linked to it because it was ridiculously silly
23:22:32 <fizzie> ehird: "For the lulz."
23:22:47 <ehird> fizzie: well I hereby revoke your lulz relating to vrms
23:23:03 <fizzie> I think I disabled the cron job, though, so it won't email me every month, but I can still run it when I feel like it.
23:23:10 <fizzie> Quite a pile of non-free stuff.
23:23:26 <ehird> fizzie: Are you lucky enough to have a gfx card with decent free drivers?
23:23:27 <fizzie> Mostly it's -doc packages, though.
23:23:37 <ehird> You’re part of a community of 4, I guess.
23:24:01 <olsner> ah, the virtual richard m stallman
23:24:04 <fizzie> No; I have nvidia-glx, nvidia-glx-dev and four nvidia-kernel-* packages in the vrms list.
23:24:07 <oerjan> ehird: i am just amazed that vrms was about exactly what i guessed it was :D
23:24:47 <ehird> fizzie: I wonder if you can buy a 100% free computer? LinuxBIOS and all?
23:25:16 <fizzie> But I have even more -doc packages: autobook (well, technically that's not -doc), autoconf-doc, automake1.9-doc, gcc-4.1-doc, gcc-4.3-doc, gcc-doc-base, gdb-doc, ocaml-doc.
23:25:27 <fizzie> I'unno, sounds unlikely.
23:25:42 <ehird> fizzie: Bloody GFDL
23:26:54 <oerjan> fizzie: why would vrms complain about -docs?
23:27:06 <ehird> oerjan: gfdl’s “invariant sections”
23:27:26 <ehird> They’re bits you can’t change; invented for emacs so that nobody could remove RMS’s masturbatory philosophical sludge.
23:27:36 <ehird> (aka the free software manifesto)
23:27:48 <ehird> This is, of course, non-free.
23:27:55 <fizzie> oerjan: Yes, VRMS is perhaps not completely aligned with RRMS.
23:28:22 -!- Judofyr_ has quit (Remote closed the connection).
23:31:10 <oerjan> what the heck my watch is playing an early AF prank on me
23:31:32 <oerjan> somehow i've set the date one day late
23:32:00 <Asztal> it's always 1 day ahead :(
23:32:58 <ehird> in 28 minutes it’ll be correct!
23:33:22 <oerjan> darn this could be a problem to fix, i may have to turn it 30 days
23:33:50 <oerjan> since the other direction changes weekday instead of turning it backwards
23:34:50 <oerjan> hm maybe 24 hours backwards is easier
23:36:16 <oerjan> i wonder if this somehow happened when i changed to DST. it cannot have been this way for the whole month because i didn't miss my dentist's appointment...
23:41:46 <oerjan> <ehird> nooga: This is the first time I.ve ever heard .circumsized. applied to a typeface. This is a historic moment.
23:41:54 <ehird> oerjan: Unicode failure
23:42:08 <oerjan> you may be thinking of circumcised
23:42:39 <oerjan> why the heck would people do funny quotes in irc anyway?
23:43:16 <ehird> “Funny” quotes? Dear sir, read a goddamn book and tell me what quotes you see.
23:43:23 <ehird> And we come to the answer: pedants.
23:44:09 <oerjan> ah yes, like mac users
23:44:20 <oerjan> <ehird> nooga: This is the first time I.ve ever heard .circumsized. applied to a typeface. This is a historic moment.
23:44:33 <ehird> oerjan: Fix yer damn terminal, peasant.
23:45:01 <fizzie> «I prefer quotes like this, mainly to be contrary.»
23:45:16 <ehird> Asztal_: Dude, your quotes got enlarged with silicone.
23:46:50 <fizzie> Also, »+« is a Perl 6 operator: (1,1,2,3,5) »+« (1,2,3,5,8); # (2,3,5,8,13)
23:47:05 <ehird> fizzie: isn't it »anything«
23:47:17 <fizzie> Maybe not exactly anything, but most things, certainly.
23:47:17 <ehird> meaning “emulation of vector operation”, I guess.
23:47:22 <ehird> fizzie: any operator
23:47:38 <ehird> I think the syntax is »op«
23:47:48 <fizzie> I don't think you can use the hyper operator with ?? ::, for example.
23:48:32 <oerjan> <ehird> btw, is anyone appreciating my .... quotes?
23:48:34 <ehird> 23:48 ehird: @type \f a b -> map (uncurry f) $ zip a b
23:48:34 <ehird> 23:48 lambdabot: forall a b c. (a -> b -> c) -> [a] -> [b] -> [c]
23:48:41 <ehird> oerjan: Fix yer terminal!
23:48:49 <fizzie> It is all so complicated. $tree.».?foo; # short for $tree.?foo, $tree.each: { .».?foo }
23:49:12 <ehird> fizzie: Perl 6 is pretty lame
23:49:22 <oerjan> i think something went wrong the time i tried
23:50:33 <ehird> I was just reimplmenting that perl6 thang
23:52:38 -!- neldoreth has quit ("leaving").
23:52:45 -!- neldoreth has joined.
23:52:56 <oerjan> also, luckily irssi seems to simplify quotes on incoming, it's just a problem when i paste from the logs
23:53:39 <Asztal_> it may be detecting utf-8 and recoding it
23:54:30 <fizzie> Well, if «this» is fine, then I prefer to quote with ‷(reversed and not) triple prime‴.
23:55:48 <oerjan> hm irssi got only half of that
23:56:11 <fizzie> There is no reversed quadruple prime, only the right way around. :/
23:56:13 <ehird> fizzie: “And”? There’s a unicode symbol for that.
23:57:02 <Ilari> Echoback (using Irssi here): < fizzie> Well, if «this» is fine, then I prefer to quote with ‷(reversed and not) triple prime‴.
23:57:05 <oerjan> at least two i would assume
23:57:28 <ehird> Ilari: that’s right
23:58:43 <Ilari> I should get better unicode font from somewhere... The one that I'm using doesn't even have less-or-equal/greater-or-equal. It also seemingly doesn't have any astral characters.