00:00:48 <Gregor> Seeing as that #line has no semantic behavior, I'm gonna go with "who cares if it supports it"
00:05:44 <olsner> bah, turned out I needed another bunch of segments for user space, I have *eight* of the bastards now
00:06:23 <olsner> and this is for long mode which was supposed to do away with segmentation
00:07:20 <zzo38> Gregor: Well, I suppose the error messages will be harder to read if it doesn't support #line command. (Even then, it might be possible to make a program which filters the error messages and reads the #line commands to convert them)
00:08:06 <zzo38> (That is, the output of Frama-C (or any other program that does something like that) is piped to the filter program)
00:15:15 <pikhq> olsner: In long mode, you cannot have segments that are smaller than the entire memory space.
00:15:54 <pikhq> olsner: Unfortunately, it still uses segmentation for its idea of kernelspace and userspace seperation.
00:16:25 * pikhq concludes that qemu's I/O is really insanely slow when it has to expand a disk image.
00:24:46 <coppro> this problem should not be hard
00:24:52 <coppro> why am I spending >\epsilon time on it
00:31:42 <zzo38> The FSF has a list of free software licenses incompatible with the GNU GPL. Some of them is because you are required to change the name of the software. Didn't they fix that with GPL v3, to allow the software to have the additional restriction that you cannot call it the same thing as the official version?
00:31:51 <zzo38> (Maybe I am wrong, but I don't know.)
00:36:03 -!- poiuy_qwert has quit (Ping timeout: 272 seconds).
00:37:07 -!- FireFly has quit (Quit: swatted to death).
00:37:14 -!- poiuy_qwert has joined.
00:38:02 <Vorpal> <olsner> bah, turned out I needed another bunch of segments for user space, I have *eight* of the bastards now <-- paging! ;P
00:51:15 <zzo38> If GCC had a -Wcast-spell option, what would it do anyways?
00:53:33 <Gregor> zzo38: Warn when you do casts which are likely wrong due to typos.
00:55:07 <zzo38> Gregor: That seems correct.
00:56:29 <zzo38> Yes, that is what it would do, I guess. (Of course it doesn't exist; I even tried it. It was in a list of joke options for GCC)
00:58:41 <Gregor> Along with -vomit-frame-pointer
01:12:18 -!- BeholdMyGlory has quit (Remote host closed the connection).
01:16:57 <pikhq> Bweheheh. Qemu's snapshotting ability lets me get around load times.
01:17:19 <pikhq> Take *that*, PC games!
01:20:07 <Gregor> OMG best idea ever. swap over nbd.
01:21:30 <Gregor> 1) sshfs (the FUSE one)
01:21:41 <Gregor> 2) losetup /dev/loop0 <whatever>/swapfile
01:21:55 <Gregor> Swapping can context-switch to a program, which may be swapped out!
01:23:21 <zzo38> pikhq: But, did you consider about the Random Number Hole?
01:37:51 <pikhq> Gregor: ... Oh dear God the consequences of that.
01:38:35 -!- Sgeo has joined.
01:38:41 <Sgeo> Random Number Hole?
01:38:44 <Gregor> Can you say KERNEL-MODE INFINITE LOOP?
01:46:18 <Sgeo> It occurs to me that places like Fark and Reddit must sound to Freepers what Freepers sound to places like Fark and Reddit
01:46:19 -!- wareya has quit (Read error: Connection reset by peer).
01:46:29 <Sgeo> I'd love to see Freeper reaction to a Fark thread, for some balance
01:46:53 -!- wareya has joined.
01:57:57 <pikhq> Sgeo: You'd have to go a *bit* further left for that.
02:04:36 -!- rodgort has quit (Ping timeout: 245 seconds).
02:04:39 <zzo38> I am writing the book "Literate Programming Hacks", it describes use and examples of Enhanced CWEB, as well as various tricks that you can make with it. You can read so far if you want to, make suggestions, contributions, questions, ideas, please.
02:09:10 -!- Wamanuz4 has joined.
02:09:33 -!- rodgort has joined.
02:09:40 -!- Wamanuz3 has quit (Ping timeout: 245 seconds).
02:13:08 <zzo38> And tell me if you found any typing mistakes, too.
02:24:59 -!- calamari has joined.
02:26:49 <Gregor> pikhq: DAMN IT ME. FYTHE IS SO AWESOMETERRIBLE. ARGH.
02:29:59 <pikhq> Gregor: DO TELL, ÞEE.
02:30:17 <Gregor> Implementing the transform engine in C is painful :P
02:32:27 <Gregor> The final transform from IR -> runnable code is part of the transform engine, so it's not knottable.
02:32:48 <calamari> so I thought my computer was halfway decent until I just spent the last few hours waiting for an ubuntu stock kernel to compile
02:33:26 <pikhq> Insufficient magic.
02:33:34 <pikhq> calamari: *Few hours*?
02:33:40 <pikhq> calamari: Dear God that's sad.
02:34:02 <calamari> I've always made custom kernels with just what I needed in them
02:34:39 -!- augur has quit (Remote host closed the connection).
02:56:58 <Gregor> To be fair, Ubuntu's stock kernel probably builds with (!ships) basically every module in the source.
02:57:32 <pikhq> Gregor: Still should only take an hour or so.
02:59:02 <Gregor> Not if I built it on my ChiPad!
03:02:21 <pikhq> How much swap does that thing have?
03:02:49 * pikhq waits as Gregor sets up a terabyte swap device via nbd before answering
03:02:49 -!- augur has joined.
03:03:43 <pikhq> Good, it can definitely build then.
03:07:39 <calamari> it's going in some kind of loop
03:12:28 <zzo38> A different kind of Monty Hall problem is as thus: The contestant selects one of three doors. One has a car, two have goats. Monty Hall opens one that the contestant didn't select which contains a goat (using a satellite to control the doors). One goat always tells the truth and one goat always lies. The problem is: Who is the liar?
03:14:23 <pikhq> Answer: develop a taste for cheese made from goat's milk, and make cheese.
03:44:00 -!- poiuy_qwert has quit (Read error: Connection reset by peer).
03:47:28 -!- poiuy_qwert has joined.
04:00:38 -!- calamous has quit (Quit: Leaving).
04:04:22 -!- Mathnerd314 has joined.
04:11:30 <pikhq> D'awww. Missed moment of awesome. Carl Sagan wanted "Here Comes the Sun" (by The Beatles) on the Voyager record. EMI refused.
04:22:51 <Sgeo> Shouldn't that be the Beatles's decision? (Assuming they were still around)
04:24:56 <pikhq> That's not how copyright law works.
04:25:09 <pikhq> Because it's written by a bunch of phalluses.
04:26:58 * Sgeo vaguely imagines Voyager landing somewhere and destroying life just as it's beginning to form
04:27:18 <Sgeo> That's probably incredibly unlikely, but it would be... disturbing
04:27:52 <Sgeo> Actually, what's Voyager's final fate likely to be?
04:28:06 <Sgeo> Hit star, black hole, drift until universe ends?
04:31:28 <pikhq> "Star Trek: The Motion Picture".
04:31:39 -!- Decarabia has quit (Ping timeout: 264 seconds).
04:31:49 <pikhq> (which is not a terrible film, but has *major* pacing issues...)
04:34:22 <Gregor> Sgeo: As dictated by Star Trek XI, its final fate is that it never existed in the first place :P
04:34:59 * Sgeo builds a Cathedral
04:35:33 <Sgeo> Wow, I suck at references
04:35:42 <pikhq> Sgeo: Nah, I got the reference.
04:35:55 <Sgeo> Yeah, but could have said something that made more sense
04:36:16 <Sgeo> Like "/me joins the campaign for real time"
04:36:25 <pikhq> Okay, true, that would've helped a lot.
04:37:08 <Sgeo> It's been so long since I read that book that the only reason I remember anything about that is because I saw it on Wikipedia :/
04:37:43 <pikhq> It's been a couple months since I last listened to the radio series.
04:38:09 <pikhq> The book... I would have reread it sometime in the past couple years, but I don't know where my copy is.
04:38:13 <Sgeo> Can I legally listen to the series for free?
04:38:39 <pikhq> Not unless you have a time machine.
04:39:24 <Sgeo> There is a campaign for real tea...
04:40:16 <Sgeo> No there isn't, but there's someone saying there should be
04:40:21 <Sgeo> http://www.telegraph.co.uk/comment/letters/3563374/Letters-to-the-Telegraph.html
04:40:37 <Sgeo> And the word Cathedral right there is why this thing came up in my google search
04:41:55 <pikhq> There is, however, a standard cup of tea.
04:43:56 <Sgeo> According to wiki, it's more for documenting the making of a cup of tea
04:44:11 <Sgeo> Rather than standardizing a cup of ... oh, hm
04:44:21 <pikhq> It still allows you to produce and ISO standard cup of tea.
04:45:46 <pikhq> Oooh. Andre Geim received the Ig Nobel Prize in Physics in 2000, and the *Nobel* Prize in Physics in 2010.
04:46:23 <Sgeo> I'm going to guess that the Ig Nobel is for the tea? What's the Nobel for?
04:46:35 <pikhq> Ig Nobel for magleving a frog.
04:47:00 <pikhq> Nobel for experiments with graphene.
04:53:00 * Sgeo manages to confirm that someone's saying ISO-0xC0FFEE was, in fact, a joke unrelated to the number 3103
04:53:10 <Sgeo> Without using anything to do the actual conversion
04:53:28 <Sgeo> (Factor was loading too slowly)
04:55:42 <Sgeo> <Sgeo> 0xC0FFEE looks a bit too large to be 3103, but I'm checkin... 1 2 4 8 | 16 32 64 128 | 256 512 1024 2048 | 4096
04:55:42 <Sgeo> <Sgeo> No way 3103 has more than 3 hex digits
04:56:04 <Sgeo> I don't know why I'm saying this
04:58:27 <pikhq> Because 0xC0FFEE should be a real standard.
04:58:32 <pikhq> Preferably for coffee.
05:06:44 -!- Decarabia has joined.
05:14:08 <zzo38> If you earn the Ig Nobel prize and Nobel prize both for the same thing, then you win.
05:14:31 <zzo38> Is there METAFONT files for Creative Commons licenses?
05:20:29 -!- calamari has quit (Quit: Leaving).
05:34:40 <pikhq> zzo38: I agree. That would be win.
05:39:11 <Sgeo> Was about to say something like "If I die in a fire tonight, it's because I left the dryer on while I slept", but I'm deciding I'll continue drying the clothes tomorrow morning
05:45:09 -!- Sgeo has quit (Read error: Connection reset by peer).
05:49:28 -!- sftp has quit (Remote host closed the connection).
05:52:49 <pikhq> Clearly Sgeo's "dryer" is a blowtorch.
05:54:40 -!- augur has quit (Remote host closed the connection).
06:20:41 -!- poiuy_qwert has quit (Ping timeout: 240 seconds).
06:22:57 -!- augur has joined.
06:23:17 -!- zzo38 has quit (Quit: Now I am going to fire my clothes manually while sleeping).
06:28:28 -!- nooga has quit (Ping timeout: 245 seconds).
07:13:10 -!- augur has quit (Remote host closed the connection).
07:15:16 -!- augur has joined.
07:22:09 -!- augur has quit (Remote host closed the connection).
07:32:25 -!- FireFly has joined.
07:47:49 -!- MigoMipo has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:13:46 -!- augur has joined.
08:18:10 -!- MigoMipo has quit (Read error: Connection reset by peer).
09:41:03 -!- nooga has joined.
10:03:03 -!- ais523 has joined.
10:13:08 -!- yiyus_ has quit (Ping timeout: 240 seconds).
10:27:12 -!- FireFly has quit (Quit: swatted to death).
10:44:15 -!- yiyus_ has joined.
10:46:54 -!- ais523 has quit (Remote host closed the connection).
10:55:54 -!- ais523 has joined.
11:59:12 -!- ais523 has quit (Remote host closed the connection).
12:51:49 -!- lilja_ has quit (Quit: ~ Trillian Astra - www.trillian.im ~).
13:14:41 <Gregor> Are there any free SSL cert providers that don't suck as hard as startSSL X_X
13:15:02 <Gregor> (Observing that CAcert sucks much, much more since nobody recognizes them)
13:26:28 -!- sftp has joined.
13:32:09 -!- aloril has quit (Ping timeout: 240 seconds).
13:34:33 <fizzie> I doubt there are any; I was in fact a bit surprised about StartSSL being as reasonable as it is. How exactly do they suck?
13:49:55 -!- nooga has quit (Ping timeout: 265 seconds).
14:03:54 -!- ais523 has joined.
14:10:31 -!- aloril has joined.
14:16:11 -!- elliott has joined.
14:17:12 <elliott> 08:00:36 <mycroftiv> elliott: I know you weren't talking to me, but by a wonderful coincidence "make it detect the client" was actually good helpful advice for a software project I'm working on, so thanks!
14:17:17 <elliott> mycroftiv: also, it's more than 15
14:17:19 <elliott> more helpful advice for you
14:18:04 <elliott> 12:36:40 <olsner> probably my own API, posix is overrated
14:18:10 <elliott> kinder words than I'd choose
14:18:31 <elliott> 12:43:50 * ais523 tries to figure out what something likely is if Vorpal describes it as a joke
14:18:31 <elliott> 12:43:57 <ais523> a... sentence?
14:19:28 <elliott> 13:42:13 <nooga> i remember that one guy from gamedev.pl wrote a game that was multiplayer and allowed to mine and set blocks, build houses from block & stuff and it looked better than minecraft
14:19:28 <elliott> 13:42:21 <nooga> it was, huh, in 2003 ?
14:19:28 <elliott> 13:42:39 <nooga> how come that such a crap gets so many players and cash
14:19:34 <elliott> 13:42:40 <Vorpal> nooga, minecraft is intentionally retro though
14:19:34 <elliott> 13:42:51 <nooga> i don't think so
14:19:43 <elliott> 13:43:14 <Vorpal> nooga, anyway good gameplay beats good graphics
14:19:44 <elliott> 13:43:24 <nooga> that's true
14:19:44 <elliott> 13:43:32 <nooga> but it might have looked better
14:20:00 <elliott> 13:44:38 <nooga> i guess it's software rendered
14:20:00 <elliott> 13:44:45 <Vorpal> nooga, no it isn't
14:20:00 <elliott> 13:44:49 <Vorpal> or well, it isn't for me
14:20:00 <elliott> 13:44:49 <fizzie> It's OpenGL.
14:20:00 <elliott> 13:44:52 <nooga> then WTF?!
14:20:01 <elliott> 13:45:00 <nooga> why do i get 10 FPS?
14:20:05 <elliott> Because the rendering depth is ridiculous.
14:20:11 <elliott> 13:46:33 <nooga> i guess it was written by some student in 2-3 days
14:20:21 <elliott> 13:47:02 <nooga> "well, i know a bit of java and we did opengl during the classes so uh, i will write funny game" he thought
14:20:30 <elliott> I wonder what uniformly-sized 3D blocks did to you as a child to inspire this hatred.
14:21:06 <elliott> 15:15:29 <nooga> what an impertinence
14:21:07 <elliott> 15:16:00 <nooga> i hate the author
14:21:10 <elliott> 15:17:46 <nooga> cheesy virtual lego for 4 packs of cigarettes
14:21:38 <elliott> Quite an advanced stage of addiction where it's bad and wrong of someone to ask for more money than it takes for N of what you're addicted to.
14:21:45 <fizzie> Wow, I didn't notice that last bit. That's some quality hate there.
14:21:48 <elliott> Also, nobody is forcing you to buy the horrible student non-retro slow game of evil.
14:21:56 <elliott> You know, if you don't want to.
14:23:12 <elliott> fizzie: Mind, if real-life lego had zombies and mining...
14:24:45 <fizzie> Also really tiny portals to hell.
14:25:00 <elliott> fizzie: I have been unable to find them. :-(
14:25:07 <fizzie> You need to build them.
14:25:08 <elliott> fizzie: I am not the good at Minecraft.
14:25:18 <elliott> fizzie: Oh. It has to be said that I have also been unable to craft a single thing.
14:25:36 <fizzie> You need to build them out of obsidian, though, and obsidian is rather rare and tricky.
14:26:17 <fizzie> You need a diamond pick to mine it with, and it's oh-so-slow with that too; and naturally occurring obsidian usually sits directly on top of lava, so it's hard-ish to collect. You can make your own, though.
14:27:03 <elliott> fizzie: Man, and all I do is hit pigs and go downwards.
14:27:30 <elliott> http://www.minecraftwiki.net/images/2/2d/Natural_obsidian.png Wait, where is the Obsidian here?
14:27:51 -!- FireFly has joined.
14:28:24 <elliott> * Note: Going past the Bedrock barrier will kill you, reason being is that the world ends and you will fall to you death and you won't be able to place block to save yourself due to being far below the closest block.
14:29:03 <elliott> fizzie: I like the "economy portal"; http://www.minecraftwiki.net/images/f/fa/Portalcomparison.png.
14:29:34 <fizzie> Yeah, all my (two) portals are the economy models.
14:29:40 <elliott> "When re-entering the normal world, any distance covered in The Nether is multiplied by 8 times, effectively making The Nether a fast travel zone that allows for greater distances to be covered in a short amount of time in the normal world."
14:29:43 <fizzie> But I filled the corners with that red hell-stone.
14:30:12 <elliott> fizzie: You should feed Minepedia into fungot.
14:30:51 <fizzie> Just about everything is "Physics: No", only sand and gravel fall down.
14:31:12 <fizzie> I should, but I doubt they have XML data dumps for me, and scraping is so impure.
14:31:16 <elliott> Bah, it's funny! So shush.
14:32:14 <elliott> fizzie: Minecraft monsters freak me out.
14:32:33 <fizzie> They are pretty scary.
14:32:40 <elliott> All those moabs (pronunciation-accurate!)
14:33:08 <elliott> fizzie: So far, every time I encounter them I just try and whack them with whatever block I'm holding. That ... that rarely goes well.
14:33:35 <fizzie> Yeah, maybe you should make a stone sword or something.
14:34:41 <elliott> fizzie: That would require learning how to craft.
14:35:22 <fizzie> You craft things up to 2x2 size in the inventory (i), and larger with the workbench (made with 2x2 planks). But of course if you have some sort of ideological opposition...
14:35:23 <elliott> [["Rana" were mobs who resembled little girls with frog hats. They were originally in-game as a test; Rana was made by "Dock", Minecraft's past artist, for another project of his. The player's character was supposed to look like Rana, albeit with a miner hat, no ponytails and a different skin. This was scrapped with Dock's removal, with a fair few people arguing for her to stay.]]
14:35:56 <elliott> fizzie: Well, it's just that I'd have to collect blocks and stuff to craft, and who likes doing that?
14:36:22 <elliott> fizzie: Also, it'd feel too much like a *game*.
14:36:41 <elliott> fizzie: Are you meant to figure out the recipes by yourself?
14:36:51 <elliott> Somehow I doubt anyone actually does :P
14:37:37 <fizzie> But I'm sure there's at least one "purist" in the world.
14:37:54 <elliott> "Used to hold Mushroom Stew. The player keeps the bowl when they eat the stew."
14:37:58 <elliott> Eat the stew, keep the bowl!
14:38:23 <elliott> fizzie: it seems that i am too disillusioned with minecraft to play it after thinking of Dwarf Fortress ONLINE
14:38:34 <elliott> best game?? no. only game.
14:39:02 <elliott> http://www.minecraftwiki.net/wiki/Creeper AAAAAAAAH
14:39:03 <fizzie> Yes, Dwarf Fortress is indubitably far more eXtreme.
14:39:38 <fizzie> http://www.minecraftwiki.net/wiki/Zombie_Pigman is a bit... too.
14:40:18 <elliott> fizzie: At least that isn't camouflaged.
14:40:25 <elliott> fizzie: Now I'm wondering what Dwarf Fortress 3D would be like.
14:40:59 <elliott> "By editing the mob's files, it is possible to make the slimes much bigger, some values creating 50 block high slimes."
14:43:36 <elliott> [[ New plan: Whenever someone tells me to "fix smp" or similar, I will work on something else for four hours!]]
14:43:44 <elliott> Assuming SMP is what I think it is, I wonder what problems people have been having.
14:44:39 <elliott> "Creepers are misunderstood creatures. Try to befriend them instead of hurting them! A friendly pat on the back should do it!" ;; oh dear god
14:46:59 <elliott> fizzie: I don't want to scare you or anything, but Creepers carry records. It is probable that they have found jukeboxes too, and have some form of civilisation and intelligence.
14:47:15 <fizzie> In minecraft parlance, SMP is Survival MultiPlayer.
14:47:22 <fizzie> And that's buggey as Nether.
14:47:31 <elliott> fizzie: Ahhhh. (Wait, how do you do that?)
14:48:27 <fizzie> Minecarts don't work there, and all mobs (as well as players) are unkillable; the server doesn't track health.
14:48:51 <fizzie> On the plus side, you can jump from top of the map to bottom for fast one-directional vertical travel.
14:49:13 <fizzie> I actually jumped down my mine-shaft in single-player mode before thinking.
14:49:43 <elliott> fizzie: I will not be satisfied until I get a portal on the very bottom of a level, somehow flat, go inside, dig a hole all the way down, get back into the portal, build a tower to the very top of a level, and ... drop
14:50:02 <elliott> <fizzie> I actually jumped down my mine-shaft in single-player mode before thinking. ;; What I do is, I destroy the block below me. Forever.
14:50:13 <ais523> elliott: this reminds me of the volcanos in Dwarf Fortress
14:50:26 <elliott> fizzie: Then I collect a bunch of blocks -- uh, you know, those really easy to destroy ones? -- and build my way back up my shaft.
14:50:33 <elliott> (Uh, except worded less like that.)
14:50:40 <ais523> apparently, they were originally a bug caused by lava falling down a bottomless pit, overflowing it, and falling back on from the top of the screen
14:50:54 <elliott> wait, *overflowing* a *bottomless* pit?
14:51:04 <ais523> elliott: it's a MAXINT pit
14:51:18 <elliott> ais523: impressive to get MAXINT lava
14:51:33 <ais523> well, I suppose it depends on the units, and the size of the integers
14:51:41 <ais523> I wouldn't be surprised if they were using shorts or even chars there to save memory
14:54:10 <elliott> anyone want to lend me a server farm?
14:54:16 <elliott> with a decent network pipe?
14:54:33 <fizzie> Just put whatever you want to the cloud.
14:54:52 <fizzie> Of course money will fly out too, but them's the breaks.
14:54:54 <elliott> fizzie: ...how... Look at the first line of http://sprunge.us/SFFO X-D
14:55:17 <elliott> fizzie: (The bot, there, is meant to be running on $some_random_server. I refuse to use the term "cloud", only make snarky references to it in nomenclature.)
14:55:41 <elliott> I hypothesise that fizzie has read my mind.
14:55:52 <fizzie> It could be the other way around, too.
14:56:13 <fizzie> Didn't you just do your psychic "more than 15" stuff.
14:56:37 <elliott> fizzie: If that was the case, you'd have asked for a server farm to implement that on first!
14:56:42 <elliott> Actually I think I've mentioned the idea to ais523 before.
14:56:52 <elliott> (fizzie's mind is in fact patented.)
14:59:50 <fizzie> "Document is currently being inspected. Please allow 7-10 days. No action is required by you at this time." 19 days already! Stop being in that state!
15:00:22 <elliott> fizzie: This is what you get for STEALING MY IDEA.
15:03:06 <elliott> Things worse then Apple regularly spamming you because you did, at one point, buy an Apple product: Apple spamming you about Microsoft Office 2011 being out for OS X.
15:09:51 <ais523> elliott: I'm not sure either
15:09:56 <elliott> ais523: "wait what?" at what?
15:10:04 <ais523> oh, Apple spamming about a Microsoft product
15:10:26 <elliott> Subject: Introducing Microsoft Office for Mac 2011.
15:10:38 <ais523> (Apple spamming about the availability of a Windows Phone 7 sync app for Mac OS X would be funnier, though)
15:10:47 <elliott> ais523: Apple and Microsoft have an agreement of some sort there, I think.
15:11:04 <elliott> ais523: They sell it in the online store and all, and advertise it on the sites of the various Mac models.
15:11:08 <ais523> where Microsoft advertises, umm, Safari for Windows?
15:11:21 <ais523> (acually, Microsoft /do/ advertise Safari for Windows, but only because the EU told them to)
15:11:36 <elliott> ais523: No, where Apple advertises office and Microsoft do nothing because Apple couldn't possibly make demands :)
15:11:49 <elliott> I bet it does get conversions out of the "...but what about my precious Office???" crowd though.
15:11:55 <elliott> Does Office for Mac 2011 support Visual Basic?
15:12:04 <elliott> [[Yes. Office for Mac 2011 comes with VBA 6.5, ported from Office for Windows. According to Macworld magazine this version of VBA will give Mac users all the latest VBA tools and be fully compatible with Office. So if you create macros on your Mac they "should work just fine" in Windows.]]
15:13:50 -!- zzo38 has joined.
15:14:37 <fizzie> I tried "Office v. X" for a bit on the iBook there; it was, well, reasonable, I guess. (Seems like 2011 will be >=10.5 must-be-Intel though.)
15:16:33 <elliott> fizzie: Do you really still have OS X on there?
15:18:00 <fizzie> I do have that Ubuntu dual-boot thing going on, but it's a lot more noisy laptop when in Linux. (Either the temperature tolerances or just something about fan control or power-saving in general could do with some tuning.)
15:18:10 <elliott> "Commerce and business will shift from offices and malls to networks and modems. [...] Baloney. Do our computer pundits lack all common sense? The truth in no online database will replace your daily newspaper, no CD-ROM can take the place of a competent teacher and no computer network will change the way government works. [...] Yet Nicholas Negroponte, director of the MIT Media Lab, predicts that we'll soon buy books and newspapers straight over
15:18:54 -!- MigoMipo has joined.
15:19:01 <fizzie> Did he say why it's baloney? (Bologna!)
15:19:36 <elliott> fizzie: No, not really. http://www.newsweek.com/1995/02/26/the-internet-bah.html
15:19:46 <elliott> [[What the Internet hucksters won't tell you is tht the Internet is one big ocean of unedited data, without any pretense of completeness. Lacking editors, reviewers or critics, the Internet has become a wasteland of unfiltered data. You don't know what to ignore and what's worth reading. Logged onto the World Wide Web, I hunt for the date of the Battle of Trafalgar. Hundreds of files show up, and it takes 15 minutes to unravel them—one's a biog
15:19:46 <elliott> raphy written by an eighth grader, the second is a computer game that doesn't work and the third is an image of a London monument. None answers my question, and my search is periodically interrupted by messages like, "Too many connectios, try again later."]]
15:19:51 <elliott> fizzie: That's pretty much the only, uh, argument in there.
15:20:08 <elliott> Of course, now we can get our unedited, inaccurate information from one place: Wikipedia!
15:20:37 <elliott> [[ Every voice is heard. The cacophany more closely resembles citizens band radio, complete with handles, harrasment, and anonymous threats.]]
15:21:02 <elliott> fizzie: It's the guy who sells http://www.kleinbottle.com/.
15:21:11 <elliott> fizzie: And also wrote The Cuckoo's Egg.
15:21:11 -!- aloril has quit (Ping timeout: 255 seconds).
15:21:19 <ais523> elliott: at least Wikipedia is usually right, despite being inaccurate and unedited
15:21:22 <elliott> Well, who sells Klein bottles from that site, not who... sells that site.
15:21:28 <fizzie> Heh, it's again that "-- the myopic glow of a clunky computer replaces the friendly pages of a book" thing.
15:21:31 <elliott> ais523: Inaccurate true statements!
15:22:04 <fizzie> The Klein bottle thing is awesome, though; I've been thinking of buying one some day as a birthday self-present or something.
15:22:25 <elliott> fizzie: Slereah_ has the mug, I think.
15:22:43 <elliott> http://www.kleinbottle.com/drinking_mug_klein_bottle.htm This'un.
15:23:48 <fizzie> It sounds a bit too useful.
15:23:59 <elliott> fizzie: Haha -- [[Stoll was a regular contributor to MSNBC's The Site.]] -- [[The Site, hosted by Soledad O'Brien, was an hour-long TV program devoted to the Internet revolution.]]
15:24:13 <elliott> Wow, he even wrote a book on how the Web is rubbish; dedication.
15:24:17 <fizzie> What I'd like to see is a restaurant that randomly served this http://www.kleinbottle.com/wine_bottle_klein_bottle.html to people ordering wine.
15:24:30 <elliott> http://upload.wikimedia.org/wikipedia/en/e/e2/Devnull.gif Oh wow, he's called "Dev Null".
15:25:24 <elliott> "For optimal aerodynamic performance, your Wine Bottle Klein Bottle has smooth, spline-like curves."
15:25:27 <fizzie> Like they say: "Guaranteed to frustrate even the most dedicated wine connoisseur: it's difficult to fill, difficult to pour, and difficult to clean."
15:27:57 <elliott> bash-4.1 egobf-0.7.1 ick-0.-2.0.29 perl-5.12.2 ruby-1.9.2-p0
15:27:58 <elliott> CLC-INTERCAL-1.-94.-2 emacs-23.2 nginx-0.8.53 Python-2.7 zsh-4.3.10
15:28:06 <elliott> fizzie: Quick, give me other software that has something like a ./configure script to install!
15:28:53 <fizzie> Sorry, I need to run to a bus. NASM has a configure script. ->
15:30:10 <ais523> elliott: does it all work, in addition to installing?
15:30:42 <elliott> ais523: well, you have to set $PERL5LIB for CLC-INTERCAL, but yes.
15:30:46 <elliott> ais523: (that's to be expected, though)
15:31:10 -!- aloril has joined.
15:31:28 * elliott tries to figure out why "aloril" rings a bell...
15:31:31 <ais523> I'm actually impressed with ick here for finding all its libaries and so in in the right place
15:31:57 <elliott> ais523: it's probably a sign of madness that I want to make it read Makefiles and detect a --prefix-esque variable and a relevant target (e.g. for programs that have platform-specific targets), right?
15:31:59 <elliott> for non-autotools programs
15:32:20 <ais523> elliott: it's the good sort of madness; flow with it, don't fight it
15:32:37 <ais523> I don't see what Python has to do with it
15:32:46 <elliott> ais523: Parsing Makefiles with Python seems odd.
15:33:01 <ais523> hmm, perhaps, mostly because it's neither make(1), sed, nor Perl
15:33:32 <elliott> The thing make's written in?
15:33:47 <elliott> I'll try nasm, because I'm crazy, ha ha.
15:33:57 <zzo38> The company that makes the klein bottle has a lifetime guarantee that you will live your entire lifetime.
15:35:11 <elliott> ais523: oh, by the way, if you ever release a tarball that ends in -src.tar.gz or -src-unix.tar.gz or such silliness, my program will hate you even more
15:35:20 <elliott> I'm going to have to work around that for other programs...
15:36:49 <elliott> elliott@dinky:~$ /opt/nasm-2.09.03/bin/nasm
15:36:49 <elliott> nasm: error: no input file specified
15:42:40 <elliott> dear god @ Inkscape's interface
15:44:32 <ais523> for using, or for compiling?
15:46:54 <zzo38> elliott: What about Inkscape's interface? (I don't use Inkscape, so I don't know)
15:47:04 <elliott> ais523: using; Debian comes with it
15:48:23 <elliott> ais523: http://imgur.com/SOvrL.png, if you care
15:52:44 <pikhq> zzo38: I love Acme Klein Bottles.
15:53:17 <elliott> pikhq: Psht; doesn't he know that e-commerce is a stillborn concept?
15:53:54 <elliott> pikhq: (If you didn't see, he wrote an article in 1995 claiming so.)
15:54:10 <pikhq> I'll grant, his website design is very very bland. Like a '95 website design done by someone sane.
15:54:36 <elliott> pikhq: (As well as seemingly rejecting the idea of Amazon a year after it was founded.)
15:54:44 <pikhq> Actually, more than "Like". It is.
15:55:54 <zzo38> I think GF-Magick is more better instead of Inkscape try it maybe, it is not cluttered?
15:56:24 <zzo38> Try this program tell me if it can work for you or not, or other question/complaint. http://zzo38computer.cjb.net/prog/gfmagick/gfmagick.zip
15:56:26 <elliott> zzo38: Are you using ImageMagick or GraphicsMagick?
15:56:27 <pikhq> zzo38: I'm pretty sure the two have completely different usecases.
15:56:42 <pikhq> Much like the GIMP and ImageMagick.
15:57:14 <zzo38> elliott: ImageMagick. Although, you might be able to get it to work with GraphicsMagick too (you can try if you want to).
15:57:32 <elliott> zzo38: You should probably use GraphicsMagick; ImageMagick development is dead and GraphicsMagick is much better maintained.
15:58:35 <elliott> zzo38: Also, ImageMagick is now under corporate control, and the corporation isn't very community-friendly.
15:59:07 <elliott> zzo38: GraphicsMagick is also a lot smaller and faster than ImageMagick and it has fewer dependencies, too.
16:00:08 <elliott> zzo38: Plus the interface is compatible (for the command-line interface, simply put "gm " in front of every command, e.g. "gm mogrify ...".)
16:00:47 <Vorpal> <elliott> fizzie: I have been unable to find them. :-( <-- not in the old version, if you were still using that
16:01:02 <Vorpal> also it seems buggy to me
16:01:07 <zzo38> Maybe I can do that. ImageMagick, even if under corporate control is still free software. But if GraphicsMagick is smaller and faster and compatible, then is probably much more better. Maybe in some time I will insteall GraphicsMagick.
16:01:17 <Vorpal> I end up in different portals different times. And so on
16:02:01 <elliott> Vorpal: I thought Nether portals were far downwards.
16:02:12 <elliott> Vorpal: Anyway doesn't it auto-update? It can tell I'm not premium so presumably it connects to the server.
16:02:20 <elliott> Or does it only update if I'm premium?
16:02:28 <Vorpal> elliott, well updating requires you to login and get a session token
16:02:38 <Vorpal> elliott, so yeah, only if premium
16:03:07 <elliott> Vorpal: Don't suppose I could -- hypothetically -- not have a couple of non-files from somewhere that isn't ~/.minecraft to not get a more inaccurate perception of the game as it isn't, hypothetical non-cough?
16:03:11 -!- Phantom_Hoover has joined.
16:03:24 <elliott> Ah! The hypothetical non-phantom hoover.
16:03:39 <zzo38> elliott: Again I will tell you, that you might be able to get GF-Magick to work with GraphicsMagick, it should not be difficult to make the change (either the .w file can be fixed to allow GraphicsMagick, or a .ch file can be written to compile the program to work with GraphicsMagick instead.)
16:03:58 <elliott> zzo38: Okay. (Again? When did you say this previously?)
16:06:33 <zzo38> Tell me if you can use this program or if you understand this program or whatever, too.
16:06:37 -!- zzo38 has quit (Quit: \\\).
16:07:42 <ais523> pikhq: sometimes it's fun wondering which of two inappropriate programs are best for a particular job
16:08:30 <elliott> ais523: would you rather edit text files with the GIMP or Rhythmbox?
16:08:55 <ais523> I can't figure out how rhythmbox would do it at all
16:09:09 <elliott> ais523: Ooh, wait, I have a better inappropriate program.
16:09:17 <elliott> ais523: Would you rather edit text files with Rhythmbox or Emacs?
16:10:12 <elliott> Phantom_Hoover: Am I not allowed to playfully hate it, too?
16:12:07 <elliott> Vorpal: Oh, just remove all the hypotheticals and the "not"s and you'll get what I ABSOLUTELY DIDN'T want to say.
16:12:10 <Vorpal> elliott, I just got back, and parse timeout on that hypothetical sentence
16:13:02 <Vorpal> elliott, possibly not, do you have somewhere to not scp them to?
16:13:10 <elliott> Vorpal: Er, I could create such a place.
16:13:15 <elliott> ais523: ubuntu.com WTF: "Originally coined in 1998, the term open source came out of the free software movement, a collaborative force going strong since the dawn of computing in the 1950s."
16:13:43 <elliott> ais523: (1) rms must hate them for associating open source and free software like that (2) oh yeah, all those FOSS programs of the 1950s
16:13:54 <elliott> Vorpal: In fact I am running an ssh server already; isn't that convenient?
16:14:01 <elliott> Vorpal: Now to figure how to jail your user.
16:14:12 <elliott> Wait, that was on the previous machine.
16:17:15 <Phantom_Hoover> elliott, wasn't programming mainly an academic venture in the 50s?
16:17:42 <Phantom_Hoover> i.e. there wasn't a *point* in restricting the code, since about 3 other people in the world could possibly run it.
16:18:09 <elliott> Phantom_Hoover: Well, it's more that it never actually got released because yours was the only computer that could run it because nobody had standardised anything yet.
16:19:29 <Phantom_Hoover> Yeah, I was about to add "and those people were all in the same room as you."
16:21:33 <elliott> "Bent Linux: small simple distro, statically linked against uClibc"
16:21:56 <elliott> Phantom_Hoover: Well, yes. Fortran too.
16:22:04 <elliott> Nobody else *modified* these programs.
16:22:07 <elliott> Even if they were in papers.
16:22:10 <elliott> So I think it's a stretch.
16:22:22 <Phantom_Hoover> True, but if you *wanted* to, noöne was going to bother doing anything about it.
16:23:02 <Phantom_Hoover> And of course, there were about 4 programs worth paying for at all, so you weren't exactly losing any profit.
16:23:17 <elliott> Which programs, exactly? :-P
16:27:21 -!- Sgeo has joined.
16:27:47 <Sgeo> Well, my professor is apparently incapable of checking her Perl programs for typos before she puts them on an exam
16:28:08 <Gregor> You ... have Perl programs ... in a college course ...
16:28:38 <Gregor> Phantom_Hoover: Yeah, but elliott is elliott :P
16:28:49 * Gregor , on the other hand, is Gregor!
16:28:53 <elliott> Gregor: His university *beyond* sucks.
16:29:11 <elliott> Gregor: It's the biggest, shittiest state university I can possibly imagine.
16:29:35 <Gregor> I go to a big state university :P
16:29:36 <Sgeo> I'm wondering if, once I graduate from here, I could get a Master's in CS from a decent university
16:29:50 <elliott> Gregor: It's called "Farmingdale".
16:29:59 <elliott> [[The Farmingdale State College, also known as SUNY Farmingdale, and also called Farmingdale State College, is the former Long Island Agricultural and Technical Institute or LIATI.]]
16:30:05 <elliott> Gregor: Agricultural and Technical Institute!
16:30:27 <Gregor> elliott: My university has two Poultry Science buildings.
16:30:39 <Phantom_Hoover> We should totally write a language in a West Country accent.
16:30:56 <elliott> Gregor: "Undergraduates: Over 6,800"; they lost count.
16:34:07 <Phantom_Hoover> Feersum Endjinn and The Algebraist are in the same universe!
16:37:55 * Phantom_Hoover wonders if he can shoehorn "Against a Dark Background" in.
16:37:58 <fizzie> Our university has a lot to do with wood. (Finland is a foresty place, they do I guess a lot of paper technology.)
16:38:15 <elliott> Phantom_Hoover: What about all his non-sci-fi too?
16:38:53 <Phantom_Hoover> elliott, that can be stuffed into either the Culture or Hoovershoehorn universe.
16:42:40 -!- elliott has quit (Read error: Connection reset by peer).
16:42:56 <Vorpal> I think he just crashed his router
16:43:13 -!- elliott has joined.
16:46:52 <Sgeo> Was the problem that Gregor was complaining about was the word _program_ to describe Perl code?
16:46:58 <Sgeo> That's entirely my wording/fault
16:47:16 <Gregor> Sgeo: No, it was the presence of Perl in a university course.
16:51:18 <elliott> https://www.yourdomaingoeshere.com/openvz.php $9.88/mo for 512 MiB of RAM, now taking bets on how shit it will be :P
16:51:41 <fizzie> Gregor: Our natural language processing course had some Perl pre-processing scripts, too; what's wrong with that?
16:52:22 <elliott> [[Bent Linux is a compact Linux distro. It's inspired by LFS, but uses Busybox, uClibc, and static linking. It's particularly suited to building dedicated servers, initrds for custom installers and rescue disks, and systems with a nice crisp mid-1980s mouthfeel to satisfy the mid-life crises of crusty curmudgeons.]]
16:53:09 <fizzie> (Perhaps not in an exam, though.)
16:54:12 <ais523> <elliott> .cpio.bz2 based :-) <--- this is an abomination against the timeline of archival formats
16:54:23 <elliott> ais523: *a wonderful, wonderful abomination
16:54:49 <elliott> ais523: IIRC Gregor said some system (HP? IRIX?) used .bz2.cpio for packages.
16:55:02 <elliott> ais523: Which he held up as a good example of random-access archives, except that he said it's useless for packages.
16:55:34 <elliott> ("$ find foo -exec bzip2 '{} \; && tar cf foo.tar foo" or whatever. :P)
16:55:51 -!- Sgeo has quit (Ping timeout: 255 seconds).
16:55:57 <ais523> .bz2.tar or whatever is entirely different from .tar.bz2
16:56:23 <ais523> elliott: I love the way you can have nostalgia from a period before you were born
16:57:09 <elliott> ais523: It's the best kind!
16:57:29 <elliott> ais523: I swear I'm 40-odd on the inside, except whingy.
16:58:10 <ais523> `addquote <Gregor> elliott: My university has two Poultry Science buildings. <Gregor> Two!
16:58:20 -!- Sgeo has joined.
16:58:26 <HackEgo> 255|<Gregor> elliott: My university has two Poultry Science buildings. <Gregor> Two!
16:58:30 <Sgeo> I assume that's expensive?
16:58:39 <ais523> see, I'm hoping that if I do this enough, a random `quote will actually be likely to return something interesting or funny
16:58:52 <elliott> ais523: it does! If you're me.
16:58:53 <ais523> also, we've almost run out of quotes, assuming HackEgo thinks in 8-bit
16:58:59 <HackEgo> 36|<Deewiant> ehird: There is no h in "honour"
16:59:10 <HackEgo> 15|<Lil`Cube> wouldn't that be considered pedophilia? <Quas_NaArt> No. They all go by stage names.
16:59:16 <HackEgo> 191|<Sgeo_>ARGARGARHAHRHARHA REDDIT ON CHROME IS FULL OF HATE
16:59:48 <Sgeo> It's still full of hate. Despite me getting a new computer.
16:59:55 <HackEgo> 29|IN AN ALTERNATE UNIVERSE: <bsmntbombdood> there is plenty of room to get head twice at once
17:00:05 <HackEgo> 54|<lacota> I guess when you're immortal, mapping your fonts isn't necessary
17:00:15 <HackEgo> 105|<ehird> Gracenotes: No I said it does 54-bit
17:00:17 <HackEgo> 40|<ehird> That'd be the fahrenheit? I'm trying to have a mental breakdown here.
17:00:19 <HackEgo> 147|<ais523> (still, whatever possessed anyone to invent the N-Gage?)
17:00:37 <HackEgo> 199|<Sgeo> Why shouldn't I just do everything in non-Microsoft-specific C#? <ais523> it's like trying to write non-IE-specific JavaScript with only Microsoft documentation and only IE to test on
17:00:49 <HackEgo> 134|<Warrigal> A person's sex is not the same thing as their penis length.
17:00:49 <HackEgo> 63|<fizzie> The thing is just to exist
17:01:20 <HackEgo> 183|* Phantom_Hoover wonders where the size of the compiled Linux kernel comes from. <cpressey> To comply with the GFDL, there's a copy of Wikipedia in there.
17:01:20 <HackEgo> 93|<Oranjer> oohhh <Oranjer> ha <Oranjer> heh <madbrain> and what are your other characteristics? <Oranjer> oh, many, madbrain <Oranjer> but it's hardly worth it to go on with listing that list here
17:01:32 <HackEgo> 244|<zzo38> catseye: Please wake up. Not recorded for this timezone. The big spider is not your dream
17:01:52 <HackEgo> 175|<AnMaster> alise, marble <AnMaster> marbelus
17:01:53 <HackEgo> 12|<Madelon> Lil`Cube: you had cavity searches? <Lil`Cube> not yet <Lil`Cube> trying to thou, just so I can check it off on my list of things to expirence
17:01:54 <HackEgo> 171|<fungot> alise: nobody is allowed to fnord me in soviet russia
17:02:04 <Sgeo> WTF is the context for the zzo thing?
17:02:26 <ais523> Sgeo: there probably wasn't any even at the time
17:02:49 <HackEgo> 98|<fungot> ehird: every set can be well-ordered. corollary: every set s has the same diagram used from famous program talisman with fnord windows to cascade, someone i would never capitalize " i"
17:02:51 <HackEgo> 243|<fungot> ais523: my nose feels like a bad heuristic
17:02:52 <HackEgo> 78|<GregorR> ??? <GregorR> Are the cocks actually just implanted dildos? <GregorR> Or are there monster dildos and cocks? <GregorR> Or are both the dildos and cocks monster?
17:03:20 <Sgeo> I really need to bring my quotes pages back up
17:03:27 <HackEgo> 167|<alise> like, just like I'd mark "Bob knob hobs deathly poop violation EXCREMENT unto;" as English <ais523> alise: that's great filler <alise> ais523: well it contains all the important words in the english language...
17:03:27 <HackEgo> 166|<oklopol> you move on the tape and shit
17:03:28 <HackEgo> 108|<Warrigal> It's not incest if you're third cousins!
17:04:05 <Gregor> <elliott> ais523: IIRC Gregor said some system (HP? IRIX?) used .bz2.cpio for packages.
17:04:09 <Gregor> It was .gz.tar, and HP-UX
17:04:31 <elliott> Gregor: Your mom's .gz.tar.
17:05:08 <Sgeo> W.. what purpose could that serve? Compress a single file for ... tar
17:05:27 <Gregor> elliott: OH GOD CONVERSATIONAL INFINITE LOOP RUN
17:05:39 <ais523> Sgeo: it makes the file smaller, obviously
17:05:51 <ais523> .gz.tar.gz would be ridiculous
17:05:53 <Sgeo> But why is it tar'd?
17:06:04 <ais523> Sgeo: because there's more than one file?
17:06:14 <elliott> I wish I could see how stupid Sgeo's being right now.
17:06:17 <ais523> also, to preserve permissions, etc
17:06:23 <Gregor> Sgeo: .tar.gz gives you a hierarchy of files but no random access. .tar gives you random access but no compression. .gz.tar gives you compression and random access.
17:06:57 <Sgeo> But how is it more than one fil.. oh, all the files in the .tar are gziped, I guess
17:07:00 <ais523> idea: truly random-access memory: you ask it for some data, and it gives you a byte back at random and lets you change it
17:07:04 * Sgeo should not take things so literally
17:07:23 <ais523> hmm, actually, perhaps it's random whether it writes or reads
17:07:51 * ais523 wonders if you could make this into a probabilistic TC system somehow, in the limit; it would be a form of analog storage
17:08:32 <elliott> Gregor: .zip is basically .gz.tar with CRC32 to boot, right?
17:08:51 <Gregor> And no UNIX file modes and such.
17:08:55 <ais523> is it the same algo as .gz?
17:09:01 <Gregor> ais523: Close enough *shrugs*
17:09:44 <elliott> ais523: gzip is "GNU zip", i.e. "like zip but we changed everything so we can make it Free".
17:10:01 <ais523> I thought it was inspired by compress(1)
17:10:05 <elliott> GNU zip implements the gzip format.
17:10:17 <Sgeo> Can gzip actually handle multiple files?
17:10:25 <elliott> ais523: Perhaps, but I'm fairly sure the zip name is based on PK-ZIP...
17:10:27 <ais523> no, it's not an archival format
17:10:35 <elliott> <ais523> no, it's not an archival format
17:10:35 <ais523> elliott: potentially indirectly, but I agree
17:10:48 <ais523> elliott: gzip isn't, it doesn't store multiple files
17:10:52 <ais523> it's just a compression format
17:10:59 <Sgeo> For the record, that's what I thought
17:11:09 <Sgeo> I just had a panick attack with the idea that I might have been wrong
17:11:10 <elliott> ais523: oh, just checked logs; (sorry, I have Sgeo on ignore... and spent 10 seconds trying to reword this sentence so I didn't mention that)
17:11:24 <elliott> hmm, "a; (b)", what odd grammar of me
17:11:49 <ais523> apparently, some zip algos (including gzip, I think) store the permissions of the file, which is acting like an archival format a bit
17:12:14 <ais523> hmm, also, why does gzip defy the normal conventions of UNIX programs, and also the normal conventions of GNU programs?
17:12:36 <elliott> I HATE unX decompressors that destroy the original file.
17:12:36 <ais523> I'd expect to use gzip as "gzip archive.tar -o archive.tar.gz", and still have the original
17:12:37 <Gregor> gzip doesn't store the permissions, it just preserves them.
17:12:58 <elliott> ais523: *My* favourite archival format is {.ar}^∞.
17:13:04 <Gregor> If you gzip foo, foo.gz has the same permissions as foo had.
17:13:10 <ais523> elliott: you'll love the syntax for using azip; "azip archive.tar > archive.tar.az" (alternatively "azip < archive.tar > archive.tar.az")
17:13:12 <Gregor> Even if those permissions make no sense for the gzipped file.
17:13:21 <ais523> Gregor: I think it stores them inside the file too
17:13:27 <elliott> ais523: You have to ask me what {.ar}^∞ is now.
17:13:33 <ais523> elliott: nope, I guessed
17:13:35 <Gregor> I know it stores the original filename, but I'm pretty sure it doesn't store the permissions.
17:13:39 <ais523> recursively convert directories to .ar files
17:13:40 <Sgeo> Well, that's useful for .gz.tar
17:13:56 <elliott> ais523: Yep! Also, you store a directory list as the file... I'm not sure yet. Either "/" or "[NUL byte]".
17:13:58 <Gregor> By default, gzip keeps the original file name and timestamp in the compressed file. These are used when decompressing the file with the -N option. This is useful when
17:13:58 <Gregor> the compressed file name was truncated or when the time stamp was not preserved after a file transfer.
17:14:11 <ais523> it's not quite a 1-1 conversion, though, because there might be .ar files there already
17:14:14 <elliott> ais523: These aren't valid paths, so you process it before continuing. (Note: ar(1) fails on this, but eh, you can process it yourself; the format is simple enough.)
17:14:17 <elliott> ais523: <elliott> ais523: Yep! Also, you store a directory list as the file... I'm not sure yet. Either "/" or "[NUL byte]".
17:14:30 <elliott> That is, in every directory, the file "/" in the archive has a list of the directory names, separated by NUL.
17:14:38 <elliott> So you only decompress foo.ar if ar is in that file.
17:14:51 <ais523> you need to store permissions and file types too, don't you?
17:15:05 <ais523> as in, whether the file's regular/socket/symlink, etc
17:15:12 <elliott> ais523: Eh, who cares about either of those.
17:15:22 <elliott> It's for real men, who don't use such silly things.
17:15:29 <ais523> otherwise you couldn't distinguish between a text file saying "/etc/passwd", and a symlink to /etc/passwd
17:15:50 <elliott> ais523: Symlinks not supported!
17:16:12 * Sgeo suddenly wonders about the physical structure of symlinks
17:16:14 <Gregor> Symlinks are for the weak. Real men use HARDlinks. ... which also aren't usefully supported.
17:16:34 <ais523> Gregor: they aren't usefully supported by anything other than some of the insane modes of rsync, IIRC
17:16:38 <elliott> ais523: FINE. / can be more complicated.
17:16:52 <Gregor> ais523: rsync -H FTW HOORAY
17:16:54 <ais523> because finding the other end of a hardlink requires brute-forcing with most file systems
17:17:22 <Sgeo> What does it mean to support a hardlink?
17:17:40 <elliott> ais523: Firstly, a NUL-separated list of "filename[NUL]X", where X is "s" for socket, and "y" for symlink. Then, a NUL. Then, a NUL-separated list of directory names.
17:17:47 <Sgeo> Isn't a hardlink essentially just another name for the same file? So if you include it in an archive, you're just including the file
17:17:48 <ais523> Sgeo: it means that if you hardlink two files, archive, unarchive elsewhere, the two files will also be hardlinked in the new version
17:18:09 <ais523> elliott: well, directories aren't a regular file either, so you can merge the two parts of that
17:18:16 <ais523> as in filename[NUL]d for a directory
17:18:38 <elliott> ais523: I love how ar is plain text.
17:18:45 <ais523> I think the man page for ls has a list of the various other things you can put there
17:18:59 <elliott> ais523: hmm, yeah, what does ls show sockets and symlinks as?
17:19:04 <ais523> character and block special devices should probably be supported
17:19:07 <elliott> I should just copy those indicators.
17:21:23 <ais523> The file type indicators are `/' for directories, `@' for symbolic links, `|' for FIFOs, `=' for sockets, `>' for doors, and nothing for regular files.
17:21:44 <ais523> that's the indicators used by ls -F, which has a very old-fashioned look to it
17:21:59 <ais523> ls -l uses a completely different style of indicators, which I can't find documented in either man ls or info ls
17:22:12 <ais523> also, I don't know what a door is in that context
17:23:09 <Sgeo> Taking the fact that it's within a certain context out of context
17:24:58 <ais523> hmm, do I care enough to attempt to understand rot13d German?
17:26:00 <Sgeo> Fvr unora jnuefpurvayvpu avpug traht Fbetsnyg nhs qvrfr orefrgmra...
17:27:10 <ais523> wow, I guessed the last word was "übersetzen" just from the first letter and context
17:27:16 <ais523> translated the rot13 mentally, and I was right
17:27:48 <elliott> ais523: also, rot13'd German? How the heck does it handle umlauts?
17:27:57 <ais523> it leaves them unchanged
17:28:20 <elliott> ais523: so, umlaut on the replaced character?
17:28:48 <ais523> mostly because ä would transform to n-umlaut, which doesn't exist
17:30:30 -!- Sgeo has quit (Ping timeout: 255 seconds).
17:30:54 <ais523> wow, I submitted a Slashdot story 5 days ago, still hasn't been accepted or rejeted
17:31:43 <ais523> interestingly, it went all the way up to yellow, was removed from public view with a copy placed into the queue at the original time it was submitted, and the copy went up to green, so the few people who actually care about Firehose voting presumably liked it
17:38:20 <fizzie> With combining umlauts, x-umlaut exists for any x; that's no reason not to do the swappery.
17:38:49 <ais523> <LWATCDR> I let the users pick their own passwords and user names but I warned them with this text. "Your username and password are case sensitive." Then one day someone wanted me to change their user name and password for them. They set them them to... Case and Sensitive!
17:38:52 <fizzie> Finnish official alphabet has an odd number of characters, which makes self-decrypting ROT-n a no-go. :/
17:39:26 <ais523> you could just leave one alone
17:39:47 <fizzie> Anyway, the gzip file format has support for multi-part things, you could have a multi-file .gz. No-one does, though.
17:39:53 <fizzie> ("A gzip file consists of a series of "members" (compressed data sets). The format of each member is specified in the following section. The members simply appear one after another in the file, with no additional information before, between, or after them.")
17:40:15 <fizzie> (And each "member" has its own original-filename fields and so on.)
17:40:33 <ais523> hmm, you could use that as a sidechannel to steganograph data into in a normal .gz, couldn't you?
17:41:04 <ais523> (when I came back to Wikipedia after over a year, I noticed that one of the CSD changes was to allow instant deletion of images with data steganographically hidden in them)
17:41:20 <fizzie> Well, yes, I guess; but it's not a very clever side-channel, it's a bit painfully obvious. I'm sure some tools will complain about "junk", too.
17:42:05 <ais523> painfully obvious side-channels can be the best ones, sometimes
17:42:27 <ais523> I think someone made a polyglot between an image format and a compression format by exploiting the fact that one ignored junk at the end, and the other ignored junk at the start
17:43:53 <fizzie> 'gunzip' seems to just concatenate the uncompressed data together, so it's not a very hidingy with that tool either.
17:44:21 <fizzie> Oh, I guess the manual goes and even says that explicitly.
17:44:53 <elliott> fizzie: You may be disappointed to know that I have learned how to craft.
17:44:58 <fizzie> "Multiple compressed files can be concatenated. In this case, gunzip will extract all members at once. For example: gzip -c file1 > foo.gz; gzip -c file2 >> foo.gz .. Then gunzip -c foo is equivalent to cat file1 file2 .."
17:45:19 <ais523> fizzie: well, you could hide data more subtly by splitting a file into lots of pieces, and having the splitpoints encode the data
17:45:46 <fizzie> That you could do, though I think you could do the same within a single-member gzip file by manipulating the chunk boundaries.
17:45:59 <fizzie> Or block boundaries, I think the spec uses that word.
17:47:56 <fizzie> 7z does a bit sillily: http://p.zem.fi/7z-multimember (for context, test.gz was made by gzipping tmp1.txt and tmp2.txt, both of which had one line)
17:48:47 <fizzie> So that you don't need to follow a link: with the "list" option, it shows info (size/name) for the first member; when uncompressing, it uses the name of the first file but writes contents of both files there.
17:49:19 <ais523> so the file it creates is larger than the size it reports it will create?
17:49:47 <elliott> "hi", then a billion terabytes of "0"
17:50:40 <fizzie> "gzip -l" does the same: it reports the size of the latter file, but decompresses into a bigger file. Not that I think anyone does "gzip -l" ever.
17:51:58 <ais523> anyone here who uses Windows: does it use .rar as the default for compressed (zipped) folders nowadays?
17:52:10 <ais523> whenever I ask someone to compress files for me, they send me .rars for no obvious reason
17:52:13 <elliott> ais523: It doesn't even support .rar.
17:52:20 <elliott> ais523: .rar is just the ubiquitous Windows compression format.
17:52:23 <ais523> I wonder why they're so popular, then
17:52:30 <elliott> ais523: .zip is the Old Rubbish, .7z is the Obscure Stuff.
17:52:33 <ais523> I assumed .zip was the ubiquitous Windows compression format
17:52:48 <elliott> ais523: Torrents are always in .rars, for instance, when they're in some kind of archive.
17:53:02 <elliott> I think the reason is: (1) Piracy, (2) Software and stuff.
17:53:39 <Gregor> Often they're in .rars even when that's totally retarded.
17:53:59 <fizzie> Sometimes there's .rars within .rars within .rars.
17:54:28 <ais523> hmm, context for my earlier random quote off Slashdot: someone hacked the UK navy's website via SQL injection
17:54:41 <fizzie> Or multi-file split .rars that extricate into a big single-file .rar.
17:54:48 <ais523> <MalHavoc> I'm not sure what is worse. The fact that they fell victim to an SQL injection attack, or the HTML source that is displayed on TFA is badly broken. A "centre" tag? And the closing HTML tag is broken. Someone put up that maintenance page in a mega hurry.
17:54:54 <ais523> I love the concept of a <centre> tag
17:55:04 <ais523> it's wrong on so many levels, well at least two
17:55:30 <Gregor> <centre> centers things only in commonwealth countries.
17:55:50 <elliott> <fizzie> Or multi-file split .rars that extricate into a big single-file .rar.
17:56:13 <fizzie> Well, sometimes it has an .arj. (Okay, so not any more, but still.)
17:56:34 <fizzie> Back when there were floppies, there used to be floppy-splitted .arj archives around.
17:56:57 <fizzie> (Since .zip didn't split.)
17:57:08 -!- dbc has quit (Ping timeout: 240 seconds).
17:57:19 <fizzie> Or, as Wikipedia says: "Generally ARJ was less popular than PKZIP, but it did enjoy a niche market during the BBS era and in the warez scene. This was largely due to ARJ's creation and handling of multi-volume archives (archives which are split into smaller files which are then suitable for dial-up transfers and floppy distribution) being more robust than PKZIP's."
17:57:27 <fizzie> They sound oh-so-sophisticated.
17:57:43 <fizzie> I wonder why that bit doesn't have a [citation needed] tag.
17:57:56 <fizzie> Smells like original research to me!
17:58:22 <ais523> fizzie: you know, you can add one if you like?
17:58:33 <ais523> the tags don't add themselves
17:58:52 <fizzie> I don't really like them, I was just expecting one.
17:58:53 <ais523> (although Wikipedia would likely be greatly improved by a bot that went around adding [citation needed] to things at random)
18:12:42 <elliott> EsoCPU in Minecraft: http://www.minecraftforum.net/viewtopic.php?f=35&t=69299
18:12:48 <elliott> Not that esoteric, I guess, but hey, 8 bytes of RAM.
18:13:14 <ais523> hmm, Oracle are apparently planning to create a paid JVM in addition to the free one
18:13:55 <ais523> people are wondering what extra features it will have; my personal guess is that it'll just be bundled with a support contract and some vague enterprisey nonsense ("uptime optimised", etc) and otherwise be identical to the free one
18:14:17 <ais523> on the basis that companies tend to pick the most expensive option, and /especially/ companies that are used to Oracle products
18:14:22 <elliott> No, wait! "I'm doubling the memory of my CPU to 16 bytes"
18:14:40 <ais523> with sufficient ROM, 16 bytes is actually quite a lto
18:14:53 <ais523> I think I've written simple games in 16 bytes of RAM before
18:14:59 <elliott> ais523: I was about to say "it might get technical updates like GC overhauls faster than the free version", but then realised that Oracle are hardly going to improve the GC.
18:15:07 <elliott> (Especially as it's already the single best GC ever.)
18:15:27 <ais523> because it would impact their hardware sales?
18:15:41 <elliott> ais523: no, just because Oracle are incompetent
18:15:59 <ais523> hmm, what leads you to that conclusion?
18:16:28 <ais523> admittedly, PL/SQL is evidence in favour, but I thought they had a reputation for working very well, just requring tens of Oracle consultants all the time and being massively overpriced
18:16:37 <elliott> ais523: Specifically the DB, but also EVERY OTHER THING EVER.
18:16:46 <ais523> hmm, I suspect I'm mentally considering the consultants to be part of the program, IRP-style
18:16:55 <ais523> in which case it works better than any other DB as it's actually intelligent
18:16:59 <elliott> ais523: Well, the DB is supposed to be (a) not all that performant compared to Postgres and the like, and (b) the STUPIDEST PROGRAM EVER.
18:18:00 <ais523> I'm sure there are other candidates for the title of STUPIDEST PROGRAM EVER
18:23:22 <elliott> Gregor: Wrong! secho is BRILLIANT.
18:23:27 <ais523> no, PSOX is far from the most stupid program ever
18:23:31 <ais523> it's misguided, but not stupid
18:40:43 <Gregor> ARGH. How is it that people can speak English perfectly and yet are incapable of writing a cogent sentence X_X
18:41:14 <ais523> Phantom_Hoover: I'm not sure I'd describe either of those as "stupid" either
18:41:17 <elliott> <Gregor> ARGH. How is it that people can speak English perfectly and yet are incapable of writing a cogent sentence X_X
18:41:17 <elliott> <Gregor> ^^^ Case in point.
18:41:24 <elliott> Gregor: don't be so hard on yourself.
18:41:38 <ais523> elliott: why are the < and > surrounded by tabs?
18:41:44 <ais523> as in tab < tab Gregor tab > tab
18:41:49 <elliott> ais523: because xchat copies the blueness of Gregor as mirc colours because it's stupid
18:41:56 <elliott> there's also [000F] after the last tab
18:42:23 <Gregor> No, tab does not mean blue :P
18:42:44 <fizzie> I haven't had xchat copy any strange tabs in.
18:43:24 <elliott> It only happens sometimes.
18:43:27 <elliott> For god knows what reason.
18:43:59 <fizzie> <fizzie> Weirdiey. ← this was some pasta. It is a bit weird in that it adds the <>s in, they're not in the "timestamp nick|line" sort of format it shows in the screen.
18:59:09 <Gregor> "LOral USA Fellowships For Women"
18:59:32 <Gregor> When I think "respectable scientist", I then immediately think "L'Oral"
18:59:48 <elliott> Gregor: http://www.youtube.com/watch?v=0I2KnFx-kb4
18:59:58 <Gregor> elliott: Very at work.
19:00:18 <Gregor> elliott: Very your MOM.
19:00:31 <ais523> hey, no insult chains in two channels at once
19:00:57 -!- kar8nga has joined.
19:06:35 -!- augur has quit (Remote host closed the connection).
19:12:39 <elliott> Go match: John Trump (yes, the binary lambda calculus) vs. COMPUTER
19:13:14 <elliott> "In early 2006 MCTS (Monte Carlo Tree Search [7]) started being used to make strong 9x9 programs. The idea was crazy - have the computer play out random games, play out a lot of them, and choose the move leading to the position where randomness wins more. I would still say it is crazy, except for the inescapable fact that it works."
19:22:34 <fizzie> The description sounds a bit simplificationary: I don't think they usually play the simulated games completely randomly.
19:25:14 <fizzie> I'm not sure it sounds crazy when elaborated. "For each move, we sample out different alternative futures of how the game could go from there, and choose the move which leads to victory more often" vs. "DURR make random moves DURR".
19:25:47 <elliott> fizzie: SIMPLIFICATIONS ARE EVIL
19:27:07 <ais523> elliott: it's basically a normal alpha-beta search, except pruning at random
19:27:21 <ais523> which I suppose is one way to look to a large depth quickly
19:27:55 <fizzie> ais523: Welll... usually they do play one game to the very end, which means no guesswork with position-evaluation heuristics.
19:29:46 <ais523> yep, it's an intriguing idae
19:30:00 <ais523> unlike most computers, it seems like it's really good at strategy yet bad at tactics
19:30:06 <fizzie> I think there's been at least one monte-carloish contestant in that AI competition I harp about year after year. I really should tag up some meta about the participants so I could look for these things, now I just have piles of messy .pdf results.
19:30:26 <ais523> hmm, we need an equivalent of indent(1) for PDFs
19:32:21 <fizzie> "for p in *.pdf; do pdftotext $p -; done | grep -i monte" reveals just two reports (in 2010) who mention Monte Carlo approach in the discussion part as an alternative.
19:34:05 <oklopol> it's impossible to make infinite memory with minecraft physics?
19:34:26 <oklopol> that is, is there a block factory
19:34:47 <ais523> you wouldn't necessarily need a factory if you could do something analog with water levels, or something like that
19:35:12 <oklopol> well all you need is a stack or line of blocks you can grow infinitely
19:35:23 <ais523> I was trying to work out if Enigma was TC without Lua scripting; I concluded that it wasn't, but only because there was no way to keep a supply of coins handy to feed coinslots
19:35:49 <ais523> (you'd use the remaining duration of the coinslots as a bignum for a Minsky machine, thus getting round the limited storage space of the level)
19:35:59 <fizzie> Only people can create blocks. And the water/lava spread is... wonky, I doubt it's very workable. Some blocks are destroyed by fire/lava, though.
19:36:20 <oklopol> i was just watching this vid and water doesn't really seem to work
19:36:31 <oklopol> actually in any of the vids
19:36:31 <elliott> "> I hope this will Go well for the computer.
19:36:31 <elliott> I personally hope not. There ought to be a few things that humans do that computers can not do.
19:36:31 <elliott> I felt pretty bad when Kasparov lost to the IBM computer."
19:36:38 <fizzie> It's different in different versions, and really pretty wonky in the latest too.
19:36:39 <ais523> obviously it's TC if you allow arbitrary Lua scripting (notwithstanding the risk of the computer running out of memory)
19:36:59 <ais523> elliott: I still feel sorry for the game designers
19:37:13 <elliott> ais523: BUT DUDE ARIMAA IS LIKE SO UN COMPUTERABLEEEE
19:37:17 <ais523> I wonder if computers will ever beat humans at Arimaa (a chess-like game designed specifically to be hard for AIs to play)?
19:37:19 <elliott> That's why we PATENTED it.
19:37:27 <elliott> ais523: Protip: Arimaa is 99.999999999999999999% marketing.
19:37:32 <elliott> ais523: And stupid copyright policies.
19:37:43 <elliott> ais523: (You can't make a program to play Arimaa, I think. Or was it that you can't sell one?)
19:37:46 <elliott> fizzie knows the gory details.
19:37:58 <ais523> but if you can memorise something, like the rules of Arimaa, it should probably be uncopyrightable
19:38:01 <ais523> (note: opinion, not law)
19:38:58 <fizzie> It was very strange. I think the designer has had some "oh I want to let SCIENCE use this thing" notions when writing the license, but had no clue how to actually do that.
19:38:58 <elliott> ais523: Most things should be uncopyrightable. :)
19:39:16 <ais523> fizzie: screwed-up custom license?
19:39:21 <elliott> <ais523> but if you can memorise something, like the rules of Arimaa, it should probably be uncopyrightable
19:39:35 <elliott> ais523: I like the idea of making a board game's rules so complex that you can't memorise them, just so it can be copyrighted.
19:39:47 <ais523> hmm, I remember the row about the JSON license
19:39:51 <elliott> Like a thousand special cases for one scenario that almost never happens, and it's almost always one of three outcomes.
19:39:58 <ais523> it's MIT with an extra condition: "please use this software for good, not evil"
19:40:04 <fizzie> "Educational, research or personal use or distribution: For this category there is basically no restrictions on the use or distribution of Arimaa related products or services." And then in the list of examples what you can do:
19:40:10 <elliott> ais523: IBM got a license saying they could use the software for Evil, too. :-)
19:40:17 <ais523> elliott: really? that's hilarious
19:40:20 <elliott> Which is great (although I agree that the license is stupid and IBM did the right thing).
19:40:22 <Gregor> Wow, I just looked for "1835 (board game)" on Wikipedia and got "Blood-vomiting game"
19:40:25 <Gregor> Oh, those silly Japanese.
19:40:25 <fizzie> "Development of Arimaa related software for educational, research or personal use. However you may not hire, contract or commision someone else to develop the software; may not sell the software or release the software publicly in any form (source code or executable)."
19:40:34 <elliott> ais523: Well, technically they got it under the MIT license, I think.
19:40:36 <fizzie> I don't really see how that's "basically no restrictions".
19:40:48 <elliott> ais523: But it's more amusing saying that they got a license that permitted them to use it for Evil, too.
19:41:10 <Gregor> elliott: http://en.wikipedia.org/wiki/18XX_games <-- board game with rules so complicated you can't memoriZe them.
19:41:17 <fizzie> But there really aren't any real conditions: it's just "basically no restrictions" and a very non-exhaustive list of things you might do, and then that inscrutable "wahh if you make software you can't give it to anyone" thing.
19:41:28 <elliott> Gregor: What @ that naming scheme. Bizarre.
19:41:34 <fizzie> Actually, it sort of makes me think if it's been added there to make it slower for computer programs to get better at the game.
19:41:49 <Gregor> elliott: The original games were all named after some year in the 19th century :P
19:42:05 <elliott> fizzie: "Arimaa has so far proven to be more difficult for artificial intelligences to play than chess." -- Wikipedia; I think what it means to say is "nobody competent enough has bothered to".
19:42:13 <fizzie> According to Wikipedia, for Arimaa's latest yearly challenge: "In 2010, Mattias Hultgren's bot Marwin edged out Clueless in the computer championship. In the Challenge match Marwin became the first bot to win two out of three games against a single human defender, and also the first bot to win three of the nine games overall."
19:42:20 <elliott> The Arimaa pieces are hard to distinguish for me.
19:42:26 <fizzie> Of course the human players aren't probably exactly as serious about it as Chess/Go players.
19:42:44 <elliott> http://en.wikipedia.org/wiki/Arimaa#Computer_performance tl;dr Brute forcing is hard! ZOMG!
19:42:58 <ais523> ideally, it should just use numbers
19:43:12 <ais523> trying to remember elephant > camel > horse > dog > cat > rabbit is the hardest part of the game
19:43:41 <Gregor> ais523: Really? "Bigger-than" over mammals is too hard for you?
19:43:46 <oklopol> "Top chess programs use brute-force searching"
19:43:52 <fizzie> The computer pessimizations have been made pretty much with existing chess-playing techniques in mind.
19:44:11 <ais523> Gregor: cats are larger than rabbits?
19:44:11 <elliott> Gregor: It's irritating...
19:44:15 <fizzie> oklopol: For smart enough values of brute.
19:44:20 <ais523> admittedly, I don't see rabbits very often
19:44:31 <elliott> ais523: I say this as someone who owns both a small and a big rabbit :P
19:44:41 <elliott> (And has, in fact, seen and owned cats in my time.)
19:44:51 <elliott> ais523: Wild rabbits, still smaller.
19:44:55 <ais523> elliott: you strike me as sorr of the person who /would/ own multiple sizes of rabbit just to be able to compare them
19:45:02 <Gregor> ais523: I think I can say with certainty that though cats and rabbits share a size range, cats are on average larger.
19:45:05 <elliott> ais523: That is now officially my reason.
19:45:11 <Gregor> ais523: Same is true of cats and dogs, so that's not much of an argument :P
19:45:53 <ais523> *as the sort of person
19:46:10 <fizzie> "Domestic cats are similar in size to the other members of the genus Felis, typically weighing between 4 kilograms (8 lb 13 oz) and 5 kilograms (11 lb 0 oz). -- The smallest adult cat ever officially recorded weighed around 1.36 kilograms (3 lb)." -- "[Rabbits'] size can range anywhere from 20 cm (8 in) in length and 0.4 kg in weight to 50 cm (20 in) and more than 2 kg."
19:46:14 <fizzie> Not very much of overlap there.
19:46:17 <elliott> Gregor: Yes, but (|rabbits > cats|/(|rabbits|*|cats)) > (|cats > dogs|/(|cats|*|dogs|)).
19:46:26 <elliott> Gregor: Er, without the *|cats|.
19:46:58 <elliott> Gregor: x = percentage of rabbits that are bigger than the average-sized cat. y = percentage of cats that are bigger than the average-sized dog.
19:47:05 <ais523> elliott: is that an attempt at a famous equality?
19:47:16 <elliott> ais523: Nope. What does it look like to you?
19:47:17 <ais523> my youth is scarred from having to memorise the things
19:47:23 <ais523> and it looks vaguely like one of them
19:47:45 <ais523> famous inequalities all look much the same
19:50:41 <fizzie> http://p.zem.fi/cats-vs-rabbits
19:50:57 <fizzie> Whoops, I got the < upside down!
19:51:15 <ais523> isn't an upside-down < a <?
19:51:25 <fizzie> Yes, I meant... uh, leftside-right.
19:52:23 <fizzie> No, I want it in the fooside-bar form.
19:53:10 <ais523> "fooside bar" would be a great name for a pub
19:53:33 <elliott> fizzie: That isn't what I said at all!.
19:53:50 <fizzie> elliott: I wasn't trying to be.
19:53:56 <fizzie> "Showing results for "offside bar". Search instead for "fooside bar"." Aw.
19:54:29 <elliott> I said |{x : x in rabbits, size(x) > (sum x in cats : size(x)) / |cats|}| > |{x : x in cats, size(x) > (sum x in dogs : size(x)) / |dogs|}|.
19:54:32 <ais523> off/foo is an interesting typo suggestion
19:55:23 <ais523> also, I love IRC for its ability to explode an offhand comment into a full scientific investigation
19:56:20 <fizzie> elliott: I didn't even bother reading it; I wasn't yet in the whole rabbits-vs.-cats vs. cats-vs.-dogs mess.
19:57:02 <elliott> Wolfram Alpha does not have the average cat size in it.
19:57:08 <elliott> Assuming "size" is referring to species data extras | Use as referring to administrative divisions instead
19:57:29 <elliott> Assuming length | Use height or [more | v] instead. More = maximum recorded trunk diameter, or weight.
19:57:37 <elliott> Maximum recorded trunk diameter of a cat: 0!
19:57:42 <elliott> (Note: Last line my own speculation.)
19:57:59 <ais523> what does it say the maximum size cat is, then, if the average size cat isn't listed?
19:58:16 <fizzie> elliott: Note that it does have photoreceptors/square millimeter stats for cats.
19:58:31 <fizzie> And olfactory epithelium surface area.
19:58:56 <elliott> ais523: (data not available)
19:59:10 <ais523> OK, so we know how dense the photoreceptors are on cats, but not how large they are?
19:59:20 <ais523> alpha reminds me of mathematica, awesome but only for things they thought of
19:59:36 <elliott> ais523: It's actually useful for basic algebra/calculus data-munging.
19:59:42 <elliott> In that you don't have to start up Mathematica.
20:00:08 <fizzie> Also if I write ungrammatically "volume of cat", it goes "Input interpretation: Caterpillar | volume" and the result is "2.981 million shares".
20:00:12 <ais523> I'm trying to remember which term in the ToS I disagreed with
20:00:17 <ais523> perhaps the one that claims it's a contract
20:00:19 <elliott> ais523: age limit? no bots?
20:00:42 <ais523> oh, the age limit is ridiculous, also self-defeating
20:01:09 <ais523> because if you think about it, they're claiming in the contract a) the ToS are a contract, b) minors aren't allowed to use the site because they can't agree to contracts
20:01:12 <elliott> I'm being corrupted by the evil pornographic Wolfram Alpha data.
20:01:26 <ais523> b) is obviously not enforced on minors, due to being in a contract they can't legally agree to
20:01:39 <elliott> are you sure that's the argument for (b)?
20:01:42 <elliott> rather than just "NO UNDER 18S K"
20:01:50 <ais523> elliott: not sure, it's a memory
20:01:59 <ais523> and the sort of thing I might plausibly misremember
20:02:14 <ais523> you can check if you like, IIRC (b) was spelled out in the ToS itself
20:02:33 <elliott> [on The Art of Unix Programming] "Yeah, there's a reason nobody likes that crap. A book on unix programming by a man who knows neither unix nor programming leaves a lot to be desired."
20:02:57 <ais523> elliott: I read the UNIX-HATERS HANDBOOK btw (as it's UNIX, I assume it's case-sensitive)
20:03:02 <ais523> I wasn't particularly impressed with it
20:03:28 <ais523> partly, beause it's dated in that many of the things that it claimed were outdated actually are outdated, as in not even used in UNIX any more
20:06:00 <elliott> ais523: many of the complaints are still true, though
20:06:03 <elliott> as well as the general complaints
20:08:28 <elliott> ais523: the email in the preface is the best, though, I think
20:08:35 <elliott> "This is the fifth day I’ve used a Sun. Coincidentally, it’s also the fifth
20:08:35 <elliott> time my Emacs has given up the ghost. So I think I’m getting a feel
20:08:35 <elliott> for what’s good about Suns."
20:08:54 <elliott> ("Another nice thing about Suns is their simplicity. You know how a
20:08:54 <elliott> LispM is always jumping into that awful, hairy debugger with the
20:08:54 <elliott> confusing backtrace display, and expecting you to tell it how to pro-
20:08:54 <elliott> ceed? Well, Suns ALWAYS know how to proceed. They dump a
20:08:54 <elliott> core file and kill the offending process. What could be easier? If
20:08:54 <elliott> there’s a window involved, it closes right up. (Did I feel a draft?)
20:08:57 <elliott> This simplicity greatly decreases debugging time because you imme-
20:08:59 <elliott> diately give up all hope of finding the problem, and just restart from
20:09:01 <elliott> the beginning whatever complex task you were up to. In fact, at this
20:09:03 <elliott> point, you can just boot. Go ahead, it’s fast!")
20:09:05 <elliott> ^ possibly my favourite paragraph
20:09:56 <ais523> elliott: with the typical quality of today's programs, effectively running them all under a debugger all the time would make your life unusable with really-continue commands
20:10:14 <elliott> ais523: well, it worked on the lisp machine because it had hardware support
20:10:17 <elliott> and also people weren't idiots
20:10:26 <elliott> why is weren't not a word according to my spell checker?
20:10:39 <elliott> ais523: anyway, a lot of the reasons today's programs suck ARE Unix and Windows
20:10:42 <elliott> so that's a bit of a circular counter-argument
20:11:57 <elliott> ais523: I really need to just buy a Lisp Machine some day. Shipping from Virginia can only be hilariously cheap, right? Sigh.
20:16:15 <elliott> -r-- 1 1005728 Jan 7 2007 bzip2-1.0.4.cpio.bz2
20:16:22 <elliott> ais523: is it physically possible to not smile at filenames like that?
20:16:51 <ais523> I assume there are people who wouldn't
20:16:54 <olsner> <elliott> 12:36:40 <olsner> probably my own API, posix is overrated | <elliott> kinder words than I'd choose
20:17:03 <ais523> also, "-r--", is this a single-user system?
20:17:24 <elliott> ais523: presumably, it lists whatever permissions apply to the user it's looking at them as
20:17:27 <elliott> nobody or www-data or whatever
20:18:44 <olsner> you may hate posix, but I just think it's an API and as all API:s it sucks more or less :)
20:19:24 <elliott> olsner: it's more a system specification
20:21:14 <olsner> yeah, definitely more than an API ... s/API/whatever it is/g or something
20:22:20 <elliott> Anyone want to donate $675 or a fraction thereof for a shell account on a Lisp Machine? :-)
20:22:38 <elliott> Bet I can afford international postage of an object weighing about 400 pounds if I just get the actual cost out of the way!
20:23:52 <ais523> how heavy can a shell account be?
20:24:13 <elliott> ais523: I'm going to assume that misinterpretation was deliberate. :P
20:24:36 <elliott> (The object is the Lisp Machine; DKS says the 36xx machines weigh "up to 400 pounds", but then factor in the monitor, keyboard, etc.)
20:24:54 <elliott> Of course the newer machines are likely to be cheaper -- and quieter! -- but also more expensive. Perhaps the postage would balance it out.
20:25:15 <ais523> elliott: oh, I've only just worked out your intended interpretation
20:25:23 <elliott> Well, actually, all the >36xx machines are either MacIvories or the WTFEXPENSIVE $3,500 XL1200/XL1201 and apparently for that one "shipping can be expensive".
20:25:36 <elliott> 400 pounds is the maximum weight.
20:25:39 <elliott> The smallest of these machines is the 3620, which is
20:25:39 <elliott> a deskside machine (9 x 18 x 25) that runs off standard 120V power and
20:25:39 <elliott> weighs about 70 pounds, not including the monitor
20:25:56 <elliott> cofiguration is 4 MWords with a 760 MB of ESDI disk and a 17" monochrome
20:25:56 <elliott> console with keyboard and 3-button mouse. You can add another 760 MB disk
20:26:07 <elliott> [[You can add additional memory for $50 per MWord up
20:26:07 <elliott> to a total of 8 MWords. You can upgrade to the 19" premium monochrome
20:26:08 <elliott> monitor for an additional $300.]]
20:26:29 <ais523> why would a 760 MB disk cost $150 nowadays?
20:26:40 <elliott> ais523: ESDI, SMD or ST506 disks.
20:26:45 <ais523> the hard disk doesn't depend on the CPU architecture, surely?
20:26:48 <elliott> ais523: also, I think they're original Symbolics disks.
20:26:55 <elliott> ais523: but: <elliott> ais523: ESDI, SMD or ST506 disks.
20:27:15 <ais523> hmm, is this a company who's come across a load of unused original Lisp machines, and is hanging onto them in the hope that someone will buy them?
20:27:21 <elliott> ais523: this *is* the Lisp Machine company
20:27:32 <elliott> and sells its copious stock of Lisp Machines
20:27:34 <ais523> ah, are they still making them, or just selling their extra stock?
20:27:43 <elliott> ais523: http://symbolics-dks.com/
20:27:55 <elliott> ais523: (he sold symbolics.com to a squatter -- SHAME on him! (oldest .com))
20:28:18 <elliott> ais523: well, more than one guy -- [[Send bug reports to Kalman Reti at reti@symbolics-dks.com]]
20:28:24 <elliott> ais523: so they still have a Macsyma developer, it seems
20:28:40 -!- wareya has quit (Read error: Connection reset by peer).
20:28:44 <ais523> elliott: you'd sort-of expect verisign.com to be the oldest, although maybe they got hold of .com after it was created rather than owning it forever
20:29:00 <elliott> ais523: verisign didn't own .com from the start
20:29:11 <elliott> ais523: here is the list and spec of lisp machines still sold, plus prices, FWIW: http://www.lispmachine.net/symbolics.txt
20:29:40 -!- wareya has joined.
20:29:42 <elliott> [[This would only be a viable
20:29:42 <elliott> option, if you live close enough to Washington, DC to pickup the machine.]]
20:30:08 <elliott> ok, conclusion: only viable option is 3620, which can pack 1.5 gigs of disk, but "barely makes for a useable environment once you load all the software and documentation files"
20:30:28 <ais523> you could always just move the documentation to a different system
20:30:41 <ais523> unless it's needed for the debugger to work
20:30:50 <elliott> ais523: no, you want it (it's hyperlinked and integrated)
20:30:56 <elliott> i.e. click a function in Emacs (yes, it had emacs), see the documentation
20:31:06 <elliott> ais523: it's a large part of the appeal of these systems -- the integration
20:31:10 <ais523> how could it not have Emacs?
20:31:17 <elliott> ais523: no, but it's not GNU Emacs
20:31:22 <elliott> ais523: it's *real* emacs :-)
20:31:22 <ais523> (I'm assuming it wouldn't be GNU Emacs, but a different dialect)
20:31:43 <elliott> Okay, decked-out 3620 with 8 MWords of memory and 1.5 gigs of storage, that's... $1025.
20:32:05 <elliott> Not bad, especially if you ditch some of the RAM.
20:32:33 <elliott> hmm, 8 MB = 1.3 MWord, apparently
20:34:00 <elliott> ais523: the real problem would be shipping 70 pounds + keyboard + monitor + CRT monitor internatioanlly!
20:34:07 <elliott> ais523: that's got to be hundreds of pounds
20:34:56 <elliott> I love how the LoseThos developer linked the last line of his Loper OS blog comment, "I’m funded from social security disability for being insane.", to the LoseThos site.
20:35:15 <elliott> Phantom_Hoover: Are they willing to pay ridiculous shipping prices? Besides, I think the actual machines are in Washington DC.
20:35:40 <Phantom_Hoover> elliott, well, I'm assuming you'll give me tonnes of money to pay them off.
20:35:50 <ais523> elliott: just ask for it to be delivered to the US Embassy and pick it up from there; that way, it technically isn't international
20:35:56 <elliott> Phantom_Hoover: Or I could just give DKS tons of money for the same effect.
20:35:59 <elliott> ais523: please tell me that works
20:36:07 <ais523> I doubt it, but it's a hilarious concept
20:37:19 <elliott> ha, Stanislav just bought one of the Tru64 UNIX-based machines and used Open Genera on it
20:37:37 <elliott> [[My place of business recently purchased a copy of Open Genera from David Schmidt of Symbolics Inc. (http://www.symbolics-dks.com/) and he included this machine as a no-cost extra.]]
20:38:25 <elliott> http://www.loper-os.org/wp-content/bolix1.png ;; whereas this looks like an actual lisp machine to me, although i'm not sure...
20:39:16 <Phantom_Hoover> I assume Stanislav is still bowing before his Lisp machine overlords before going to sleep every night?
20:45:32 -!- kar8nga has quit (Read error: Connection reset by peer).
20:46:09 <elliott> Ergo, God is in his heaven.
20:48:25 -!- ais523 has quit (Quit: Page closed).
20:48:33 <elliott> " as the laws progress they quickly devolve into a delusional hacker wonderland ... A debugger opening on a program error? ... really? I can't think of anything much more use-hostile."
20:52:14 <elliott> olsner: it seems unfortunately likely at this point that Kitten will be linux-based
20:52:39 <olsner> don't be like the others :(
20:52:47 <elliott> olsner: I don't want to be! But have you *seen* NetBSD?
20:52:56 <elliott> olsner: Especially its hardware support. I do not think it will speak to my network card.
20:53:20 <olsner> they use *CVS*, I'm not sure it will even speak with this century
20:54:07 <Gregor> elliott: Can you at least make it a true No-GNU Linux?
20:54:28 <elliott> Gregor: That is the idea -- well, a consequence of the idea -- although somehow I suspect you sincerely want that.
20:54:45 <elliott> Gregor: Although I will likely ship GNUmacs, at least, and -- I forget the prominent other piece of GNU-only software.
20:55:13 <Gregor> All the other Linux libcs are based on glibc.
20:55:29 <elliott> Gregor: Is not glibc-based :P
20:55:33 <elliott> Gregor: (And works just fine for desktop systems)
20:55:47 <elliott> Development on uClibc started around 1999.[3] uClibc was mostly written from scratch,[4] but has incorporated code from glibc and other projects.[5]
20:55:54 <elliott> Gregor: A few bits are from glibc, but no, it's mostly own code.
20:56:05 <olsner> oh, kitten would be BSD/Linux?
20:56:08 <elliott> Gregor: Also, I think that newlib technically works for Linux.
20:56:12 <elliott> olsner: Well. Kitten/Linux.
20:56:19 <Gregor> elliott: The Linux bits are lifted from glibc.
20:56:33 <Gregor> I think you can only call it a No-GNU Linux if there are no [L]GPL'd pieces other than the kernel itself :P
20:56:37 <elliott> olsner: But mostly cobblings of software, plus I'm not sure which set of core utilities.
20:56:59 <elliott> Gregor: Anyway, WRONG: dietlibc.
20:57:02 <Gregor> elliott: I was kidding :P
20:57:09 <Gregor> elliott: dietlibc! POYFECT
20:57:12 <elliott> Gregor: (dietlibc is GPL and thus utterly unsuitable for my purpose, as it means I could not distribute binaries, because the author is a fuckwit.)
20:57:25 <Gregor> GPL is unsuitable is unsuitable for any purpose due to that license, yes :P
20:57:38 <Gregor> No, I stand by my original statement :P
21:00:00 <elliott> Gregor: Best sentence ever :P
21:00:10 <elliott> Gregor: It's a real shame because dietlibc is *perfect* code.
21:00:20 <elliott> Gregor: But... source-based distro no.
21:02:01 <elliott> Gregor: (Although it also warns about stdio, which is fucking annoying.)
21:02:14 <elliott> Gregor: dietlibc, upon linking.
21:02:17 <elliott> stdio/printf/etc. = warning.
21:02:32 <Gregor> That's pretty lameawesome :P
21:02:42 <elliott> Gregor: Well, no true djbite would use them :-)
21:02:51 <elliott> [[Q: GPL sucks! Now I can't compile my BSD programs with the diet libc!
21:02:51 <elliott> A: Wrong. You can compile them, and you can use them. You just can't
21:02:51 <elliott> redistribute the binaries. If you are a distribution vendor and want
21:02:51 <elliott> to use the diet libc to make BSD licensed binaries for the install
21:02:51 <elliott> or rescue floppy which you sell commercially, please talk to me.]]
21:02:57 <elliott> What about the install I want to not sell freely...
21:03:39 <elliott> Gregor: If I use .xz.tar for Kitten packages, do I have to pay you royalties?
21:03:44 <Gregor> elliott: You just have to provide sources secondarily to the install media, which puts the onus of continuing that redistribution on the people who download it :P
21:03:53 <Gregor> elliott: I stole it from HP-UX, so, ... yes.
21:03:57 <elliott> Gregor: ...??? <Gregor> elliott: You just have to provide sources secondarily to the install media, which puts the onus of continuing that redistribution on the people who download it :P
21:04:15 <Gregor> elliott: The install media can just be downloadable alongside the sources.
21:04:56 <elliott> Gregor: No, you can't redistribute a BSD-licensed program linked with dietlibc.
21:05:33 <Gregor> In spite of what that Q&A says, you can, so long as you redistribute the sources too.
21:05:54 <elliott> Gregor: ...I don't see how that's true at all.
21:05:59 <Phantom_Hoover> elliott, incidentally, what awesome things is Kitten going to have?
21:06:48 <elliott> Phantom_Hoover: Simplicity. Speed. No-bullshit administration. Snarky curmudgeonness. Like being in a goddamn time machine and setting it to SANE AD.
21:06:56 <elliott> Phantom_Hoover: Also: Lots and lots of breakage when I break stuff!
21:07:17 <Gregor> elliott: Dude, the GPL is not MS Shared Source. When you want to redistribute something linked to something under the GPL, that just means you have to redistribute the whole thing under the terms of the GPL; that just means including sources.
21:07:32 <elliott> Gregor: Isn't this the CLISP bullshit?
21:07:37 <elliott> Gregor: If I believe rms, I can't do that.
21:07:55 <Gregor> elliott: If you believe the GPL FAQ on gnu.org RIGHT FRIGGIN' NOW, you can.
21:08:20 <elliott> Gregor: Cool -- I'll go email fefe and tell him, so that he changes the license.
21:08:40 <Gregor> elliott: BSD is GPL-//compatible//
21:09:08 <olsner> I still don't get how you can redistribute BSD stuff under the GPL (which you must if you link it with GPL), without violating the BSD license
21:09:44 <Gregor> olsner: Because the BSD license allows you to add further restrictions.
21:10:04 <elliott> Gregor: Permission to quote you in lieu of restating what you said to fefe?
21:12:26 <olsner> but it would still mean that the result of combining BSD and GPL code is effectively GPL:d - this may not be acceptable for whatever elliott wants to build using dietlibc
21:13:53 <Gregor> Yes, that's the (only) issue.
21:16:23 <elliott> olsner: I don't see how that's an issue; distributing GPL'd binaries is pretty much irrelevant, is it not?
21:16:39 <elliott> As long as I have the source to the program -- as BSD -- and the source to dietlibc -- as GPL -- available too.
21:16:43 <elliott> <elliott> Gregor: Permission to quote you in lieu of restating what you said to fefe?
21:17:03 <elliott> Phantom_Hoover: [[So why another fortune program? Because the BSD one sucks. It needs a separate file and a program called "strfile" to create that said file.]]
21:17:10 <olsner> well, it wouldn't be irrelevant for anyone wanting to fork kitten
21:17:39 <elliott> Gah, tar(1) is so limited.
21:17:45 <elliott> It can't run a program for each file it extracts.
21:18:18 <olsner> because they would be forced to release the source
21:18:39 <elliott> Those are already source-available.
21:19:01 <elliott> If you mean the administration utilities... they'd only have to release the source if they distributed them linked to dietlibc.
21:19:06 <elliott> So they could just use uClibc or something.
21:19:14 <elliott> Or, you know, I could just use uClibc :P
21:19:20 <olsner> they would need to release the source of anything linked to the GPL:d dietlibc, including the "BSD-licensed" parts
21:19:27 <elliott> pipe extracted files to another program
21:22:53 <fizzie> That made me wonder how it passes the file name and other stuff like that, but it seems it's actually the somewhat sensible thing: a pile of TAR_FOO envvars.
21:23:22 <elliott> fizzie: (Also, does it not save the output of the command to the filename?)
21:23:41 <elliott> --transform, --xform EXPRESSION
21:23:41 <elliott> use sed replace EXPRESSION to transform file names
21:25:01 <elliott> Gregor: Are the .gz.tars actually called that on HP-UX?
21:25:30 <elliott> I'm considering .xzt and .xtr at this point. :P
21:26:24 <fizzie> TAR_{FILENAME,REALNAME,SIZE,GID,GNAME,UID,UNAME,MODE,ATIME,CTIME,MTIME} are the ones that are probably most useful. I don't know what's the difference between REALNAME and FILENAME.
21:26:42 <elliott> Where are you getting this? Hand-written program?
21:26:49 <fizzie> I don't know of a real list, I just "tar xf foo.tar --to-command env | grep TAR_" 'd a single-file foo.tar.
21:28:10 <fizzie> Name of the file as stored in the archive.
21:28:14 <elliott> fizzie: Hmph, it seems to not bother creating a directory structure or anything if you do that.
21:28:15 <fizzie> Who knows what the difference is.
21:28:28 <fizzie> (That last one was from http://www.gnu.org/software/automake/manual/tar/Writing-to-an-External-Program.html )
21:29:19 <fizzie> You will probably (perhaps) get the directories first with TAR_FILETYPE=d, you could mkdir at those times.
21:29:35 <elliott> fizzie: "Non-regular files (directories, etc.) are ignored when this option is used."
21:29:45 <elliott> fizzie: So in fact I can't even extract empty directories like this.
21:29:53 <elliott> fizzie: I think I will just use find(1) after the fact to unxz.
21:31:15 <fizzie> Strange that there's TAR_FILETYPE at all if it just does regular files. I guess it could be for future-proofing.
21:34:06 -!- Sgeo has joined.
21:34:35 <Vorpal> fizzie, strange minecraft sighting: http://sporksirc.net/~anmaster/minecraft/screenshots/2010-11-08_22.31.01.png (yes this was natural...)
21:35:10 <Vorpal> fizzie, see the other screenshots in that dir too for some weird bugs and some nice scenery
21:36:05 <fizzie> Trees sometimes do some weird stuff.
21:36:33 <Vorpal> fizzie, I guess some leaves touched the lava and then that happened
21:37:32 <fizzie> In multiplayer it quite often happens that it gets into some sort of lag w.r.t. retrieving blocks, and then you can see lava-caves. I'm not explaining this very well, but unfortunately I don't have a screenshot.
21:42:33 <Vorpal> ooh not floating islands but the largest and most improbable overhang ever
21:42:47 <Vorpal> must be several chunks alrge
21:43:29 <Vorpal> fizzie, do you know what the FOV is of minecraft? this might be an interesting application of hugin
21:43:46 <Vorpal> hm I think you rotate with no parallax right?
21:44:21 <Vorpal> damn there is parallax
21:44:56 <Vorpal> yes just checked very close and very far objects
21:45:10 <Vorpal> the camera is slightly in front of the rotation point I think
21:46:34 <fizzie> http://zem.fi/~fis/minecraft-hugin.jpg -- made this earlier.
21:47:22 <fizzie> The HFOV on my system is something very close to 53.8 degrees, I'd guess: v53.7996384046093, says the optimizer.
21:47:47 <fizzie> It's not quite correctly aligned, there were control points in the clouds and whatnot.
21:48:44 <fizzie> I've seen minecraft "movie"-style motion videos without the crosshairs and with programmed-like camera motions; you could probably mess around with the client to get better images.
21:48:51 <Vorpal> fizzie, you already did? hm not many close objects there.
21:49:06 <fizzie> That one was just from 16 random cropped screencaps, to see how well it goes.
21:49:14 -!- sebbu2 has quit (Read error: Connection reset by peer).
21:49:41 -!- sebbu has joined.
21:50:26 <fizzie> Stopping the clouds at least would help. Even if you avoid putting control points there, they do move.
21:50:58 <fizzie> (I was wondering if I should upgrade from my current OpenTTD-themed N900 panorama-background to a minecraft one.)
21:51:29 <Sgeo> http://newyork.cbslocal.com/2010/11/08/l-i-woman-dies-after-marrow-donors-refuse-to-show/
21:52:01 <Vorpal> wow there are caves in these almost-floating-islands
21:52:29 <Vorpal> fizzie, elliott: new screenshots in http://sporksirc.net/~anmaster/minecraft/screenshots/
21:56:33 <fizzie> The terrain generator does make some interesting-looking scenery.
21:57:13 <Vorpal> fizzie, so it does. Also the straight edges are probably due to borders between old and new generator
22:00:10 -!- Sasha has quit (Read error: Connection reset by peer).
22:01:42 -!- Sasha has joined.
22:02:51 -!- Sgeo has quit (Read error: Connection reset by peer).
22:03:12 <elliott> <Vorpal> wow there are caves in these almost-floating-islands
22:03:24 <elliott> imagine what a mindfuck it'd be if you found out the ground you started on is actually some huge overhand
22:03:35 <elliott> of some immensely larger island on top of the actual world
22:04:28 <fizzie> Vorpal: Just for the record, current boring background: http://zem.fi/~fis/openttd-bg.png
22:06:04 <Vorpal> one waterfall lands on the top of a tree
22:06:35 <fizzie> I'd like to get snowfall back in snowy biomes, though.
22:06:42 <fizzie> Currently there's snow, but if you scoop it up, it's gone.
22:06:46 -!- Sasha has quit (Read error: Connection reset by peer).
22:08:05 -!- Sasha has joined.
22:11:45 <fizzie> Vorpal: Even when it's not doing floating islands, it doesn't always go for realism: http://zem.fi/~fis/spire.png
22:15:50 -!- Phantom_Hoover has quit (Remote host closed the connection).
22:16:12 -!- Sasha has quit (Quit: NO U).
22:16:32 <elliott> http://sporksirc.net/~anmaster/minecraft/screenshots/2010-11-08_22.14.22.png
22:16:52 -!- Sasha has joined.
22:16:53 <elliott> http://sporksirc.net/~anmaster/minecraft/screenshots/2010-11-08_22.19.57.png this also
22:17:06 <elliott> Vorpal: I beg you to play in 1366x768 so I can wallpaper :P
22:17:16 <elliott> (thought: screenshots need to be Minecraft-voxel based.)
22:17:31 <Vorpal> time to upload some more
22:17:36 <elliott> Vorpal: you're mistaken; in http://sporksirc.net/~anmaster/minecraft/screenshots/2010-11-08_22.31.01.png, those are giant torches!
22:18:02 <elliott> http://sporksirc.net/~anmaster/minecraft/screenshots/2010-11-08_22.39.18.png This is Holland.
22:18:23 <elliott> http://sporksirc.net/~anmaster/minecraft/screenshots/2010-11-08_22.41.35.png This isn't an overhang; it's resting on that tree!
22:18:32 <fizzie> elliott: For terrain features, you only need the seed and coordinates to regenerate them. I think there are (and/or have-been/were) some features to get the seed out and in.
22:18:56 <elliott> fizzie: Apart from all the things that have happened to it since the start... and any changes Vorpal might have made...
22:19:06 <Vorpal> I'm in the progress of uploading another bunch
22:19:16 <elliott> fizzie: Really though, I'd like a replay format for Minecraft; records the seed and all actions taken.
22:19:20 <elliott> (Including rotating the camera, I guess.)
22:19:46 <Vorpal> elliott, also as I predicted: net plus on coal
22:19:50 <fizzie> You could possibly just tcpdump-capture the multiplayer connection stuff.
22:19:55 <Sasha> feels good to be a legit owner
22:20:01 <Vorpal> fizzie, this is single player
22:20:14 <Vorpal> uploaded another bunch
22:20:24 <Vorpal> refresh index and look for different coloured links ;P
22:20:36 <elliott> http://vinhboy.com/blog/2010/03/25/how-to-get-free-comcast-internet/
22:21:12 <elliott> Sasha: I obtained it from an evil malicious legit owner because I'm horrible and kill kittens regularly. I would like to see Notch starve to death.
22:21:15 <elliott> (Note: 99% of above line false.)
22:21:25 <fizzie> Your upload-place doesn't show visited-links in my firefox, which is a bit strange and annoying. (Or maybe the browser is being snarky.)
22:21:28 <elliott> Vorpal is evil and malicious! AND I REFUSE TO STATE THE REASONS
22:21:34 <Sasha> elliott: It doesn't matter
22:21:51 <Vorpal> elliott, because I placed torches so I could track my path?
22:21:53 <Sasha> There's a little tiny feeling of satisfaction of owning my own account
22:21:59 <elliott> I have an account, too. :P
22:22:15 <Vorpal> mine is premium since yesterday
22:25:08 <elliott> Gregor: Permission plz? :P
22:26:26 <elliott> A program actually using write(2) instead of stdio. Impressive...
22:27:21 <Gregor> elliott: The .gz.tars are not called .gz.tars on HP-UX, they were just part of the crazy package system, called .pkg or something.
22:27:34 <elliott> Gregor: Permission to quote your statements about BSD/GPL to fefe? :P
22:27:52 <Gregor> elliott: Tell me A) whoTF is fefe and B) how exactly you're going to quote me first.
22:28:00 <elliott> Gregor: I told you: fefe is the dietlibc developer.
22:28:02 <elliott> I told you multiple times.
22:28:11 <Gregor> You told me after I was gone :P
22:28:13 <elliott> Gregor: Also, I'll just quote your lines that say "BSD" or "GPL" and not our argument.
22:28:19 -!- augur has joined.
22:28:49 <olsner> elliott: plain read/write is much nicer than stdio IMO
22:29:02 <fizzie> Someone should specify some sort of "tar-based certificate private key storage" thing; they could then call those files ".key.tar"s.
22:29:04 <elliott> olsner: Agreed, but... printf, dude.
22:29:16 <elliott> I don't even need printf really, as long as I have separate functions for the printf formats, I guess.
22:29:26 <elliott> olsner: printf is the bloatiest part of stdio :P
22:29:35 <Gregor> elliott: I don't think that grep -E 'Gregor.*(GPL|BSD)' produces a cogent stream of words, I want to see the actual quote(s) you intend to use :P
22:29:41 <elliott> olsner: but yeah, i have never understood why stdio isn't just write() and read() + strlen() :P
22:29:41 <olsner> but printf could just as well be mapped to read/write and no-one would notice
22:29:41 <Vorpal> <elliott> fizzie: Really though, I'd like a replay format for Minecraft; records the seed and all actions taken. <-- would be huge and slow down things. Especially since I tend to move view a lot to look for minerals
22:29:55 <elliott> Vorpal: Just record those at N fps.
22:29:59 <elliott> The actual keypresses and stuff will be tiny.
22:30:06 <olsner> well, except that printf probably makes good use of output buffering that write wouldn't give you
22:30:09 <Vorpal> elliott, mouse movements
22:30:35 <Vorpal> printf is quite useful
22:30:56 <Vorpal> write, read, printf, mmm
22:31:11 <fizzie> 50 fps of a couple of floating-point camera-position values isn't very many kilobytes/sec even as-is like that.
22:32:00 <Vorpal> <Gregor> elliott: The .gz.tars are not called .gz.tars on HP-UX, they were just part of the crazy package system, called .pkg or something. <--- where did we get *.gz.tar!?
22:32:16 <Gregor> Oh god INFINITE LOOPING CONVERSATION
22:32:32 <Vorpal> Gregor, please tell me though
22:33:08 <Vorpal> fizzie, do you use nice or fast rendering?
22:33:11 <Gregor> Vorpal: .tar has random access, .gz has compression, .tar.gz sacrifices random access in the name of compression, but you can get nearly the same compression ratio with file-level random access by gzipping files then tarring that.
22:33:15 <Gregor> Vorpal: Hence, .gz.tar.
22:33:33 <fizzie> The Alpha multiplayer protocol sends a 9-byte (float, float, bool) packet every time you turn your head. (Don't quite know how often.)
22:33:40 <olsner> elliott: one iffy part of read/write is that they may randomly decide to do only part of the buffer because it's convenient or because the system call was "interrupted" (well, for the "traditional" variants of read/write at least)
22:33:41 <Gregor> Vorpal: Yes, .zip or .7z is better, but they're so WINDOWSY :P
22:33:48 <Vorpal> I use fast, anything else is too slow for me
22:33:59 <elliott> olsner: Nobody actually does that :P
22:34:17 <Vorpal> elliott, mostly for sockets
22:34:18 <elliott> Vorpal: In my nice world, nobody actually does that.
22:34:26 <Vorpal> elliott, for sockets it happens sometimes
22:34:26 <elliott> Vorpal: In my nice world, nobody actually does sockets X-P
22:34:46 <Vorpal> elliott, mostly because kernel buffer is full though
22:34:57 <olsner> you do run into the partial read/write even on sensible modern systems like linux
22:35:21 <elliott> Gregor: [[[on distributing BSD-licensed programs linked with the GPL'd dietlibc]
22:35:21 <elliott> <Gregor> elliott: You just have to provide sources secondarily to the install media, which puts the onus of continuing that redistribution on the people who download it :P
22:35:21 <elliott> <Gregor> elliott: The install media can just be downloadable alongside the sources.
22:35:21 <elliott> <Gregor> elliott: Dude, the GPL is not MS Shared Source. When you want to redistribute something linked to something under the GPL, that just means you have to redistribute the whole thing under the terms of the GPL; that just means including sources.]]
22:35:25 <elliott> Gregor: Is this acceptable to you?
22:36:05 <elliott> Gregor: Oh, wait, I forgot <Gregor> I [...] f[...]u[...]ck go[...]at[...]s [...] a lot.
22:36:34 -!- augur has quit (Remote host closed the connection).
22:36:53 <elliott> Vorpal: It's totally true, I just elided some rubbish!
22:37:07 <olsner> when I get around to implementing file I/O I will just make my kernel do the whole operation except when it encounters an error or if non-blocking IO is explicitly enabled
22:37:11 <Gregor> elliott: Add "in this case" after "GPL;" and before "that", and that's fine.
22:37:15 <elliott> #define writes(fd,s) write(fd, s, sizeof(s))
22:37:15 <elliott> #define writev(fd,s) write(fd, s, strlen(s))
22:37:22 <elliott> olsner: You don't need files! :-)
22:37:27 -!- MigoMipo has quit (Read error: Connection reset by peer).
22:37:50 <olsner> hmm, did you swap writes and writev there?
22:37:54 <elliott> Gregor: Sure. Can I also remove "elliott:" and "Dude,"?
22:37:59 <elliott> olsner: writes writes a constant string
22:38:02 <elliott> for which sizeof(s) is the length
22:38:07 <elliott> writev writes a variable length string
22:38:15 <elliott> for which strlen(s) is the length (and sizeof == sizeof(char *))
22:38:16 <Vorpal> elliott, isn't writev to do scatter-io?
22:38:24 <Vorpal> readv, writev - read or write data into multiple buffers
22:38:29 <Vorpal> elliott, it is awesome :P
22:38:36 <Vorpal> for some high performance network stuff
22:38:36 <elliott> Consider it *writevs or something.
22:38:53 <elliott> Gregor: What about ":P"? You want that removed too, Mr. Uptight? :-P
22:38:55 <fizzie> olsner: I don't see how that's a problem with read/write as opposed to fread/fwrite; fread is allowed to return partial results too if the underlying read gets interrupted. (It's specified in terms of fgetc, and fgetc may return with errno=EINTR.)
22:39:03 <Gregor> elliott: No, :P is my signature :P
22:39:17 <elliott> Gregor: Dude, no, it's mine, I have :P on almost every line :P
22:39:28 <elliott> fizzie: Quick, SQL our logs; :Ps/msgs of both of us, who has more?
22:39:34 <Vorpal> elliott, Gregor so do I :P
22:39:44 <elliott> You don't count as a person.
22:40:07 <olsner> fizzie: the problem I have with fread is that fread is somehow supposed to be "higher" level - but it doesn't actually provide any useful abstraction over plain read
22:40:17 <Vorpal> elliott, hey you! [ACTIVATE MINECRAFT_TROJAN]
22:40:49 <elliott> Gregor: How would you like to be referred to? "Associate"? "Cow-orker"? "Fiancée"? "Lord"? "Murderer"? "Mother of my son"?
22:41:08 <olsner> fizzie: oh, and I meant that as a generic thing that is iffy with read/write without considering fread/fwrite at all
22:41:23 -!- cheater99 has quit (Ping timeout: 245 seconds).
22:41:38 <olsner> wasn't quite aware that fread/fwrite were allowed to do that except (obviously) on EOF
22:43:09 <olsner> this almost makes me want a pet antelope: http://www.google.se/images?q=royal+antelope
22:43:09 <elliott> Gregor: I already mentioned it was on IRC :P
22:44:48 <elliott> Gregor: I'll just avoid referring to you at all.
22:44:52 <olsner> speaking of butlers: http://www.youtube.com/watch?v=uPVcxy1zreA :)
22:46:36 <Vorpal> elliott, Gregor http://sprunge.us/cTEd
22:46:40 <fizzie> elliott: Your :P-factor is .022655 (7003/309109) while Gregor's is .132536 (6265/47270), sorry. (These numbers have a significant margin of error, they were quite quick-and-dirty, but anyway.)
22:46:48 <elliott> fizzie: NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
22:47:11 <elliott> I STILL WIN ON ABSOLUTE COUNT
22:47:26 <Vorpal> fizzie, can you give me the query and I can do them on the nick merged table?
22:47:29 <elliott> Vorpal: Hey, zuff is me too.
22:47:38 <Gregor> Vorpal: FFFFFFFFFFFFFFFFUUUUUUUUUUUUUUUUUUUUU :P
22:48:28 <Vorpal> elliott, then http://sprunge.us/bORK
22:48:31 <fizzie> Vorpal: I did some hand-made nick-merging there: http://p.zem.fi/ymnz -- and I counted "%:P%", not just end-of-line; and I counted the factors manually out of those two numbers.
22:48:44 <elliott> It is the termination that matters.
22:48:52 <elliott> Vorpal: Now divide by total messages :P
22:49:04 <elliott> fizzie: YOUR MERGE IS WOEFULLY INADEQUATE
22:49:06 <olsner> what, tusho was also elliott?
22:49:08 <elliott> I have many billions of nicks.
22:49:16 <elliott> fizzie: Dude... you forgot "alise".
22:49:20 <elliott> fizzie: Pretty big omission there :P
22:49:30 <Vorpal> elliott, are you *sure* that "<elliott> %:P% does not count." ?
22:49:46 <olsner> elliott: figured that from elliott matching the first name of ehird
22:50:04 <Vorpal> elliott, well bad for you. In absolute number that increases your lead even more
22:50:05 <fizzie> elliott: Oh, right. You should stop with all that nick-changing. The automagical regexp-remapping I have only works for the Python log-plots, I don't have a messy sqlite view like that.
22:51:25 <olsner> I wonder if I ever talked to tusho, but I remember the nick
22:51:43 <elliott> olsner: I am ehird, tusho, alise, zuff, estoppel, ... :P
22:52:09 <olsner> "estoppel", hmm, vague recollection there
22:52:21 <Vorpal> elliott, working on it
22:52:35 <elliott> olsner: I was estoppel for all of a few days :P
22:52:38 <Vorpal> elliott, you want to write the INNER JOIN?
22:52:52 <Vorpal> elliott, then don't complain
22:53:12 <fizzie> You don't need any joins, just two subqueries and then a / on them.
22:53:16 <elliott> olsner: WE SHOULD WRITE NONSTD-IO!!11127238947293487234
22:53:30 <elliott> (Does SQL taint minds so?)
22:53:40 <fizzie> Vorpal: Alternatively, just do it like in http://p.zem.fi/ymnz and do the single division manually.
22:53:54 -!- cheater99 has joined.
22:53:58 <fizzie> (That one gives you "is-:P" and "is-not-:P" counts for one nick.
22:54:36 <fizzie> Oh, unless you want to sort by the normalized :P ratios; that's trickier. But here you just needed the numbers for elliott and Gregor.
22:55:49 <elliott> Quick -- favourite libc/POSIX function. Name it!
22:55:59 -!- sebbu has quit (Read error: Connection reset by peer).
22:56:35 <Vorpal> > select a.nick as nick, b.cnt / a.total as ratio from (select nick, count(*) as total from irc.logs_na where type in (0,1) group by nick) as a, (select nick as nick,count(*) as cnt from irc.logs_na where type in (0,1) and body like '%:P%' group by nick) as b WHERE a.nick = b.nick order by ratio desc limit 20;
22:56:53 <elliott> Vorpal: That will run on every single nick.
22:56:56 <Vorpal> elliott, why? I dropped the inner join
22:56:59 <Vorpal> elliott, and correct, it will
22:57:12 <Vorpal> I got integer division
22:57:30 <Vorpal> fizzie, how do you get non-integer division?
22:58:04 <elliott> fizzie: "case body like '%:P%' when 1 then 1 else 0 end"
22:58:10 <elliott> fizzie: Couldn't you do int(bool) or something?
22:58:37 <Vorpal> elliott, arvid=> select a.nick as nick, b.cnt::real / a.total::real as ratio from (select nick, count(*) as total from irc.logs_na where type in (0,1) group by nick) as a, (select nick as nick,count(*) as cnt from irc.logs_na where type in (0,1) and body like '%:P%' group by nick) as b WHERE a.nick = b.nick order by ratio desc limit 20;
22:58:39 <fizzie> Vorpal: Also +0.0 works.
22:58:49 <Vorpal> it is postgresql specific syntax I think
22:58:58 <fizzie> If you want the joined version, here's it in sqlite:
22:59:05 <Vorpal> elliott, you are not in the top 20
22:59:07 <fizzie> sqlite> select nick, (p.pcnt+0.0)/np.cnt as pf from (select nick, count(*) as pcnt from logs where type = 0 and body like '%:P' group by nick) as p natural join (select nick, count(*) as cnt from logs where type = 0 group by nick) as np order by pf desc limit 20;
22:59:10 <fizzie> Svenstaro|0.666666666666667
22:59:25 -!- sebbu has joined.
22:59:25 -!- sebbu has quit (Excess Flood).
22:59:26 <elliott> fizzie: I think ordering by absolutes is the best X__X
22:59:31 <elliott> And just displaying the ratio.
22:59:48 <fizzie> elliott: Filtering with number-of-lines is good enough. Like this:
22:59:54 <fizzie> sqlite> select nick, (p.pcnt+0.0)/np.cnt as pf from (select nick, count(*) as pcnt from logs where type = 0 and body like '%:P' group by nick) as p natural join (select nick, count(*) as cnt from logs where type = 0 group by nick) as np where np.cnt > 10000 order by pf desc limit 20;
22:59:54 <fizzie> GregorR|0.118136747077105
22:59:54 <fizzie> ehird`|0.0351198287174529
22:59:54 <fizzie> psygnisfive|0.0350286077558805
22:59:55 <fizzie> pikhq|0.0294485036190346
22:59:55 -!- sebbu has joined.
22:59:55 -!- sebbu has quit (Excess Flood).
22:59:57 <fizzie> tusho|0.0246700781039591
23:00:00 <fizzie> Then just run that in the nick-merged table.
23:00:03 <elliott> Vorpal: Run that on the merged one kthx :P
23:00:21 -!- sebbu has joined.
23:00:21 -!- sebbu has quit (Excess Flood).
23:00:42 <Vorpal> different column names
23:00:47 <elliott> Hey -- I think BSD tar doesn't accept options without - in front of them.
23:01:06 <fizzie> Well, it needs nick, body and type, where type = 0 tests for msg-ness.
23:01:27 <Vorpal> select * from (select a.nick as nick, b.cnt as cnt, b.cnt::real / a.total::real as ratio from (select nick, count(*) as total from irc.logs_na where type in (0,1) group by nick) as a, (select nick as nick,count(*) as cnt from irc.logs_na where type in (0,1) and body like '%:P%' group by nick) as b WHERE a.nick = b.nick order by b.cnt desc limit 20) as foo order by foo.ratio;
23:01:47 <fizzie> You're doing %:P% there again.
23:02:09 <elliott> Regexp matching isn't sh-portable, right?
23:02:26 <elliott> i.e. regexp matching might not work in non-bash shells.
23:02:27 <Vorpal> elliott, you mean [[ =~ ]] ?
23:02:34 <Vorpal> then that is bash specific
23:02:52 <Vorpal> elliott, http://sprunge.us/eMRN
23:03:09 <Vorpal> is that order the wrong way?
23:03:24 <Vorpal> elliott, JUST READ IT BOTTOM UP!
23:03:33 -!- sebbu has joined.
23:03:36 -!- sebbu has quit (Excess Flood).
23:03:40 <elliott> Vorpal: Please tell me that my number of %:P messages exceeds Gregor's number of total messages.
23:03:47 <elliott> Or, wait, fizzie measured both, didn't he?
23:04:02 -!- sebbu has joined.
23:04:02 -!- sebbu has quit (Excess Flood).
23:04:12 <Vorpal> elliott, you could calculate the total number from the ratio...
23:04:21 <olsner> elliott: 8947*0.03 looks like a lot less than 6652
23:04:29 -!- sebbu has joined.
23:04:30 -!- sebbu has quit (Excess Flood).
23:04:38 <Vorpal> elliott, okay other way around: http://sprunge.us/TJgX
23:04:57 -!- sebbu has joined.
23:04:57 -!- sebbu has quit (Excess Flood).
23:05:08 <Vorpal> olsner, I do different nick merging
23:05:20 <Vorpal> because he had so many nicks
23:05:22 -!- sebbu has joined.
23:05:22 -!- sebbu has quit (Excess Flood).
23:05:49 <Vorpal> fizzie, you have +o can you ban sebbu for now so he stops this spam?
23:05:52 -!- sebbu has joined.
23:05:53 -!- sebbu has quit (Excess Flood).
23:06:20 -!- sebbu has joined.
23:06:21 -!- sebbu has quit (Excess Flood).
23:06:25 <elliott> poiuyutjyyj_qwert, whatever his nick is, did it yesterday too :P
23:06:29 <Vorpal> olsner, dumped the view: http://sprunge.us/WGYW
23:06:36 <Vorpal> elliott, well I wasn't here then
23:06:42 <Vorpal> so I didn't know about it
23:06:50 -!- sebbu has joined.
23:06:50 -!- sebbu has quit (Excess Flood).
23:06:56 <elliott> Vorpal: Um, you don't have estoppel.
23:07:16 -!- ChanServ has set channel mode: +o fizzie.
23:07:19 -!- sebbu has joined.
23:07:20 -!- sebbu has quit (Excess Flood).
23:07:26 <olsner> Vorpal: that's an SQL query, as opposed to something actually useful for me
23:07:29 -!- fizzie has set channel mode: +b sebbu!*@*.
23:07:44 <fizzie> How long should I keep that thing up? A few minutes? Will it stop autojoining then?
23:08:06 <olsner> not that any of this is "useful" to begin with, but the results I can at least read :)
23:08:10 <Vorpal> elliott, here: http://sprunge.us/BaGc
23:08:13 <elliott> fizzie: If you op me I'll sort it out for you.
23:08:25 <Vorpal> fizzie, unlikely to help since it is auto quitting
23:08:49 <Vorpal> olsner, sql query is all you get
23:09:03 <fizzie> I think I'll -- not now, but later -- have that Python script push in merged nicks to another table. It's not as elegant as a modifiable view, maybe, but might be fastener.
23:09:10 <elliott> fizzie: If you op me I'll sort it out for you!1121212121111
23:09:19 <Vorpal> olsner, FINALLY NOTHING BETWEEN ME AND WORLD DOMINATION!
23:09:33 <elliott> `addquote <Vorpal> olsner, FINALLY NOTHING BETWEEN ME AND WORLD DOMINATION!
23:09:50 <elliott> I should maybe have added the olsner thing.
23:09:53 <HackEgo> 256|<Vorpal> olsner, FINALLY NOTHING BETWEEN ME AND WORLD DOMINATION!
23:10:00 <elliott> `addquote <olsner> DAMN YOU, I'm leaving <Vorpal> olsner, FINALLY NOTHING BETWEEN ME AND WORLD DOMINATION!
23:10:01 <HackEgo> 257|<olsner> DAMN YOU, I'm leaving <Vorpal> olsner, FINALLY NOTHING BETWEEN ME AND WORLD DOMINATION!
23:10:09 <Vorpal> elliott, your revert failed
23:10:15 <Vorpal> elliott, now sort it out
23:10:16 <HackEgo> 256|<Vorpal> olsner, FINALLY NOTHING BETWEEN ME AND WORLD DOMINATION!
23:10:52 <elliott> A pittance of, uh, hugs, to anyone who writes xzt(1).
23:11:11 <elliott> (find -type d -exec xz '{}' \; + tar, and the reverse of that)
23:11:18 <elliott> Vorpal: .xzt = .xz.tar, except not confusingly-named if taken literally.
23:11:22 <elliott> (Also more 8.3 compatible.)
23:11:33 <olsner> elliott: like... tar with compression on the inside instead of the outside?
23:11:41 <elliott> olsner: Yes, as discussed 89573495834734589 times with Gregor.
23:11:57 <olsner> I wasn't listening then!
23:11:58 <elliott> olsner: You can see the file structure, and only decompress some files if you want to, and the like.
23:12:05 <Vorpal> elliott, .zip.tar.bz2 inside a tarball
23:12:08 <elliott> olsner: And the compression lost is roughly zero, since tars hardly have much overhead :P
23:12:18 <elliott> Vorpal: No, not seriously :P
23:12:29 <Vorpal> elliott, it was something from Xilinx
23:12:48 <Vorpal> elliott, so it does. ais would not be surprised
23:12:52 <olsner> orly? I always thought of tar as having massive overhead
23:12:59 <fizzie> Overhead is not why it loses in compression ratios: not being able to compress across files is.
23:13:06 <elliott> olsner: Try it for yourself (say, the Linux kernel).
23:13:20 <Vorpal> elliott, I think the outer tar was not compressed. But it was made without an index...
23:13:25 <Vorpal> elliott, and it was around 3 GB
23:13:25 <elliott> fizzie: Well, yes; obviously, the Ideal Format would handle that, too, but -- hey, zip and rar do it.
23:13:29 <elliott> And they're not exactly unpopular.
23:13:46 <elliott> Going to do the kernel (with bz2, so it's fair) to compare.
23:14:17 <Vorpal> elliott, not xz outside and xz inside?
23:14:33 <fizzie> Intel outside and Intel inside.
23:14:36 <elliott> Vorpal: That would be immensely pointless and overhead-inducing :P
23:14:46 <fizzie> Let's see if the noise is still there.
23:14:48 <Vorpal> elliott, not at the same time duh
23:14:52 <Vorpal> elliott, I meant to compare the sizes
23:15:05 -!- fizzie has set channel mode: -b Sebbu!*@*.
23:15:09 <olsner> I don't remember which block size tar uses, but it pads to even blocks and stuff, which seems like something you might want to skip if you don't compress on the ourside
23:15:39 <elliott> olsner: ...that's beyond stupid.
23:15:53 <Vorpal> elliott, no it isn't. zip does that by default too
23:15:54 <elliott> olsner: LET'S JUST USE 7ZIP.
23:16:10 <elliott> http://p7zip.sourceforge.net/
23:16:25 <Vorpal> elliott, random seeking is not important in the kernel though
23:16:27 <elliott> Open format, portable compressor.
23:16:32 <elliott> Vorpal: Yes, but it's a matter of principle :P
23:16:36 <olsner> elliott: yep, afaik 7zip is perfectly sensible
23:16:59 <fizzie> Sensible, and therefore off-topic for us.
23:17:18 <elliott> I hope .7z stores file permissions and symlinks and the like.
23:17:23 <elliott> If it does, fuck it, I'm just going to use it forever.
23:17:51 <elliott> http://en.wikipedia.org/wiki/Solid_compression
23:18:04 <elliott> olsner: LOL 7ZIP IS LIKE .TAR.FOO
23:18:40 <olsner> elliott: istr that's an option when you create the archive, if you want it solid or not
23:18:58 <elliott> Fuck it, we'll do it live.
23:19:09 <elliott> <elliott> Quick -- favourite libc/POSIX function. Name it!
23:19:13 <olsner> "*Support for* solid compression" emphasis mine
23:19:38 <olsner> but: "The 7z format does not store UNIX owner/group permissions"
23:20:26 <elliott> Vorpal: ...wh...you stole mine.
23:20:26 <fizzie> olsner: You can both toggle it on/off as well as set the maximum block size when it's on.
23:20:30 <elliott> I was going to pull out mmap() at the end.
23:20:42 <elliott> Vorpal: It's my favourite!
23:20:44 <fizzie> (So you sort of get the benefits of both.)
23:20:45 <elliott> Vorpal: The closest thing to orthogonal persistence this side of Torsion!
23:20:47 <Vorpal> elliott, it is mine too
23:21:04 <Vorpal> elliott, it is microoptimising of course ;P (j/k)
23:21:21 <fizzie> There's also a flag which makes it start a new solid block for each new file extension.
23:21:47 <fizzie> The assumption being that you want across-file compression for file format headers and such.
23:21:53 <Vorpal> <olsner> but: "The 7z format does not store UNIX owner/group permissions" <-- nor symlinks?
23:22:17 <Vorpal> elliott, idea: squashfs archives
23:22:19 <olsner> if you bother implementing symlinks I'm pretty sure you'll make space for a few bits of owner/group info too
23:22:39 <Vorpal> elliott, they are non-solid
23:22:45 <Vorpal> elliott, also very convenient to use
23:22:53 <elliott> Vorpal: Why convenient? 'Cause you can mount them?
23:23:31 <Vorpal> elliott, hey I put mathematica on one and icc on another, nwn static files on a third. In total I saved over 7 GB on my laptop. :P
23:23:32 <olsner> hmm, if you can only have exactly one system call, I think you should make it mmap
23:23:47 <Vorpal> elliott, and yes you can mount them
23:23:53 <Vorpal> elliott, then out of tree build becomes easy
23:23:57 -!- nooga has joined.
23:24:04 <Vorpal> mount gcc.squash gcc-src
23:24:20 <Vorpal> ../gcc-src/configure --whatever
23:24:23 <elliott> Vorpal: I need to write a mount-loopback(1) with FUSE sometime; stupid mount(1) is root-only...
23:24:37 <elliott> Also, how is that easier than just untarring it?
23:24:57 <elliott> Vorpal: Why? Saves disk space only?
23:25:01 <Vorpal> elliott, and you don't use up inodes
23:25:24 <Vorpal> elliott, that used to be a problem historically
23:26:49 <fizzie> There's a thing called "mountlo" which I think is a fuse-driven loopback device.
23:26:50 <Vorpal> elliott, because it was like 16 bit counter and it would overflow.. With serious effects
23:26:54 <olsner> hmm, or a system call for writing a single byte on some kind of "system command stream" might be another nice single-syscall os
23:27:06 <fizzie> It seems that the autojoin/quit noise did in fact stop.
23:27:12 -!- fizzie has set channel mode: -o fizzie.
23:27:20 <olsner> but it seems a bit too easy to just add multiplexing and wrap it in nice libraries so that it isn't esoteric anymore but just slow :/
23:28:16 <olsner> same thing for mmap-only, you'd just use mmio to "call" hundreds of system functions by writing commands into designated pages
23:28:30 <olsner> *something similar to mmio
23:28:58 <Vorpal> elliott, weird yellow text in minecraft
23:29:12 <Vorpal> it doesn't change either
23:29:38 <elliott> Notch - Minepedia - The Minecraft Wiki!
23:29:38 <elliott> He lives just outside of Stockholm, Sweden. Notch is 31. He enjoys working on Minecraft. He is engaged to his fiancée, Elin aka "ez"; He is SCUBA certified ...
23:29:54 <fizzie> [freenode] -!- There is no such nick ez
23:30:14 <fizzie> Not in Freenode == doesn't exist.
23:30:15 <elliott> Boy this is beyond tedious...
23:30:20 <Vorpal> elliott, what was his real name now again?
23:30:27 <Vorpal> Some very very generic one iirc
23:30:31 <elliott> Vorpal: Markus ALEXEJ Persson.
23:30:47 <Vorpal> elliott, not very common middle name
23:30:53 <olsner> you missed a few on the end :)
23:31:00 <Vorpal> but Markus Persson, that is very common
23:31:24 <Vorpal> olsner, beat you to it
23:31:43 <olsner> Vorpal: DAMN YOU, and this time I'm really leaving!
23:31:55 <fizzie> http://p.zem.fi/minecraft-splash has the regulat splash-texts, though it was extricated from a pre-Halloween Alpha.
23:31:57 -!- nooga has quit (Ping timeout: 276 seconds).
23:32:29 <fizzie> The "This text is hard to read if you play the game at the default resolution, but at 1080p it's fine!
23:32:51 <Vorpal> elliott, Legal in Finland?
23:32:52 <fizzie> Yes, most things are illegal here.
23:32:56 <Vorpal> huh, someone better explain that
23:33:17 <Vorpal> fizzie, "Now supports ���!"?
23:33:40 <fizzie> Vorpal: Encoding problem, maybe.
23:33:55 <fizzie> /v/ could be a image-board thing.
23:34:01 <elliott> Vorpal: http://boards.4chan.org/v/
23:34:08 <elliott> Fits in with the other surrounding "Woo"s.
23:34:28 <Vorpal> elliott, I just closed browser and opened minecraft, what is it about?
23:34:29 <elliott> As they're all game-related (well, Something Awful only semi-related, but.)
23:34:34 <elliott> Vorpal: It's a 4chan board.
23:34:40 <Vorpal> elliott, yes and the topic is?
23:34:51 <elliott> "The Vidya"; or, less opaqueley, video games.
23:36:37 <fizzie> Current list: http://p.zem.fi/minecraft-splashes-2
23:37:26 <fizzie> The chars in "now supports" one actually seem to be valid utf-8 for the missing-char glyph in the source .txt, unless I mislooked.
23:38:06 <elliott> "Eple (original edit)!" -- http://www.youtube.com/watch?v=NBiRcAmmJVY
23:38:12 <elliott> (Unless that isn't the original edit.)
23:44:32 <oklopol> and then i decided to dig up but ran outta ladders and then i came up and i died.
23:45:52 <elliott> oklopol: you dig staircase-style
23:45:57 <elliott> oklopol: also, classic or alpha?
23:47:36 <oklopol> but wanted to go straight up
23:49:02 <fizzie> You can go staight up with a 4x4 shaft with no ladders by leaving a circular-staircase thing going around and around the shaft. (It's more boring to climb, though.)
23:49:39 <fizzie> W.r.t. ladders, have you noticed that you only need to put those to every other square in order to be able to climb?
23:49:52 <fizzie> That's 50 % savings right there.
23:52:14 <fizzie> s/4x4/3x3/ up there if you prefer.
23:52:17 * elliott considers forking http://dmr.ath.cx/net/darkhttpd/ and adding CGI/SCGI support
23:52:29 <elliott> pikhq: shell account is gone i would guess
23:52:34 <elliott> pikhq: (not in here and email account on the same server is dead)
23:52:47 <elliott> pikhq: but he's commented on Gödel's Last Letter and P=NP so he's fine :P
23:52:48 -!- elliott has left (?).
23:52:50 -!- elliott has joined.
23:58:07 <Vorpal> fizzie, the happy birthday, ez one is not there?
23:58:25 <Vorpal> fizzie, do you need one at the top square?
23:58:49 <fizzie> Possibly, or maybe you can jump.
23:58:57 <oklopol> fizzie: obviously i didn't start playing without watching hours of tutorials and reading the wiki for hours first
23:59:19 <fizzie> oklopo: But of course!
23:59:46 <Vorpal> nether is really quite efficient if you only care approximately where you end up