00:33:52 -!- KingOfKarlsruhe has quit (Remote closed the connection).
00:34:48 -!- oerjan has quit ("I'll take that as a 'No'").
00:45:45 <AnMaster> psygnisfive, about "fedora or bowler hat"
00:48:51 -!- olsner has joined.
00:52:00 -!- AnMaster has quit ("ERC Version 5.3 (IRC client for Emacs)").
00:57:42 -!- olsner has quit ("Leaving").
01:53:13 -!- Enki-][ has joined.
01:53:50 -!- Enki-][ has left (?).
02:32:08 -!- GreaseMonkey has joined.
03:26:54 -!- lifthrasiir has quit (calvino.freenode.net irc.freenode.net).
03:26:54 -!- dbc has quit (calvino.freenode.net irc.freenode.net).
03:27:03 -!- dbc has joined.
03:27:03 -!- lifthrasiir has joined.
03:51:28 -!- CO2Games has joined.
04:04:56 <CO2Games> Hmm I should make a brainfuck compiler
04:08:38 -!- calamari has joined.
04:17:27 -!- Enki-][ has joined.
04:23:31 <ihope> Everyone should make a brainfuck compiler.
04:24:38 <Enki-][> if anyone has a bot or can grab one
04:24:49 <Enki-][> #sumisu is full of bots chatting to one another
04:25:15 <Enki-][> i figure we should get some non-markov-chain bots (elizas or alices maybe) to try to put some sense into the mix
04:33:25 <calamari> "Having the Markov property means that, given the present state, future states are independent of the past states. In other words, the description of the present state fully captures all the information that could influence the future evolution of the process. Future states will be reached through a probabilistic process instead of a deterministic one."
04:49:31 <Asztal> I had a bot, but being lazy I used a Winsock component that came with VB6 instead of using Sockets properly, and now I don't have VB6 :(
04:50:16 <Enki-][> visual basic... talk about an esolang!
04:51:02 <Asztal> the bot itself was C#, thankfully, which is a slight step up
04:52:50 -!- CO2Games has quit ("And I invented doors, no joke!").
05:46:26 -!- Sgeo has quit (Success).
05:49:02 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | Would you like to believe I was ?.
06:11:12 -!- GregorR-L has joined.
06:12:24 <GregorR-L> For Plof syntax naysayers: Plof's 'if' function is now called like so: if (condition) (code) else (code) (presumably with some newlines in there). The only way you could have complaints about this syntax is if you're hyper-insistent on using curly-braces. If you are, go away :P
06:19:10 <Asztal> what was it like before?
06:20:55 <GregorR-L> if((condition), (code), else, (code));
06:21:05 <GregorR-L> In fact, it's still like that, but I've swizzled the function-call syntax.
06:47:12 <GregorR-L> Now that I'm /officially/ working on PL it just makes me want to work on Plof that much more :P
06:52:24 <GregorR-L> Yes. PL/I is an extremely popular, modern programming language :P
06:58:11 <calamari> nice http://pl1gcc.sourceforge.net/
07:00:43 <GregorR-L> "There is still no code generation taking place, so don't run out and uninstall your production PL/I compiler just yet :-)"
07:03:20 <GregorR-L> Wikipedia claims that PL/I is still actively used today.
07:05:56 -!- Enki-][ has quit (Read error: 104 (Connection reset by peer)).
07:11:11 <pikhq> You're officially working on programming languages?
07:11:14 <pikhq> How bizarrely appropriate.
07:12:38 <calamari> we still use it to write mainframe software
07:17:11 <pikhq> OJ Simpson found guilty.
07:20:23 <GregorR-L> pikhq: Yuh, I'm a grad student now.
07:20:40 <pikhq> Didn't realise that was exactly what you were working on.
07:20:54 <pikhq> Though that's because I didn't think about it.
07:21:30 <pikhq> It's the most natural thing for you to be working on...
07:22:21 <GregorR-L> I could see myself in networking, but I'm hoping I can leverage that in PL instead.
07:23:18 <pikhq> Take off the tie; it's controlling you.
07:24:20 <GregorR-L> Yesterday I was proctoring an exam wearing a Do Not Put the Baby T-shirt and a fez :P
07:24:39 <pikhq> Dammit; makes me wish I was at Purdue.
07:25:19 * GregorR-L is also forming Purdue Extreme Croquet.
07:26:09 <GregorR-L> Are the people on the front page of mst.edu pointing in random directions?
07:27:40 <GregorR-L> Ohyeah: Everybody go buy a Pandora (www.openpandora.org), it would suck if they didn't make their preorder max.
07:27:51 -!- GregorR-L has quit ("Leaving").
07:28:43 <pikhq> Can't say; that picture is randomised.
07:34:59 -!- calamari has quit ("Leaving").
07:41:47 -!- moozilla has quit (Read error: 104 (Connection reset by peer)).
07:45:20 -!- moozilla has joined.
07:52:58 -!- kar8nga has joined.
07:55:27 <Asztal> I want a pandora now :(
07:55:41 <Asztal> but it's £199 compared to £129 for the GP2X Wiz
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:12:25 -!- Mony has joined.
08:31:27 -!- GreaseMonkey has quit ("So, how much do you love noodles?").
09:02:12 -!- kar8nga has left (?).
10:40:52 -!- AnMaster has joined.
10:46:40 -!- kar8nga has joined.
11:49:03 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | Define tonight.
12:40:24 -!- olsner has joined.
12:52:43 -!- slereah has joined.
12:52:44 -!- Slereah_ has quit (Read error: 54 (Connection reset by peer)).
13:20:20 -!- Hiato has joined.
13:26:21 -!- slereah has quit (Read error: 104 (Connection reset by peer)).
13:26:28 -!- slereah has joined.
13:39:43 -!- oerjan has joined.
13:41:26 -!- kt3k has joined.
13:50:56 <oerjan> GregorR: the count on your hats page has not been updated :D
13:54:26 <AnMaster> anyone know how to find out the size of the stack from inside gdb?
13:55:47 -!- Hiato has quit ("Leaving.").
14:15:08 -!- kar8nga has quit (Read error: 60 (Operation timed out)).
14:55:32 -!- kar8nga has joined.
15:05:31 -!- kt3k has quit (Read error: 104 (Connection reset by peer)).
15:06:57 -!- lifthrasiir has quit (calvino.freenode.net irc.freenode.net).
15:06:57 -!- dbc has quit (calvino.freenode.net irc.freenode.net).
15:07:01 -!- ihope has quit (calvino.freenode.net irc.freenode.net).
15:07:04 -!- Asztal has quit (calvino.freenode.net irc.freenode.net).
15:07:04 -!- pikhq has quit (calvino.freenode.net irc.freenode.net).
15:07:06 -!- rodgort has quit (calvino.freenode.net irc.freenode.net).
15:07:08 -!- Mony has quit (calvino.freenode.net irc.freenode.net).
15:07:08 -!- puzzlet_ has quit (calvino.freenode.net irc.freenode.net).
15:07:11 -!- ais523 has quit (calvino.freenode.net irc.freenode.net).
15:07:11 -!- GregorR has quit (calvino.freenode.net irc.freenode.net).
15:07:11 -!- bsmntbombdood has quit (calvino.freenode.net irc.freenode.net).
15:07:11 -!- cmeme has quit (calvino.freenode.net irc.freenode.net).
15:07:13 -!- moozilla has quit (calvino.freenode.net irc.freenode.net).
15:07:13 -!- psygnisfive has quit (calvino.freenode.net irc.freenode.net).
15:07:13 -!- Ilari has quit (calvino.freenode.net irc.freenode.net).
15:07:15 -!- oklopol has quit (calvino.freenode.net irc.freenode.net).
15:07:16 -!- olsner has quit (calvino.freenode.net irc.freenode.net).
15:07:18 -!- sebbu has quit (calvino.freenode.net irc.freenode.net).
15:07:18 -!- danopia has quit (calvino.freenode.net irc.freenode.net).
15:07:18 -!- optbot has quit (calvino.freenode.net irc.freenode.net).
15:07:18 -!- ehird has quit (calvino.freenode.net irc.freenode.net).
15:07:22 -!- SimonRC has quit (calvino.freenode.net irc.freenode.net).
15:09:12 -!- kt3k0 has joined.
15:09:12 -!- olsner has joined.
15:09:12 -!- Mony has joined.
15:09:12 -!- moozilla has joined.
15:09:12 -!- lifthrasiir has joined.
15:09:12 -!- dbc has joined.
15:09:12 -!- ihope has joined.
15:09:12 -!- puzzlet_ has joined.
15:09:12 -!- ehird has joined.
15:09:12 -!- sebbu has joined.
15:09:12 -!- psygnisfive has joined.
15:09:12 -!- Ilari has joined.
15:09:12 -!- danopia has joined.
15:09:12 -!- Asztal has joined.
15:09:12 -!- pikhq has joined.
15:09:12 -!- optbot has joined.
15:09:12 -!- ais523 has joined.
15:09:12 -!- oklopol has joined.
15:09:12 -!- GregorR has joined.
15:09:12 -!- bsmntbombdood has joined.
15:09:12 -!- rodgort has joined.
15:09:12 -!- cmeme has joined.
15:09:12 -!- SimonRC has joined.
15:11:28 -!- Slereah_ has joined.
15:11:28 -!- slereah has quit (Read error: 104 (Connection reset by peer)).
15:15:55 -!- M0ny has joined.
15:18:33 -!- slereah has joined.
15:18:48 -!- Slereah_ has quit (Read error: 113 (No route to host)).
15:23:26 -!- kt3k0 has quit (Remote closed the connection).
15:34:16 -!- Mony has quit (Read error: 110 (Connection timed out)).
16:10:42 -!- Slereah_ has joined.
16:11:15 -!- ENKI-][ has joined.
16:16:01 -!- slereah has quit (Read error: 113 (No route to host)).
16:16:09 -!- KingOfKarlsruhe has joined.
16:19:52 -!- oklopol has quit (Connection timed out).
16:28:25 -!- oerjan has quit ("Large road-based collective transport").
16:51:25 <GregorR> Asztal: Sure, but it's also substantially superior to the Wiz :P
16:53:34 <ehird> I submit the following proposal, titled "Have listing" (AI=1):
16:53:34 <ehird> [Major philosophical change: shorten "30 days" to self-installation to prevent
16:53:35 <ehird> the stayed Order has been judged".
16:53:35 <ehird> A message is public, the only mechanism by which rules can be required, so may
16:53:35 <ehird> as well, but after taking all other rules.
16:55:44 <GregorR> When did I become "gregorBR" ...
16:55:54 <ehird> gergorBR: No, gergorBR.
16:57:53 <ehird> GregorR: Who is GERGOR?
16:57:55 <ehird> I only know a gergor.
17:00:13 -!- kar8nga has left (?).
17:02:57 -!- oklocod has joined.
17:05:27 -!- M0ny has changed nick to Mony.
17:07:41 <oklocod> that's the greatest greeting i've *ever* gotten
17:14:25 <ehird> see, i addressed you by name
17:28:04 <AnMaster> psygnisfive, about who it was that designed that language thing
17:28:09 <AnMaster> I don't have the scrollback any more
17:28:46 <AnMaster> psygnisfive, the convo of what "or" meant in English
17:29:03 <psygnisfive> uh.. i dont get what you mean by who designed it
17:29:20 <AnMaster> psygnisfive, someone suggested making a language anywhere
17:29:26 <AnMaster> my question was: did that get anywwhere
17:29:38 <psygnisfive> i suggested we make a language with disjunction scope indicators
17:29:47 <psygnisfive> no it didnt get anywhere since we only mentioned it last night :P
17:29:59 <AnMaster> psygnisfive, but what plans do you have for it?
17:30:22 <psygnisfive> probably none but i'd like to experiment with it
17:30:56 <psygnisfive> like i was saying to oklociod, i think it'd work nicely along side the quantification and predication ideas i had a few months ago
17:37:35 <psygnisfive> id rather have your stuff if you know what i mean
17:39:23 -!- jix has joined.
17:45:43 -!- Hiato has joined.
17:48:26 -!- Hiato has quit (Client Quit).
17:49:03 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | bbl.
18:47:39 <AnMaster> how much overhead does a malloced block have on average on a 32-bit platform
18:48:04 <ehird> there should be a channel for hardware questions like that
18:48:09 <ehird> its kind of hardware
18:48:12 <ehird> and kind of software
18:48:16 <ehird> AnMaster: dependant on the malloc impl
18:48:32 <ehird> AnMaster: I don't think there's any standard.
18:48:46 <ehird> psygnisfive: The least helpful reply to a question is 'why?'.
18:48:52 <AnMaster> ehird, surely there is some average? Like "probably 8-16 bytes" or whatever
18:49:02 <ehird> In #esoteric we can at least assume the people have a reason for doing something.
18:49:10 <psygnisfive> ehird: it wasnt an answer to that question :P
18:49:10 <ehird> AnMaster: I don't really think so...
18:49:13 <AnMaster> and reason why is because I consider implementing a memory pool system
18:49:24 <ehird> psygnisfive: why did you ask why then
18:49:43 <psygnisfive> i was asking why he was asking about the thing earlier
18:49:50 <AnMaster> because valgrinds massif says I got almost half a MB of overhead, and the total memory usage is around 7 MB
18:49:53 <ehird> that's not very clear :P
18:50:25 <ehird> void mem[big_number]; size_t top = 0; void *malloc(size_t foo) { top += foo; return mem + top; } void free(void *foo) { }
18:50:45 <ehird> void mem[big_number]; size_t top = 0; void *malloc(size_t foo) { void *ptr = mem + top; top += foo; return ptr; } void free(void *foo) { }
18:51:03 <AnMaster> ehird, anyway I suspect I could reduce overhead here and yes I need low memory usage since I expect to operate on even larger data sets, so I could end up with overhead like 50 MB just for the bookkeeping data
18:51:06 <AnMaster> and that wouldn't be fun at all
18:51:11 <ehird> AnMaster: what's the project out of curiosity
18:51:49 <AnMaster> ehird, kind of closed currently, it will be open source in due time, but not for some time due to various circumstances out of my control
18:51:58 -!- oerjan has joined.
18:52:11 <ehird> AnMaster: I'm surprised you'd ever agree to an NDA. :-P
18:52:42 <AnMaster> ehird, however it is only temporary until certain other things are completed
18:52:56 <AnMaster> anyway since most of the allocations are fixed using a mempool would have less overhead I think
18:53:48 <ehird> AnMaster: Well I'm finding it hard to believe AnMaster ever agreeing to an NDA or similar so now I'm intrigued :-P. I'll be interested to see what it is when it's opened.
18:56:27 <AnMaster> I think I will have an overhead of sizeof(void*) for each object in the array, since I need to find free objects easily. The only way I can think of is having a single linked list embedded in the array. Freed objects are added to a list, The pool header contains a pointer to the first item in this linked list.
18:56:40 <AnMaster> for allocated objects, and for the last in the free list, this pointer is NULL
18:57:07 <AnMaster> only issue is I would need to initially add all objects to that free list
18:57:39 <AnMaster> unless I do something like switching allocation strategy when the last block is used
18:57:56 <AnMaster> keep a pointer to last allocated block
18:58:10 <AnMaster> allocate from that unless we reached the end of the array
18:58:21 <AnMaster> if we reached the end, then switch to allocate from the free list
18:58:28 <AnMaster> if freelist is empty, allocate a new pool
18:58:37 <AnMaster> does this sound like a good idea?
18:58:58 <AnMaster> This is the first time I try to do something like this so advice is welcome :)
18:59:08 <ehird> AnMaster: it sounds good but i have no idea about this stuff
18:59:24 <ehird> i can't think of a channel that might have people who know this kind of stuff, though
18:59:38 <ehird> its not C, it's not Linux... i mean, what is it, really
18:59:53 <AnMaster> ehird, also I had very bad memory fragmentation with malloc/free, due to allocating differently sized objects and freeing/mallocing is more or less random order
19:00:11 <AnMaster> so I will instead have mempools for the two sizes of objects I need
19:00:23 <AnMaster> and of course the smaller overhead
19:00:25 <ehird> couldn't you peek at some other memory pool system perhaps
19:00:28 <ehird> there are a lot of them
19:00:36 <AnMaster> ehird, hm like boehm-gc and such?
19:00:38 <ehird> yours sounds a bit overcomplicated to me but as i said i don't really know this stuff
19:00:48 <ehird> AnMaster: well, i know one quite often used app uses it
19:00:51 <ehird> but i do not recall its name
19:01:11 <AnMaster> however many try to be general to handle not exactly of size x but of range x-y
19:01:42 <ehird> AnMaster: i think yours sounds kind of more complicated than theirs but again i don't really know this stuff :-)
19:01:45 <AnMaster> <ehird> its not C, it's not Linux... i mean, what is it, really <-- memory allocation!
19:02:00 <ehird> AnMaster: somehow i doubt #memory-allocation would get many people :-P
19:02:39 <ehird> AnMaster: very observant
19:02:49 -!- metazilla has joined.
19:04:03 <AnMaster> ehird, just had to check in case
19:04:14 <oerjan> i never metazilla i didn't like
19:06:34 -!- kt3k has joined.
19:07:58 * oerjan wonders about the An in AnMaster's nick
19:11:31 <fizzie> I haven't read of all the context, but to me it would sound somewhat cleaner to always just give out the first item in the free-list; or if the list is empty, the next free entry in the last block; or if the last block is full, allocate a new one. That way your free-list will be marginally shorter than in the "fill the last block first" case.
19:12:02 -!- moozilla has quit (Connection timed out).
19:12:15 <AnMaster> fizzie, the issue is that I need to prepare freelist. Setting up the pointers initially is O(n) after all
19:12:27 -!- moozilla has joined.
19:12:54 <AnMaster> fizzie, free list is basically a linked list *embedded* in the array that these are allocated from
19:13:03 <AnMaster> since I want to avoid overhead of malloc
19:13:33 <fizzie> Yes, I know you can keep it there, but I see no reason why it needs to be prepared in advance if it starts out empty and you keep a separate "we have allocated this many objects from the last block" count.
19:14:07 <fizzie> When you free() the object you just need to stick the value of the current free-list pointer to wherever the free()d pointer points to, and update your current "start of free list" pointer to point there.
19:14:16 <AnMaster> fizzie, the array used for memory pool is malloced (of course), That means memory is undefined
19:14:36 <AnMaster> I would need to set the pointers of the "next free" to null
19:15:45 <fizzie> I don't see why. When you start, you set your top-level "next free" pointer to NULL, which means it will allocate from the end of the already-allocated blocks. When you free() a block, just stick the current "next free" value to the place you freed, and update "next free" to point there.
19:16:02 <fizzie> That way you'll end up with a singly-linked list of pointers, terminated by a NULL entry.
19:16:12 <AnMaster> fizzie, but that is more or less what I said :)
19:16:30 <fizzie> Huh? There's no setting-up pointers in advance, only when free()ing the element.
19:16:53 <AnMaster> but I said I would do basically what you said first
19:17:07 <AnMaster> fizzie, so how is mine more complex
19:17:53 <AnMaster> only difference is that I would allocate from end of the used blocks until I hit the end of the memory area, while you use free list as soon as possible
19:17:55 <ehird> http://www.google.com/trends <- Why is the Mormon "church"'s domain the #1 trend...???
19:18:14 <fizzie> AnMaster: Yes, that _is_ the only difference I was mentioning there.
19:18:23 <ehird> http://www.google.com/trends/hottrends?sa=X And with a www. in front, #20.
19:18:30 <ehird> Did they ask their members to googlebomb them or something?
19:18:44 <AnMaster> fizzie, So the only difference in the code is what entry in the struct I test for NULL ;P
19:20:00 <fizzie> AnMaster: I just think it's -- as I said, marginally -- cleaner to allocate from the pointers-all-around-the-place free list so that it goes away, instead of filling the last memory block completely first.
19:20:42 <fizzie> But the idea itself sounds good, though terribly non-esoteric.
19:20:57 <AnMaster> fizzie, however for point of memory fragmentation it doesn't matter, since all objects in the array are the same sizer
19:21:07 -!- metazilla has quit (Read error: 110 (Connection timed out)).
19:21:16 <AnMaster> fizzie, anyway yes but I couldn't think of another channel to ask
19:21:46 <fizzie> I think mooz, who used to hang around here writing befunge stuff, wrote a very nice fixed-size memory pool in C. Don't remember the details, but at least there were some similiarities.
19:21:58 <fizzie> I don't think I have a copy of it any more, though.
19:22:02 -!- timmytiptoe has joined.
19:23:03 <ehird> fizzie: being here since the start, what year would you say #esoteric was most active in?
19:24:36 <fizzie> Compared to the long-term average, it has certainly felt pretty active these last few months.
19:24:47 <AnMaster> ehird, easy to find using logs
19:25:01 <ehird> i mean actual activity
19:25:04 <ehird> not 'ooh, this place is dead'
19:25:20 <AnMaster> ehird, just check actual messages in the log
19:25:30 <ehird> AnMaster: 'ooh, this place is dead' is an Actual Message.
19:25:33 <AnMaster> but that doesn't fix the "what a dead place"
19:25:35 <ehird> actual activity is subjective
19:27:30 -!- timmytiptoe has quit.
19:29:20 <oerjan> ooh, this place is dead
19:29:41 <fizzie> Well, the byte sizes of the logs _do_ indicate _something_ about actual activity, and here's a quick-and-dirty GNUplot plot, even though the default options suck a bit: http://zem.fi/~fis/eso.png
19:29:48 <AnMaster> oerjan, full of life, bacterias living on rust
19:30:09 <ehird> fizzie: Wow, so we are living in the golden age of #esoteric?
19:30:12 <AnMaster> oerjan, there are bacterias in nuclear reactors... so?
19:30:49 <AnMaster> fizzie, there seems to be a pattern, more active during the summers?
19:31:22 <fizzie> ehird: As far as amount of content goes, maybe. I can't really meaningfully quantify the quality.
19:31:35 <ehird> fizzie: Oh our quality is certainly down.
19:31:37 <ehird> AnMaster: duh, summer holidays
19:31:41 * oerjan feels nervous about a scale using e notation without being logarithmic
19:31:59 <AnMaster> if it is, holidays is the likely reason yes
19:32:04 <fizzie> The data is so noisy I can't really tell.
19:32:23 <fizzie> Anyway, amount of bytes in my monthly logfile might not be the best measure anyway.
19:32:35 <AnMaster> fizzie, you could filter join/parts
19:32:43 <AnMaster> that would be a *bit* more correct
19:32:45 <ehird> fizzie: If you switch to wc -l, and then make it so that it draws lines between the points
19:32:48 <ehird> that'd be reasonable
19:32:50 <ehird> and hopefully not hard?
19:33:06 <AnMaster> ehird, yes and sed away anything but messages and /me
19:33:09 <fizzie> Not hard, nope. Although I think I'll also grep it so that only those so-called actual messages are in.
19:33:30 <AnMaster> fizzie, don't forget CTCP ACTIONs
19:33:51 <oerjan> making the dates on the X axis actually readable might help too *duck*
19:33:56 <AnMaster> if it is raw log then I suggest grepping for PRIVMSG would work
19:34:09 <ehird> AnMaster: it isn't raw
19:34:13 <ehird> I know because I have seen his 2002 logs
19:34:29 <ehird> AnMaster: Pretty typical-looking.
19:34:33 <ehird> Let me get you a line
19:34:52 <ehird> [18:05:22] -!- lament [~lament@h24-78-145-92.vc.shawcable.net] has joined #esoteric
19:34:52 <ehird> [18:10:45] < lament> my tarantula molted!
19:34:53 <ehird> [18:10:49] < shapr> yay!
19:34:53 <ehird> [18:14:02] -!- lament [~lament@h24-78-145-92.vc.shawcable.net] has quit ["PROSECUTORS WILL BE TRANSGRESSICUTED."]
19:34:53 <ehird> --- Log closed Fri Jan 03 18:47:53 2003
19:35:11 <AnMaster> I suspect irssi behind that log
19:35:13 <ehird> AnMaster: Humans sometimes read things :P
19:35:16 <oerjan> ehird: great excerpt :D
19:35:35 <ehird> Incidentally, mooz is in that log.
19:35:42 <oerjan> wait, shapr was here? that must be #haskell i think
19:35:50 <ehird> [04:52:55] * andreou is feeling REALLY GOOD
19:35:55 <AnMaster> oerjan, it says "-!- lament [~lament@h24-78-145-92.vc.shawcable.net] has joined #esoteric"
19:35:57 <fizzie> Okay, I think I've got a suitable expression; and the timestamps have changed since those earliest logs.
19:36:03 <ehird> AnMaster: yeah, and then shapr talks.
19:36:20 <ehird> [04:52:55] * andreou is feeling REALLY GOOD
19:36:22 <ehird> is how the /mes look
19:36:28 <AnMaster> ehird, since it says "foo joined #esoteric" that was pretty obvious
19:36:40 <ehird> [2002-12-15 01:17:38] < navigator> 27M
19:36:43 <ehird> is from the second log part
19:36:47 <ehird> but that's even older
19:36:57 <ehird> that is an iso date
19:40:14 <fizzie> The file is now updated, although the date labels are very messed up; gnuplot is really user-unfriendly when it comes to time data and I don't remember the magic settings.
19:40:31 <fizzie> At least the labels are now readable, but the tickmarks don't hit the months correctly.
19:41:00 <fizzie> Well, the points are months, and you can just count the from the nearest tickmark, which seems to be using the day/month/year format maybe. I think.
19:41:11 <ehird> fizzie: Does it draw lines between the plots?
19:41:36 <ehird> That last huge peak.
19:41:53 <fizzie> Well, the last point is this October, it's not really comparable.
19:42:11 <ehird> So, essentially, "#esoteric is dying" has never been true.
19:42:15 <ehird> It's always been gaining steadily.
19:42:28 <AnMaster> yes over 5000 lines in 4 days in October?
19:42:51 <AnMaster> ehird, also no, look at the low before that
19:43:02 <AnMaster> which was way way lower than so far this month
19:43:12 <ehird> it goes up and down BUT
19:43:15 <ehird> in the big picture
19:43:23 <AnMaster> ehird, the peaks are always larger
19:43:25 <ehird> so #esoteric has never been dying... it's been expanding
19:43:35 <ehird> we just need to figure out how to sustain the peaks :-P
19:44:08 <oerjan> more pasted code, clearly
19:45:18 <Mony> >___O< Koin Koin
19:45:55 <AnMaster> On 32-bit: 12 bytes overhead per memory pool. 4 bytes overhead per memory block. Double both on 64-bit. Still I think I beat malloc/free in the long run
19:46:31 <AnMaster> or maybe the name of some of that anime crap or whatever
19:46:46 <fizzie> No per-object overhead usually means directly that you will beat a generic malloc.
19:47:00 <AnMaster> fizzie, well that is impossible for free list
19:47:08 <oerjan> also, koin would be approximately qua... with a nasal vowel
19:47:15 <AnMaster> fizzie, and object == memory block in this case
19:47:29 <AnMaster> fizzie, but the main reason is that I got really bad memory fragmentation
19:47:31 <oerjan> but i'm googling to be sure
19:47:44 <fizzie> Oh; I though memory block == one page or so.
19:47:58 <fizzie> What do you need four bytes there for?
19:47:59 <AnMaster> fizzie, anyway I figured out a way to make that less.
19:48:20 <oerjan> http://www.eleceng.adelaide.edu.au/personal/dabbott/animal.html claims "coin, coin"
19:48:32 <AnMaster> fizzie, for the pointer for next free object Or do you want me to allocate the memory for the linked list of them from *another* memory pool?
19:48:41 <fizzie> No, I mean, the free-list only contains unallocated objects, which means that the pointers can be "inside" the objects there.
19:48:44 <AnMaster> anyway I could use offset in array
19:49:28 <AnMaster> fizzie, well since the objects are *less* than 8 bytes that wouldn't work on amd64 at least
19:50:04 <AnMaster> that would be truely awesome idea
19:50:21 <fizzie> I thought the "pointers use the space normally allocated for objects" was pretty much the "standard" way of doing that, at least when object size >= pointer size.
19:50:25 <AnMaster> which means 4 bytes overhead on x86_64 and 0 bytes on x86
19:51:29 -!- Slereah_ has quit (Read error: 104 (Connection reset by peer)).
19:51:53 <fizzie> If you don't mind the "more instructions involved in free/malloc", I guess you could easily manage to fit into 32 bits some sort of "block index + offset" value instead of a raw pointer.
19:51:58 <AnMaster> and if I have at most 2^32 objects in each memory pool then I could use a 32-bit index instead of a pointer
19:52:17 <AnMaster> fizzie, you may have hit enter first, but I thought of it first ;P
19:53:04 <AnMaster> fizzie, anyway I use "memory pool" here in the meaning "block header (3 * sizeof(void*)) + the relevant array"
19:53:40 <AnMaster> anyway each such block would have it's own free list I think... Or maybe I should use a global freelist
19:57:55 <oerjan> the fool, er, the pool
19:58:24 <AnMaster> oerjan, that pun totally failed
19:58:51 <oerjan> are you saying it was puny?
19:59:18 <AnMaster> err assuming puny means "has the quality of a pun"
19:59:28 <AnMaster> I guess it could mean something else
19:59:39 <AnMaster> oerjan, does it mean something else?
20:00:52 <oerjan> http://www.google.no/search?hl=no&q=define%3Apuny&meta=
20:01:20 -!- Slereah_ has joined.
20:02:51 <AnMaster> though the pun with "puny" was quite good
20:09:48 -!- ENKI-][ has quit (Read error: 104 (Connection reset by peer)).
20:14:08 -!- calamari has joined.
20:14:36 -!- hakware has joined.
20:15:04 -!- hakware has changed nick to ENKI-][.
20:15:25 <AnMaster> growing with realloc() may fail
20:15:47 <AnMaster> can't see any reason why it wouldn't
20:16:31 <oerjan> it _could_ be just a NOP couldn't it
20:18:20 <AnMaster> fun fact: realloc(ptr, 0); is same as free(ptr);
20:19:11 <AnMaster> and realloc(NULL, n); is same as malloc(n);
20:19:17 <AnMaster> for all values (including 0) of n
20:19:33 <AnMaster> so basically we could do away with malloc and free
20:20:53 <fizzie> However, C99 guarantees only realloc(NULL, n) doing the same thing as malloc(n), not the "size 0 does free" thing.
20:21:36 <fizzie> And in fact my "realloc" man page says "If size was equal to 0, either NULL or a pointer suitable to be passed to free() is returned."
20:21:42 <AnMaster> "If size is 0 and ptr is not a null pointer, the object pointed to is freed."
20:21:43 <oerjan> oklocod: yllillä ollon ällä
20:21:54 <AnMaster> so that is from POSIX 2001.whatever
20:22:02 <fizzie> I don't have posix man pages installed on this system.
20:22:02 * oerjan now wonders if he actually said anything
20:22:32 <AnMaster> fizzie, so we could still do away with malloc
20:22:41 <AnMaster> and rename realloc to alloc basically
20:22:45 <oklocod> AnMaster: do you know what ollon means?
20:22:53 <oklocod> or oerjan, i'm sure one of you should
20:23:14 <oklocod> aaaanyway, oerjan, you didn't say anything meaninful, but it was definitely finnish
20:23:26 <AnMaster> oklocod, it is the fruit of a type of tree
20:23:41 <oklocod> that's not the only meaning, but yeah
20:23:45 <oerjan> well i did ensure vowel harmony
20:23:52 <AnMaster> 2. Ollon - Wikipedia, the free encyclopedia
20:23:52 <AnMaster> Ollon is a municipality in the district of Aigle in the canton of Vaud in Switzerland, sited in the foothills
20:23:58 <oklocod> oerjan: ällä is the letter l
20:24:26 <oklocod> AnMaster: well it should also mean a glans
20:24:28 <fizzie> Also according to wiktionary: "2. den översta delen på penis eller/och klitoris; glans penis/clitoris"
20:24:39 <oklocod> i didn't even remember the other meaning
20:24:53 <AnMaster> oklocod, actually more like shine
20:25:04 <AnMaster> not sure of how to translate "glans"
20:25:10 <oklocod> it's the tip of the cocker
20:25:20 <fizzie> WordNet quote: 1. glans -- (a small rounded structure; especially that at the end of the penis or clitoris)
20:25:26 <fizzie> So it might be any small rounded structure.
20:25:50 <AnMaster> oklocod, no, it is the shine from, for example, a well polished metal surface.
20:25:57 <AnMaster> that is what the adjective glans means
20:26:05 <AnMaster> actually that is the noun form
20:26:12 <AnMaster> glänser would be the adjective
20:26:33 <oklocod> ah, that's what he was blabbering about
20:26:37 <AnMaster> psygnisfive, ah ok, but he was using Swedish before
20:26:42 <oklocod> yeah i only know obscene swedish
20:26:44 <psygnisfive> i propose a new convention: whenever referencing words from specific languages
20:27:10 <AnMaster> psygnisfive, It is needed for English too
20:27:20 <ehird> psygnisfive: <span lang="en">hello</span>
20:27:20 <oklocod> psygnisfive: i used what AnMaster used in lingobot, and it seems standard for some reason
20:27:32 <oklocod> (lingobot was a bot of mine that translated words to 150 other languages)
20:27:38 <fizzie> Why not <span xml:lang="en">like this</span>, it's nice and verbose.
20:27:42 <AnMaster> psygnisfive: English is English not English synonymous
20:27:57 <AnMaster> English that English is English how English the English correct
20:28:16 <psygnisfive> are you saying you'd prefer just en:glans?
20:28:27 <AnMaster> psygnisfive, English nicks English doesn't English need English it
20:28:31 <oklocod> fizzie: that's not verbose enough, have a separate block for each word
20:28:57 <AnMaster> psygnisfive, en:you en:don't en:get en:the en:point! en:you en:need en:it en:for en:every en:word
20:29:17 <AnMaster> <psygnisfive> ok fine :P <-- en:should en:have en:been: <psygnisfive> en:ok en:fine :P
20:29:29 <ehird> AnMaster: Use-mention distinction.
20:29:32 <ehird> Plz to be learning it.
20:29:44 <AnMaster> ehird, sure ok, I was just trying to make fun of en:this
20:29:50 <fizzie> oklocod: <phrase xml:lang="en"><word xml:lang="en">Something</word><word xml:lang="fi">kuten</word><word xml:lang="sv">det</word><word xml:lang="sv">här</word></phrase>?
20:29:57 <ehird> AnMaster: Haha! You made fun of an entirely different, unjustifiably different thing!
20:30:15 <oklocod> fizzie: how about a question block too?
20:30:19 <psygnisfive> i suggested the standard natural-language version :P
20:30:26 <fizzie> oklocod: Maybe as an attribute to 'phrase'.
20:30:46 <psygnisfive> ehird, you'd do good in the semantics class i was in
20:30:48 <AnMaster> fizzie, what does <word xml:lang="fi">kuten</word> mean?
20:30:59 <fizzie> AnMaster: It's close to en:like.
20:31:04 <psygnisfive> it took the students like two weeks to get the use-reference distinction
20:31:05 <AnMaster> fizzie, that wasn't my question
20:31:19 <AnMaster> fizzie, if I had wanted that I would have used > and such
20:31:19 <psygnisfive> tho there it was called "object language" and "meta language"
20:31:33 <ehird> psygnisfive: i've always got it intuitively
20:31:41 <AnMaster> I thought he meant the syntax was like it
20:31:57 <fizzie> The whole phrase was trying to be "something like this?"
20:32:21 <AnMaster> psygnisfive, hm? you mean pointer vs object?
20:32:36 <ehird> oh look, AnMaster can only think in C
20:32:49 <ehird> AnMaster: yeah uh that'd be worse.
20:33:06 <ehird> i'm surprised you didn't take the time to respond in obscure erlang to flaunt your skillz in it, though
20:33:23 <AnMaster> ehird, if you don't stop attacking me I shall begin to use C++ with boost!
20:33:33 <oklocod> the use-reference distinction? err... the fact you can quote strings?
20:33:36 <ehird> AnMaster: if you begin to use C++ with boost then I'll just /ignore you.
20:34:16 <AnMaster> ehird, and xerces-c or whatever that horrible xml library is
20:35:09 -!- calamari has quit ("Leaving").
20:35:37 <oerjan> wow xerces is not a misspelling of xerxes
20:36:18 <oklocod> oerjan: i noticed you've blurted out two quite low quality puns today; are you feeling alright?
20:36:43 <oerjan> oklocod: the weight of duty must be getting to me
20:37:02 <oerjan> 97% is just so hard to acheive, even with bogus accounting
20:37:30 <psygnisfive> anmaster: no, i mean code that operates on data
20:37:45 <oerjan> and what do you mean _two_?
20:37:48 <psygnisfive> e.g. quotations of the language you're speaking/coding in
20:38:40 <oerjan> maybe if i higher some recently jobless bankers...
20:38:46 <oklocod> oerjan: or perhaps just one... i don't remember the other one, i just vaguely recall there was another
20:38:52 <psygnisfive> Chicago is a major city, 'Chicago' is a 7 letter word.
20:39:00 <AnMaster> psygnisfive, Hm I'm not sure exactly what you mean. Do you mean like: Code that operates on another LISP expression vs. running that LISP expression?
20:39:04 <oerjan> my spelling is off, i think i must be tired
20:39:28 <oklocod> AnMaster: he means '(code here) versus (code here)
20:39:30 <psygnisfive> i mean the difference between calling someone a nigger, and saying that there is this word 'nigger'
20:39:39 <oklocod> the first one is data that is code, the second is just code
20:39:51 <AnMaster> oklocod, yes I know that much of lisp
20:40:12 <psygnisfive> now see, i was gonna reference lisp but i figured it'd be too easy to miss
20:40:32 <oklocod> yeah that'd've been prettier
20:40:43 <AnMaster> psygnisfive, It should be possible to make a language without that distinction
20:41:01 <psygnisfive> its possible to make a language that doesnt have reference, as such
20:41:18 <psygnisfive> in the sense that you can't talk about strings as strings-in-the-language
20:41:43 <AnMaster> psygnisfive, not if it isn't TC!
20:42:03 <AnMaster> psygnisfive, there are some interesting non-tc languages
20:42:04 <GregorR> Who was it that argued (quite accurately) that C isn't TC? :)
20:42:30 <oklocod> everyone has realized that at some point in their life
20:42:40 <oklocod> and confirmed it @ #esoteric
20:42:42 <AnMaster> GregorR, I think perl regex may be tc
20:42:52 <AnMaster> but it should be possible to extend it to me
20:42:55 <oerjan> part of gödel's theorem is essentially that in any sufficiently powerful logical system, you _can_ do reference
20:42:59 <GregorR> oklocod: Amazing since the vast majority of people know neither C nor what "TC" means :P
20:43:13 <AnMaster> oerjan, how is "sufficiently" defined?
20:43:34 <oklocod> GregorR: are you sure about the majority not knowing what C is?
20:43:36 <AnMaster> oerjan, maybe: "a logical system where you can do reference"?
20:43:37 <oerjan> predicate logic + a tiny bit of arithmetic
20:43:58 <GregorR> oklocod: Your mom is the mom of somebody who knows what C is :P
20:44:24 <GregorR> oklocod: If I go ask some random art student what C is, they'll say "UHH, THE LETTER AFTER BEEEEEE"
20:44:38 -!- kt3k has quit ("CHOCOA").
20:44:46 <AnMaster> GregorR, ais523 said it was thanks to the file IO
20:45:04 <oklocod> GregorR: i loved that BEEEEEEE :P
20:45:08 <AnMaster> psygnisfive, you need infinite memory, C doesn't allow that. sizeof(char*) must be finite
20:45:12 <GregorR> AnMaster: C minus libraries is not TC. C plus libraries with hardware access (which eventually leaves C) is TC.
20:45:18 <AnMaster> psygnisfive, so memory size must be finite
20:45:31 <GregorR> Naturally if you had a libInfiniteTape, C would be TC, but libInfiniteTape can't be written entirely in C.
20:45:47 <psygnisfive> but why does sizeof char* have to be non-finite?
20:45:55 <AnMaster> GregorR, well the file IO is part of the standard
20:46:01 <oklocod> psygnisfive: otherwise only a finite amount of memory can ever be addressed
20:46:03 -!- Slereah_ has quit (Read error: 113 (No route to host)).
20:46:22 <GregorR> AnMaster: When we talk about languages in #esoteric, we're not talking about libraries ^^
20:46:29 <GregorR> AnMaster: Even if those libraries are a standard part of the language :P
20:46:29 <psygnisfive> im not sure what sure, but does the C SPEC say that pointers have to be of a specific size?
20:46:46 <oklocod> psygnisfive: no, but they must be of *some* size
20:46:51 <AnMaster> GregorR, so you talk about freestanding C? As used for kernels
20:46:51 <psygnisfive> or does the fact that C is on a finite machine require that?
20:47:08 <GregorR> AnMaster: Shore, but they always have some ASM too.
20:47:09 <AnMaster> GregorR, and it was ais523 who first mentioned the file IO argument
20:47:15 <psygnisfive> oklocod, if that's the case, then _all_ programming languages are non-TC :P
20:47:28 <GregorR> psygnisfive: No, because most languages don't have pointers.
20:47:33 <oklocod> in many languages, there is no need to have an address for an object
20:47:34 <GregorR> (Most *modern* languages anyway)
20:47:46 <psygnisfive> sure but the C spec doesnt say that pointers have to be of some specific size does it?
20:47:59 <oklocod> psygnisfive: no, but they must be of some finite size when execution starrts
20:48:02 <AnMaster> psygnisfive, the problem is the size of the pointer itself have to be finite
20:48:39 <psygnisfive> since every memory address on an infinitely long tape is also a finite number
20:48:42 <oklocod> psygnisfive: the set of all usable numbers is infinite extendable
20:48:48 <oklocod> if you don't have pointers
20:48:59 <psygnisfive> there are an infinite set of usbale numbers
20:49:03 <oklocod> turing-completeness doesn't need infinite memory, just infinitely extendable
20:49:04 <oerjan> psygnisfive: sizeof(int *) has to be a finite number, no int pointer can be larger than that
20:49:07 <AnMaster> psygnisfive, "finite size when execution *starts*"
20:49:07 <psygnisfive> but NONE of those numbers themselves are infinitely large
20:49:15 <AnMaster> psygnisfive, you need to be able to grow it forever
20:49:19 <oklocod> psygnisfive: nothing needs to be infinitely large
20:49:30 <GregorR> psygnisfive: With a tape you don't need to absolutely address any of those finite numbers.
20:50:08 <oklocod> psygnisfive: turing completeness is not about actually ever being able to allocate infinite memory, just that for any finite amount of memory the program may request at runtime, that amount of memory will be accessible
20:50:13 <psygnisfive> you mean that because you have to be able to talk about the size of specific pointers in C
20:50:20 <oklocod> for any pointer size, this is not enough.
20:50:23 <psygnisfive> because those pointers, being crucial to C's TCness
20:50:48 <GregorR> Mind you, it's a strawman argument since C is defined for finite machines :)
20:50:54 <psygnisfive> im not sure how pointers are crucial to TCness but
20:51:21 <oerjan> hm wait a minute does C say anything about the unit of sizeof?
20:51:31 <AnMaster> oerjan, only that it must be finite
20:51:34 <psygnisfive> sure but i dont have to allocate memory myself when doing, say, int five = 5
20:51:45 <oklocod> i think we've went over this
20:51:51 <GregorR> oerjan: An type with infinite range can't actually store the pseudonumber "infinity" anyway.
20:51:53 <psygnisfive> and im fairly certain that you can get TCness with just that
20:52:07 <AnMaster> and size_t is as large as pointers are
20:52:10 <GregorR> psygnisfive: But everything in C must be addressable: That is &var must always be defined.
20:52:11 <psygnisfive> without any reference to pointers or pointer tizes
20:52:16 <GregorR> psygnisfive: Whether you use it or not.
20:52:18 <oklocod> is whether size_t could be abstract, and actually a bignum
20:52:26 <psygnisfive> well then that ruins the argument, GregorR:
20:52:31 <oklocod> i think that's what i'm asking
20:52:40 <AnMaster> oerjan, sizeof() returns a size_t, sizeof(size_t) == sizeof(int*)
20:52:41 <psygnisfive> because the argument was that the size of the pointer was crucial to TCness
20:52:58 <AnMaster> since it needs to be finite when execution starts
20:52:59 <psygnisfive> but if you can build a TC bit of code without referencing the size of a pointer
20:53:06 <GregorR> psygnisfive: No, it was that /because/ C lets you address any variable, the size of pointers is crucial to the definition of C.
20:53:07 <psygnisfive> then the size of a pointer ISNT crucial to TCness
20:53:18 <AnMaster> psygnisfive, err see what I said
20:53:18 <GregorR> psygnisfive: You could make a subset of C that didn't have that property and would be TC, yes.
20:53:49 <AnMaster> psygnisfive, and you can't access memory without pointers
20:54:07 <AnMaster> psygnisfive, so you can't malloc() a block larger than a pointer
20:54:18 <AnMaster> larger than the range of a pointer*
20:54:28 <psygnisfive> im not sure you'd need to do malloc() to make something TC in C.
20:54:43 <AnMaster> psygnisfive, or access offsets in a static array either
20:54:54 <AnMaster> psygnisfive, however it is TC with file IO
20:54:55 <psygnisfive> im not sure you'd need ARRAYS to make TCness in C.
20:55:14 <psygnisfive> i just dont see how the size of something unrelated to TCness can affect TCness.
20:55:33 <AnMaster> psygnisfive, you need infinite memory, You can't access memory outside the range of pointers in C
20:55:35 <psygnisfive> yes, ok, the C spec requires size(int*) be finite, meaning that it requires finite memory, meaning its not TC
20:55:46 <psygnisfive> sure, fine. that i can see as a sort-of-argument
20:56:06 <AnMaster> psygnisfive, prove it is TC even without file IO then
20:56:51 <psygnisfive> but thats more about how pointers are implemented in C, not about C itself.
20:57:09 <oerjan> what i mean is, couldn't sizeof(int *) = 1, and _still_ int pointers have infinite range because the sizeof unit is infinite
20:57:19 <oklocod> psygnisfive: the finite pointer size basically means, you have a turing machine, but there is a finite amount of cells it can ever reach.
20:57:50 <AnMaster> oerjan, sizeof(char) == 1 by definition. char must be a finite number of bits (the define CHAR_BIT iirc)
20:57:57 <oklocod> it is about C, because C guarantees you need to be able to address a variable.
20:58:06 <AnMaster> oerjan, pointer must be whole bytes
20:58:26 <AnMaster> oerjan, could be CHAR_BITS btw, not sure about the name
20:58:27 <psygnisfive> and you can address an INFINITE number of variables in C! you just need enough memory to store that many variables
20:58:28 <oerjan> GregorR: did you fix your hat count? :D
20:58:40 <psygnisfive> and a c-compiler to know how big the memory addresses are for that memory.
20:58:43 <oklocod> psygnisfive: you need to be able to access them all with a finite pointer.
20:58:51 <oerjan> GregorR: the count on your hats page is outdated
20:58:57 <GregorR> I'll just remove the count.
20:59:18 <oklocod> psygnisfive: finite, as in there will always be a program that allocates a greater amount of memory
20:59:24 <psygnisfive> just because C guarantees you can address all the pointers doesnt mean that being ABLE to address all pointers is relevant to TCness
20:59:55 <psygnisfive> does C dynamically adjust pointer sizes to handle memory differences?
21:00:00 <GregorR> AnMaster: http://codu.org/hats.php
21:00:06 <oklocod> psygnisfive: pointer sizes are static.
21:00:13 <AnMaster> psygnisfive, "finite and fixed when program starts"
21:00:14 <psygnisfive> if my machine has more memory than yours, does C know this, and alter its pointer size?
21:00:38 <psygnisfive> the question is does the size depend on what machine you start the program on
21:00:40 <oklocod> psygnisfive: C doesn't say anything about the pointer size
21:00:46 <oklocod> that has nothing to do with this argument
21:00:49 <AnMaster> other than it needs to be finite
21:01:07 <AnMaster> psygnisfive, you can't create/use memory that can't be accessed with a pointer in C
21:01:11 <psygnisfive> because C DOES let you address any and all variables you want
21:01:17 <AnMaster> so every variable needs to be accessible with a pointer
21:01:40 <psygnisfive> theres no such thing as memory that cant be accessed by a C pointer
21:01:42 <AnMaster> and every variable need to have an unique address
21:02:04 <AnMaster> psygnisfive, you could have a 32-bit C on a 64-bit machine
21:02:05 <GregorR> I BLAH YOUR BLAHS UNTIL BLAH BLAH
21:02:10 <oklocod> psygnisfive: even with an infinitely large memory C wouldn't be tc
21:02:38 <oerjan> fungot: do you blah about this?
21:02:38 <fungot> oerjan: something like scheme48 ( upon which scsh was based) would be
21:02:42 <ehird> psygnisfive: You are wrong, C is not turing complete, end of.
21:02:46 <AnMaster> psygnisfive, no I didn't. I said every variable must be addressable
21:03:02 <AnMaster> psygnisfive, and a C program can't access any memory that is not addressable with a pointer
21:03:04 <psygnisfive> all computations require only finite, but indefinitely large amounts of momory
21:03:09 <ehird> psygnisfive: Of course I am, because you have continually shown that your attitude is that you cannot possibly be wrong, especially your intuitions.
21:03:49 <oerjan> psygnisfive: the point is you cannot calculate the needed size in advance
21:04:07 <AnMaster> psygnisfive, pointer size can't change at runtime
21:04:15 <psygnisfive> thus proving that there is no computation that cannot be performed in C
21:04:25 <psygnisfive> so long as you are given the appropriate amount of memory
21:04:30 <ehird> way to go psygnisfive, whenever someone explains when you are wrong ignore them
21:04:32 <psygnisfive> thus proving that C is, despite your idiocy, Turing Complete
21:04:41 <psygnisfive> anmaster, i didnt say change it at runtime
21:05:01 <oerjan> psygnisfive: but a single C program run isn't Turing Complete
21:05:06 <AnMaster> psygnisfive, restarting the program on another system is not valid for TC
21:05:27 <oklocod> psygnisfive: that's a valid point, yes
21:05:43 <psygnisfive> because you're talking about individual RUNS of a program in C
21:06:10 <psygnisfive> but that too is a problem with computers being finite
21:06:19 <ehird> we're talking hypothetical
21:06:23 <AnMaster> psygnisfive, no you are wrong, since the program is basically another one if you change pointer size
21:06:36 <oerjan> psygnisfive: the problem here is that C is then not a single language in the CS theoretical sense
21:06:38 <ehird> we are talking about C running on a machine with actual, real, infinite tape
21:06:51 <oerjan> it becomes a family of languages indexed by pointer size
21:06:55 <oklocod> psygnisfive: ignore AnMaster and ehird, and listen to oerjan
21:07:02 <psygnisfive> oerjan: that is the first sensible response.
21:07:22 <AnMaster> oklocod, well s/AnMaster and// ;P
21:07:33 <oklocod> psygnisfive: your arguments weren't exactly sensible until recently either :P
21:07:43 <psygnisfive> and to that i'd say, in that case, sure. but then it makes no sense to say the C language is not TC since there is no such thing as the C language, merely particular C languages with specific pointer sizes
21:08:15 <oerjan> Hurrah! Icecream to everyone!
21:08:34 <oklocod> AnMaster: no, i win. i'm the winner
21:08:35 <oerjan> also, when is oklocod and psygnisfive going to marry?
21:09:00 <ehird> oerjan: hopefully soon so they can stop spamming #esoteric with it.
21:09:24 <oklocod> also, i'm still waiting for your proposal
21:09:27 * oerjan likes icecream with chocolate bits
21:10:48 <oerjan> has anyone done a wedding proposal on Agora yet, i wonder
21:11:47 <AnMaster> oerjan, I prefer vanilla icecream
21:11:54 <ehird> oerjan: that would be awesome
21:12:06 <ehird> oerjan: 'Proposal: Marriage (AI=1) { ... }'
21:12:13 <oerjan> well vanilla icecream with chocolate sauce is also a favorite
21:12:25 <psygnisfive> oerjan: vanilla icecream plain is decicious
21:12:43 <AnMaster> oerjan, I prefer with maple syrup
21:12:51 * ehird rips AnMaster's and psygnisfive's head off for insulting the Great Mighty 15-Year-Old Agora
21:13:14 <oerjan> ehird: i think AnMaster was discussing icecream
21:13:27 <psygnisfive> im just confused by the popularity of nomics in general
21:13:33 <ehird> psygnisfive: why not
21:13:59 <psygnisfive> i dont like games, so thats partially why ;)
21:14:09 <AnMaster> oerjan, ever tried it? Oh and Ice cream made from fresh vanilla pods. Not just some vanilla-flavoured sugar.
21:14:15 <ehird> yes, games are trivial!
21:14:36 <ehird> also, ice cream in any form, shape or anything is amazing
21:14:41 <psygnisfive> oh, im not saying people dont find them to be fun
21:14:52 <psygnisfive> i'm just not one of the people that does. :P
21:15:06 <AnMaster> ehird, sure but some forms is tastier than other ones
21:15:19 <ehird> AnMaster: well... it's kind of like bacon
21:15:25 <ehird> there's not much room for suckitude :-P
21:15:37 <oerjan> psygnisfive is italian?
21:15:44 <ehird> SOMEONE MAKE IT, NOW
21:16:05 <AnMaster> <ehird> you are a GENIUS <-- well thank you
21:16:14 <AnMaster> I shall remember that for the future
21:16:15 * oerjan thought it:gelato = en:icecream
21:16:17 <ehird> AnMaster: Well only on the subject of bacon ice cream.
21:16:31 <GregorR> There's a donut shop in Portland that makes Bacon Maple Bars
21:16:32 <ehird> oerjan: http://en.wikipedia.org/wiki/Gelato sez wp
21:16:48 <AnMaster> ehird, please say it again with "AnMaster:" in front (without quotes), so optbot could put it in topic!
21:16:52 <psygnisfive> there are slight differences in how its made
21:17:04 <ehird> AnMaster: no, optbot strips those off
21:17:04 <optbot> ehird: Screen brightness. Turn it down. :P
21:17:15 <ehird> GregorR: Hmm. Say, an #esoteric meetup in Portland. YES THAT SOUNDS GOOD
21:17:17 <oerjan> ehird: that mad english cook has an egg and bacon icecream, was mentioned in the Ig Nobel news recently
21:17:24 * ehird plots to steal all the bacon maple bars
21:17:26 <GregorR> ehird: I don't live in Portland now :P
21:17:36 <psygnisfive> in my experience, gelato is smoother and heavier
21:17:38 <AnMaster> ehird, Portland in what country?
21:17:40 <ehird> GregorR: Well... fine it'll be a very lonely meetup
21:17:57 <ehird> AnMaster: US i'm guessing.
21:18:22 <GregorR> Is there a Portland, UK? I can't imagine there's a Portland anywhere else ...
21:18:27 <GregorR> But yeah, I was referring to Portland, OR, USA.
21:18:41 <ehird> 5000000000000 portlands
21:18:44 <oerjan> psygnisfive: maybe it's like it:pizza /= us:pizza
21:18:46 <ehird> http://en.wikipedia.org/wiki/Portland
21:19:23 <AnMaster> GregorR, what about Australia?
21:21:16 <GregorR> Portland, OR, USA is the only Portland of significance :P
21:21:29 <oerjan> there's a Sortland, Norway at least
21:22:09 <oerjan> i've never been to a portland
21:22:42 <psygnisfive> oh btw europeans, especially french and germans:
21:23:40 <ehird> i am kind of european. i'm in europe but my country cries whenever anybody says europe
21:24:04 <psygnisfive> which country? england? they dont like being part of europe.
21:25:09 <ehird> AnMaster: our government keep weaseling out of european union stuff :-P
21:25:48 <ehird> AnMaster: why? I haven't seen actual objections to the EU beyond the beauocracy
21:25:52 <psygnisfive> but who's going to check America's international influence? CHINA? RUSSIA?
21:26:00 <psygnisfive> not that americans influence is so hot these days but
21:26:26 <AnMaster> ehird, well that is one part, and the other part is that, while for some countries stuff improved with EU, it went the other way for Sweden. We used to have better social security before EU
21:26:35 <oerjan> a beauocracy would be something
21:26:41 <ehird> oerjan: bearocracy
21:26:48 <ehird> the government consists of bears
21:26:51 <AnMaster> ehird, it seems everything goes to some average
21:26:52 <ehird> and the bears decide everything.
21:27:08 <ehird> AnMaster: I guess.
21:27:29 <AnMaster> ehird, so for Sweden it really been a bad thing. For some other countries it has been a good thing
21:35:51 -!- lilja has joined.
21:55:01 <ehird> THIS IS A BEAR HELLO
21:55:03 <ehird> ( http://www.albinoblacksheep.com/flash/bearhello )
21:56:45 <ehird> all of Somebody's toons are like that
21:56:47 <ehird> but its not a series
21:57:01 <ehird> bear hello is the masterpiece though
21:57:51 <ehird> actually i think bear hello makes some sort of sense if you recognize that its not in chronological order
21:58:22 <psygnisfive> non-linear story telling is also pretty awesome
21:58:34 <ehird> http://shii.org/knows/Bear_Hello <- a scholarly interpretation of bear hello
21:58:35 <oerjan> sense it no! make cannot
22:11:31 <ehird> psygnisfive: google. use it
22:14:04 <AnMaster> oerjan, Talk like Yoda day it isn't
22:16:22 <oerjan> 21 May, Talk like Yoda day is.
22:17:58 <oerjan> Är det inte Kim Jong-Il som sitter der borta?
22:22:06 <ehird> psygnisfive: shrug
22:25:17 <ehird> link? ive seen one more of his
22:25:28 <ehird> also there is no tusho in #esoteric
22:25:31 <psygnisfive> http://www.albinoblacksheep.com/flash/misc/
22:25:49 <ehird> huh, he made puppy whirl?
22:26:31 <ehird> psygnisfive: also because i felt like it
22:28:19 <psygnisfive> no, i just liked "tusho" better. it sounded cooler.
22:28:38 <ehird> it also had a quota of 1 'tush' joke a day
22:28:56 <psygnisfive> ::hug:: well i liked it and i never made such crude jokes
22:29:24 <ehird> should i grep to find it
22:29:32 <psygnisfive> so i never noticed that "tush" reading at all
22:30:02 <AnMaster> ehird, btw GCC got something called "objective-c++"
22:30:14 <ehird> AnMaster: thats not gcc specific
22:30:18 <ehird> its for interfacing C++ and obj-c code
22:31:01 <ehird> AnMaster: probably, but you gotta use c++ stuff somehow
22:31:06 <ehird> http://en.wikipedia.org/wiki/Objective-C#Objective-C.2B.2B
22:31:28 <ehird> its just the objective- transformation applied to c++ instead of c
22:31:39 <ehird> no actual interaction
22:32:09 <AnMaster> ehird, not GCC specific you said?
22:32:16 <ehird> AnMaster: originated in gcc.
22:32:47 <AnMaster> ehird, what other compilers have it?
22:33:03 <ehird> gcc is like the only obj-c out there
22:33:05 <ehird> apart from that one guys'
22:33:13 <ehird> he has a vendetta against apple
22:33:20 <ehird> and thus no actual obj-c program compiles with his impl
22:33:23 <ehird> because it is totally different
22:33:29 <ehird> AnMaster: theres not anything in that that is _specific_ to gcc
22:33:32 <ehird> but i think gcc is the only current impl
22:33:38 <ehird> but then gcc is the only real obj-c impl
22:33:42 <ehird> so only as much as obj-c is gcc specific
22:34:20 <AnMaster> what do you think of the language "Dylan"
22:34:33 <ehird> AnMaster: its a lisp derivative
22:34:35 <AnMaster> however I ran into it a few times recently
22:34:40 <ehird> it originally wasn't syntaxful
22:34:45 <ehird> but it was made syntaxful to appeal to a wider market
22:35:06 <ehird> AnMaster: not an improvement, but it DOES show that a lisp can have added-syntax and not break
22:35:17 <ehird> define method factorial(n :: <integer>)
22:35:18 <ehird> n * factorial(n - 1)
22:35:46 <ehird> yes, pascally languages generally are very easy to read
22:35:49 <ehird> but not easy to write
22:36:07 <AnMaster> and it prevents the best thing with lisp
22:37:12 <AnMaster> btw in "R5RS" what does the R and the RS stand for?
22:37:22 <ehird> Revised^5 Report on the Algorithmic Language Scheme
22:37:29 <ehird> Revised Report on the Algorithmic Language Scheme
22:37:31 <ehird> Revised Revised Report on the Algorithmic Language Scheme
22:37:33 <ehird> Revised Revised Revised Report on the Algorithmic Language Scheme
22:37:36 <ehird> Revised^4 Report on the Algorithmic Language Scheme
22:37:44 <ehird> because nobody could be arsed to write out that many "Revised"s
22:38:09 <ehird> AnMaster: but that's less fun
22:38:29 <ehird> its up to Revised Revised Revised Revised Revised Revised Report on the Algorithmic Language Scheme
22:38:31 <ehird> although as i said
22:38:36 <ehird> R6RS wasn't really passed in
22:38:39 <ehird> by any sane vote counting method
22:39:02 <ehird> AnMaster: pretty much, yes
22:39:22 <ehird> AnMaster: it adds a base standard library to scheme, which is cool, but its not structured very schemey
22:39:28 <ehird> and it also bloats the language
22:39:32 <ehird> with some unneccessary stuff
22:39:40 <AnMaster> ehird, a standard library *is* a good idea however
22:39:53 <ehird> but r6rs isn't the answer
22:39:54 <AnMaster> would make portable scheme programs actually be possible
22:40:13 <ehird> AnMaster: no, because that'll be a revision of r7rs
22:40:17 <AnMaster> probably wants to be compatible with r5 hm...
22:40:24 <ehird> and most of the scheme community has disavowed the committee
22:40:41 <AnMaster> you said r7 would be a revision or r7
22:40:52 <ehird> [[On 29 August 2007, the Revised Revised Revised Revised Revised Revised Report on Scheme was ratified by the Steering Committee. This has made a lot of people quite angry and has been widely regarded as a bad move.
22:40:52 <ehird> Many programmers believe that it was created by some sort of community process, though the Jatravartid people of Viltvodle VI believe that the entire Standard was in fact sneezed out of the nose of a being called the Great Green Arkleseizure. This theory is not widely accepted outside Viltvodle VI, and so, standards being the puzzling documents that they are, other standards are being designed. And this wiki, which is called SchemePunks, is definitely not part
22:40:56 <ehird> Which is very odd, because without that fairly simple piece of knowledge, nothing that is written on here could possibly make the slightest bit of sense. We hope to develop an alternative specification for the Family of Programming Languages known as Scheme. Watch this space.]]
22:41:00 <ehird> that likely got cut off
22:41:30 <ehird> oerjan: yes, from scheme-punks.org
22:41:36 <ehird> the second paragraph got cut off
22:41:47 <AnMaster> which is called SchemePunks, is definitely not par
22:42:01 <ehird> t of the Scheme Underground, even if it is, which it isn't.
22:42:08 <oerjan> ... definitely not part
22:42:20 <ehird> "R6RS must die." -- Chicken lead developer Felix Winkelmann
22:42:52 <ehird> http://lists.r6rs.org/pipermail/r6rs-discuss/2007-October/003351.html
22:43:00 <ehird> the whole list of people who ain't gonna implement r6rs
22:43:01 <AnMaster> lisp should have module name spaces
22:43:04 <ehird> (Spoiler: all of them)
22:43:07 <ehird> AnMaster: common lisp does
22:43:25 <AnMaster> ehird, it makes code easier to organise
22:43:32 -!- jix has quit ("CommandQ").
22:43:36 <ehird> AnMaster: plt has modules and such
22:43:48 <ehird> plt is as featureful as common lisp, really, just with a more schemish (generally cleaner) attitude
22:43:55 <AnMaster> ehird, yep. However non-portable code troubles me
22:44:01 <AnMaster> call it a character flaw if you want
22:44:12 <ehird> AnMaster: i don't like the scheme situation either
22:44:16 <ehird> common lisp isn't any more portable
22:44:22 <ehird> common lisp has no portable networking etc
22:44:38 <ehird> AnMaster: except that attempts to reinvent lisp have been almost universally poor
22:45:44 <ehird> im considering doing something with plt scheme sometime
22:45:48 <ehird> just to kind of show my support for it
22:45:51 <ehird> reach out to more languages
22:46:34 <ehird> AnMaster: oh, also
22:47:18 <ehird> all the reviews of Chez Scheme i've read are _very_ highly praised, it sounds like its IDE is state of the art (really good analysis, refactoring and such cools) and apparently its library set is really good
22:47:26 <ehird> also it was first released in 1985
22:47:27 <ehird> and uses incremental native-code compilation
22:47:31 <ehird> (read: really really fast)
22:47:47 <AnMaster> if ((pool->first_free - pool->base) >= (POOL_ARRAY_COUNT * sizeof(memory_block))) <-- GCC complains that I compare signed and unsigned, but I can't figure out which side it thinks is signed..
22:47:51 <ehird> so i guess PLT isn't *the best* but it's the best to *use*
22:48:07 <ehird> indeed scheme.com (chez scheme site) doesn't even list the price
22:48:13 <ehird> just a 'contact us for licensing information'
22:48:17 <ehird> which is code for '$$$$$$$$$$$$$$$$$$$$'
22:50:41 -!- Slereah_ has joined.
23:13:14 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | S pushed 647201.
23:21:12 <fizzie> Re R6RS, note that "all of them" does not include PLT.
23:26:52 <ehird> there are odd unicode chars in that line
23:27:04 <ehird> its not even unicode
23:27:41 <ehird> fizzie: do you know
23:30:27 <fizzie> 0xc2, 0xa0 -> 0b11000010 0b10100000 -UTF8-> 0b00010100000 -> U+00A0 NO-BREAK SPACE
23:30:59 <fizzie> Or maybe a "-[de-UTF8]->" notation would be more appropriate.
23:31:41 <ehird> fizzie: sqlite3.OperationalError: Could not decode to UTF-8 column 'text' with text 'Wooble is a coauthor of this proposal.
23:31:43 <ehird> no, its not utf-8/
23:31:56 <fizzie> Well, 0xc2 0xa0 _is_ UTF-8 encoding for no-break space.
23:34:42 -!- lilja has quit ("KVIrc 3.2.0 'Realia'").
23:42:07 <AnMaster> valgrind: the 'impossible' happened:
23:42:14 <AnMaster> I think my code is really fucked up atm
23:49:03 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | i would be really happy if someone checked if the update is ok. :-).
23:58:54 -!- moozilla has quit (Read error: 110 (Connection timed out)).