00:17:15 -!- Vorpal has quit (Ping timeout: 252 seconds).
00:22:41 -!- calamari has quit (Quit: Leaving).
00:23:51 -!- derdon has quit (Ping timeout: 252 seconds).
00:25:03 -!- derdon has joined.
00:28:54 -!- augur has joined.
01:21:38 -!- derdon has quit (Remote host closed the connection).
01:39:50 -!- sully has joined.
01:40:06 -!- hagb4rd2 has quit (Ping timeout: 276 seconds).
01:42:53 -!- hagb4rd has joined.
02:31:00 -!- oerjan has quit (Quit: Good night).
02:36:07 -!- Methead has joined.
02:43:38 <Methead> WEIRDEST CHEESE - what is it?
02:45:53 -!- Methead has quit.
03:10:40 -!- pikhq_ has joined.
03:11:06 -!- pikhq has quit (Ping timeout: 276 seconds).
03:40:29 -!- hagb4rd has quit (Ping timeout: 246 seconds).
03:40:45 -!- hagb4rd has joined.
03:52:54 <Sgeo> Yes, apt-get update after adding a ppa is in fact a good idea
03:53:08 <Sgeo> I think the original program was in the main repo
03:53:29 <Sgeo> But an older version
04:11:12 -!- pikhq has joined.
04:11:28 -!- pikhq_ has quit (Ping timeout: 265 seconds).
04:12:37 -!- MDude has changed nick to MSleep.
04:39:41 <tswett> Is "arxiv" properly pronounced like "arfourteen"?
04:41:41 <shachaf> That's what I get for typing random Unicode codepoints into my IRC client to find χ.
04:42:39 <Sgeo> Don't look very random
04:42:46 <Sgeo> Alpha beta gamma
04:42:51 <shachaf> Sgeo: That's the thing about randomness.
04:49:12 <quintopia> suppose we reconstructed Turing from the remains of his corpse and frog DNA (a process known as "Turing-completion"). Jurassic Park tells us such a Turing would be able to switch sexes if the situation called for it.
04:49:48 <quintopia> Question: Given this possibility, would Turings, as a species, choose to remain male? or would some choose to switch to catch just for the novelty?
05:10:30 -!- asiekierka has joined.
05:43:01 -!- hagb4rd has quit (Ping timeout: 252 seconds).
06:00:26 -!- hagb4rd has joined.
06:09:11 -!- asiekierka_ has joined.
06:09:26 -!- asiekierka_ has quit (Client Quit).
06:10:59 -!- hagb4rd has quit (Ping timeout: 246 seconds).
06:11:48 -!- hagb4rd has joined.
06:49:29 <pikhq> There is precisely one way to avoid tearing on X11.
06:50:15 <pikhq> If you do otherwise and *don't* tear, it is only because you're getting lucky.
06:50:27 <pikhq> And probably not updating the screen often.
06:50:49 <pikhq> Try updating the screen 60 times a second? It'll tear like a mofo.
06:53:24 <pikhq> My hate for X11 has grown 3 sizes this day.
06:55:56 <pikhq> That either tears or is using OpenGL!
06:56:13 <pikhq> The composite manager *itself* has to render via X, you know.
07:04:58 <pikhq> Wayland can't come fast enough. Then maybe my code using SDL would be reasonable, instead of tearing like crazy.
07:05:09 <pikhq> As-is, I need to rewrite it to use OpenGL. :(
07:05:50 <fizzie> How about Xv? That's not OpenGL, programs that use it update the screen a lot, and I haven't *noticed* any tearing.
07:06:09 <Deewiant> I notice a lot of tearing with it, but maybe that's just AMD.
07:06:14 <pikhq> fizzie: Okay, you either need to use OpenGL or be rendering in a *completely different colorspace*.
07:06:40 <pikhq> Oh, and BTW, the driver support for it is either reasonable or moronic.
07:06:48 <Jafet> pikhq, all the cool kids use OpenGL.
07:07:22 <pikhq> The proprietary AMD drivers, for instance, do the colorspace conversion to RGB in software, don't vsync, *and* don't even do the colorspace conversion right.
07:08:25 <pikhq> (the free AMD drivers just route it through OpenGL)
07:09:56 <pikhq> fizzie: Just to elaborate: your screen is in the RGB colorspace, with color values of 8-bit red, green, and blue samples. Xv does the YUV colorspace, with color values of 8-bit luminance, U chrominance, and V chrominance. And those values are in the range [16..235] rather than [0..255].
07:10:08 <fizzie> I know about colorspaces.
07:10:15 <fizzie> Anyway, there's a regular 8-bits-per-sample packed RGB visual in my xvinfo for the "NV05 Video Blitter" adapter, that's not a different colorspace. (The "NV17 Video Texture" adapter doesn't have one, though.)
07:10:32 <pikhq> Why, you might ask? Because broadcasters are utterly mad and hate programmers.
07:12:15 <fizzie> No RGB visuals on the "Intel(R) Textured Video". I suppose it's not such a common thing.
07:12:17 <Deewiant> Oh hey, http://wiki.x.org/wiki/RadeonFeature seems to finally say something decent about Evergreen. Maybe the free drivers are usable nowadays.
07:12:51 <pikhq> I've been using them for months.
07:13:24 <pikhq> About my only complaint is slightly lower FPS in OpenGL, and a bug that only seems to trigger when I try running a Source game in WINE.
07:13:38 <Deewiant> Last I checked, which was probably several months ago, the Evergreen column was too red for my liking.
07:13:53 <pikhq> And given that that only even works if you are lucky *anyways*, that's not too big a concern.
07:14:11 <Deewiant> How's the power saving i.e. fan loudness?
07:14:34 <pikhq> No noticable difference.
07:14:36 <Deewiant> That was one reason to go with the proprietary, for me.
07:14:45 <pikhq> Though, I've got a desktop system.
07:15:09 <Deewiant> Until I start up X, the fan is ridiculously loud.
07:15:25 <Deewiant> And, previously, with the free driver. But maybe I'll give it another shot when I have the time.
07:15:46 <pikhq> I've also found that for anything 2D, the free drivers beat the proprietary ones soundly.
07:16:02 <Deewiant> Yes, that was the case even several months ago. :-P
07:16:07 <pikhq> Mind you, Flash still looks like shit.
07:18:11 -!- MoALTz_ has joined.
07:20:50 -!- MoALTz has quit (Ping timeout: 250 seconds).
07:25:38 <hagb4rd> hi! i need to build a catalog-db storing information associated to any kind of file. neither i want to store any path/adresses nor the files themself, but some kind of a unique identifier (like a checksum). would you recommend any methods (like md5/sha-1 or whatever) when the primary goal is high performance (speed)
07:26:24 <hagb4rd> or maybe a totally different idea then a checksum?
07:28:16 -!- pikhq_ has joined.
07:31:09 -!- pikhq has quit (Ping timeout: 252 seconds).
07:47:14 <kmc> http://www.youtube.com/watch?v=Xk_XaJ7gE4Q
08:02:38 -!- Madoka-Kaname has quit (Remote host closed the connection).
08:04:38 -!- Madoka-Kaname has joined.
08:04:38 -!- Madoka-Kaname has quit (Changing host).
08:04:38 -!- Madoka-Kaname has joined.
08:09:17 -!- hagb4rd has quit (Read error: Connection reset by peer).
08:18:09 -!- itidus21 has quit (Quit: Leaving).
08:31:27 -!- elliott has joined.
08:36:06 <kmc> "By default, Java Ciphers (at least in Sun's implementations) are constructed in what is called Electronic Codebook (ECB) mode."
08:50:41 -!- monqy has quit (Quit: hello).
09:10:53 -!- cheater has quit (Ping timeout: 246 seconds).
09:21:10 -!- hagb4rd has joined.
09:24:32 <elliott> http://www.reddit.com/r/haskell/comments/r6idy/inmemory_keyvalue_store_in_c_go_and_python/
09:24:40 <elliott> This link has so much to do with Haskell!
09:25:40 <shachaf> elliott: Well, it's about the other three languages that exist other than Haskell.
09:26:12 <olsner> but it's relevant! because "it seems Go's RTS implements things similar to GHC's RTS, which means we could get similar performance in Haskell"
09:27:00 -!- calamari has joined.
09:27:18 <fizzie> It's relevant because there's the letter H in both Haskell and PytHon.
09:28:09 <kmc> <shachaf> elliott: Well, it's about the other three languages that exist other than Haskell.
09:28:17 <kmc> ^^^ ah, I see it's proposing a problem to be solved by the Haskell community
09:28:19 <shachaf> How many letters H are there in this here sentence.
09:28:20 <kmc> the existence of other languages
09:28:29 <shachaf> If you found all six, you = genius
09:28:41 <kmc> YOU'RE WINNER!
09:29:04 <Deewiant> > filter (`elem` "Hh") "How many letters H are there in this here sentence."
09:29:20 <fizzie> Deewiant: Apparently lambdabot is not a genius. :/
09:29:20 <kmc> > length "HHhhh"
09:29:30 <shachaf> lambdabot: you /= genius :-(
09:29:39 * shachaf curses his broken compose key.
09:30:00 <olsner> nice, C and Python versions create one thread per request, but the post also notes "None of my implementations are thread safe"
09:30:00 <fizzie> ≠ hey there is a sequence for it wow.
09:30:01 <kmc> shachaf used U+EF6614A1 'LATIN CAPITAL LETTER H ALTERNATE NON-PRINTING PRESENTATION FORM'
09:30:23 <fizzie> The "really silent h".
09:30:26 <olsner> > length $ filter (`elem` "Hh") "<shachaf> How many letters H are there in this here sentence."
09:30:37 <kmc> i wish i could say "i was using Compose before it was cool" but that was probably 10 years before i was born
09:30:48 <shachaf> kmc: Compose was always cool.
09:31:25 <shachaf> Wow, my compose key just broke.
09:31:32 <shachaf> I think it's because I ran dist-upgrade.
09:31:46 <shachaf> I don't want to close my X session. :-(
09:31:58 <shachaf> Before I dist-upgraded it warned me that I might have to close my X session.
09:32:02 -!- cheater has joined.
09:32:07 <olsner> hurr durr, "The C and Python versions perform similarly. They are spending all their time in the recv system call."
09:32:08 <kmc> rebooting X is obsolete
09:32:20 <kmc> you just need to port ksplice to patch X
09:32:23 <kmc> that should be Trivial right?
09:34:48 <fizzie> There's an actual compose key in the Sun keyboard, maybe I should try hooking it up to a regular computer. I'm sure I recall seeing a driver for it in Kconfig and all.
09:35:24 <elliott> olsner: We just need IP over ansible.
09:35:40 <olsner> just fork a new X, send the relevant state over, then let the new process take over all the existing connections
09:35:47 <elliott> Then we can make recv O(1).
09:37:11 <elliott> "Nice concept. We've completely done away with email registrations. Just get the user to like your Facebook page or follow you on twitter. That takes away the additional burden of verification, cause someone's already done it for you. No issues with promotion emails going to the spame folder. Facebook/ Twitter don't have them."
09:37:26 <elliott> Finally, someone has invented the most evil method of verification ever.
09:37:56 <shachaf> What about the verification method where...
09:38:04 <shachaf> I can't think of a way to complete that sentence. :-(
09:38:20 <elliott> ...you have to kill a kitten and send in videographic proof?
09:38:44 <shachaf> elliott: And then like them on Facebook?
09:39:03 <elliott> That's MARGINALLY more evil.
09:39:19 <olsner> kill a kitten, share the video on facebook, tag yourself
09:39:41 <elliott> I just call that Saturday.
09:40:12 <olsner> also known as kill-a-caturday
09:55:01 -!- ais523 has joined.
10:01:08 <HackEgo> KMC: WELCOME TO THE INTERNATIONAL HUB FOR ESOTERIC PROGRAMMING LANGUAGE DESIGN AND DEPLOYMENT! FOR MORE INFORMATION, CHECK OUT OUR WIKI: HTTP://ESOLANGS.ORG/WIKI/MAIN_PAGE
10:02:42 <shachaf> http://publib.boulder.ibm.com/infocenter/wsdatap/v3r8m1/index.jsp?topic=/xs40/convertingbetweenjsonandjsonx05.htm
10:03:53 <kmc> WIILKOMMEN AUF DER INTERNATIONALHUB FÜR EZOTERIKPROGRAMMINGSPRACHE DEZIGNUNG UND DEPLÖYMENTZE
10:04:28 <fizzie> They really should've made an uppercase ß.
10:04:28 <kmc> ZO RELAXEN UND WATSCHEN DER BLINKENLICHTEN
10:04:59 -!- calamari has quit (Quit: Leaving).
10:05:22 <kmc> shachaf, http://software.intel.com/en-us/articles/xml-parsing-accelerator-with-intel-streaming-simd-extensions-4-intel-sse4/
10:06:50 <shachaf> kmc: Pft. IBM will just sell you an XML coprocessor.
10:07:05 <shachaf> I think I first learned of this fact from you.
10:07:12 <kmc> "XML coprocessor" meaning they enable one of the processors already in your zSeries mainframe
10:07:28 <kmc> but the service of enabling it is cheaper than usual because they tell it to only process XML
10:07:40 <kmc> well, they also sell dedicated rackmount XML processing hardware
10:08:03 <kmc> keep fuckin that chicken
10:11:19 <kmc> fizzie, there sort of is one http://en.wikipedia.org/wiki/Capital_%C3%9F
10:11:39 <lambdabot> <no location info>: not an expression: `'
10:11:49 <kmc> oh right, :t doesn't understand unicode
10:11:58 <kmc> > generalCategory 'ẞ'
10:13:04 <kmc> "Dutch hip-hop in Atlanta / Tight new R&B in Estonia / And just another sick American in Yoshinoya"
10:13:16 <kmc> you were expecting LowercaseTwoLetters?
10:13:40 <kmc> "i accidentally entered the German dictionary twice and now I'm... two letters"
10:17:33 <elliott> "JSONx is an IBM® standard format to represent JSON as XML." awesome
10:21:48 <mroman> "Let's invent something so we do not need to parse XML!"
10:22:05 <shachaf> Why is it that some people come into #haskell and just don't get it?
10:22:10 <mroman> "Hey, cool. We can also display that thing we invented to not use XML in XML so we can parse it as XML!"
10:22:11 <kmc> don't get... what
10:22:40 <mroman> Patternmatching with lists?
10:23:23 <ais523> fizzie: uppercase ß is SS
10:23:24 <shachaf> Maybe I just know the syntax well enough that I get annoyed at people for not seeing the obvious which isn't really obvious, or something.
10:23:47 <shachaf> It seems like there've been a large number of people who've come into the channel and kept asking little variations on the same question.
10:24:11 <kmc> that's called homework
10:24:22 <mroman> > :t (\[x:xs] -> undefined)
10:24:23 <lambdabot> <no location info>: parse error on input `:'
10:24:38 <kmc> that's called it's nearly spring break and they gotta pass finals or midtterms or some shit
10:25:12 <mroman> @type (\[x:xs] -> undefined)
10:25:30 <mroman> If you can't derive it from the syntax at least ghci can tell you .
10:25:55 <shachaf> Maybe that's it. How do you make a channel where people don't ask you to do their homework?
10:26:26 <kmc> make a language so obscure nobody teaches it in school
10:26:33 <kmc> /join #agda
10:27:04 <fizzie> ais523: Sure, but that's just boring. The thing kmc linked to is much better, and in fact a bit like whatI thought it'd be.
10:27:32 <kmc> there are germans in mexico, germans in mexico, taking over tonight, falling in love with your daughter
10:27:39 <ais523> I accidentally discovered there was such a thing as uppercase schwa recently
10:27:46 <ais523> elliott: hmm, debatable; I've seen SS more commonly
10:28:12 <kmc> sing everybody deutsche deutsche vaya con Dios amigos
10:28:14 -!- sebbu has quit (Ping timeout: 246 seconds).
10:28:56 -!- itidus21 has joined.
10:29:53 * kmc reads core(5) manpage
10:30:14 <kmc> it has an ad hoc list of 7 different situations where corefiles will not be created, for security
10:30:19 <kmc> that's great
10:30:26 <kmc> i'm 100% sure they thought of them all
10:31:34 <shachaf> kmc: What would you prefer?
10:31:53 -!- sebbu has joined.
10:31:54 -!- sebbu has quit (Changing host).
10:31:54 -!- sebbu has joined.
10:32:10 <kmc> whitelisting > blacklisting
10:32:16 <cheater> the linux kernel is not an esolang
10:32:25 <kmc> there's a confused deputy problem here too
10:32:30 <shachaf> I don't see how whitelisting would be different in this situation.
10:32:40 <shachaf> They have a bunch of predicates for what makes a core file "safe".
10:32:57 <kmc> shachaf, it would be more conservative, e.g. corefiles only created in a particular directory or something
10:33:03 <kmc> i don't know, i don't have a specific proposal
10:33:29 <kmc> in fact i think the canonical ancient example of a confused deputy is something akin to dumping core
10:33:58 <kmc> http://cap-lore.com/CapTheory/ConfusedDeputy.html
10:34:23 <kmc> there was a linux exploit some years back which involved getting a setuid program to dump core in /etc/crontab.d/
10:35:04 <kmc> turns out cron will ignore "malformed lines" in a crontab
10:35:24 <kmc> where by "malformed lines" we mean most of an ELF coredump file
10:58:16 <ais523> elliott: fun fact: Verity turned out to have capability security by accident
10:58:21 <ais523> we didn't put it in there deliberately, it just happened
10:58:45 <mroman> http://www.reddit.com/r/haskell/comments/r5dxn/please_critique_my_very_simple_bmi_calculator/ <- srsly
11:00:18 <elliott> ais523: well, capability security is more about /not/ doing certain things, really
11:00:41 <ais523> want to light up an LED? someone has to pass you a function that lets you do that
11:01:01 <ais523> we have a linker where you just do "import <library>" and it'll automatically set up the required calls behind the scenes, though
11:03:25 <elliott> ais523: hey, what's the equivalent of "bibliography" for programming languages?
11:03:57 <ais523> hmm, what exactly do you mean by that? I can think of two possible meanings and am having problems expressing either of them
11:04:53 <elliott> well, what would you title the list of languages created by a person?
11:05:02 <elliott> ah, catseye uses "Lingography"
11:05:03 <ais523> oh, that wasn't either of them
11:05:14 <ais523> we can go with catseye's version, I think
11:05:20 <elliott> "(What is a "lingography", you ask? Well, if bands have discographies and directors have filmographies...)" -- right, possibly I should have listed more examples of the form
11:06:37 <elliott> ais523: (I realised [[Chris Pressey]] should probably have a comprehensive list of the languages we have articles on)
11:07:01 <elliott> except ideally I should include /all/ the esolangs he's created, so we have some nice redlinks, but that involves sorting his esoteric languages from his non-esoteric ones
11:07:06 <elliott> which is probably impossible
11:10:35 -!- hagb4rd has quit (Ping timeout: 246 seconds).
11:11:34 <elliott> hmm, cpressey's favourite esolangs are Homespring, Muriel, and Please Porige Hot
11:12:02 <ais523> admittedly, I like Homespring and Muriel too
11:12:03 <elliott> "What are my favourite esolangs, you ask? Well, the three I listed earlier (Homespring, Muriel, and Please Porige Hot) hold special places in my heart. The favourite esolangs of my own is a much harder question to answer. I will say that, in terms of striking a balance between "challenge to code in" and "beauty", I think Mascarpone is one of my best. That doesn't necessarily mean it's my favourite, though."
11:12:08 <elliott> (http://catseye.tc/cpressey/retrospective.html)
11:12:38 <ais523> actually, I should start using Homespring rather than Haifu as an example of a good thematic language
11:12:44 <ais523> they both are, but Homespring is better
11:12:56 <ais523> it's an acronym, after all
11:13:18 <elliott> it's Homespring or HOtMEfSPRIbNG
11:13:19 <ais523> that's the expanded version
11:14:03 <elliott> hmm, the "documentation" link on http://esolangs.org/wiki/Homespring demands a password
11:14:09 <elliott> and gives an email address to ask for the password if you don't know it
11:15:07 <ais523> there's also documentation in the esoteric files archive, though
11:15:20 <elliott> i'm wondering if i should just remove the link
11:17:31 <elliott> "Okapi is a language I designed as a present for my true love. In it, the only means of control flow is throwing exceptions, and as if this wasn't enough, there are two restrictions on exceptions that are thrown — they must be divide-by-zero exceptions, and they must be caught in a lexically enclosing block. Nor is there any facility to "retry" after an exception is caught. The language is nonetheless Turing-complete."
11:17:41 <elliott> i wish there were more details than that available, that's amazing
11:19:53 <elliott> RocketJSquirrel: BTW, I believe an Eightebed-like language can avoid having most of a garbage collector and still not leak memory
11:20:04 <ais523> elliott: what are Linux DEs like nowadays?
11:20:23 <elliott> RocketJSquirrel: hmm... or at the very least, avoid running the garbage collector machinery more than once in e.g. 2^64 allocations
11:20:44 <ais523> I'm still on Gnome 2, but I'm going to have to upgrade eventually
11:20:52 <ais523> and am wondering if there are any I'll find usable
11:21:04 <ais523> (note: I consider Windows 7 usable; not on default settings, but it only requires small changes)
11:21:26 <elliott> it's fairly likely you'll find Xfce tolerable. I'm happy with xmonad.
11:21:39 <elliott> GNOME and KDE haven't spontaneously fixed themselves, obviously
11:21:50 <ais523> pity, I was rather hoping they would
11:22:03 <ais523> or at least, that someone had got gnome-panel working properly in gnome 3
11:22:08 <ais523> that's what I'd want, really
11:22:30 <elliott> ais523: gnome 3 does have a gnome-panel, the "fallback" mode
11:22:34 <ais523> keep the applications, use the old window manager and panel application
11:22:52 <ais523> crazy idea: is it possible to use gnome 2's gnome-panel in gnome 3?
11:23:39 <ais523> I guess I'll try Unity first, and probably decide I don't like it
11:24:20 <elliott> ais523: why not just try Xfce first?
11:24:52 <ais523> I guess trying the options that you expect not to succeed first is likely to produce a better decision in the end
11:25:02 <ais523> also note that I used an earlier version of KDE 4 for a while
11:25:07 <ais523> and it was mostly usable
11:30:04 -!- hagb4rd has joined.
11:31:54 <cheater> the only means of control flow is throwing exceptions < has he mentioned whether it's just prolog
11:33:50 <ais523> a fail isn't really quite the same thing as an exception, although they're similar
11:33:55 <ais523> and a cut is not the same thing as an exception at all
11:34:05 <ais523> and a nonlocal cut is not at all related
11:34:45 <cheater> i realize, i just think they're similar ideas.
11:35:18 <cheater> you visit an AST and backtrack.
11:35:59 <ais523> it's possible to implement fails as exceptions
11:36:18 <ais523> I was showing my students how to do that in Java, for if they wanted to write a parser that wasn't LL(1)
11:36:32 <hagb4rd> good thing bout throwing excpetions is that you can decide how to handle them in the calling routine
11:36:40 <ais523> (which was a surprisingly possible thing, as they had to write a parser, and had been shown how to write LL(1) parsers which was sufficient for the exercise)
11:38:51 <hagb4rd> which enables you to provide fair encapsulation of your libs
11:41:49 -!- Jafet has quit (Quit: Leaving.).
11:42:48 <ais523> hagb4rd: don't all error handling mechanisms do that?
11:46:39 <hagb4rd> well, i'm not sure about this, because i don't know them _all_
11:48:20 <hagb4rd> but there is a difference between just handling a return value (in which you also can place an errorcode or sth) and throwing an exception
11:49:12 <hagb4rd> or even a greater difference between handling a possible error _inside_ of the function/class
11:52:58 <hagb4rd> also the return value and an exeption _can_ have a different type, which allows you pass valueable information inside the exception-object
11:53:37 -!- Patashu has joined.
11:59:10 <ais523> oh right, I forgot about the antipattern where when anything goes wrong, you print it to stdout and keep on going
12:04:39 -!- Jafet has joined.
12:06:58 -!- augur has quit (Remote host closed the connection).
12:11:23 <hagb4rd> lets keep things simple for an easy example: imagine i use one of your classes (i don't care how they work ;) in which a method needs to open a file. for the case the file does not exist you could throw a FileNotExists exception (or sth) allowing me to decide how to handle it; i could retry with another file or depending on the runmode write a log or whatever (not so when you handle it by your own)
12:13:09 <hagb4rd> or you could check the permissions and throw a FilePermissionNotAllowed Exception allowing me to handle it another way
12:14:14 <elliott> That isn't related to exceptions. Returning an error code has the same behaviour.
12:14:24 <elliott> What exceptions do is propagate automatically.
12:15:26 <hagb4rd> when FileNotExists Exception and FilePermissionNotAllowed Exception both inherit from Exception (which at least has an errorcode and errormessage), i can just handle the base class properties and output them .. halt the program ..and and and
12:19:23 <fizzie> http://en.wikipedia.org/wiki/Currying#Example_in_Java + Java 8 = http://p.zem.fi/m0yi
12:19:41 <fizzie> One more step and it's pretty much Haskell, right, guys?
12:21:25 <kmc> need moar monads
12:21:26 -!- nortti has joined.
12:21:52 <kmc> you can't add three numbers in Haskell without using at least five monads
12:21:54 <kmc> everyone knows that
12:21:55 <hagb4rd> sure but its absolutely not what java is written for. ugly
12:24:48 -!- azaq23 has joined.
12:24:59 -!- azaq23 has quit (Max SendQ exceeded).
12:25:12 -!- azaq23 has joined.
12:25:21 <fizzie> For some unfathomable reason the horrible Java thing is the only example in the article. Based on the talk page there's been a whole pile of them, earlier.
12:29:29 <elliott> fizzie: Wikipedians love piling on ten thousand language examples.
12:29:47 <elliott> fizzie: Take a look at this: http://en.wikipedia.org/wiki/Fork_bomb#Examples
12:30:28 <elliott> That Java example in Currying is terrible though, how bad must the other examples have been for that to be the one spared?
12:30:58 <fizzie> Based on the history it seems that first the "Profusion of examples" (to quote talk page title) was removed, and then recently-ish that one was added, perhaps because there were no examples.
12:31:35 <fizzie> Also seems that from 16 Dec to 12 Mar the Java example that was there was a broken one.
12:31:50 <elliott> <!-- NB: creating a bunch of threads does not constitute a forkbomb, it must somehow call the system's fork() function! -->
12:31:55 <elliott> WHO THE FUCK IS ADDING COMMENTS TO THIS SHIT
12:32:12 <kmc> const int main[] = { 14776, 3942977280, 247 };
12:33:01 <fizzie> I like the example that's written "in C standard library"; not e.g. using the library, but in it.
12:33:05 -!- ais523 has quit (Ping timeout: 248 seconds).
12:33:46 <elliott> And of course the majority of revisions are messing with the examples: http://en.wikipedia.org/w/index.php?title=Fork_bomb&action=history
12:34:20 <kmc> breaking news: people waste time on the internet
12:34:36 <kmc> film at 11, just after cute videos of a cat and before some more cute videos of a cat
12:34:57 <elliott> fizzie: For the second time. The last time I did so when it came up in here, another person in the channel decided to revert-war to keep it.
12:35:05 <elliott> I was hoping you'd stay silent. :'(
12:35:54 <elliott> <elliott> RocketJSquirrel: BTW, I believe an Eightebed-like language can avoid having most of a garbage collector and still not leak memory <elliott> RocketJSquirrel: hmm... or at the very least, avoid running the garbage collector machinery more than once in e.g. 2^64 allocations
12:36:34 <elliott> "fork Fork Bomb fork fork fork fork wabbit" -- http://zh.wikipedia.org/wiki/Fork%E7%82%B8%E5%BC%B9
12:36:39 <nortti> kmc: how does your forbomb even work? It says "Illegal instruction" on my machine
12:36:50 <kmc> nortti, it's for amd64 Linux specifically
12:37:11 <hagb4rd> it only runs on his washing machine
12:37:19 <nortti> kmc: ok. I am using PPC 7450 Mac OS X
12:37:20 <kmc> on which it's an array of three 4-byte little endian integers
12:37:54 <elliott> RocketJSquirrel: How about this: You represent every pointer as (realptr, id). Whenever an allocation is made, the global id counter is incremented by one, stored just before the allocated memory, and returned with the pointer. Whenever you dereference a pointer, it checks that the identifier matches that of the pointer's; if it doesn't, it safely fails. Whenever the id counter is about to wrap-around, you do the standard GC thing to reassign all
12:37:59 <kmc> if you write it down as such and then decode the bytes as amd64 machine code, you get
12:38:19 <kmc> foo: mov $0x39, %eax; syscall; jmp foo
12:38:26 <kmc> and 0x39 is the syscall number for fork()
12:38:27 <elliott> RocketJSquirrel: Admittedly that just puts off the problem :P
12:39:12 <kmc> the array is 'const' because many distros forbid executing writable memory (unless you explicitly ask to)
12:39:24 <kmc> and fewer (but still some) forbid executing read-only data
12:39:37 <elliott> RocketJSquirrel: Sure, or you could do the same "if valid" thing as Eightebed does.
12:39:51 <elliott> Or define the dereferencing of an invalidated pointer to always return 0, or whatever; point is, no corruption.
12:40:40 <elliott> RocketJSquirrel: Actually this is basically the same as never freeing any memory and GCing when you run out, except that the resource to run out of increases much slower, because it's independent of the allocation size.
12:40:51 <elliott> Still, doesn't truly solve the problem :P
12:41:02 <RocketJSquirrel> elliott: Wrapping around is a bigger issue than you've let on, it actually has to rewrite all existing IDs.
12:41:16 <elliott> Whenever the id counter is about to wrap-around, you do the standard GC thing to reassign all
12:41:19 <elliott> <elliott> the identifiers.
12:41:51 <RocketJSquirrel> The question is, is that more a GC than the previous strategy?
12:41:58 <elliott> It's a rarer but stronger GC :P
12:42:53 <elliott> RocketJSquirrel: OK, my next idea is that you don't have any global counter. Memory is write-once: you allocate a slab of memory, populate it (under restricted conditions), and then freeze it, which is when you get a pointer back. That pointer comes bundled with a cryptographic hash of the data. When you dereference it, it checks the data matches the hash, and when you free it, it's released to be reused as normal.
12:43:04 <elliott> RocketJSquirrel: Now corruption is just rare! (Rare = probability 1)
12:43:53 <elliott> RocketJSquirrel: More seriously, I'm PRETTY sure you could solve it with linear types or the like.
12:44:00 <RocketJSquirrel> You Haskell people and your solving problems with immutability.
12:44:41 <elliott> RocketJSquirrel: In fact it's pretty much the problem all the applications of linear logic to programming are designed to solve :P
12:44:51 <elliott> i.e. deterministic, explicit resource allocation with safety.
12:44:57 -!- oerjan has joined.
12:47:51 -!- oerjan has set topic: This is CERTAINLY the MOST TERRIFYING GAMALOST you will RUN AWAY FROM all OLSOK: http://championofbirds.com/?p=4991 | http://codu.org/logs/_esoteric/.
12:48:01 <oerjan> <Methead> WEIRDEST CHEESE - what is it?
12:48:07 <elliott> Is it (provably) true that any function can be curried? For example, I'd expect that y = f(a,b) = ((a+b) / (a*b)) could not be curried. — Preceding unsigned comment added by 87.194.171.29 (talk) 23:19, 25 October 2011 (UTC)
12:48:10 <oerjan> I THINK THIS SHOULD NOW BE MORE CLEAR
12:53:26 -!- Phantom_Hoover has joined.
12:53:27 <oerjan> <pikhq> Try updating the screen 60 times a second? It'll tear like a mofo. <-- i always knew the path of programming was full of tears.
12:55:19 <elliott> oerjan: hey how do you add articles to your WP watchlist. do i have to use the raw interface? it doesn't have the "watch" link in the dropdown next to view history like Esolang does...
12:55:37 <oerjan> elliott: the star button
12:56:30 <oerjan> also, there's a checkbox when you edit.
12:56:59 <elliott> yeah, i knew that, but didn't have any edits to make :)
12:57:11 <oerjan> just thought i'd mention.
12:57:55 <oerjan> i was a bit confused when it initially changed to a star too
12:58:49 <elliott> wow, how does [[Befunge]] still exist?
12:58:59 <elliott> i have a hard time believing it meets any notability standard.
12:59:12 <elliott> let's see... the citations are the Jargon File, and an Esolang talk page.
12:59:26 <elliott> "however, it has been shown that Befunge-93 is Turing Complete with unbounded stack word size"
12:59:36 <elliott> it's not original research, it was published on another wiki's talk page!
12:59:50 <oerjan> hey we do _science_ on esolang!
12:59:59 <Phantom_Hoover> "the Befunge-93 language is, unlike most machine languages, not Turing-complete"
13:01:33 <oerjan> elliott: you could add http://scientopia.org/blogs/goodmath/2009/09/08/two-dimensional-pathology-befunge/
13:02:01 <elliott> yay now i have 10 pages on my watchlist.
13:02:25 <elliott> oerjan: i don't think adding a citation to a blog will bolster the notability case :P
13:02:30 <elliott> oerjan: although maybe good math, bad math is prominent enough for it to.
13:02:42 <elliott> i am rather sceptical that Befunge /is/ notable, anyway
13:03:08 <oerjan> i _may_ have more, but it's quite an eclectic bunch. i have [[Gorse]] and [[Ben Murphy]] as a result of some spammer who hit them once...
13:03:52 <oerjan> elliott: ok but the good math/bad math features were probably some of our best exposition
13:04:02 * elliott assumed oerjan must have hundreds of pages on his watchlist.
13:04:10 <elliott> since I keep hearing "oh yes, I have it on my watchlist" :P
13:04:15 <oerjan> although he rarely linked to the wiki, iirc
13:07:24 <elliott> i wish you could somehow watchlist a page and every page "like" it.
13:07:36 <elliott> say, it would watchlist that page and every page it links to that shares a certain number of categories, or such
13:08:01 <elliott> so i could just manually watch a few central pages and have the rest filter through.
13:08:36 <elliott> i suppose the simpler version of that is just watching every article in a certain wikiproject.
13:09:35 -!- augur has joined.
13:11:09 <elliott> oerjan: hm about the featured esolang process. would it be better if no language suggested last time could be suggested post-clearing? rather than just not by the same person
13:11:26 -!- hagb4rd2 has joined.
13:11:28 -!- hagb4rd has quit (Ping timeout: 252 seconds).
13:11:31 <elliott> i suppose that has the problem that people might propose a large fraction of the good esolangs on a particularly active list and thus starve the following list of candidates.
13:11:43 <elliott> but otoh it'd stop the same suggestions coming up again and again and promote some diversity.
13:21:30 <oerjan> <HackEgo> KMC: WELCOME TO THE INTERNATIONAL HUB FOR ESOTERIC PROGRAMMING LANGUAGE DESIGN AND DEPLOYMENT! FOR MORE INFORMATION, CHECK OUT OUR WIKI: HTTP://ESOLANGS.ORG/WIKI/MAIN_PAGE
13:21:35 <oerjan> LINK STILL NOT WORKING
13:24:39 <hagb4rd2> i think you should write a step by step tutorial on howto implement an esoteric language, so maybe someone less susceptible to forces of the one ring may show up and breed something completely new fresh and different
13:25:59 -!- Madoka-Kaname has quit (Quit: Hug~♪).
13:27:09 <oerjan> 1. DON'T USE ANY OF THE FOLLOWING AS COMMANDS: +-<>.,[] 2. DON'T USE COMMANDS MEANING ANY OF THE FOLLOWING: INCREMENT, DECREMENT, READ CHARACTER/BYTE, WRITE CHARACTER/BYTE, BEGIN LOOP, END LOOP, MOVE LEFT, MOVE RIGHT 3. DON'T USE A TAPE OR ARRAY AS YOUR MAIN STORAGE.
13:27:15 <oerjan> Phantom_Hoover: does that about cover it?
13:27:37 <elliott> hagb4rd2: like a step-by-step guide on how to create a beautiful painting.
13:27:48 <elliott> or a brilliant composition.
13:27:57 <hagb4rd2> more on how to make a pencil and get some colors
13:29:03 <Phantom_Hoover> No, you forgot "if you do any of these things just kill yourself and save us the bother".
13:29:32 <oerjan> elliott: hey he killed himself already. twice.
13:32:02 -!- MSleep has changed nick to MDude.
13:33:16 <oerjan> <elliott> ais523: Ss, no? <-- since it doesn't actually occur at the start of words it would probably always be all caps, so SS as ais523 says
13:33:20 <fizzie> That wiped out quite a few people; especially the second bit, which implicated every language that has a command like "increment" or "begin loop".
13:33:52 <elliott> oerjan: yay i am now the second-top editor on esolang :P
13:33:57 -!- Patashu has quit (Quit: MSN: Patashu@hotmail.com , Gmail: Patashu0@gmail.com , AIM: Patashu0 , YIM: patashu2 , Skype: patashu0 .).
13:34:17 <Phantom_Hoover> fizzie, you could be merciful and say that most non-BF derivatives wouldn't really do that with instructions.
13:34:56 <elliott> fizzie: I have gnuplot, and a big ol' sorted list of numbers; what's the minimum necessary gnuplot incantations I need to know to plot the distribution?
13:35:04 <elliott> I want to see what Esolang's distribution of edit counts looks like.
13:35:45 <fizzie> elliott: Ho-hum, so did you want gnuplot to compute the histogram for you?
13:35:56 <elliott> fizzie: Well, I don't really care, I just want something that looks sciency.
13:37:18 <oerjan> fizzie: hm that _is_ a point, we shouldn't have rules that exclude Deadfish.
13:37:43 <fizzie> I mean, if you have something like "1 1 1 2 3 4 5 5 5 5 5 5", do you want to have 5 points at heights 3, 1, 1, 1 and 6, respectively?
13:38:24 <elliott> Yes, I know what a histogram is. (Except the ranges are going to have to be bigger than that, considering how tiny the data is. :p)
13:42:08 <fizzie> It needs a bit of manual fiddling, but something like plot "your.data" using (floor($1)):(1) smooth freq with boxes
13:42:28 <fizzie> If you want wider-than-1 bins, K*floor($1/K).
13:42:40 <elliott> What manual fiddling would that be, roughly?
13:44:03 <elliott> elliott@solidity:~$ mysql esolang_wiki <<< 'select user_editcount from user order by user_editcount desc' >editcounts
13:44:30 <fizzie> Well, I just mean that you need to manually convert your numbers from x into (z,1) where 'z' is the bin center for that point; the 'using' spec achieves that. (And then "smooth freq" means it sums up all points with identical 'x' coordinates.)
13:46:17 <elliott> fizzie: Aww. That produces exceedingly boring results thanks to the sheer number of 0s.
13:46:40 <elliott> Can I get some logarithms or something? :p
13:47:27 <fizzie> Sure, or you could just ignore the 0s by setting the viewport suitably.
13:47:45 <fizzie> plot [xmin:xmax] [ymin:ymax] "your.data" ... sets the area which it draws.
13:48:20 <fizzie> And, err, I guess you'll need to change the "using" spec if you want 0s to end up in a bin of their own.
13:48:32 <elliott> fizzie: Well, the thing is that I'm grouping them into bins of 50. And ignoring all users with 50 or less edits will result in... few users.
13:49:27 <fizzie> Right. Though if you group into bins of 50, you will in any case only have a single number for "<50".
13:50:18 <elliott> fizzie: I could just grep -v out the 0s from the raw data.
13:50:28 <elliott> Can gnuplot take data from a shell line? :p
13:51:29 <fizzie> You could plot using (50*floor(($1-1)/50)+26):(1):(20) so that 0 ends up in a single bar (at -24), and 1..50 are in one bar at 26, and 51..100 in another, and so on.
13:52:47 <fizzie> The file name '-' is special, but I can't quite recall how it was special.
13:52:51 <elliott> That results in a slightly less boring box. I think I want to increase the size of the bins as the numbers get larger, or something, since otherwise it's all going to clump into basically one bin.
13:52:51 <oerjan> <elliott> And of course the majority of revisions are messing with the examples: http://en.wikipedia.org/w/index.php?title=Fork_bomb&action=history
13:53:04 <oerjan> what you don't realize is that the examples _are_ a fork bomb
13:54:29 <fizzie> You can do log() for natural log in the using line, but you'd need to calculate the proper constants if you want some particular bin edges.
13:56:21 <itidus21> DMM has the best wiki picture ever
13:56:32 -!- Vorpal has joined.
13:56:41 <elliott> Maybe I'll just plot the editcounts as-is. :p
13:56:52 <itidus21> oh i see harry potter behind him
13:57:11 <nortti> `run apropos www-browser
13:57:14 <HackEgo> apropos: can't open the manpath configuration file /etc/manpath.config
13:59:06 <HackEgo> apropos: can't open the manpath configuration file /etc/manpath.config
13:59:26 <Vorpal> someone who knows what nick Gregor currently uses should poke him about that.
13:59:43 <elliott> I don't think it's broken.
13:59:54 <elliott> Gregor is currently TodPunk.
14:00:15 <fizzie> One alternative is to start octave and then do d = load("your.data"); hist(d, [0 logspace(1, max, n)]); -- that will produce a histogram with bins where there's a bin at 0, and then 'n' more bins with centers between 1 .. 'max', logarithmically spaced.
14:00:29 <elliott> fizzie: You forgot step 0: install Octave.
14:00:41 <fizzie> It probably works in MATLAB too. :p
14:00:49 <elliott> Hey, Octave uses FLTK. Or one of its dependencies does.
14:01:32 <fizzie> Maybe it's the plotting. I think it has supported several things.
14:01:53 <fizzie> The polish on Octave's plotting system is... slightly lacking, compared to MATLAB. You do get some windows, but that's about it.
14:02:20 <azaq23> I am not one of the people that were named
14:02:22 <elliott> It can use gnuplot. Or so sayeth the optional dependencies list; "alternative plotting".
14:02:48 <elliott> octave:1> d = load("editcounts")
14:02:48 <elliott> error: load: failed to read matrix from file `editcounts'
14:02:51 <elliott> It doesn't work, fizzie. :(
14:02:57 <fizzie> It worked for *my* data.
14:03:05 <fizzie> What's the file like? Anything extra in there?
14:03:12 <elliott> Oh, the first line is "user_editcount".
14:03:14 <fizzie> Just plain numbers, one per line, should work.
14:03:19 <elliott> Now I wonder how gnuplot managed to plot it.
14:03:33 <fizzie> It has some smarts about automatically skipping header-looking things.
14:03:45 <elliott> octave:2> hist(d, [0 logspace(1, max, n)]);
14:03:45 <elliott> error: Invalid call to max. Correct usage is:
14:03:45 <elliott> -- Loadable Function: max (X)
14:03:45 <elliott> -- Loadable Function: max (X, Y)
14:03:45 <elliott> -- Loadable Function: max (X, [], DIM)
14:03:45 <elliott> -- Loadable Function: max (X, Y, DIM)
14:03:47 <elliott> -- Loadable Function: [W, IW] = max (X)
14:03:59 <fizzie> That was "max" as in "put some value there". :p
14:04:04 <fizzie> But you can put max(d) in.
14:04:21 <fizzie> Also you need to specify an 'n'.
14:04:48 -!- nortti has quit (Quit: ChatZilla 0.9.88.1 [Firefox 11.0/20120312200651]).
14:04:54 <fizzie> It doesn't probably widen the bars when plotting, so it might look a bit silly.
14:05:31 <elliott> Well, that fails horribly.
14:05:44 <elliott> http://sprunge.us/CINY Many errors like this.
14:06:14 <elliott> Then a chart with one point at 0 and some very improbable things on the axes (e.g. 1e+212).
14:06:57 <fizzie> You could try saying backend("fltk") first.
14:07:36 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: backend': not found
14:07:46 <fizzie> octave:16> available_backends
14:07:46 <fizzie> { [1,1] = fltk [1,2] = gnuplot
14:08:30 <elliott> `available_backends' undefined
14:08:32 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: available_backends': not found
14:08:43 <fizzie> This is Octave 3.2.4 that we have here, for the record.
14:08:56 <fizzie> Well, it's obviously too new. :p
14:09:09 -!- hagb4rd2 has quit (Read error: Connection reset by peer).
14:10:03 <Vorpal> fizzie, is that "available_backends" a 2D array?
14:10:11 <Vorpal> those indices look rather weird
14:11:05 <fizzie> It's an 1x2 cell array, yes.
14:11:28 <fizzie> I'm not sure if octave treats all cell arrays as at least two-dimensional, though.
14:11:44 <fizzie> elliott: Apparently it's available_graphics_toolkits() nowadays.
14:12:06 <fizzie> Well, maybe graphics_toolkit("fltk") will then work.
14:12:11 <Vorpal> what is it you are trying to plot?
14:12:38 <fizzie> Edit count distribution histogram with logscale on the X axis, I think.
14:13:19 <Vorpal> can't gnuplot do the job? Pretty sure there are logscale options in there
14:13:41 <fizzie> Even just getting it to do the summing involved for a histogram is quite a manual job.
14:16:08 <fizzie> elliott: You want hist(d, [0 logspace(0, log(max(d)), n)]) or some-such.
14:16:15 <elliott> Just as I switch back to gnuplot to try and plot the raw data with some exponentially axes. :p
14:16:49 <elliott> fizzie: That still produces some fairly improbable axes.
14:17:10 <Vorpal> the wiki has been very active
14:17:56 <fizzie> I don't know how I managed to forget the thing where logspace(a, b) means "from 10^a to 10^b except if b is pi".
14:18:03 <oerjan> Vorpal: we now have a featured article!
14:18:12 <Vorpal> fizzie, what does it mean if b is pi?
14:18:25 <fizzie> Vorpal: Then it means "from 10^a to pi".
14:18:31 <oerjan> Vorpal: only got up and running this week
14:18:41 <fizzie> Vorpal: But it must be the single single- or double-precision float that's closest to pi.
14:18:55 <fizzie> Vorpal: For compatibility with MATLAB 'logspace' function.
14:19:12 <elliott> fizzie: Okay, that... "works".
14:19:14 <Vorpal> that just raises the new question: Why does MATLAB do it that way?
14:19:49 <elliott> oerjan: let's put it this way: if we had a superb, Wikipedia-featured-article-level article on Snack, would you want it to be featured on the main page as exemplary?
14:19:52 <fizzie> Vorpal: "y = logspace(a,pi) generates the points between 10^a and pi, which is useful for digital signal processing where frequencies over this interval go around the unit circle."
14:20:00 <fizzie> Vorpal: It's a rather curiously specific application.
14:20:24 <Vorpal> fizzie, what if you actually want from 10^a to 10^pi ?
14:20:28 <elliott> fizzie: The problem now is that it still looks like a really tall line and then a tiny pile of moss surrounded in a sea of nothingness. :p
14:21:12 <fizzie> Vorpal: Possibly you just then add an epsilon to it or something.
14:21:13 <Vorpal> elliott, isn't that what the logarithmic bit is supposed to deal with?
14:21:18 <oerjan> elliott: maybe some April :P
14:21:26 <Vorpal> elliott, and surely you could nest it in another logspace call
14:22:04 <oerjan> admittedly, we have much better _jokes_ than that still
14:22:29 <fizzie> elliott: You can always try doing set(gca,'yscale','log') after the plot to see if it's any better.
14:22:36 <itidus21> its my plotting of the function
14:22:51 <elliott> fizzie: That blanks the entire chart.
14:23:05 <elliott> warning: axis: omitting non-positive data in log plot
14:23:19 <Vorpal> elliott, non-positive edit counts?
14:23:34 <fizzie> elliott: Well, that's impressive. I guess it doesn't like the negative infinities that the bars have.
14:23:37 * oerjan relatedly recalls that the Wumo comic has both a best of week and worst of week link. sometimes they are the same comic.
14:25:05 <fizzie> elliott: If you want just an outline of the shape with logscale y, try [y,x] = hist(...); semilogy(x,y); or something.
14:25:06 <oerjan> (note: sometimes/frequently NSF[WL])
14:25:50 <elliott> fizzie: Can I just hire you to science it?
14:26:03 * Sgeo is up early for reasons
14:26:32 <fizzie> If it looks boring, then it looks boring. But if you want to dump the numbers somewhere, I can take a look, except only after some three hours or so. I'll have to go do a thing soon, first.
14:27:02 <elliott> Don't I have to ANONYMISE THEM for PRIVACY or something?!?!?!??!
14:28:15 <oerjan> elliott: _technically_ all the information is publically available, isn't it? just not conveniently.
14:28:48 <elliott> oerjan: It's also not tied to the user IDs, so there's very little privacy being violated. :p
14:29:19 -!- hagb4rd has joined.
14:29:57 <elliott> But, you know. *Someone* has 1901 edits.
14:30:14 <elliott> That violates oerjan's privacy by revealing the number, even if the fact that it's oerjan's edit count isn't revealed at all.
14:30:21 -!- derdon has joined.
14:30:37 <elliott> It would be irresponsible of me to reveal that data, therefore, as it would be a serious violation of oerjan's trust.
14:31:03 <elliott> Don't worry, if I wanted to be malicious I'd start logging all your passwords or something. Er, did I say something?
14:31:33 <Vorpal> elliott, have you tried the technic mod pack for minecraft?
14:31:35 <elliott> Remember, folks, never use the same password on multiple sites!
14:31:51 <oerjan> AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
14:32:46 <elliott> I get the feeling oerjan reuses passwords.
14:33:00 <Vorpal> it is kind of cool, but the main problem is that towards the mid game you are 1) want resources 2) have tools that can do significant damage to the landscape. Which means that whatever place you decide to mine at will end up looking like a wasteland.
14:34:18 <Vorpal> which you realise didn't look so good next to your nice house
14:34:48 <elliott> "Church encoding (aka Visitor Pattern)"
14:35:05 <oerjan> but it's a sound policy: i reuse a _different_ password for net banking and public services. the kind that just yesterday managed to give thousands of people access to a single person's tax reporting account.
14:35:38 <Vorpal> elliott, btw, does mediawiki store passwords hashed?
14:37:13 <oerjan> (yesterday was supposed to be the day of the year most people in norway can start doing tax reports, and the system is generally horribly flaky that day.)
14:37:19 <fizzie> oerjan: These are the (length-sorted) lines you've said in my logs where there's just ten or more A's and nothing else: http://p.zem.fi/ca24
14:37:21 <Vorpal> btw I think I only reused passwords once in recent years, and that is the same local login passwords to a couple of old, rarely used computers I have. Mostly use them for when I happen to need serial ports
14:37:25 <oerjan> *reports via the internet
14:37:43 <Vorpal> oerjan, wait, net banking and password?
14:37:49 <oerjan> fizzie: that's surprisingly
14:37:54 <Vorpal> don't you use hard certificates over there?
14:38:00 <oerjan> Vorpal: there's a password + a hard certificate.
14:38:34 <oerjan> *fizzie: that's surprisingly few
14:38:50 <Vorpal> fizzie, since the channel started?
14:39:05 <Vorpal> (well, since you started logging)
14:39:24 <elliott> oerjan: A hard certificate? Really?
14:39:32 <fizzie> Vorpal: Well, it should be. It could be missing some.
14:39:36 <elliott> I am not aware of those being widespread.
14:40:18 <Vorpal> elliott, for bank? Usually using challenge response. Like you get a code on screen, enter it into a small device with a keypad, and enter your personal code, then get a reply code back and enter it on the login screen
14:40:35 <fizzie> We just have boring passwords and one-time password lists.
14:40:50 <Vorpal> in my case I need my bank card in the device for it to work, the device itself is not user-specific
14:40:57 <fizzie> Though I think some bank had some sort of a beta test of a device.
14:41:01 <elliott> Vorpal: No, not "usually".
14:41:08 <elliott> Approximately nobody does that.
14:41:43 <oerjan> elliott: i was using it in a general sense: it's a small device into which my bank/visa card enters and then you push a button and it gives a series of digits.
14:41:50 <Vorpal> elliott, I remember that a few years back there was a lot of articles in the news papers about the bad security with one time code pads and such and that prompted most banks to switch to hard certificates
14:41:59 <elliott> fizzie: oerjan: From the glogbot logs: http://sprunge.us/HOZC
14:42:17 <elliott> Vorpal: Maybe in Scandinavia. It's just passwords + security question here.
14:42:21 <Vorpal> oerjan, so not challenge response?
14:42:34 <fizzie> Well, it looks rather similar.
14:43:00 <elliott> I don't really see what the point of an involved security system for your online banking is if you use a credit card.
14:43:15 <elliott> Like leaving the door open but reinforcing your windows with unobtanium.
14:43:22 <oerjan> Vorpal: no. no information passes from the web to the device.
14:43:32 <Vorpal> elliott, the chipsets are supposed to be safer than the old magnetic strip thingy
14:44:02 <oerjan> they're announcing some new system to replace that, anyhow. _before_ this we just had a card with a heap of one-time numbers on it.
14:44:19 <elliott> Vorpal: ?! Dude, it's not the cards themselves that are the problem.
14:44:37 <Vorpal> and nowdays it is all chipsets in Sweden. They passed a law that said that if a shop used magnetic strip and that led to issues then the shop is responsible for any losses. That made shops switch over rather fast.
14:44:45 <Vorpal> not sure how much better it actually is
14:44:50 <elliott> If you pay someone with a credit card, they know your credit card details.
14:45:03 <elliott> Therefore whenever you pay anyone with a credit card, you're giving them your credit card.
14:45:10 <elliott> There is no security whatsoever.
14:45:52 <oerjan> here's a norwegian newspaper article on yesterday's scandal btw http://www.dagbladet.no/2012/03/20/nyheter/innenriks/altinn/selvangivelsen/20770039/
14:46:15 <fizzie> That's just a backwards compatibility thing. It'd go away if you'd actually need to properly use the card when buying stuff with it.
14:46:37 <fizzie> Anyway, I've seen more MasterCard SecureCode / Visa VerifiedByVisa things on the web-web lately.
14:46:40 <Vorpal> elliott, what if you moved the keypad and display to the card itself, wouldn't that be safer? Then you could just sign a request on the card and the bank could verify it
14:47:18 <elliott> fizzie: Of course it's a fixable problem.
14:47:24 <elliott> That doesn't make it not a problem.
14:47:45 <elliott> I mean, obviously you can just make a mini-device with a private key and the like.
14:48:12 <Vorpal> oh for web transaction some banks here offer to create a special bank card number for just a single transaction. So there won't be more available than what you specify on it.
14:48:48 <fizzie> The single-use numbers I think quite many places offer.
14:49:23 <fizzie> I think I remember some bank (probably not Finnish) having a test of the system where you actually see the amount you're authorizing when you're getting the authentication code for a net-bank transaction. That's part of some of the standards, IIRC.
14:51:06 <Vorpal> anyway with the "verified by visa" thing it redirects me to my internet bank and I have to use the hard certificate I think.
14:51:38 <Vorpal> yeah, pretty sure it was with "verified by visa" stuff
14:51:45 <oerjan> elliott: oh also by net banking i'm not referring to buying things on the web, as i never do that (i think i've bought a plane ticket _once_). i'm referring to actually logging into my bank account for paying bills and stuff.
14:52:24 <oerjan> elliott: not any more, no.
14:52:47 <Vorpal> elliott, actually since internet banking offers a lot more than just taking money out of the account your card is tied to, your analogy with the reinforced windows is not quite accurate
14:53:08 <oerjan> it was back in 2002-2004 or so, before the security insanity had really hit here
14:54:42 <Vorpal> you can access all accounts from the internet bank, and stocks (if you have any, I seen a link for it in the menu, never used that). Probably even more stuff.
14:56:34 <fizzie> elliott: Anyway speaking of credit card security, we did in fact have to buy a plane ticket via a travel agency recently (all purchases of the university have to go through it for businessy reasons, and we wanted my wife's ticket in it for the same order), and it was certainly very safe: they wanted the CVV2 number sent in a *different* email than rest of the card details. (Or faxed.)
14:57:27 <oerjan> Vorpal: yeah like that in my bank too, i sold off my stocks that way.
14:57:52 <fizzie> elliott: I'm not entirely sure, it was handled by our department secretary.
14:57:57 <oerjan> yeah i did. well fund options.
14:58:12 <oerjan> fund shares, i guess it is.
14:58:29 <elliott> oerjan: I doubt the "security insanity" has even vaguely hit Norway.
14:58:48 <Vorpal> oerjan, oh and insurances and borrowing money too
14:59:08 <oerjan> elliott: well relatively speaking, but i think you can get part of your luggage confiscated for stupid reasons here too.
14:59:37 <Vorpal> <elliott> oerjan: I doubt the "security insanity" has even vaguely hit Norway. <-- what?
15:00:02 <oerjan> although i wouldn't really _know_, since i don't use planes any more :P
15:00:44 <oerjan> i suspect it mostly gets bad only if your plane trip leaves the country
15:00:55 <fizzie> oerjan: Also here's a less picky list: http://p.zem.fi/o1ug
15:01:04 <oerjan> (and of course if you get anywhere near the US, all bets are off)
15:02:16 <elliott> fizzie: Now do all lines with "A" in them.
15:03:21 <Vorpal> btw the security if you are not flying with airlines is basically nil. Was up in a small Cessna a couple of years ago (2? 3?) and there was no security to speak of.
15:03:43 <Vorpal> (didn't leave the country though)
15:03:46 <fizzie> Quite a lot of it is just "AnMaster:" prefixes.
15:04:00 <Vorpal> fizzie, what if you filter those out?
15:04:15 <sebbu> Vorpal, normal, they put security in airports because it's more easy and get seen easily, but elsewhere... don't do much about it
15:04:52 <fizzie> 4933 with 'A' but not 'AnMaster'.
15:04:57 <Vorpal> sebbu, well, it was at an airport. But with the local flight club. Happens to be located right next to the airport, and share the runways with it.
15:05:26 <elliott> <Vorpal> btw the security if you are not flying with airlines is basically nil. Was up in a small Cessna a couple of years ago (2? 3?) and there was no security to speak of.
15:05:41 <elliott> Vorpal: You realise this is because people don't try to hijack small Cessnas, right?
15:06:14 <Vorpal> elliott, of course, but someone could hijack a business jet. That could still cause significant damage.
15:06:20 <Vorpal> not nearly as much of course
15:06:28 <Vorpal> of course the chance is pretty low
15:06:57 <elliott> They don't let anyone on to business jets in return for a fee. :p
15:07:13 <Vorpal> elliott, err I meant private business jets
15:08:10 <fizzie> The dreaded mad millionaire problem, yes.
15:08:23 <elliott> Vorpal: Yes, my point is that terrorists don't generally find themselves with a good reason to be on a business jet.
15:08:40 <elliott> You don't need a reason to be on a commercial flight.
15:09:49 <oerjan> i'd sort of figured the reason they don't put business jets through the security stuff is because there are rich people involved
15:10:09 <Vorpal> probably that helps as well yes
15:11:15 <elliott> But it's not surprising that the "security" is done for commercial flights in airports; if it was effective, that's just by far the most reasonable place to put it.
15:11:27 -!- pikhq has joined.
15:11:34 -!- pikhq_ has quit (Ping timeout: 252 seconds).
15:11:35 <Vorpal> elliott, I wonder how effective it actually is?
15:12:40 <elliott> Vorpal: The TSA stuff? It isn't.
15:13:13 <fizzie> It's effective for giving people cancer. (Okay, not really.)
15:13:22 <elliott> Vorpal: Does the Swedish government never do stupid things?
15:13:38 <Vorpal> elliott, of course it does
15:13:55 <oerjan> fizzie: supposedly the screening has less effect than being exposed to cosmic rays high in the atmosphere during the flight itself
15:13:56 <elliott> I mean, you can thank 9/11.
15:15:19 <fizzie> oerjan: Right; so it's not exactly effective. I'm sure if they'd contact some proper engineers they could make a far more efficient cancer machine.
15:21:00 <elliott> fizzie: They did that once: http://en.wikipedia.org/wiki/Therac-25
15:21:18 <fizzie> I remember that thing.
15:21:23 <elliott> fizzie: Sadly it also failed to give anyone cancer, seeing as they already had it.
15:21:35 <fizzie> It did manage to kill people, though.
15:22:09 <elliott> Non-real-time software is nice. Lower risk, and all that.
15:22:16 <fizzie> The details of the OS that you get from the report were interesting; it was quite the spaghetti code, with their own scheduler and so on.
15:23:20 <elliott> I'd hate to be one of the engineers on that project.
15:23:29 <elliott> Not that the guilt isn't deserved.
15:24:48 <fizzie> I'd hate to be the guy who's fault is it that when you beg the killer robot for mercy, it will just go all "PLEASE ENUNCIATE MORE CLEARLY" and resume killing you.
15:25:14 <fizzie> I'd hate to be the guy who writes "who's" like that.
15:26:12 <quintopia> but you are the guy BWAHAHAHA shun the nonbeliever shuuuuuunnnnnnn
15:42:15 -!- ais523 has joined.
15:44:35 <Vorpal> how is the mouse cursor drawing handled in X11? It is after all a form of compositing really (especially with antialiased cursors with smooth edges), but it was supported long before more general compositing was supported. So at which layer is it handled?
15:45:44 <elliott> I don't think there are any "smooth"-edged cursors without compositing.
15:45:54 <ais523> Vorpal: the GPU renders the cursor, AFAIK
15:46:07 <ais523> the reason I think this is that I've seen GPU crashes that render nothing on the system working /but/ the mouse cursor
15:46:17 <fizzie> At least I recall a couple of graphics drivers having cursor-specific options.
15:46:27 <elliott> ais523: That's a bit modern-day for Vorpal's question.
15:46:34 <Vorpal> so a sprite of some sort
15:46:44 <elliott> But OTOH I suspect that was how it was done in the olden days, too.
15:46:49 <fizzie> E.g. 'man radeon' -> "SWcursor boolean: Selects software cursor. The default is off."
15:46:51 <elliott> Perhaps not the stretch in-between.
15:47:36 <Vorpal> well, that seems reasonable
15:49:15 <fizzie> "X provides a set of standard cursor shapes in a special font named cursor. Applications are encouraged to use this interface for their cursors because the font can be customized for the individual display type. The shape argument specifies which glyph of the standard fonts to use." That's the Xlib interface view of it, from man XCreateFontCursor. (But you can also create cursors from a Pixmap ...
15:49:22 <fizzie> ... or any regular X font.)
15:49:36 <fizzie> Presumably there's also something more modern.
15:50:15 <Vorpal> anyway, I seem to remember smooth cursors back ages ago. Pretty sure it was possible back on xfree86 even
15:52:54 <elliott> Vorpal: Yes, but "smooth" != "translucent".
15:53:10 <fizzie> The 'Render' extension -- the thing that does composition -- has cursor loading functions too.
15:53:16 <elliott> I'm pretty sure the standard GNOME 2 white cursor thing had sharp edges, it was just antialiased inside.
15:53:26 <elliott> And it was rounded off a bit so it looked quite smooth.
15:53:34 <Vorpal> trying to remember what it looked like in KDE 2
15:53:40 <fizzie> "When using images loaded from files, Xcursor prefers to use the Render extension CreateCursor request if supported by the X server. Where not supported, Xcursor maps the cursor image to a standard X cursor and uses the core CreateCursor request."
15:53:41 <elliott> http://3.bp.blogspot.com/_JSR8IC77Ub4/TSSKKbXaWCI/AAAAAAAABlY/ch99AiNX2Sc/s1600/Ubuntu-desktop-2-410-20080706.png -- it can be seen here
15:53:51 <elliott> I guess there is a shadow.
15:53:59 <elliott> fizzie: Ah, maybe it's that then.
15:54:46 <fizzie> XRender dates back to about 2000, "first released with XFree86 version 4.0.1".
15:54:52 -!- oerjan has quit (Quit: Later).
15:56:46 <ais523> bleh, my zoom-in-and-see-the-individual-pixels window manager trick doesn't work on the cursor
15:57:42 <fizzie> From the Xcursor lib again: "XcursorBool XcursorSupportsARGB (Display *dpy): Returns whether the display supports ARGB cursors or whether cursors will be mapped to a core X cursor."
15:58:06 <Vorpal> ais523, a screen magnifier? Or something else?
15:58:25 <ais523> Vorpal: it's one of the compiz features
15:58:32 <ais523> most of them, I'm not even sure if they have an intended use
15:58:36 <ais523> they were just added because they could
15:58:53 <ais523> e.g. I can make a window black-and-white using super-shift-mousewheel
15:59:04 <ais523> which is nice on occasion, if something's too brightly colored
15:59:10 <Vorpal> heh, I don't use compiz, it messed up with opengl stuff for me in the past. And nowdays I just use xfwm4 or whatever it is called.
15:59:13 <fizzie> So it does sound like on the protocol level, you have a non-translucent CreateCursor, and then the XRender extension CreateCursor that can do ARGB images; and then it is -- or at least was -- the graphics driver's job to keep the cursor rendered.
15:59:34 <Vorpal> ais523, anyway, what happened to the cursor when you used that trick?
15:59:41 <ais523> it just kept moving around as normal
15:59:43 <elliott> xfwm4 does compositing too.
15:59:48 <ais523> oh, you mean zooming in?
15:59:51 <ais523> it stayed the same size
15:59:57 <Vorpal> elliott, right, but it doesn't do the silly stuff compiz does
15:59:59 <ais523> and moving it changed the point the screen was zoomed around
16:00:19 <Vorpal> as far as I know at least
16:00:22 <fizzie> I was helping a guy at the department few weeks back, I think he had ALL THE EFFECTS on. At least when I moved a terminal window around, it wiggled like it was made out of jelly.
16:01:46 <ais523> fizzie: that's actually in the default set when you ask for high effects
16:02:01 <Vorpal> what do you mean by wiggled?
16:02:09 <Vorpal> like... shakes when you move it?
16:02:47 <ais523> Vorpal: there's a bit of a timelag on pixels further from the mouse
16:02:53 <ais523> so it looks like the window's slightly elastic
16:03:08 <Vorpal> utterly pointless as far as I can tell
16:03:31 <ais523> Vorpal: it looks nicer than the window staying rigid
16:03:44 <ais523> but yes, there's no real need for it
16:03:57 -!- ais523 has quit (Remote host closed the connection).
16:04:04 <Vorpal> I mean, if it ends up looking slightly wider than it is, you can't really do precise placement
16:04:36 <fizzie> It settles down relatively fast if you don't move the mouse. But still.
16:04:44 <elliott> hmm, Wikipedia (generally) bans linking to forks of Wikipedia
16:04:55 <fizzie> Personally I think it just looks like the window is somehow broken or melted. Makes me think "you should probably get that checked out".
16:05:04 <Vorpal> that means waiting, any special effect that takes away time that I could have spent doing other stuff is bad.
16:05:09 <elliott> what about wikis with substantial content with a specific focus but that are based on Wikipedia's articles to start with?
16:05:15 <elliott> I guess that would count as one of the exceptiosn
16:06:23 <Vorpal> like minimizing, OS X has one of the worst minimizing effects I ever seen (is it default btw? I know you can turn it off but still, why even put that in there)
16:06:46 <fizzie> The genie is I think the default.
16:06:55 <Vorpal> right, and it is horrible
16:07:10 <fizzie> It had a second one, at least in 10.4.
16:07:59 <fizzie> I suppose you know of the easter egg that you can slow down the genie by a factor of ten or so? And the not-so-easter-egg thing that you can freeze windows in the middle of the effect so that they're stuck in that shape, and then keep using it?
16:08:12 <Vorpal> nope, don't know either of those
16:08:25 <fizzie> ctrl-shift-minimize or something does the slowdown.
16:08:30 <Vorpal> my experience with OS X is limited to helping people (family) who have macs.
16:08:31 <elliott> You can't turn off the effect (through the standard interface).
16:08:35 <elliott> fizzie: It's not an easter egg.
16:08:39 <elliott> You can do that for anything that animates.
16:08:54 <elliott> It was used to demonstrate Expose and the like, IIRC.
16:09:03 <elliott> The freezing thing is fun, though.
16:09:15 <Vorpal> <elliott> You can't turn off the effect (through the standard interface). <-- thought there was an alternative effect?
16:09:17 <elliott> Unfortunately it doesn't quite map the geometry.
16:09:19 <Vorpal> or have they removed that?
16:09:30 <Vorpal> the alternative was significantly better at least
16:10:07 <Vorpal> still, on the computer I'm writing this on I have no effect at all when minimizing/maximising.
16:10:15 <Vorpal> anyway, how do you do the freezing?
16:11:34 <Vorpal> does that mean the dock handles window manager style tasks?
16:12:52 <elliott> Vorpal: Well, minimised windows go into the Dock.
16:13:03 <elliott> So it makes sense for the Dock to handle it.
16:13:04 -!- hagb4rd has quit (Ping timeout: 276 seconds).
16:13:29 <Vorpal> indeed, but I would have assumed the actual minimising part would be handled by the equivalent of the window manager.
16:13:42 -!- hagb4rd has joined.
16:13:46 -!- MoALTz_ has quit (Quit: brb).
16:14:09 <elliott> Well, the Dock also enlargenifies to make a space for it while the animation is happening.
16:14:17 <elliott> So you'd need quite some coordination with WindowServer to manage that.
16:14:22 <elliott> Arguably WindowServer should display the dock.
16:14:53 <Vorpal> well you need that already I assume? Or does the dock handle the compositing of the window as well?
16:15:47 <elliott> Why does that matter? The Dock just has to morph it.
16:16:45 <Vorpal> how is the actual morphing done? The obvious choice to me is doing it by a matrix on the GPU
16:17:17 <Vorpal> which seems to be the realm of the window manager
16:18:16 -!- MoALTz has joined.
16:18:51 <elliott> Vorpal: Core Animation, one presumes.
16:19:28 <elliott> "Core Animation first appeared in Mac OS X v10.5" -- never mind, must be something else.
16:20:07 <Vorpal> also can any program mess with the window of another program like that? Or is it just the dock?
16:20:21 <Vorpal> the security model seems rather bad if any program can do that
16:20:33 <Vorpal> not that the X11 security model is any better
16:21:11 <elliott> Vorpal: There are rather bigger problems with the security model.
16:21:21 <elliott> For instance, any windowing system that lets any application take a screenshot has no security.
16:21:35 <Vorpal> indeed, that should require special privileges
16:23:16 <Vorpal> is there any secure windowing system at all?
16:23:24 <Vorpal> (that currently exists I mean)
16:23:32 <hagb4rd> you want to modify the animation on the layer vorpal?
16:23:45 <Vorpal> elliott, do you happen to know of one?
16:23:48 <elliott> I'm sure one of the capability-based OSes has had a windowing system.
16:24:14 <elliott> NeWS might have been secure, too? All I know about NeWS is that it's about 15x better than X.
16:24:28 <Vorpal> that isn't difficult to achieve
16:25:42 <elliott> Sorry, I left off "thousand".
16:26:02 <fizzie> If you canvas.drawImage(img) where img.src is cross-origin, it clears the origin-clean flag, and thereafter you can never call .toDataURL() or .getPixelData() on that canvas.
16:26:23 <Vorpal> btw I found something interesting recently, it seems steam automatically adds exceptions to the windows firewall for installed games. To allow them to listen on any port on either tcp or udp
16:26:27 <fizzie> Browsers are so security-minded.
16:26:40 <Vorpal> it does this without asking the user or there even being an option to disable the behaviour as far as I can tell
16:27:17 <elliott> fizzie: A known exploit, I take it?
16:27:42 <elliott> (Also a textbook example of how ad-hoc security is completely unworkable.)
16:27:57 <Vorpal> hm, has it been used in any exploit?
16:28:02 <elliott> Vorpal: Why would you use the Windows Firewall?
16:28:14 <Vorpal> elliott, well, it is there by default in windows.
16:28:38 <Vorpal> (and it isn't quite as bad in windows 7 as it was in XP)
16:28:41 <elliott> Vorpal: So is that fucking animated dog.
16:29:08 <hagb4rd> would you mind to drop me a short line on what you want to do vorpal? would be so much more fun to get a point
16:29:11 <Vorpal> the machine is behind a real firewall (and NAT too), but still it is interesting behaviour
16:29:25 <Vorpal> hagb4rd, uh? Did I want to do something?
16:29:48 <Vorpal> well, I do want to breath, quite often in fact.
16:30:01 <Vorpal> (wth are you going on about?)
16:31:22 <hagb4rd> you catched my attention while discussing this animation thing.. so i thought you're out to experiment with this or sth
16:31:32 <Vorpal> hm, I just sprinted into a bucket in Skyrim by mistake, it went flying several hundred meters, over a rather tall mountain, before I lost sight of it. Trying to find where it landed now...
16:31:57 <Vorpal> hagb4rd, I believe we discussed terrible animations of compiz, then OS X?
16:33:18 <Vorpal> To be fair it was a magical "whirlwind sprint" (lowest "level"), but even so that result was way out of proportion to the force I hit the object with.
16:33:55 <elliott> It's a: new user account on the: esoteric languages wiki.
16:34:41 <Vorpal> or was that just a weird way of announcing it?
16:50:58 <elliott> No, I was feeling very: fizzie.
16:53:10 <Vorpal> RocketJSquirrel, ah hi Gregor. Looks like apropos doesn't work on hackego
16:53:20 <HackEgo> apropos: can't open the manpath configuration file /etc/manpath.config
16:53:35 <elliott> Neither does man. Oh noes.
16:53:50 <elliott> fizzie: Hey, can you use SCIENCE to impale some colours?
16:54:06 <Vorpal> presumably whatis doesn't work either
16:54:11 <HackEgo> whatis: can't open the manpath configuration file /etc/manpath.config
16:54:21 <HackEgo> cs \ de \ es \ fi \ fr \ fr.ISO8859-1 \ fr.UTF-8 \ gl \ hu \ id \ it \ it.ISO8859-1 \ it.UTF-8 \ ja \ ko \ man1 \ man2 \ man3 \ man4 \ man5 \ man6 \ man7 \ man8 \ nl \ pl \ pl.ISO8859-2 \ pl.UTF-8 \ pt \ pt_BR \ ru \ sv \ tr \ zh_CN \ zh_TW
16:54:40 <Vorpal> well the man pages are there, just not the config I guess
17:02:58 -!- hagb4rd has quit (Quit: hagb4rd).
17:05:30 -!- Taneb has joined.
17:06:33 <elliott> https://upload.wikimedia.org/wikipedia/commons/a/af/AmsterdamIsl_Map.png "La Grande Marmite"
17:07:00 <Taneb> That's suspiciously close to wikipedia
17:07:22 <Taneb> I'm gonna be so glad come Easter
17:08:19 <Taneb> Only God and he who challenged by this task to begin with.
17:08:24 -!- hagb4rd has joined.
17:08:29 <Taneb> WHO WEARS A PICTURE OF JESUS ON HIS CROTCH
17:10:51 * sebbu ressussicate himself using post-mortem necromancy
17:11:01 <elliott> fizzie: Kick sebbu for unlicensed self-revival.
17:11:13 <elliott> You need a permit for that kinda shit.
17:11:40 <sebbu> my god/religion allows it
17:12:16 <elliott> fizzie is your new god now.
17:13:22 <sebbu> i'm not apostate, just from a different religion
17:13:47 <sebbu> didn't abandoned something i didn't belonged to :p
17:14:30 <sebbu> there's lots of different religions in the world :p
17:14:34 <elliott> The only valid religion is fungotism, dammit!
17:14:35 <fungot> elliott: now that one is ' no' isn't concrete enough for me
17:15:25 <fungot> elliott: i've heard sarahbot is r5rs's bitch.
17:15:37 <sebbu> Phantom_Hoover, my true faith was different
17:15:55 <sebbu> which proove yours is wrong, aka not THE true faith
17:16:06 <sebbu> at most A true faith
17:16:16 <RocketJSquirrel> Sarahbot, thy code art in Scheme, for thou art the bitch of R5Rs.
17:16:54 <elliott> The Revised Revised Revised Revised Revised Report on the Algorithmic Language scheme
17:17:59 <elliott> sebbu: Have you converted to fungology yet?
17:18:05 <elliott> fungot: Tell sebbu of the true path.
17:18:06 <fungot> elliott: has anyone here ever tried to parse that
17:18:15 <sebbu> nop, ressussicated again
17:18:29 <sebbu> forgot to plug in some pain nerves that time
17:18:37 <elliott> Phantom_Hoover: Use the POWER OF FUNGOT to replace sebbu with a fnord already.
17:22:03 -!- Taneb has quit (Quit: Goodbye).
17:22:16 <sebbu> ( http://www.fnord.org/occult/discordia/scanned/fnord.html )
17:26:02 -!- ais523 has joined.
17:31:03 <itidus21> esoteric is occasionally esoteric
18:02:03 -!- calamari has joined.
18:08:51 -!- Nisstyre has quit (Ping timeout: 246 seconds).
18:20:34 <elliott> fizzie: Yes, COLOURS. With the U, my friend. With the U.
18:22:30 <elliott> fizzie: No, "U", as they say.
18:22:34 -!- Nisstyre has joined.
18:22:55 <elliott> fizzie: I'm commissioning you to produce corresponding background colours for all the section header colours on http://esolangs.org/wiki/Main_Page, you see.
18:23:07 <elliott> By commissioning, I mean "delegating the job to you on IRC so I can sit back and do nothing".
18:23:25 -!- FireFly has quit (Changing host).
18:27:37 <fizzie> So, what, same hue and saturation but a lot lighter? (Or maybe same a*/b* chroma but more L*, or something, since CIE (L*,a*,b*) is like so perceptual.)
18:28:48 <elliott> Yes, I was hoping for something fancy and scientific.
18:29:20 <elliott> Even if that kills my dream to implement it within the MediaWiki template.
18:45:12 <fizzie> http://zem.fi/~fis/esocolor.html -- that's quickly done in HLS, with top color giving the lightness. I suppose the ones that are >= 0.90 are lighter than your already-rather-light headers.
18:45:25 <fizzie> I can do L*a*b* but I think it's dinnertime first.
18:46:23 -!- sebbu2 has joined.
18:46:23 -!- sebbu2 has quit (Changing host).
18:46:24 -!- sebbu2 has joined.
18:46:33 <elliott> Does COLOUR SCIENCE have anything to say about a nicer choice than that peach everyone dislikes?
18:46:50 <elliott> Well, the green is a bit off too. I did not: apply colour science to create these section header colours.
18:47:17 <fizzie> I can ask gcolor for "readers", "creators", "meta" and "featured". (But they'll probably all just turn out brown.)
18:49:48 -!- sebbu has quit (Ping timeout: 255 seconds).
18:50:40 <elliott> fizzie: I was hoping for something more like "these colours are of equal perceptual intensity and all have the same 'distance' from each other" or something. :p
18:50:41 -!- monqy has joined.
18:51:54 -!- augur has quit (Remote host closed the connection).
18:53:33 -!- derdon has quit (Remote host closed the connection).
18:57:20 <fizzie> I guess you could e.g. put a circle of some radius around the white point in cielab space, aligned in the equal-L* plane, and then pick the points at angles k, k+pi/2, k+pi, k+3pi/2 where k is arbitrary. Lab's *supposed* to be "perceptually uniform", in that distances in it equal perceived amount of differentness. But I'm not sure how true that actually is, and it probably won't give colors ...
18:57:27 <fizzie> ... that a graphics designer would pick.
18:57:41 <fizzie> Also if you have a too large circle you'll end up with colors not in sRGB.
18:59:04 -!- augur has joined.
19:00:07 -!- augur has quit (Remote host closed the connection).
19:00:36 <elliott> fizzie: I guess HSL would be ACCEPTABLE? I don't know, the idea of one colour being less "far apart" or "more intense" than the others because of arbitrariness upsets me.
19:02:27 <olsner> how far apart can one color be?
19:02:42 <fizzie> I guess you could also just some sort of equal-saturation-and-lightness in HSL, but that's not so perceptual.
19:03:27 <elliott> fizzie: Yes, perceptuality is nice. There should be a tool for dicking about in Lab. Maybe there is.
19:04:07 <fizzie> Maybe a very good color picker dialog somewhere. I don't know of any offhand.
19:07:53 -!- sebbu2 has changed nick to sebbu.
19:07:55 -!- hagb4rd has quit (Ping timeout: 276 seconds).
19:10:05 <fizzie> Current esocolors in HLS are (100°, 83.3%, 60%), (60°, 83.3%, 100%), (45°, 86.7%, 100%) and (240°, 90%, 33.3%), but I don't really know if that tells one anything. It's really still quite closely tied to the "raw" RGB values, it's just pretending to have meaningful coordinates. (Same applies to HSV.)
19:13:13 <fizzie> Possibly in the majority of cases.
19:13:43 <fizzie> Sometimes it ends up being HLS, e.g. in Python's "colorsys" module, and that's what gcolor used, so I've been a bit stuck with it.
19:13:53 <fizzie> "colorsys" has HLS, but it also has HSV. Go figure.
19:14:50 <elliott> Well, you know what they say: haughty sandwich vixens haggle less sexily.
19:15:40 <fizzie> Maybe you need to figure out a way to put HSI in there too. (It's yet another thing with 'I' for intensity.)
19:16:09 <fizzie> They all call the 'S' part saturation, but it means different things in each three.
19:16:38 -!- Phantom__Hoover has joined.
19:17:34 <elliott> It *would* be nice to figure out what the True Fourth Colour is in #CEB, #FFA, ?, #DDE. But perhaps colour science is not yet ready for such problems, as they say.
19:17:58 <fizzie> You may need to ask a humanist.
19:20:17 <itidus21> if one can read the word humans one can already understand all that human's collective knowledge on the subject is, sadly
19:20:55 -!- Phantom_Hoover has quit (Ping timeout: 276 seconds).
19:21:04 -!- pikhq_ has joined.
19:21:10 -!- pikhq has quit (Ping timeout: 244 seconds).
19:24:36 <ais523> use Gregor's neural network?
19:34:56 -!- litejk has joined.
19:34:59 -!- litejk has left ("61 This is CERTAINLY the MOST TERRIFYING GAMALOST you will RUN AWAY FROM all OLSOK: http://championofbirds.com/?p=4991 | http://codu.org/logs/_esoteric/").
19:40:25 <itidus21> it's the attack of the bot who uses the topic and 61 as it's quit message, when it does it's dastardly quick join and exit
19:41:42 <ais523> is there even any point?
19:46:59 <itidus21> ok im liking the new look wiki to the extent that my opinion really matters
19:49:05 <itidus21> or maybe its my imagination playing trick
19:49:50 -!- augur has joined.
19:53:48 -!- Phantom__Hoover has quit (Read error: Connection reset by peer).
19:55:32 -!- Phantom_Hoover has joined.
20:00:47 -!- sebbu2 has joined.
20:00:47 -!- sebbu2 has quit (Changing host).
20:00:47 -!- sebbu2 has joined.
20:04:18 -!- sebbu has quit (Ping timeout: 252 seconds).
20:07:09 -!- asiekierka has quit (Remote host closed the connection).
20:12:05 -!- augur has quit (Remote host closed the connection).
20:16:32 -!- nortti has joined.
20:24:10 -!- sebbu2 has changed nick to sebbu.
20:25:12 -!- hagb4rd has joined.
20:26:19 -!- oerjan has joined.
20:28:06 <itidus21> torrents are like a library where all the books are on reserve
20:28:45 <itidus21> it's only natural that people can't resist such a thing
20:29:06 <oerjan> wait, what does on reserve mean again
20:29:36 <oerjan> *wait, what does on hold mean again
20:30:35 <oerjan> isn't it more like, you know, the complete opposite
20:30:45 <nortti> 01:20 <@MakeGho> what... some big guy just walked here and said, "Hi, are you a jew?" "...No" "I am." *handshake*
20:31:15 <itidus21> it means when you tell the library to keep a book off the shelf for you for you to pick it up
20:32:06 -!- elliott has quit (Read error: Operation timed out).
20:32:11 <oerjan> itidus21: but that usually means you _cannot_ get it right now, which seems like the opposite of a torrent. but then i have no experience with torrents.
20:32:55 <olsner> it's like being in the library and the book is on hold, and you are sitting in the "books on hold" room, and your lap is the storage for the particular book you want to read next
20:32:58 <HackEgo> itidus21 just made some instant coffee.
20:33:06 <HackEgo> itidus20's entry has been censored.
20:33:23 <olsner> but actually, it's more like torrents
20:35:54 <itidus21> can safely disregard my previous nonsense.. will be replaced with..
20:36:55 <itidus21> i guess it's an engineering problem when it is necessary to add artificial constraints to a product in order to enable the producers to turn a profit
20:37:02 <oerjan> `learn itidus20 is a 7th dan of broken metaphors.
20:37:56 <olsner> oerjan: now write that as a broken metaphor
20:38:08 <oerjan> olsner: wait, it wasn't already broken?
20:38:27 <oerjan> what else would it be?
20:38:35 <olsner> hmm, but maybe we ought to leave it in the competent hands of our 7th dan broken metaphorist
20:39:19 <itidus21> If people were all already contributing to society as best as they could, would money still be necessary at all, or could people with their freetime just merrily use whatever they encounter which would probably be in abundance
20:39:51 <MDude> That depends on how you define contribution to society.
20:40:29 <itidus21> it sounds awful though i know.. the thought of a whole life without any hope of lawn chairs in hawaii sipping alcohol and watching bikini clad women dance with maracas
20:41:01 <oerjan> MDude: highest amateur rank in many martial arts
20:41:53 <oerjan> or wait, scratch the highest, that varies widely.
20:42:40 <itidus21> MDude: lets say.. one time of day for production.. one time of day for consumption
20:42:47 <oerjan> in other news, i hate it when my laptop decides to burn cpu just as i'm trying to google something.
20:43:02 <itidus21> during production time the only consumables are things which are used to produce things
20:43:20 <MDude> That sounds a lot like having a job.
20:43:51 <itidus21> except theres no money... so noone is selling anything during the day
20:43:59 <itidus21> and its illegal to consume during working hours
20:44:20 <itidus21> except to the extent you need to work optimally
20:44:25 <MDude> I'd say most economic theories work if you presume human nature magically conform to thier expectations.
20:44:34 <nortti> did you just describe marxist socislism?
20:44:42 <MDude> No I mean all of them.
20:44:50 <oerjan> itidus21: congratulations for reinventing 1918-1920 communism, i think. (even the communists had to scrap not using money)
20:44:53 <MDude> Oh wait, you weren't talking to me.
20:45:29 <MDude> But yeah, it's a metter of determining what you need.
20:46:09 <MDude> If you can presume everyone jsut knows what they need to do, and does it, then you don't need much of an economy.
20:46:27 <itidus21> MDude: but that makes thinking boring
20:46:40 <itidus21> if things have to actually apply
20:46:59 <oerjan> itidus21: the ability to decentralize economic decisions is what makes money a brilliant, if imperfect idea.
20:47:06 <nortti> oerjan: actually that is not marxist-leninist communism, but marxist socialism (I have read communistic manifesto)
20:47:38 <oerjan> nortti: i wasn't going for accuracy there
20:47:49 <oerjan> the dates are probably wrong too
20:47:58 <itidus21> it is best if one does not give too much attention to my posts
20:48:54 <itidus21> any system where everyone has to contribute would make my life more difficult honestly
20:49:29 <nortti> what you described was almost exactly like marxist socialism, but it lacked common ownership
20:49:45 <oerjan> <MDude> I'd say most economic theories work if you presume human nature magically conform to thier expectations. <-- as i and probably others have said before, every form of government works if all people are perfect, including nazism.
20:50:35 <itidus21> nortti: actually i hinted at that.. in the second half of one post :-D but didnt quite fit it in, in my ramble
20:50:38 <itidus21> "or could people with their freetime just merrily use whatever they encounter which would probably be in abundance"
20:50:40 <oerjan> nortti: nazism makes it a better joke
20:50:54 <itidus21> that would be for the consumption period
20:51:08 <oerjan> Phantom_Hoover: the nazis only persecuted people who weren't considered perfect, right?
20:51:11 <Phantom_Hoover> It's not like people never invent logically inconsistent systems, after all.
20:51:11 <MDude> How would you measure productivity?
20:52:18 <nortti> itidus21: so you basicaly reinvented narxism
20:52:48 <itidus21> well i have chatted with a pro-marx guy before who explained that communism was a distortion of marx.. and another guy who said marx is unworkable
20:52:54 <oerjan> narxism, the best form of government
20:53:07 <itidus21> and maybe figured out the gaps
20:53:26 <MDude> Narxism: Marxism for narcisists?
20:54:06 <olsner> narczism, nazism for narcotics
20:54:20 <itidus21> in any case we hve the best bodies on earth.. out of all animals and plants.. so we should accordingly have the best quiality of life
20:55:17 <nortti> I am socialist myseft and I think that every "Communistic' country is just dictatorship, but Marxist Communism might work if executed properly
20:56:10 <itidus21> i find that when we consider things like long life as needs then it eats away a piece of the labour pie
20:56:34 <itidus21> but we don't want to just let each other die young or from diseases
20:57:04 <oerjan> <itidus21> in any case we hve the best bodies on earth.. out of all animals and plants.. <-- ok that's enough. go take a timeout and read about human evolutionary flaws.
20:57:33 <oerjan> our _brains_ may be the best, but the rest of our bodies have paid for it.
20:57:47 <MDude> We are the only animal with such ready access to candy, cartoons, and air conditioning.
20:58:07 <MDude> Maybe cats have it better, but they can't taste sugar.
20:58:47 * Phantom_Hoover remembers the time his family visited a museum and his mother (an obstetrician) spent all her time examining animal pelvises and pointing out how comparatively large the birth canal is to humans'.
20:58:59 <nortti> even if our brains are best we must remember that no other species has nerly blown up the entire world...
20:59:32 <itidus21> oerjan: yeah.. anything more i say about this will be downhill from here
21:00:35 <oerjan> nortti: hm are you sure? see: the oxygen catastrophe
21:01:05 <fizzie> @tell elliott FWIW, the esocolors in L*a*b* are (138.4, -29.2, 30.6), (149.6, -17.9, 59.0), (143.8, -2.7, 38.9) and (135.5, 4.6, -12.0). http://zem.fi/~fis/esocolor.png shows the slice of L*a*b* for L* = 141.8 (mean of those) that's representable in sRGB. The red dots are the colors' a* and b* coords.
21:01:55 <nortti> oerjan: oxygen didn't came even close to blowing up the planet. It just poisoned most of life on it
21:02:30 <oerjan> nortti: we aren't close to blowing up the planet either. you think nuclear weapons will do more than scratch the surface?
21:03:20 <oerjan> http://qntm.org/destroy
21:04:30 <fizzie> @tell elliott Oh, and the black dot is the sRGB white point's chroma.
21:04:45 <nortti> oerjan: what about if we drill a hole and put all nukes and bombs to it and then launch all of them at the same time
21:04:45 -!- oerjan has quit (Quit: Lost terminal).
21:05:52 <olsner> nortti: probably almost nothing happens
21:06:08 <fizzie> That will just cause oerjan to lose eir terminal, nothing more.
21:06:08 <MDude> Maybe a lot of earthquakes.
21:07:18 <nortti> olsner: but it will propably do more than sractch the surface
21:08:09 <itidus21> plants would survive it i hope..
21:08:23 <olsner> nortti: I think it will just barely scratch the surface :) but enough fallout might make the earth uninhabitable without needing to destroy much of it
21:08:32 -!- augur has joined.
21:08:35 -!- nortti has quit (Quit: nortti).
21:09:26 -!- hagb4rd2 has joined.
21:09:27 -!- hagb4rd has quit (Ping timeout: 246 seconds).
21:11:26 -!- oerjan has joined.
21:15:55 <hagb4rd2> are you familiar with the nadsat slang? you're interested in langs iirc
21:16:26 <hagb4rd2> the clockwork orange soviet-american teen language
21:16:26 -!- oerjan has quit (Quit: Lost terminal).
21:24:42 -!- Nisstyre has quit (Quit: Leaving).
21:25:21 -!- Nisstyre has joined.
21:26:52 -!- oerjan has joined.
21:27:55 * oerjan knows only vaguely that nadsat is supposedly from a clockwork orange, a movie he has never seen.
21:30:06 -!- asiekierka has joined.
21:30:11 -!- asiekierka has quit (Remote host closed the connection).
21:30:36 <olsner> oerjan: you haven't seen it? but it's like the most famousest classic movie
21:31:09 <oerjan> well i've heard enough to think it's not the kind of movie i'd like to see.
21:32:34 <Vorpal> yeah, from what I heard the plot is quite nasty, don't think I'll ever watch that
21:33:13 <Jafet> That movie is a menace to civilised society. It encourages ultraviolence, shameless sexual depravity, and Deutsche Grammophon. I recommend it to everyone.
21:34:32 <Vorpal> does Deutsche Grammophon still exist?
21:34:50 <olsner> if not, how could the movie encourage it?
21:35:10 <Vorpal> "It is now part of Universal Music Group (UMG) since its acquisition and absorption of PolyGram in 1999, and it is also UMG's oldest active label. It is also the oldest surviving established record company."
21:35:20 -!- asiekierka has joined.
21:35:55 <Vorpal> olsner, anyway when is the movie supposed to be set? "In the future" I know, but it was a while ago that movie was made
21:36:23 <olsner> Vorpal: in the year of the Deutsche Grammophon
21:36:48 <olsner> it's like the year of the rat, but different
21:36:55 <oerjan> Jafet: been reading http://tvtropes.org/pmwiki/pmwiki.php/Main/ArsonMurderAndJaywalking ? (EVERYONE ELSE: DON
21:37:23 <Jafet> http://tvtropes.org/pmwiki/pmwiki.php/Main/RuleOfThree
21:37:41 <oerjan> i _tried_ to warn you, but my fingers thwarted me
21:37:54 <Vorpal> oh I read that before, I'm immune thus
21:38:00 <Jafet> http://tvtropes.org/pmwiki/pmwiki.php/Main/AndItsTerrible
21:38:19 <olsner> Vorpal: re-read it, it's been updated since then
21:38:44 <Vorpal> besides I use w3m -dump to read tv tropes. So I get plain text, no links
21:39:04 <olsner> sounds like derptropes
21:39:12 <Vorpal> and I /always/ go for the laconic version if one exists
21:39:42 <Vorpal> anyway I have an exam tomorrow. So good night.
21:39:46 * Vorpal puts computer to sleep
21:40:53 <hagb4rd2> i like the fact that the language in develops in a way where russian and american are melting togeter
21:41:42 <oerjan> Jafet: your last link doesn't exist
21:41:51 <hagb4rd2> who actually has written in during the cold war
21:41:58 <Jafet> http://tvtropes.org/pmwiki/pmwiki.php/Main/AndThatsTerrible
21:42:22 <oerjan> i thought you might mean that. we've been introduced before.
21:42:28 <HackEgo> Gregor took forty cakes. He took 40 cakes. That's as many as four tens. And that's terrible.
21:43:51 -!- asiekierka has quit (Remote host closed the connection).
21:44:16 -!- Vorpal has quit (Ping timeout: 260 seconds).
21:44:51 <oerjan> shachaf: is that some kind of reverse arrow notation?
21:45:31 <shachaf> It's as many as seven 111s.
21:49:10 -!- azaq23 has quit (Quit: Leaving.).
22:04:27 -!- nortti has joined.
22:05:49 -!- tzxn3 has joined.
22:13:43 <HackEgo> ln: `../': hard link not allowed for directory
22:15:02 <nortti> `run ln /hackenv ./hackenv
22:15:05 <HackEgo> ln: `/hackenv': hard link not allowed for directory
22:15:39 <nortti> `run ln -s /hackenv ./hackenv
22:15:53 <HackEgo> UNDELETE.EXE \ bin \ canary \ hackenv \ karma \ lib \ paste \ quotes \ share \ wisdom
22:16:58 <HackEgo> UNDELETE.EXE \ bin \ canary \ hackenv \ karma \ lib \ paste \ quotes \ share \ wisdom
22:17:52 <fizzie> What's that all about?
22:18:25 <HackEgo> http://codu.org/projects/hackbot/fshg/index.cgi/raw-file/tip/paste/paste.18411
22:18:26 <ais523> fizzie: someone tried to run it and complained that it didn't work
22:18:39 <fizzie> `run rm UNDELETE.EXE #you will be missed
22:19:03 -!- Phantom_Hoover has quit (Remote host closed the connection).
22:19:29 -!- ais523 has quit (Remote host closed the connection).
22:20:25 <nortti> so that's what rm UNDELETE.EXE does
22:21:30 <nortti> it makes Phantom_Hoover and ais523 quit
22:27:51 <HackEgo> bin \ canary \ hackenv \ karma \ lib \ paste \ quotes \ share \ wisdom
22:28:36 <HackEgo> HackEgo, also known as HackBot, is a bot that runs arbitrary commands on Unix. See `help for info on using it. You should totally try to hax0r it! Make sure you imagine it's running as root with no sandboxing.
22:28:55 <HackEgo> Runs arbitrary code in GNU/Linux. Type "`<command>", or "`run <command>" for full shell commands. "`fetch <URL>" downloads files. Files saved to $PWD are persistent, and $PWD/bin is in $PATH. $PWD is a mercurial repository, "`revert <rev>" can be used to revert to a revision. See http://codu.org/projects/hackbot/fshg/
22:29:22 -!- tzxn3 has quit (Quit: Leaving).
22:31:17 -!- derdon has joined.
22:37:45 -!- hagb4rd2 has quit (Quit: hagb4rd2).
22:41:29 <shachaf> No worries, we know what to do when a file is rmed.
22:41:39 <shachaf> `run ./UNDELETE.EXE ./UNDELETE.EXE
22:41:42 <HackEgo> bash: ./UNDELETE.EXE: No such file or directory
22:42:28 -!- Patashu has joined.
22:42:30 <fizzie> `run ./UNDELETE.EXE "Gregor's porn stash"
22:42:33 <HackEgo> bash: ./UNDELETE.EXE: No such file or directory
22:49:09 <fizzie> `run ./UNDELETE.EXE .gregors_sneakier_sicko_porn_stash #Nice trick hiding with a dot like that, too.
22:49:11 <HackEgo> bash: ./UNDELETE.EXE: No such file or directory
23:03:46 <pikhq_> I think I hate SDL's audio.
23:04:10 <pikhq_> *Callback* based audio?
23:13:11 -!- nortti has quit (Quit: nortti).
23:35:21 <HackEgo> bash: -c: line 0: syntax error near unexpected token `{:' \ bash: -c: line 0: `:(){:|:&}:;'
23:35:39 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: :(){:|:&}:;: not found
23:35:54 -!- hagb4rd has joined.
23:35:57 <ion> `run :(){:|:&};:
23:36:00 <HackEgo> bash: -c: line 0: syntax error near unexpected token `{:' \ bash: -c: line 0: `:(){:|:&};:'
23:36:08 <ion> `run foo(){foo|foo&};foo
23:36:11 <HackEgo> bash: -c: line 0: syntax error near unexpected token `{foo' \ bash: -c: line 0: `foo(){foo|foo&};foo'
23:36:21 <ion> `run foo(){ foo|foo&};foo
23:36:41 <HackEgo> GNU bash, version 4.1.5(1)-release (x86_64-pc-linux-gnu) \ Copyright (C) 2009 Free Software Foundation, Inc. \ License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> \ \ This is free software; you are free to change and redistribute it. \ There is NO WARRANTY, to the extent permitted by law.
23:36:51 <ion> `run printf '%s\n' 'foo(){ foo|foo&};foo'
23:37:10 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: init: not found
23:37:51 <sebbu> damn, were it 0 i would have tried using dd ^^"
23:39:23 <ion> You can still use dd. :-P
23:40:25 <HackEgo> rootfs on / type rootfs (rw) \ none on /bin type hostfs (ro,nosuid,relatime,/bin/) \ none on /usr type hostfs (ro,nosuid,relatime,/usr/) \ none on /dev type hostfs (ro,nosuid,relatime,/dev/) \ none on /opt type hostfs (ro,nosuid,relatime,/opt/) \ none on /lib type hostfs (ro,nosuid,relatime,/lib/) \ none on /sbin type hostfs (ro,nosuid,relatime,/sbin/) \ none on /lib64 type hostfs (ro,nosuid,relatime,/lib64/) \ none
23:40:39 <sebbu> `dd if=/dev/zero of=/dev/sda
23:40:42 <HackEgo> dd: opening `/dev/zero of=/dev/sda': No such file or directory
23:42:44 <HackEgo> Linux umlbox 3.0.8-umlbox #2 Sun Nov 13 21:30:28 UTC 2011 x86_64 GNU/Linux
23:44:57 <pikhq_> Who was the moron that designed SDL audio, and can I please cause bodily harm on them?
23:45:34 <sebbu> pikhq_, can't you use sdl mixer or something else ?
23:46:21 <pikhq_> sebbu: I doubt that'll help, given that my problem is I have a single stream of samples and want them to play.
23:46:52 <sebbu> and it doesn't ? :o
23:47:06 <pikhq_> And what I'd like to do is some hypothetical SDL_PlaySamples(samples, len);. But nooooooo, it has to have some cockamamy poorly-thought-out callback scheme.
23:47:46 <pikhq_> So, I've got a ring buffer that I add samples to, and read from in the callback... And it sounds really fucking weird.
23:48:29 <sebbu> maybe the reading and the cycling of the ring isn't synchronized
23:49:08 <pikhq_> I lock the audio thread whilst writing to the ring buffer.
23:49:36 <pikhq_> (BTW, WHY THE HELL SHOULD I HAVE TO DEAL WITH A THREAD JUST TO FEED SAMPLES TO THE AUDIO PLAYBACK WHY WHY WHY WHY)
23:49:59 <sebbu> pikhq_, i was more saying "maybe it reads less/more than it shoul"
23:53:23 <sebbu> or it pause too long
23:54:22 <fizzie> SDL_mixer does exactly that; you can just give it a "chunk" i.e. an in-memory sample, and it'll handle feeding that to SDL. (Admittedly it's slightly overkill, maybe, and has audio file loading and whatnot.)
23:55:30 <fizzie> (Disclaimer: haven't bothered to use it.)
23:55:38 -!- derdon has quit (Remote host closed the connection).
23:58:22 <pikhq_> That I'd have to use SDL mixer for that is... Obnoxious.