00:02:55 <fax> I dont know sed as well as I should :p
00:02:58 * fax looks somthing up
00:05:00 <fax> my sed doesnt support \| :/
00:06:31 <oerjan> what would you need that for?
00:06:48 <fax> for like this\|that, to match "this" or "that"
00:06:54 <fax> I suppose ill have to use two expressions then
00:06:57 <fax> thats a shame
00:07:43 <oerjan> if you are matching single characters, just use []
00:07:57 <fax> its either a single space, or 3 characters
00:08:11 -!- sebbu2 has quit ("@+").
00:10:33 <oerjan> does your sed support the -r option?
00:10:45 <fax> what a crap sed
00:11:02 <fax> i think ill just take out spaces
00:11:30 -!- Sgeo has quit ("Ex-Chat").
00:15:03 <fax> why can I have a fucntion named : but not a variable: |
00:16:02 <fax> x=sed`echo '0ヰ1東2京3ス4ス5オ6ラ7カ8ス9ケaトbダcガd㒵eジfヤ' | sed -e 's/\(.\)\(...\)/ -e s\/\1\/\2\/g/g'` && while true ; do dd if=/dev/urandom bs=1200 count=1 2>/dev/null | md5 ; done | $x
00:16:06 <fax> a lot shorter! thanks :D
00:16:44 * fax tries to find a secretive name for the variable
00:21:48 <bsmntbombdood> while true ; do dd if=/dev/urandom bs=1200 count=1 2>/dev/null | md5 ; done | sed`echo '0ヰ1東2京3ス4ス5オ6ラ7カ8ス9ケaトbダcガd㒵eジfヤ' | sed -e 's/\(.\)\(...\)/ -e s\/\1\/\2\/g/g'`
00:22:21 <fax> bsmntbombdood: thanks a lot
00:22:26 <fax> I didnt actually know that you could do that
00:25:35 -!- pikhq has joined.
00:28:25 <fax> md5 generates a fixed length string from the characters 0-9a-z
00:28:31 <fax> so just hash random data then transform it
00:28:34 <fax> makes a matrix effect lol
00:31:52 -!- Sgeo has joined.
01:07:43 -!- oerjan has quit ("leaving").
01:08:12 <fax> so hey this is a good place to ask :D
01:08:18 <fax> I was wondering what the most simple languages are
01:08:43 <fax> lambda calculus has 3 cases to consider, brainfuck has 7 (if you ignore inputs) and life has 8 cases
01:08:50 <fax> but am I missing anything?
01:09:00 <fax> wht about 2 4 5 and 6?
01:13:48 <fax> that sounds very interesting
01:18:05 <fax> how can I find out more?
01:19:34 <pikhq> I can't remember the name, but there's a functional language with 2.
01:20:21 <pikhq> Although I'd argue that as far as instruction set goes, Life is the simplest. :p
01:20:47 <pikhq> Life has 2 states.
01:21:03 <pikhq> And instructions for transformations between these 2 states.
01:21:06 <fax> yeah, but the rules of evaluation are somewhat more complex I think
01:21:14 <pikhq> Not all that complex, really.
01:21:56 <pikhq> Really just a 2 dimensional analog of a Turing machine.
01:36:51 <bsmntbombdood> pikhq: iota is one that's written up, but there are better ones
01:37:46 <fax> http://ling.ucsd.edu/~barker/Iota/
01:55:38 <ihope> Insert enthusiastic shout of "SevenInchBread!" here.
01:56:35 <lament> fax: are you aware of OISC?
01:56:49 <pikhq> One instruction computing.
01:57:13 <pikhq> The options for that one instruction range from SBN to MOV.
02:00:34 <pikhq> It's Turing complete.
02:00:38 <fax> thats weird
02:00:39 <pikhq> And yeah, it is sweet.
02:01:09 <pikhq> MOV is Turing complete by having special addresses in the CPU. . .
02:01:24 <pikhq> And SBN just gets one instruction that does multiple things.
02:01:34 <pikhq> It subtracts *and* conditionally branches.
02:01:36 <pikhq> bsmntbombdood: Yeah.
02:02:25 <fax> bloody hell!
02:02:28 <fax> thats really crazy
02:02:41 <fax> I mean about MOV
02:02:48 <fax> I never thought of thta
02:02:52 <fax> I mean like
02:02:55 <fax> impressive
02:03:06 <pikhq> A MOV OISC has been implemented in Wireworld, BTW.
02:03:21 <pikhq> (thereby proving Wireworld's Turing completeness)
02:05:35 <fax> thats neat
02:06:24 <pikhq> That kind of stuff is why we're here in #esoteric. :)
02:53:02 -!- goban has quit ("Leaving").
03:01:56 <SevenInchBread> so... what are all the options that a context switch system call might need?
03:02:48 <SevenInchBread> but you could do things like "until system call" or "until sends a messge" or something
03:04:38 <ihope> Keep the system calls to a minimum?
03:04:46 <RodgerTheGreat> erm, what precisely do you mean by a "context switch system"?
03:05:30 <SevenInchBread> a system call that switches the context of CPU execution to another processes state
03:06:23 <SevenInchBread> ihope, I don't care about the number of system calls -- that doesn't determine how small your kernel is.
03:06:57 <SevenInchBread> you NEED a context switch system call if you don't want the scheduling algorithm in the kernel itself.
03:07:46 <SevenInchBread> but hardware interupts are implied... and I've got a way to handle them.
03:09:55 <RodgerTheGreat> do you have a method for flagging processes (or whatever) at different priority levels?
03:10:02 -!- RainbowTrout has joined.
03:10:28 <SevenInchBread> RodgerTheGreat, I've got a way to set attributes... so the scheduler can set a priority attribute if it wants.
03:10:48 <SevenInchBread> RainbowTrout, YOU CAN ALREADY TELL THIS CONVERSATION IS A BLAST
03:25:36 <RodgerTheGreat> ah, then you should have some interesting/painful hardware features to play with for this
03:29:40 <RodgerTheGreat> I find the "barrel processor" approach to multitasking fascinating- with well designed software, very efficient. http://en.wikipedia.org/wiki/Barrel_processor
03:30:07 <SevenInchBread> .....why do I twitch whenever I hear the word "efficient"
03:30:28 <RodgerTheGreat> probably because it should *always* be conditionalized
03:30:59 <SevenInchBread> I just think of annoying business people and uptight engineer nerd people and SUICIDE
03:31:52 <bsmntbombdood> RodgerTheGreat: what if there's less than enough threads?
03:32:33 <bsmntbombdood> and how do you put proccesses at different priorities?
03:32:50 <RodgerTheGreat> you might be able to design your OS to prevent that from happening, or at least make use of empty threads
03:32:51 <SevenInchBread> ...I don't think it's intended to be the... high-level part.
03:33:35 <bsmntbombdood> it would be nice if they just had a array of pointers to threads
03:33:49 <bsmntbombdood> so you could fill that array fully with one thread to run it constantly
03:35:12 <RodgerTheGreat> the nice thing is that a CPU like that would be shit-tons harder to deadlock
03:35:13 <bsmntbombdood> SevenInchBread: same way you use them in software?
03:35:36 <bsmntbombdood> RodgerTheGreat: the os breaks in every timer interupt...
03:36:02 <bsmntbombdood> if the kernel wants to stop you from running, it only has to wait 1/60 second
03:38:49 <SevenInchBread> hmmm, you could do a dual core processor... one being a single-thread "master" CPU... and a second barrel processor.
03:40:18 -!- bd_ has joined.
03:40:20 -!- RProgrammer has joined.
03:40:42 <ihope> SevenInchBread: meet RProgrammer. :-)
03:41:13 <ihope> And bd_ too, for that matter.
03:41:24 <fax> Does anyone use J?
03:41:39 <fax> its kinda based on APL
03:41:42 <ihope> RProgrammer: do you know much about programming for the x86?
03:41:51 <fax> its not really esoteric..
03:41:55 * pikhq needs a true test of geekiness. . .
03:42:00 <fax> http://www.jsoftware.com/
03:42:05 <fax> I found it cool
03:42:06 <pikhq> Knowledge of Brainfuck?
03:42:07 <ihope> bd_: do you know much about programming for the x86?
03:45:25 * pikhq much prefers less esoteric architectures than IA32.
03:47:00 <GregorR> pikhq: Your definition of "esoteric" is clearly incorrect. Regardless of how sensible or nonsensical an architecture is, if it's popular, it's not esoteric.
03:47:20 <pikhq> GregorR: I'm obviously kidding.
03:47:42 <pikhq> What I really mean is, of course, that IA32 is a nonsensical architecture, worthy of being esoteric.
03:49:22 <pikhq> The current IA32 system I'm using supports IA16, IA32, and AMD64. . . Somehow.
03:54:55 <ihope> My opinion is objective and is therefore perfect.
03:57:00 * calamari creates a subclass of ihope, adds functions, and is now more perfect than ihope
03:58:00 <ihope> Perfection is achieved not when nothing can be added but when nothing can be removed.
03:58:13 <ihope> Therefore, I must be JObject or TObject or whatever fancy thing it is.
03:59:18 <RodgerTheGreat> If I were part of the Java API, I'd want to be java.util
03:59:41 <RodgerTheGreat> it contains so many of my good friends, like Vector and Random.
04:03:19 <bsmntbombdood> But I suppose a nuclear reactor would be best for power
04:05:54 <RodgerTheGreat> so this is becoming more of an underground complex than a room.
04:06:16 <RodgerTheGreat> underground complexes are awesome for a number of reasons.
04:07:31 <RodgerTheGreat> I've always thought it'd be pretty awesome to live in a renovated nuclear bunker.
04:15:48 * pikhq wants an underground complex for house/server warehouse
04:16:47 <pikhq> Sure; just pay me for the power usage.
04:16:57 <pikhq> And bandwidth, of course.
04:17:06 <pikhq> Should give you some rock-solid stability.
04:17:23 <RProgrammer> Pfft, bandwidth, just hack some guy's satellite internet
04:17:28 <pikhq> Survive the nuclear war, it will.
04:17:39 <pikhq> RProgrammer: We're talking "500 T3 lines"
04:17:53 <bsmntbombdood> pikhq: but the internet and power distribution networks won't, so it doesn't matter
04:18:17 <pikhq> bsmntbombdood: Nuclear power in-bunker. . .
04:18:23 <pikhq> The Internet link?
04:18:44 <pikhq> Well, if there's a nuclear war, I'm going to be more about *surviving* than about watching Homestar.
04:18:57 <pikhq> RProgrammer: Fine. In the same mountain.
04:19:13 <pikhq> . . . Hell; I'd just like to take over Cheyenne Mountain. :p
04:19:44 <RProgrammer> No, I mean by definition, unless you want to put every network node in your mountain, Internet will always be the first to go
04:19:45 <pikhq> Don't have to leave town, get an underground bunker. . .
04:19:53 <pikhq> RProgrammer: Yeah.
04:20:10 <pikhq> Obviously, the idea at that point is to live, not to be online.
04:20:48 <pikhq> Your server will *still* have rock-solid stability. . . Although nobody else will be able to tell, since they'll be dead.
04:20:52 <bsmntbombdood> We just have to lock ourselves in, and they'll never take it back from us
04:21:02 <RProgrammer> So, before it goes down, you download 73% of the internet into your 3 Petabytes of in-bunker storage
04:21:09 <pikhq> bsmntbombdood: Sure; only maybe an hour from here. :p
04:21:18 <pikhq> RProgrammer: Screw 3 petabytes.
04:21:35 <pikhq> I could afford at *least* 100 with all of the co-op deals I'd be getting.
04:21:36 <bsmntbombdood> http://en.wikipedia.org/wiki/Image:NORADBlast-Doors.jpg
04:22:19 <pikhq> RProgrammer: That, um, doesn't matter anymore. . .
04:22:35 <RodgerTheGreat> if there's a nuclear war, I'm going to want a couple good novels and an offline copy of Wikipedia
04:22:45 <pikhq> The addressing limit was gotten around for hard drives on 32 bit systems. . . In the late 90s.
04:23:15 <pikhq> And there aren't exactly many OSes which mmap the *entire drive*.
04:23:54 <RodgerTheGreat> RProgrammer: that's my point- it's rather simple to do, and it'd be *awesome* to have in such a situation
04:24:16 <pikhq> RProgrammer: Who said anything about NFS? I can use a *much* better system than that. . .
04:24:41 <RodgerTheGreat> plus maybe a complete dump of Bash.org for entertainment purposes and remembering the good 'ol days of IRC
04:24:50 * pikhq hems and haws. . . RAID using network block devices?
04:25:03 <RProgrammer> I despise ambiguous brand names (HD DVD, ...)
04:25:32 <pikhq> Hell. Just get an *orbiting* bunker. . . Large enough to contain everyone you care about.
04:26:01 <RodgerTheGreat> that's not at all the same and you know it. Underground is where it's at, unless by "orbiting", you mean "on the moon"
04:26:19 <pikhq> And one needn't worry about that, anyways; that only applies to addressing.
04:26:32 <pikhq> RodgerTheGreat: The moon is orbiting, isn't it?
04:26:45 <pikhq> Although I may want an orbiting *Mars-sized* bunker.
04:26:58 <pikhq> Carry the half of the planet you care about. :p
04:27:10 <pikhq> It's a very, very large one.
04:27:15 <RProgrammer> With that you wouldn't have to worry about the internet...
04:27:32 <RProgrammer> But you would have to worry about your own nuclear wars
04:27:35 <pikhq> Of course, it'd defeat the purpose by that point.
04:27:38 <RodgerTheGreat> and at that scale, you'd have a very large risk of nuclear war *inside* the bunker, which just creates the same problem all over again
04:28:17 <RProgrammer> Unless you rule a race of genetically engineered little people whose sole religion is to pay tribute in server space...
04:29:24 <pikhq> Or if I pick out only those who aren't a nuclear war risk.
04:29:37 * pikhq will avoid Vi users. :p
04:31:55 <RodgerTheGreat> "first rule of DynaMars station: Do not talk about the implausibility or inefficiency of DynaMars station."
04:32:16 <bsmntbombdood> http://static.wikipedia.org/downloads/November_2006/en/
04:32:53 <pikhq> Bit more for MySQL and MediaWiki.
04:34:03 <bsmntbombdood> ah crap, loading the list of pages crashed firefox
04:34:06 <pikhq> I'd still prefer a database dump.
04:34:11 <pikhq> Bit more flexible.
04:34:36 <RodgerTheGreat> the database dump would be more useful long-run, but the HTML would be handy for portable viewing
04:34:38 * pikhq is the kind of person who already *has* MySQL running, just for the hell of it.
04:35:03 <RProgrammer> 5.4 GB Compiled to HTML! And is that only Deflate, not even GZip?!
04:35:24 * RodgerTheGreat is the kind of person that'd want to view the archive from a Zaurus or something
04:35:38 <pikhq> Gzip is a compression format which uses the DEFLATE algorithm. . .
04:36:18 <RProgrammer> pikhq: I thought it was slightly better than LZ77
04:37:53 <RodgerTheGreat> I've found it *extremely* handy in the past having a copy of the Java API docs on my laptop
04:38:19 <fax> just noticed you can download wikipedia
04:38:51 <RProgrammer> Wiki markup uses templates, ultimate compression
04:39:03 <fax> information entropy ;(
04:39:09 <fax> think of the entropy?!
04:39:15 <RodgerTheGreat> I prefer to think of templates as "pre-compressing" data
04:39:19 <bsmntbombdood> I'm having trouble finding the sizes for the db dumps
04:39:28 <pikhq> And the Wiki database, IIRC, has compression on the live database.
04:39:32 <RodgerTheGreat> they do some of the same things a normal compression engine would do
04:39:48 <pikhq> (I could, of course, be wrong as fuck)
04:40:30 <RProgrammer> I'm having trouble uncloging my tubes (it's not a truck!), I accidently clicked on the 261M file and, note to others: Firefox loads it as text
04:41:58 <RProgrammer> Browsers should consider Content-Length, not just Content-Type when choosing to view or download
04:42:17 <RProgrammer> I would file a feature request if I wasn't lazy
04:42:52 <RodgerTheGreat> "Did you want to save this file to disk, instead of viewing it in your browser? [view] [download]"
04:45:44 <pikhq> "This file is 2000 Petabytes. Would you like to: [view] [download] [cancel request]"
04:46:08 <GregorR> Feels like a good moment to link http://www.codu.org/rxml.php
04:47:54 <bsmntbombdood> wow, "As of February of 2007, the entire collection of images produce a compressed tar.gz file of over 213 GB (gigabytes)."
04:47:57 <fax> thats like img2<Table>
04:49:03 <pikhq> bsmntbombdood: I think it uses something like:
04:49:21 <pikhq> <image src="base64:BASE64-STUFF" />
04:49:38 <pikhq> RProgrammer: That's a tarball of the JPEGs, PNGs, and SVGs.
04:49:41 <fax> why dont they just use SQL
04:49:53 <pikhq> fax: They also offer MySQL dumps.
04:50:16 <bsmntbombdood> http://meta.wikimedia.org/wiki/Data_dumps#What_happened_to_the_SQL_dumps.3F
04:50:24 <pikhq> . . . Apparently not.
04:50:56 <RProgrammer> Ok, my question is answered now, that site has monthly wikipedia HTML dumps. Up to April 2007
04:52:09 <pikhq> That's part of the idea.
04:56:31 <RodgerTheGreat> well, you could automate culling of stubs, but they often contain at least a small amount of information on the most esoteric topics on the Wikipedia
05:15:24 <bsmntbombdood> read line by line, searching the <title>...</title> lines, scan ahead to the <text ...>, display until </text>
05:16:02 -!- RProgrammer has left (?).
05:16:27 <fax> i have a HTML parser in CL if you want
05:16:43 <fax> well its not a full parser :P
05:18:50 <pikhq> And you need an XML parser, anyways.
05:22:45 -!- Sgeo has quit ("Ex-Chat").
05:26:53 -!- RainbowTrout has quit ("Chatzilla 0.9.77 [Firefox 1.5.0.11/2007031202]").
06:19:43 -!- calamari has quit ("Leaving").
06:56:45 -!- fax has changed nick to operator.
07:03:25 -!- operator has changed nick to fax.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
09:40:30 -!- sebbu has joined.
09:46:07 -!- calamari has joined.
09:51:56 -!- goban has joined.
10:32:12 -!- calamari has quit ("Leaving").
11:04:21 -!- sebbu2 has joined.
11:23:55 -!- sebbu has quit (Read error: 110 (Connection timed out)).
12:56:48 -!- sebbu2 has changed nick to sebbu.
15:43:50 -!- SevenInchBread has quit ("haaaaaaaaaa").
16:46:44 -!- jix__ has joined.
16:51:16 -!- jix__ has changed nick to jix.
16:58:20 -!- ihope__ has joined.
16:59:24 -!- pikhq has quit (Read error: 54 (Connection reset by peer)).
17:06:19 -!- ihope has quit (Read error: 60 (Operation timed out)).
17:39:37 -!- nazgjunk has joined.
18:26:06 -!- Sgeo has joined.
18:27:28 * SimonRC imagines the cartoon parkland scene from Mary Poppins, but with the music _Poisoning Pigeons in the Park_
18:49:46 -!- ihope__ has changed nick to ihope.
18:54:44 -!- oerjan has joined.
18:59:49 -!- fax has quit (" ").
20:01:39 -!- ihope has quit (Read error: 110 (Connection timed out)).
20:08:27 -!- pikhq has joined.
21:06:45 -!- ihope__ has joined.
21:06:56 -!- ihope__ has changed nick to ihope.
21:32:07 -!- jix has quit ("Bitte waehlen Sie eine Beerdigungnachricht").
21:46:17 -!- bd_ has left (?).
22:32:30 -!- ihope has quit (zelazny.freenode.net irc.freenode.net).
22:32:30 -!- sebbu has quit (zelazny.freenode.net irc.freenode.net).
22:32:44 -!- ihope has joined.
22:32:44 -!- sebbu has joined.
22:35:29 -!- ihope__ has joined.
22:53:30 -!- ihope has quit (Connection timed out).
23:02:43 -!- ihope__ has changed nick to ihope.
23:03:32 -!- sebbu has quit ("@+").
23:09:04 -!- SevenInchBread has joined.
23:43:48 -!- calamari has joined.
23:59:54 -!- Sgeo has quit ("Ex-Chat").