00:04:09 -!- tombom has quit (Quit: Leaving).
00:10:27 <alise> "The new Mac Pro. With up to 12 cores of processing power, it’s the fastest Mac ever."
00:10:31 <alise> Abuh. Resist urge to purchase.
00:13:55 <alise> (Okay, that took very little effort actually.)
00:14:03 <alise> (Looks like my disassimilation went quickly.)
00:14:30 * pikhq both wants to eat and cannot compel self to make something
00:14:55 <alise> pikhq: eat plastic!
00:14:59 <alise> it's nutritious and NUTRITIOUS!
00:19:10 <nooga> pikhq: eat raw food
00:21:03 <AnMaster> alise, Get a Intel server board, a case, some disks, a few fans, and a GPU and you get the same performance for less than half the price
00:21:36 <AnMaster> still less than half the price
00:21:46 <alise> AnMaster: Well, yeah, if all you care about is raw hardware performance. (Which is, in this case, true; but my point stands.)
00:21:52 -!- Wamanuz2 has joined.
00:22:02 <AnMaster> alise, why would you care about anything else than that?
00:22:12 -!- Wamanuz has quit (Ping timeout: 260 seconds).
00:22:19 <alise> I refuse to answer ...
00:22:19 <AnMaster> what else is there to care about? Being a mac hippie?
00:22:31 <alise> Now you're just trolling: you know I dislike Apple.
00:22:49 <alise> So the only possible point of that statement -- to ridicule my like of Apple -- does not exist; you are trolling.
00:22:56 <AnMaster> alise, http://www.gizmodo.com/gadgets/images/iProduct.gif
00:23:09 <AnMaster> alise, no, I'm not trolling you
00:23:28 <AnMaster> alise, which is a lot more fun
00:24:00 <alise> Yes; you're trolling Apple by saying "mac hippie" somewhere entirely other than Cupertino. Do you even understand how this sort of stuff works? You can't troll someone who isn't listening. Do you even know what "trolling" means?
00:24:41 <AnMaster> alise, a far out idea for something else than raw performance: the case design
00:24:54 <AnMaster> and um, not having to put together the parts on your own
00:25:13 <alise> Shut up. You're not making intelligent socio-political commentary about Mac users.
00:25:13 <AnMaster> probably buying a comparable HP server or such would still be way cheaper
00:25:27 <AnMaster> alise, ... that's an impossible task :P
00:26:07 <AnMaster> alise, btw are those apple xserver thingies equally over-priced?
00:26:16 -!- augur has quit (Read error: Connection reset by peer).
00:26:23 -!- augur has joined.
00:26:23 <AnMaster> and is that why you never see them in data centers?
00:26:56 <alise> I've made a New Midnight's Resolution; and that is to not respond to anything blatantly provocative and/or trolling in a very stupid way from you.
00:27:11 <alise> So, good luck with that.
00:27:14 <AnMaster> alise, "<AnMaster> alise, btw are those apple xserver thingies equally over-priced?" was actually a serious question
00:27:21 <alise> They're expensive.
00:27:35 <alise> Whether they are overpriced is up to you; they're certainly for institutions with a lot of cash. They /are/ in use.
00:27:45 <alise> Just not in data centres; more in processing labs and stuff.
00:28:03 <CakeProphet> Python -- derived from the verb pythein, "to rot"
00:28:09 <AnMaster> alise, because most people who 1U servers probably care a LOT more about raw power than someone who buys a mac pro.
00:28:42 <alise> CakeProphet: Oh snap.
00:29:10 <AnMaster> CakeProphet, how fitting for Monty Python was my first thought. How fit for the programming language python was my second.
00:29:12 <alise> Hmm, \#esoteric produces a distinctly unsatisfactory rendering of the channel name with LaTeX.
00:29:23 <CakeProphet> quick, someone tell me what style of columns the Parthenon had.
00:29:59 <alise> AnMaster: Minion Pro.
00:30:07 <alise> But it's more the...
00:30:08 <AnMaster> alise, try it in CMR, does it look better there?
00:30:10 <alise> I think it needs small caps.
00:30:19 <AnMaster> CakeProphet, I am however curious as to why you wondered about that
00:30:23 <alise> AnMaster: it won't; it's not a font issue. Besides, all the glyphs are perfectly reasonable and professional. This is not a font issue.
00:30:27 <alise> This is a typesetting issue.
00:30:39 <alise> AnMaster: No. There is no point.
00:30:46 <alise> \textsc{\#esoteric} is almost correct, but the letter spacing is too large.
00:31:04 <alise> You just hate my apparent hate of CMR and thus try to push CMR even when it's irrelevant.
00:31:27 <AnMaster> alise, in the context of latex CMR is quite relevant
00:31:39 <alise> But there's no reason to say "try CMR" if it's nothing to do with font.
00:31:39 <pikhq> IIRC, you felt that CMR is the only acceptable Didone font...
00:31:50 <alise> pikhq: Which it, of course, is. An opinion I believe you share.
00:31:55 <pikhq> (which of course it is)
00:31:57 <CakeProphet> I don't really get why typesetting is interesting.
00:31:57 <AnMaster> alise, I never claimed you should use it as a font for GUI elements
00:32:24 <alise> CakeProphet: Then I kindly suggest you don't bother those of us who do think it is.
00:32:30 <alise> I probably don't consider many of your hobbies interesting.
00:32:41 <pikhq> CakeProphet: You need to read more well-typeset text.
00:32:49 <alise> {\large\#}esoteric % this is good
00:32:56 <AnMaster> CakeProphet, I think LaTeX with microtype does a good job at it. I leave the rest to it
00:33:05 <AnMaster> it looks good and I don't want to fiddle with details
00:33:52 <pikhq> CakeProphet: Yes, well-typeset text will use serif fonts for most of it.
00:34:04 <AnMaster> pikhq, unless for screen reading
00:34:16 <AnMaster> pikhq, I consider type setting a solved problem: LaTeX + CMR + Microtype
00:34:40 <alise> I love how AnMaster thinks that one type fits /all purposes/; that it's spelled "type setting", and that one type fits /all purposes/.
00:34:41 <pikhq> AnMaster: Yeah, for low-DPI usage a sans serif font is often preferable.
00:34:53 <AnMaster> alise, it fits everything I type
00:35:00 <alise> It's like there's some retarded part of his brain that really doesn't understand what typefaces do to text. Wait, that's all of his brain.
00:35:17 <AnMaster> alise, also I think it doesn't matter if I spell it "type setting" or "typesetting" on IRC. Hell English is inconsistent already
00:35:22 <pikhq> Though I've found that Deja Vu Serif, hinted, is *very* readable.
00:35:30 <alise> AnMaster: i totly agre LOL fuk u english
00:35:38 <alise> pikhq wtf??? r dese esolangs
00:35:38 <pikhq> (as a screen font)
00:35:43 <alise> u dont no fuckshit bout languages
00:35:48 <AnMaster> alise, you fail at trolling however
00:36:00 <alise> you fail at knowing what trolling is and when it is intended.
00:36:06 <AnMaster> alise, like you fail at trolling augur for how he types :P
00:36:18 <pikhq> CakeProphet: I kill people for MS Comic Sans.
00:36:22 <augur> AnMaster: thats because i type awesome
00:36:41 <pikhq> I will (barely) make an exception for use in comic lettering.
00:36:57 <CakeProphet> http://julianhansen.com/index.php?/alternative-type-finder/
00:36:59 <pikhq> (where it is *appropriate*, though not the best choice of font, IMO)
00:37:07 <AnMaster> <pikhq> Though I've found that Deja Vu Serif, hinted, is *very* readable. <-- agreed. the whole Dejavu family, when hinted, is very good for on-screen reading
00:37:23 <pikhq> AnMaster: Shit for printing, but quite nice on-screen.
00:37:35 <AnMaster> pikhq, yes, I use CMR for printing ;P
00:40:04 <alise> Frutiger is a nice font but the NHS use it.
00:40:11 <AnMaster> pikhq, to me it makes no difference
00:40:26 <alise> AnMaster: and that is why you do not have the qualifications to call typesetting a "solved problem".
00:40:35 <alise> which is btw a huge insult to the very skilled and careful typographers through the ages.
00:40:36 <nooga> national health service?
00:40:42 <alise> *national hell service
00:40:58 <nooga> mauhaha, you should check it here in PL
00:41:04 <CakeProphet> alise: I think you're one of the easiest people to troll ever.
00:41:25 <pikhq> alise: Well, the boring parts of typesetting *are* a solved problem. :)
00:41:43 <alise> CakeProphet: i've not been trolled.
00:41:48 <alise> i know exactly what AnMaster is trying to do
00:41:49 * AnMaster forces alise to typeset whatever he is typesetting in Zapfino
00:41:55 <alise> i'm just talking shit about him like always
00:42:08 <augur> anyone a big bang theory fan?
00:42:18 <AnMaster> alise, yes you usually jut talk shit
00:42:23 <CakeProphet> The entrance to the Acropolis that used both Doric and Ionian columns was the...
00:43:51 <coppro> woot I finally have irssi up
00:44:04 <alise> i don't hate niggers just 'cuz they be niggers man
00:44:50 <alise> no that's the faggots
00:45:27 <nooga> oh no, please stop
00:46:41 <nooga> I'm allergic after one weird (luckily not close) encounter
00:46:56 <augur> has anyone reminded you guys recently that alise is like a small version of David Deutsch?
00:47:03 <cheater99> well go on, nothing to be ashamed of
00:47:29 <AnMaster> <coppro> woot I finally have irssi up <-- what about it?
00:47:33 <alise> nooga was /trying/ to have sex with a cat covered in peanuts
00:47:48 <augur> how would you cover a cat in peanuts
00:47:55 <augur> they'd just fall off the cat
00:47:57 <nooga> besides, i hate cats
00:48:00 <AnMaster> coppro, I mean, after getting printing files to cups from opengenera to work, that feels like an insignificant feat :P No offence meant.
00:48:37 <alise> coppro: don't worry AnMaster is just so tech-superior to you
00:48:42 <alise> he solved a trivial problem too but his was more pointless!
00:48:45 <augur> alise: that would be rather difficult
00:48:56 <AnMaster> alise, actually I expect coppro to be able to pull this off too
00:49:55 <coppro> why, just the other day I broke a computer by joining a club
00:50:29 <cheater99> nooga got scared to talk about his life
00:52:44 <AnMaster> alise, besides, it wasn't trivial. Had to translate the ppd to settings that opengenera understands. Which is in a custom format of course
00:52:53 <AnMaster> and various other strange stuff
00:53:01 <nooga> AnMaster: what for?
00:53:11 <AnMaster> nooga, what for? To get printing to work?
00:53:19 <AnMaster> or do you mean "why get printing to work"
00:53:30 <AnMaster> or "why opengenera" or "why computers"
00:53:46 <nooga> why get printing to work under opengenera
00:54:27 <AnMaster> nooga, why not. And I wanted to be able to do hardcopies of my interactive lisp session easily ;P
00:54:44 <AnMaster> however, I haven't yet got that to work, since it adds a cover page. And that breaks things
00:55:08 <nooga> ffs, why are you playing with opengenera?! don't you have emacs? :|
00:55:47 <AnMaster> ooh I just had an idea. However it has the issue that documents will print in reverse
00:56:02 <AnMaster> could add a post processing step for that though
00:56:15 <nooga> ah, i see, you're the opengenera fan, like alise is a plan9 fan
00:56:30 <AnMaster> nooga, oh yes, I quite like plan9 too as well
00:56:49 <nooga> wait a second, plan9 is actually quite useful and at least boots natively on a modern pc
00:56:57 <AnMaster> nooga, both systems were commercial failures. Both were innovative.
00:57:10 <alise> opengenera was not innovative.
00:57:29 <alise> CakeProphet: also true
00:57:34 <AnMaster> alise, but I don't have a physical lisp machine, so I will have to take the second best
00:57:53 <cheater99> http://www.tomshardware.com/news/russia-processor-collection-persian-rugs,10975.html
00:57:54 <AnMaster> and god damn, I never ever seen such a well documented system as genera
00:58:10 <CakeProphet> I have never heard of genera, but I have heard of plan 9
00:58:22 <AnMaster> CakeProphet, you heard of lisp machines?
00:58:29 <nooga> i like the idea of tty
00:58:43 <AnMaster> nooga, tty as in ttys on unix?
00:59:19 <AnMaster> nooga, why? I mean, it seems to me like something obvious that is just there.
00:59:29 <nooga> and i don't like plan9's idea of console
00:59:33 <AnMaster> it's like "I like the idea of RAM"
00:59:38 <nooga> but the rest of plan9 is okay
01:00:21 <AnMaster> I like many parts of plan9, but the user interface is not one of them. I like genera/opengenera's user interface though. You very quickly get used to scroll wheel doing nothing
01:00:27 <AnMaster> and then very quickly you stop missing it
01:00:36 <AnMaster> scrolling works quite differently but very well
01:01:12 <AnMaster> scroll wheel as well would be nice though.
01:01:24 <nooga> i tried to write a replacement for rio and, suprisingly, i even managed to display colorful rectangles on the screen
01:02:14 <AnMaster> nooga, I think there is a bit more than that to it :P
01:03:41 <alise> rio is the window multiplexer
01:04:06 <AnMaster> alise, does it draw the window borders?
01:04:12 <CakeProphet> so, anyone else know a lot about DSP? and also Haskell?
01:04:23 <AnMaster> alise, and handle dragging of borders and such?
01:04:28 <CakeProphet> because I want to figure out how to do DSP in Haskell, but so far I haven't been able to comprehend DSP libraries for Haskell.
01:04:36 <alise> AnMaster: but it is not structured like an x manager.
01:04:39 <CakeProphet> and think maybe I could do it myself pretty well?
01:04:49 <AnMaster> still, it is a window manager in my book
01:05:50 <alise> the basic strategy is pointing /dev/{screen,mouse,keyboard} of children to its own file then painter's algorithm
01:05:59 <alise> so in a way it is more involved than an x manager by completely engulfing windows
01:06:25 <nooga> http://www.facebook.com/ehird
01:06:28 <AnMaster> alise, besides I only asked if it was the WM. By pure chance both Window Manager and Window Multiplexer happens to fit that
01:06:31 <nooga> hmmmmmmmmmmmmmmmmmmmmmmmmmm
01:06:39 <nooga> hmmmmmmmmmmmmmmmmmmmmm
01:07:00 <AnMaster> alise, so it is like a combination of a X WM and the actual X server?
01:07:22 <nooga> p9's gui is nothing like X
01:07:27 <alise> AnMaster: in a way. but the graphics hardware is handled by the drivers still.
01:07:30 <CakeProphet> I'm thinking Arrow would be a useful abstraction for digital signal processors?
01:07:39 <AnMaster> alise, and by this design hw acceleration OpenGL performance would suck under plan9, right?
01:07:42 <alise> it multiplexes a bunch of fake /dev/screens to the real /dev/screen with painter's algorithm basically.
01:07:50 <alise> AnMaster: yeah -- because summaries reveal all details.
01:08:05 <AnMaster> alise, indeed, that's a universal truth :P
01:18:12 -!- Behold has joined.
01:18:32 -!- BeholdMyGlory has quit (Ping timeout: 240 seconds).
01:18:43 -!- Behold has changed nick to BeholdMyGlory.
01:22:57 -!- oerjan has quit (Remote host closed the connection).
01:23:38 -!- oerjan has joined.
01:38:49 <alise> http://julianhansen.com/files/infographiclarge_v2.png
01:39:43 <alise> READABILITY? -> NO -> Didot
01:40:10 <nooga> AnMaster: just quoting Helmer from Von Trier's Kingdom
01:43:10 <Gregor> alise: I love the link to comic sans X-D
01:44:22 <alise> The Kingdom (Danish title: Riget) is an eight-episode Danish television mini-series, created by Lars von Trier in 1994, and co-directed by Lars von Trier and Morten Arnfred.
01:45:56 * oerjan recalls his first collaborator was a big fan of Riget
01:46:32 <Gregor> I'm a big fan of scratching my head and wondering why people care so much about fonts.
01:46:43 <alise> Typefaces, dammit, typefaces.
01:47:19 <oerjan> Gregor: i guess you need to care about esthetics first, right? *badum-tish* (/me doesn't really either)
01:47:19 <alise> A font is a component of a typeface: for instance Times New Roman is a typeface, where as 10pt Times New Roman, italic is a font.
01:47:35 <Gregor> I have zero capacity for aesthetics.
01:47:48 <Gregor> This is why today I was wearing a pink shirt with a peacock blue tie and a white hat :P
01:49:00 <nooga> Gregor: you're research fellow, right?
01:49:11 <Gregor> No, I don't have a fellowship.
01:49:16 <Gregor> But I'm usually a research assistant.
01:49:24 <Gregor> And presently I'm a research intern.
01:49:38 <alise> Research Fellowship of the Pseudo-Ring.
01:49:55 <Gregor> ... no. Not justified :P
01:49:58 <nooga> Gregor: your fashion
01:50:42 <Gregor> When I showed up for my first day at MSR, all the first-day interns (in all divisions) had to herd together first. When I went through one line, the woman checking things went "I'll bet you're in research. Go that way."
01:50:48 <AnMaster> <Gregor> This is why today I was wearing a pink shirt with a peacock blue tie and a white hat :P
01:52:25 <AnMaster> Gregor, however. if you are able to note that this means you don't have a capacity for aesthetics, then that means you either were told this, or realise that it isn't aesthetic. The latter would mean that you do in fact have a capacity for aesthetics!
01:52:51 <Gregor> The way I dress is regularly described as "garish" or "gaudy"
01:52:56 <alise> MMM. SPIKY SERIFS ARE NICE -> OK -> Swift
01:53:00 <alise> That's some multiple choice.
01:53:06 <Gregor> I figure I probably haven't figured it out by now :P
01:53:25 <alise> DO YOU LIKE FUTURA? -> YES -> Futura
01:53:29 <alise> That was unexpected.
01:55:15 -!- Zuu has quit (Ping timeout: 276 seconds).
01:55:27 <AnMaster> Gregor, iirc you said you use your colour matcher to generate non-matching colours right?
01:55:45 <Gregor> I don't actually do that.
01:56:18 <AnMaster> Gregor, that would have been so cool
01:56:32 <oerjan> Gregor: YOU'VE BEEN LYING TO US
01:56:54 <Gregor> The problem is clothing matching is more complicated than colors, reflective properties can make two things that seem to match from one photo actually go terribly together.
01:57:41 <oerjan> hm so you actually need a ray tracer to do it properly? :D
01:58:13 <AnMaster> Gregor, as long as they don't turn out to go nicely together when they shouldn't
01:58:40 <Gregor> Fair enough when I'm aiming NOT to match I suppose :P
01:59:42 <AnMaster> Gregor, as with your clothes that is?
02:01:03 <AnMaster> Gregor, still you do have a sense for aesthetics. If you didn't you would be unable to even comment upon dressing without aesthetics!
02:01:32 <AnMaster> Gregor, the best you can manage is thus a limit towards zero
02:02:24 -!- CakeProphet has quit (Ping timeout: 276 seconds).
02:03:29 -!- CakeProphet has joined.
02:04:09 <AnMaster> alise, what I don't get is how it can suggest typesetting a whole book in Baskerville
02:04:50 <AnMaster> okay, it is better than Zapfino or Bodoni by far, Still horrible for long sections of text
02:05:07 <alise> a /good/ digitised baskerville -- a GOOD one, bad ones are by far the majority --
02:05:28 <alise> for instance, Douglas Adams' works I suspect would be superb in a /good/ digitised baskerville.
02:05:47 <alise> http://upload.wikimedia.org/wikipedia/en/1/16/BaskervilleSpec.svg <-- this is a good baskerville
02:05:52 <alise> AnMaster: Umm ... unless you're using metal type.
02:06:01 <alise> Then you're using a digitised font family.
02:06:05 <AnMaster> <alise> http://upload.wikimedia.org/wikipedia/en/1/16/BaskervilleSpec.svg <-- this is a good baskerville <-- this is like unreadable
02:06:12 <alise> You're like stupid.
02:06:20 <AnMaster> alise, no, I'm on like 114 dpi
02:06:32 <Gregor> "Pants goes against the Word of God"
02:06:36 <Gregor> -- http://amazinggracebaptistchurchkjv.com/subpage209.html
02:06:53 <alise> Anyway, /any/ typeface will have a range of digitisations from crap to superb.
02:07:05 <alise> Baskerville is a subtle, old typeface; so it's a lot easier to get wrong.
02:07:10 <alise> And it certainly will never fit some works.
02:07:19 <AnMaster> alise, okay those straight-up lines in the B (you freaks probably have a special name for them) are much better than in the flow chart
02:07:21 <alise> But it's wonderful if tamed.
02:07:39 <alise> You know, I could call you a freak for having a Linux/OpenGenera obsession.
02:07:44 <alise> Gregor: it's true.
02:08:16 <pikhq> alise: My goodness Baskerville is a beaut.
02:08:38 <alise> AnMaster: The Baskerville in http://julianhansen.com/files/infographiclarge_v2.png is bold, by the way.
02:08:41 <alise> Just like all the other samples.
02:08:48 <alise> And, of course, it's crammed into very few pixels.
02:08:56 <alise> So it's understandable that it looks so ugly there.
02:09:12 <AnMaster> alise, that could explains why the more readable alternative to didot is in fact less readable
02:09:17 <alise> pikhq: The italic serifs are awesome.
02:09:52 <AnMaster> alise, it doesn't look ugly there. Just unreadable for long text
02:09:58 <alise> AnMaster: Didot can be readable at very small pixel sizes because the renderer runs out of ways to make the horizontal lines ridiculously thin so it ends up making them more equal; you know, the reasonable thing to do.
02:10:11 <alise> Well, yeah, setting a book in all-bold, low-resolution type would be dumb. :P
02:10:34 <AnMaster> alise, I wouldn't claim Zapfino is ugly. Yet I would claim it is unreadable for long text.
02:11:36 <alise> Zapfino is actually quite ugly though,
02:11:53 <AnMaster> alise, not compared to the alternative
02:12:02 <AnMaster> and I don't think it is excessively ugly
02:12:44 <alise> Gregor: amazinggracebaptistchurchkjv is a joke, right?
02:12:52 <alise> http://www.amazinggracebaptistchurchkjv.com/Download99.html
02:13:03 <Gregor> alise: It's never clear :P
02:13:04 <pikhq> If you want a vaguely calligraphic logo, *hire a bloody calligrapher*.
02:13:18 <pikhq> I guarantee it'll look better than whatever calligraphic font you find.
02:14:08 <AnMaster> alise, what do you think of Myriad?
02:14:28 <alise> AnMaster: It's... Apple Sans. It's a nice font, but...
02:14:34 <alise> You can never, ever use it without someone thinking "oh, Apple".
02:14:57 <alise> Apple use it for everything.
02:15:05 <alise> All the headings on their site, the "iPod" in their adverts, ...
02:15:13 <AnMaster> alise, I'm unable to tell that
02:15:25 <AnMaster> alise, that Eurostile looks nice
02:15:33 <AnMaster> reminds me of startrek logos hehe
02:15:50 <alise> http://upload.wikimedia.org/wikipedia/commons/3/3b/EurostileSpec.svg
02:15:56 <alise> Proletariats are futuristic.
02:16:16 <AnMaster> not quite as much in that image
02:17:24 <AnMaster> Bodoni makes me think of Discworld. Think there was a dwarf with that name in "The Truth" (working at that press)
02:18:03 <AnMaster> alise, what is the type face used for the non-typeface boxes in that flow shart
02:18:09 <AnMaster> like the choices and the middle box
02:18:33 <alise> The answer is in the flowchart.
02:19:01 <alise> "I MUST SAY THAT THIS FLOWCHART IS LOOKING HOT".
02:20:00 <AnMaster> alise, I watched that movie once. Several years ago
02:20:30 <alise> Typeface, dammit, typeface.
02:20:48 <AnMaster> alise, font is shorter to type that typeface
02:21:44 <AnMaster> except I would seem like a typography geek then
02:22:04 <alise> The term is "typophile". ;)
02:52:34 -!- BeholdMyGlory has quit (Remote host closed the connection).
02:56:18 -!- derdon has quit (Ping timeout: 245 seconds).
03:20:38 <alise> "For containers, there is a lot to say. So, I will not say anything for the moment."
03:24:01 -!- Wamanuz2 has quit (Remote host closed the connection).
03:25:08 <nooga> how to set up ssh so that the user connects with a chrooted jail instead of main system hosting sshd?
03:26:54 <Gregor> nooga: Run sshd in the chroot
03:27:02 <nooga> that's what i thought
03:28:11 <nooga> chrooted system runs using a 80MB ramdisk :D
03:29:24 <Gregor> I've been trying to get a Debian chroot on my phone.
03:32:47 -!- Gregor-P has quit (Read error: Connection reset by peer).
03:43:23 -!- nooga has quit (Ping timeout: 245 seconds).
03:50:49 <alise> stallman calls iphone/ipad igroan/ibad
04:00:03 -!- Sgeo has quit (Ping timeout: 245 seconds).
04:08:05 <Gregor> Ooooooh, I may actually have a working Debian chroot 8-D
04:08:17 -!- Gregor has left (?).
04:08:20 -!- Gregor has joined.
04:08:27 <Gregor> Again, ctrl+w != shift+w
04:09:45 <Gregor> What's a good WM for a very constrained screen size?
04:18:50 <Gregor> What's a good WM for a very constrained screen size WITH NO KEYBOARD
04:20:08 <Gregor> I don't want to use Matchbox just because you're all useless :P
04:21:02 <pikhq> On a more serious note: how do you feel about Twm?
04:21:55 <Gregor> I use twm pretty much whenever I can't have XFCE.
04:22:08 <Gregor> But that would suck for a constrained screen.
04:23:04 <pikhq> Hmm. Get a keyboard?
04:23:21 <pikhq> Actually. Are there any mouse-driven tiling WMs?
04:23:24 <pikhq> And if not, why not?
04:23:48 <Gregor> A thing I've checked in the past.
04:23:53 <Gregor> There aren't, and it's bizarre that there aren't.
04:24:04 <pikhq> Gregor, you've got another project!
04:24:14 <pikhq> You really need a patron.
04:24:17 <Gregor> I'm hypothetically adding VFS to Microcosm.
04:24:38 <pikhq> I'd offer, but I suspect I have fewer funds available than you.
04:24:52 <Gregor> I have loads of money, but can a person really have ENOUGH money?
04:25:04 <pikhq> Bill Gates has enough money.
04:26:26 <Gregor> I don't think Bill Gates has sufficient money to shuttle himself to the moon for a summer vacation.
04:27:21 <oerjan> stop there, before you start using diagonalization on the question
04:29:16 <pikhq> Hmm. How much would a moon mission cost?
04:29:30 <pikhq> More than $13 billion?
04:29:59 <Gregor> For a month or so of just hangin' out on the moon?
04:30:09 <pikhq> I think for that he could just about pay to have Apollo rebuilt and launched.
04:30:11 <Gregor> It's not a vacation if you go there, collect some rocks, hit a golf ball, and come back.
04:33:53 <Gregor> You've gotta have some time to relax poolside!
04:38:01 <alise> Sea of Tranquility!
04:38:03 <pikhq> 'know what has enough money?
04:38:10 <pikhq> The US Department of Defense.
04:38:39 <pikhq> We should convince them that there's oil on the moon so we can set up a military base there.
04:39:03 <pikhq> It'll be permanently settled until the Moon ceases to be.
04:49:38 -!- CakeProphet has quit (Ping timeout: 245 seconds).
04:51:11 -!- CakeProphet has joined.
04:56:11 -!- CakeProphet has quit (Ping timeout: 260 seconds).
04:57:00 <alise> http://abstrusegoose.com/strips/oh_but_they_do_give_quite_alotta_fucks.PNG
05:03:38 <pikhq> Hrm. What's with the Chinese signature?
05:05:50 <Gregor> Shockingly, this phone does not make an ideal Debian system ...
05:06:09 <pikhq> Shocking. Positively shocking.
05:15:22 -!- alise has quit (Quit: Leaving).
05:42:18 -!- Deewiant has quit (*.net *.split).
05:42:18 -!- jcp has quit (*.net *.split).
05:42:18 -!- yiyus has quit (*.net *.split).
05:42:19 -!- ineiros has quit (*.net *.split).
05:44:45 -!- yiyus has joined.
05:45:50 -!- ineiros has joined.
05:48:38 -!- jcp has joined.
05:54:34 -!- Deewiant has joined.
06:19:30 -!- oerjan has quit (Quit: leaving).
07:48:35 -!- GreaseMonkey has quit (Remote host closed the connection).
07:55:13 -!- kabatta has joined.
07:55:42 -!- Sgeo has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:32:17 -!- kabatta has left (?).
08:57:57 -!- MigoMipo has joined.
09:08:52 -!- MigoMipo has quit (Remote host closed the connection).
09:14:10 -!- pikhq has quit (Read error: Connection reset by peer).
09:40:47 -!- CakeProphet has joined.
09:59:23 -!- tombom has joined.
10:19:34 -!- atrapado has quit (Quit: Ex-Chat).
10:49:23 -!- MigoMipo has joined.
10:50:49 -!- Phantom_Hoover has joined.
11:00:40 -!- MigoMipo_ has joined.
11:03:45 -!- MigoMipo has quit (Ping timeout: 252 seconds).
11:05:10 -!- augur has quit (Remote host closed the connection).
11:05:19 -!- augur has joined.
11:05:45 -!- MigoMipo_ has quit (Remote host closed the connection).
11:06:05 -!- MigoMipo has joined.
11:29:03 -!- cheater99 has quit (Ping timeout: 240 seconds).
11:39:54 -!- MigoMipo_ has joined.
11:43:40 -!- MigoMipo has quit (Ping timeout: 265 seconds).
11:46:04 -!- cheater99 has joined.
11:50:33 -!- Lymia has joined.
12:01:27 -!- MigoMipo_ has changed nick to MigoMipo.
12:20:13 -!- relet has quit (Quit: Leaving.).
12:25:11 -!- relet has joined.
12:29:35 -!- tombom has quit (Quit: Leaving).
12:41:28 <Phantom_Hoover> Although it's just a specific case of poor website design.
12:44:04 <AnMaster> Phantom_Hoover, but as one who have searched for course lists on my university website a LOT...
12:44:25 <AnMaster> Phantom_Hoover, well they tend to have some "product list" easily available
12:44:40 <AnMaster> which might not be a lot of consumer products
12:44:50 <AnMaster> now, contact info, might be harder
12:45:09 <Phantom_Hoover> I had a horrible time trying to find contact info for Conway a while ago.
12:45:18 <AnMaster> Phantom_Hoover, his game of life?
12:45:29 <Phantom_Hoover> The address Princeton finally gave had been deactivated.
12:46:35 <AnMaster> I want a symbolics keyboard, with the usb mod
12:46:48 <AnMaster> I seen a picture of someone modifying it to connect over usb
12:47:17 -!- nooga has joined.
12:50:11 -!- Quadrescence has quit (Ping timeout: 246 seconds).
12:51:40 -!- Quadrescence has joined.
13:12:47 -!- MigoMipo has quit (Remote host closed the connection).
13:26:09 -!- Quadrescence has quit (Read error: Connection reset by peer).
13:26:13 -!- Quadresce` has joined.
13:28:39 -!- Quadresce` has changed nick to Quadrescence.
13:28:53 -!- Wamanuz has joined.
13:58:43 -!- Flonk has joined.
14:18:44 <Phantom_Hoover> Where did you ask Sgeo that question in the first place?
14:21:05 -!- Lymia has quit (Quit: Lymia).
14:23:19 <Flonk> I only asked about it here on #esoteric
14:25:54 <Flonk> oh, right, I asked if someone could help me with C++ on #IRP, and he redirected me to here
14:35:31 -!- BeholdMyGlory has joined.
14:50:14 <AnMaster> wtf. I used genera a bit too much recently, I wondered why delete didn't work. Delete acts like backspace there and backspace acts like left-arrow
14:50:28 <AnMaster> result: when it didn't work in linux first thought was "wtf, computer is frozen"
15:16:12 <nooga> they were powerful workstations
15:33:52 <AnMaster> does anyone know zzo's web server?
15:34:52 -!- alise has joined.
15:39:44 -!- andrewnth has joined.
15:43:23 -!- Phantom_Hoover has quit (Ping timeout: 245 seconds).
15:45:06 -!- pikhq has joined.
16:03:05 -!- pikhq has quit (Read error: Connection reset by peer).
16:08:13 -!- pikhq has joined.
16:57:14 <alise> http://www.cacaovm.org/ Huh, /another/ free JVM.
17:08:11 <alise> 02:34:38 <Phantom_Hoover> Kurt Gödel looks delightfully evil...
17:08:11 <alise> 02:37:04 <Phantom_Hoover> He would be an ideal candidate for a Halloween costume.
17:08:15 <alise> I'm afraid he's dead.
17:08:33 <alise> 04:18:30 <AnMaster> vrml. How does one view it? For historical computing purposes.
17:08:33 <alise> 04:19:08 <AnMaster> tools available to me are: opengenera, ubuntu 7.10, ubuntu 10.04, current arch linux
17:08:38 <alise> yeah opengenera came with a vrml viewer.
17:19:15 <alise> On Perl 6: "I have never heard of a piece of software whose memory consumption dropped by 96% and startup time by 98% from the first alpha release to the first production release. That's what it would take to match perl5."
17:19:28 <alise> Looks like Rakudo Star is more like Rakudo LAME. :|
17:26:10 <AnMaster> alise, it did? I haven't found it then
17:26:26 <alise> AnMaster: From the FUTURE.
17:26:46 <AnMaster> alise, cl-http server contains vrml examples though
17:26:50 <alise> AnMaster: Hey, could you clone the vm, change the password then put up the files somewhere?
17:26:57 <alise> Rather than making us poor people go through the setup instructions >_>
17:27:52 <AnMaster> alise, I can't, you could steal it then
17:28:11 <alise> Just put a file in it saying "This can only be used if you legally own an OpenGenera license!" That provides you the same legal protection as anything (i.e. barely anything) :P
17:28:36 <AnMaster> alise, anyway I need to redo the VM. I did several things wrong. Oh and I keep the sys.sct directory in hg now
17:28:46 <AnMaster> alise, makes reverting stupid things easier
17:28:59 <AnMaster> needed it a few times when I messed up
17:29:15 <alise> I feel like using QEMU will be better than VirtualBox. VirtualBox is for emulating more conventional stuff without much ... you know, techy tweaking junk.
17:29:26 <alise> I'd say Bochs, but Bochs emulates at like half a Hz.
17:29:36 <alise> AnMaster: More realistic, for OpenGenera :-)
17:29:48 <pikhq> alise: Bochs is faster than *that*...
17:30:01 <AnMaster> alise, anyway I have no clue if this VM will work on OSE virtualbox
17:30:02 <pikhq> Still, it's not exactly speedy.
17:30:09 <AnMaster> alise, oh and I have nowhere I could put 2.4 GB
17:30:09 <alise> pikhq: Sorry; three quarters of a hz.
17:30:15 <AnMaster> which is what df -h in the VM reports
17:30:19 <alise> Whence the 2.4 GiB?
17:30:25 <alise> AnMaster: inaccurate
17:30:29 <alise> VirtualBox lazily allocates disk
17:30:36 <alise> most of that won't actually be on your disk, I bet.
17:30:41 <AnMaster> alise, correct. The actual image is larger
17:30:42 <alise> Unless you mean used.
17:30:58 <alise> Oh well. Link me to your guide -- is it the Moast Korrekt version?
17:31:20 <AnMaster> alise, perhaps. I haven't documented getting hardcopy with cups-pdf though.
17:31:27 <AnMaster> alise, reason is that it only works partly
17:31:34 <AnMaster> like, you must disable cover page
17:31:35 <alise> I don't care about that.
17:32:04 <alise> Incidentally, this is the craziest chain of emulation ever: VM running x86 Ubuntu running a Tru64 Unix on Alpha emulator running Genera.
17:32:08 <AnMaster> alise, while http://sprunge.us/fAfd is correct afaik, there could be errors
17:32:13 <alise> And Genera is an operating system in and of itself.
17:32:16 <AnMaster> and I have not proof read it much
17:32:28 <alise> That's alright, I can sling it.
17:33:04 <alise> Anyway, I think DKS is such a rabid salesman that he can't get angry at people downloading it.
17:33:17 <alise> dkschmidt at 2007-08-15 02:24 CET:
17:33:17 <alise> Congratulations on downloading the finest software development environment ever created. If you want to find out more about Genera or would like to have a Symbolics Lisp Machine, check out the Symbolics website at www.symbolics.com or contact sales@symbolics.com.
17:33:19 <alise> -- comment on the torrent
17:33:33 <alise> Just another opportunity for a sale1
17:33:41 <alise> I think he realises nobody's going to buy Open Genera anyway.
17:33:47 <AnMaster> alise, oh and to load cl-http, when everything else in there is done do: Load System CL-HTTP
17:34:01 <AnMaster> you will get 500 Internal Error after that
17:34:04 <alise> AnMaster: *splurts drink everywhere*
17:34:05 <alise> The VLM is a product for DEC Alpha / Tru64 Unix. There are experimental VLM versions for Linux/x64 and, newer, Mac OS X (x64).
17:34:15 <alise> Oh my god oh my god oh my god.
17:34:19 <alise> there's a native 64-bit linux version
17:34:31 <AnMaster> alise, um. that Linux/x64 is the one I used
17:34:31 <alise> andrewnth: http://www.reddit.com/r/lisp/comments/c8sjz/what_are_your_favorite_and_mostused_cl_projects/c0qxfjv
17:34:41 <alise> you're using the tru64 version
17:34:44 <alise> andrewnth: change your nick :|
17:34:52 <AnMaster> alise, it is running in a x86-64 VM though
17:34:52 <alise> AnMaster: emulated on linux/64
17:34:58 <alise> AnMaster: yes, but it's another layer of emulation
17:35:03 -!- andrewnth has changed nick to andynth.
17:35:03 <alise> this VLM is a native emulator for 64-bit linux
17:35:09 <alise> you could just run it in a chroot
17:35:16 <alise> andynth: it's more the starting letters that are bothersome :D
17:35:29 <alise> last-spoke order for nick completion now
17:35:43 <alise> AnMaster: so actually my chain of emulation was wrong, if it's Virtual Lisp Machine (VLM)
17:36:01 <alise> Incidentally, this is the craziest chain of emulation ever: VM running x86 Ubuntu running a Tru64 Unix on Alpha emulator running Genera on a Symbolics Lisp Machine.
17:36:14 <alise> Open Genera doesn't run natively on Tru64 Unix; it emulates a Lisp Machine. VLM, Virtual Lisp Machine.
17:36:19 <alise> "The VLM is a product for DEC Alpha / Tru64 Unix."
17:36:22 <alise> AnMaster: Er, right.
17:36:44 <AnMaster> alise, besides I don't know if the source to snap4 is public
17:36:47 <alise> Incidentally, this is the craziest chain of emulation ever: VM running x86-64 Ubuntu running a Tru64 Unix on Alpha emulator running Genera on a Symbolics Lisp Machine.
17:37:02 <alise> AnMaster: My point is: You're running /three layers of emulation/.
17:37:20 <alise> No, we thought it ran Genera natively on Tru64.
17:37:24 <alise> Not emulating a Lisp Machine.
17:37:38 <AnMaster> what I did think was that we skipped the Tru64 step
17:37:38 <alise> Okay, you're misunderstanding.
17:37:58 <alise> I don't /think/ we do.
17:38:02 <AnMaster> alise, I read open genera docs. It mentions emulation
17:38:09 <alise> ...Anyway, with the x86-64 Linux VLM we could just run it in a chroot.
17:38:13 <alise> For, you know, moar speed.
17:38:22 <AnMaster> alise, but you _don't_ wan't to mess up your system with NFS and NIS
17:38:23 <alise> And better mouse/keyboard stuffs.
17:38:37 <alise> AnMaster: Just serve it on a random port on just 127.0.0.1.
17:38:37 <AnMaster> but sure, skipping that would be nice
17:38:44 <AnMaster> alise, do you have a url though?
17:39:24 <alise> "For example in the snap4 version some arithmetic is broken and GC plus I/O often leads to crashes. Brad has fixed the complex arithmetic, but the error on the Genera side with GC has not been fixed yet." --lispm
17:39:31 <alise> AnMaster: It's commercial software.
17:39:35 <alise> You buy it. From Symbolics.
17:39:39 <alise> An experimental release.
17:39:42 <AnMaster> I know that the arithmetic is broken in part
17:39:50 <alise> They might give it away for a reduced price, of course, since it's an experimental release.
17:40:01 <alise> If you're REALLY smooth with the ladies^W^W^W you might be able to get it free.
17:40:04 <AnMaster> alise, I would like a link to the snap5 then
17:40:16 <alise> AnMaster: Contact D.K.Schmidt.
17:40:19 <alise> You'll probably get a quote.
17:40:36 <alise> sales@symbolics-dks.com
17:40:43 <alise> the contact is D. K. Schmidt
17:40:47 <alise> he handles all communication
17:40:56 <AnMaster> alise, I was pretty sure it was a third party project
17:40:57 <alise> "Send bug reports to Kalman Reti at reti@symbolics-dks.com" -- but it's a known, fixed bug, so.
17:41:04 <alise> AnMaster: It's a Symbolics project.
17:41:15 <AnMaster> alise, they are a god damn shell company!
17:41:27 <alise> AnMaster: Well, they still develop.
17:41:33 <AnMaster> alise, anyway, if you buy it, send me a copy
17:41:42 <alise> Do you have a license?
17:42:06 <alise> No, but you're depriving me. :)
17:42:40 <alise> In which Stanislav uses Open Genera on a real Symbolics Tru64 Alpha machine: http://www.loper-os.org/?p=186.
17:42:46 <AnMaster> alise, I have nowhere to put 2.4 GB
17:43:05 <AnMaster> reverse engineering the whole thing would take less time!
17:43:32 <alise> Error 503 Service Unavailable
17:43:32 <alise> Service Unavailable
17:43:34 -!- Mathnerd314 has joined.
17:43:40 <alise> AnMaster: At least seed the torrent :P
17:43:54 <AnMaster> alise, will take a few minutes
17:44:10 <AnMaster> I guess I could somehow import the file into ktorrent
17:44:15 <alise> Last time I downloaded it it ran at like 1 KiB/s.
17:44:29 <alise> AnMaster: Yeah: download the .torrent file, tell it to download where the file already is.
17:44:33 <alise> ( http://thepiratebay.org/torrent/3769989/Symbolics_Open_Genera_2.0_for_Alpha_-_complete_package_with_Lisp )
17:44:33 <AnMaster> alise, /msg me the url to the torrent
17:44:44 <alise> Well, I did /msg you; just a lot of other people, too.
17:46:00 <alise> AnMaster: I don't suppose you're the peer I'm downloading from at 600 KiB/s, are you.
17:46:25 <alise> AnMaster: scps what? o_O
17:46:49 <AnMaster> alise, opengenera2.tar.bz2 from laptop to desktop
17:47:03 <AnMaster> since I don't have port forwarding set up on laptop
17:47:35 * AnMaster looks at the kick peer and ban peer options
17:47:41 <AnMaster> alise, I always wondered what those did
17:48:01 <AnMaster> alise, oh I expect you to seed this to a ratio of 2 then
17:48:25 <alise> I'm sure collectively, all the times I start Transmission and then take a second to pause it, it will eventually add up to 2. >_>
17:48:34 <alise> I am the reason democracy doesn't work!
17:48:39 <alise> No, really, I will seed it.
17:48:52 <alise> You are looking at the wrong guy.
17:49:01 <alise> Almost 800 KiB/s download. Pretty nice.
17:49:06 <alise> Oop, went down there.
17:49:08 <AnMaster> alise, some very very fast other peer then
17:49:19 <alise> Yeah; I'm getting 400-600 KiB/s from them.
17:49:31 <alise> Someone else using Transmission. 18.238.1.85.
17:49:48 <AnMaster> could be D, K. Smitch (sp?) himself? ;P
17:50:03 <AnMaster> 85.1.238.18.in-addr.arpa domain name pointer UNCLE-ENZO.MIT.EDU.
17:50:31 <alise> AnMaster: Host Name: UNCLE-ENZO.MIT.EDU
17:50:40 <alise> So it's that legendary university internet.
17:50:47 <alise> I'm seeding to 0 people. Blame them.
17:50:55 <AnMaster> alise, well keep it ready there
17:51:00 <AnMaster> I seeded this when people hadn't got it
17:51:51 <alise> Tell you what, I'll seed aliseOS to you when it's out. :P
17:52:39 <alise> AnMaster: Nobody likes me enough to take my seed.
17:52:46 <AnMaster> alise, I have a better idea, since it will take less time
17:52:47 <alise> Uh... that was an awkward way to say that.
17:53:01 <AnMaster> alise, seed DNF when it is out
17:53:13 <alise> Or Chinese Democracy! Wait... that /did/ come out.
17:53:32 <alise> I mean the album. :P
17:55:05 <alise> Guns N' Roses released Use Your Illusion I & II in 1991, then toured for something like a few billion years, then finally released Chinese Democracy, promised since 1992, at the end of 2008.
17:55:13 <alise> It's more about the ridiculously long wait and anticipation than the music.
17:55:55 <alise> It took longer than DNF took to die; so, yeah.
17:56:01 <alise> Whoa, apparently someone picked up the DNF torch.
17:56:07 <alise> Developer(s) 3D Realms (1997–)
17:56:07 <alise> Unknown external developer (2009–)
17:57:23 <alise> http://nedroid.com/comics/2010-07-26-bonustholomew.gif
18:04:56 <Gregor> This spam has introduced me to my new favorite word ever:
18:06:08 <alise> Is that a ... lesbian ... man?
18:06:52 <Gregor> But it's my favorite word ever.
18:07:29 -!- Phantom_Hoover has joined.
18:07:36 <alise> AnMaster: Why doesn't it work with newer Ubuntus, again?
18:10:18 -!- Phantom_Hoover has left (?).
18:10:25 -!- Phantom_Hoover has joined.
18:11:46 <alise> Phantom_Hoover: Weird or AWESOME?
18:13:33 <Phantom_Hoover> The Chronojournal has been spammed so much it's not even funny.
18:14:29 <AnMaster> alise, I tried the first 8.x version
18:14:31 <alise> AnMaster: I am an orphan and consequently unable to read )(
18:14:47 <AnMaster> alise, well then what I write here has no use either.
18:14:55 <alise> My computer reads words to me
18:14:55 <AnMaster> since you can't read IRC either
18:15:03 <alise> In other places there is no talking
18:15:07 <alise> All I see are pictures. Sometimes they are nice
18:39:31 -!- oerjan has joined.
18:40:13 <alise> I wonder how much RAM Open Genera needs.
18:42:25 <Phantom_Hoover> alise, does anyone else think writing a Lisp OS that can run on modern hardware is a sensible idea?
18:42:37 <alise> You mean apart from me? Or apart from you?
18:42:55 <oerjan> 09:08:11 <alise> 02:37:04 <Phantom_Hoover> He would be an ideal candidate for a Halloween costume.
18:42:59 <oerjan> 09:08:15 <alise> I'm afraid he's dead.
18:43:05 <alise> I think it's a good idea -- well, not Lisp, but the same idea as Lisp Machine O-- anyway, yes; the Loper OS guy [warning: controversial half-crank, but clever]; me; ...
18:43:06 <oerjan> that is not _necessary_ a disadvantage.
18:43:42 <alise> Phantom_Hoover: make sure it can persist lisp objects rather than a silly filesystem!
18:44:12 <alise> now how much ram does open genera need
18:44:50 <alise> I'll give the vm the default
18:44:54 <alise> Phantom_Hoover: much less
18:44:56 <alise> Phantom_Hoover: 80s hardware
18:45:05 <alise> but it has to run ubuntu, running an Alpha emulator, running a Lisp Machine emulator, running Genera
18:45:07 <alise> so there is overhead
18:45:11 <alise> but 384 should be plenty. absolutely plenty.
18:45:29 <pikhq> alise: That's a lot of layers.
18:45:32 <alise> "You should have a 300 mHz or greater Alpha workstation with at least
18:45:32 <alise> 500 MB RAM, 4 MB cache and 1 GB of available disk space."
18:45:37 <alise> I'll give it 512 MiB.
18:45:50 <alise> This is probably excessive, but dammit, I have over 3 GiB free.
18:46:15 <alise> Failed to create a new virtual machine Open Genera.
18:46:15 <alise> Runtime error: -102 (File not found.).
18:46:37 <alise> Oh, it uninstalled VirtualBox.
18:47:48 <alise> Phantom_Hoover: Anyway, if you do make a Lisp OS: You are awesome. If you make a Lisp OS that, instead of a filesystem, automatically persists the Lisp objects in the system with a certain marker on them -- say, with a "persist" flag set -- to disk, and automatically reloads them on bootup (note: this is /easier/ than implementing a filesystem) -- you are quadruply awesome and should breed immediately.
18:48:32 <Phantom_Hoover> alise, it's a lovely idea, but I enjoy programming, and as such am a terrible developer.
18:48:48 <alise> Yeah, x86-64 is a drag.
18:49:01 <alise> Phantom_Hoover: The main issue is the Lisp implementation.
18:49:11 <alise> You don't really want to write it in C because then you get all the baggage of, you know, C, which is sort of against the point.
18:49:21 <alise> But writing a JIT compiler in assembly? [suicide.]
18:49:40 <pikhq> JIT... Assembly... (seppuku)
18:49:40 <alise> Or... LowLevelLisp macro DSL in whatever Lisp you are implementing!
18:50:08 <nooga> writing jit compiler in assembly is not so hard
18:50:28 <alise> Yeah, but... not for us high-level guys.
18:50:30 <nooga> esepcially if you're targetting simple processors without MMU and other useless shit
18:50:34 <alise> Whoa, VirtualBox supports EFI.
18:50:35 <Phantom_Hoover> alise, use an existing CL (or Scheme) compiler and modify it?
18:50:46 <alise> Phantom_Hoover: Those will be heavily existing-OS-oriented.
18:51:02 <alise> Phantom_Hoover: You could base it on Movitz, but nobody likes Movitz.
18:51:54 <nooga> http://fc04.deviantart.net/fs70/f/2010/160/c/5/Talk_to_Me_1_by_humon.jpg :D
18:53:08 <alise> Phantom_Hoover: No point to support it for a new OS.
18:53:25 <alise> Phantom_Hoover: 64-bit is very helpful because you can map the entire object space to address space.
18:53:37 <alise> So, say, some large 64-bit pointer will actually be a serialised object on disk, but of course this is transparently handled.
18:53:49 <alise> That way you don't have to actually directly deal with mapping things; it can be transparent, in the paging.
18:54:01 <alise> Phantom_Hoover: Also, it means that you have more space for type tags.
18:54:34 <alise> So if you identify, say, 10 types in a tag, that's four bits, so you still get 60 bits.
18:54:44 <alise> And today's CPUs don't even /use/ all 60 bits to look up memory.
18:54:57 <alise> Although IIRC there's some restriction on the higher bits; pikhq?
18:55:00 <alise> But you can always use the lower bits.
18:55:05 <alise> Phantom_Hoover: No, I don't mean just pass that pointer.
18:55:12 <alise> Obviously if it's a pointer, you >> 4.
18:55:23 <alise> So you tag the lower bits.
18:55:45 <Phantom_Hoover> alise, an instruction for each memory access seems wasteful.
18:56:30 <alise> Phantom_Hoover: Nah, not really.
18:56:38 <alise> Phantom_Hoover: Memory access is very slow compared to CPU anyway.
18:56:40 <Phantom_Hoover> alise, OTOH there's CCL's strategy, which is to align all objects.
18:56:43 <alise> So the overhead is unnoticeable.
18:56:51 <alise> Phantom_Hoover: That also works, yes, but you probably want more than just integers vs objects.
18:57:10 <alise> Phantom_Hoover: Booleans, characters, symbols... identifying these in the pointer is useful.
18:57:18 <alise> Especially type of number, since that makes polymorphic arithmetic a lot quicker.
18:57:23 <alise> Phantom_Hoover: Ah, so it aligns it even more.
18:57:47 <alise> Phantom_Hoover: That's 3 bits, so...
18:58:03 -!- Flonk_ has joined.
18:58:10 <alise> Phantom_Hoover: That would work, yes.
18:58:15 <alise> Phantom_Hoover: Probably the best idea.
18:58:23 <nooga> are there some papers about machine lisp?
18:58:43 <nooga> why not implement machine lisp on some spartan core
18:59:16 <alise> machine lisp on lisp machine core.
18:59:38 -!- Flonk has quit (Ping timeout: 265 seconds).
18:59:43 -!- Flonk_ has changed nick to Flonk.
19:00:18 <nooga> lisp machine on FPGA
19:00:35 <alise> fpga is slow though :D
19:00:43 <alise> Phantom_Hoover: do you have project ADHD?
19:01:00 <alise> Phantom_Hoover: dammit.
19:01:03 <alise> I almost got excited.
19:01:22 <nooga> probably emulating lisp machine on x86 is slow as well
19:01:29 <Phantom_Hoover> alise, OTOH I've never actually done a project with anyone else's knowledge.
19:02:02 <alise> I guess I can help with everything that isn't actually writing assembly.
19:02:09 <alise> "Organising an object model in memory" is my idea of fun.
19:04:29 <alise> Phantom_Hoover: Of course.
19:04:46 <alise> Phantom_Hoover: Basically, every allocated object gets a unique position in address space. Obviously. The thing is, this does not correlate to physical memory.
19:04:54 <alise> You can get addresses in the very high numbers, theoretically.
19:05:03 <alise> In the Lisp, in the object system underlying everything or whatever we have,
19:05:08 <alise> Every object has a flag -- "persist".
19:05:17 <alise> If it is set, then every N intervals,
19:05:23 <alise> It is saved to disk with its address.
19:05:31 <alise> Whenever an object is used, it is loaded into RAM (or at least, if it can all be).
19:05:44 <alise> Objects that have not been used for a while give way to new objects to be loaded.
19:05:52 <alise> Tada: We remove the filesystem, replace it with orthogonal object persistence, AND
19:05:54 <alise> objects never change addresses
19:06:10 <alise> Also, you can't lose data more than N intervals old if it's set to persist; it will reappear on the next boot.
19:06:17 -!- nooga has quit (Ping timeout: 260 seconds).
19:06:28 <alise> So: You never have to touch the disk. All you ever do is set "persistent" and never have to worry.
19:06:35 <alise> Some reinitialisation to do with sockets, yes, but that's easy enough to handle.
19:07:26 <Phantom_Hoover> alise, if N is high you'd presumably want to be able to force storage.
19:07:38 <alise> N would not be high.
19:07:49 <alise> We'd keep a list of "dirty" (changed) objects, so the writes would never be large.
19:08:01 <alise> So we can do them, say, every .5 seconds.
19:08:09 <alise> It could be configurable from inside, I guess.
19:08:37 <alise> This is called "orthogonal" persistence. It is quite a well-known thing and has had a bit of research done.
19:09:14 <alise> Phantom_Hoover: In fact, I even suspect you could do deduplicative storage and thereby /version every change to a persisted object/. But that's much more theoretical, and harder.
19:09:22 <alise> Fragmentation on disk?
19:09:57 <Phantom_Hoover> When an object is GCed, presumably you remove it from the disc?
19:09:59 <alise> It's not a huge problem any more, what with the speed of disks and the rising popularity of SSDs. While this is a toy, of course it will never make a difference; when it becomes practical... well, that'll take so long we'll all be using SSDs anyway.
19:10:09 <alise> Phantom_Hoover: Yes, but an object with the persist flag set is never GC'd.
19:10:27 <alise> We don't serialise every old object (you can almost do this automatically, but again that's much more theoretical)
19:10:42 <pikhq> One could presumably design the OS based around having both an SSD and a platter.
19:10:54 <alise> Yeah, but ... I'd just treat disk seeking as instant.
19:11:12 <alise> Remember, /all/ your RAM will be used to store all the objects it can.
19:11:28 <alise> Sure, it won't be totally optimal, but once SSDs become very popular (soon) it'll be just fine.
19:11:34 <alise> In fact, scheduling and the like on SSDs only hurts performance.
19:11:39 <alise> And since this would initially just be a toy, well...
19:12:10 <alise> Phantom_Hoover: But are you going to do your everyday computing on this soon?
19:12:26 <alise> Phantom_Hoover: Aspire to own an SSD, then.
19:12:37 <alise> They'll be affordable in a few years, when this thing might actually be able to edit its own code. :P
19:13:07 -!- pikhq has quit (Quit: leaving).
19:13:34 <alise> You /could/ use GRUB... but if we're even not doing C, hell; just write our own.
19:13:38 <alise> We'll have to switch into long mode anyway.
19:13:41 <Phantom_Hoover> Wait, make partitions and the bootloader lisp objects!
19:13:44 <alise> It's not hard to write a bootloader.
19:13:47 <alise> Phantom_Hoover: That makes no sense, good sir.
19:15:28 <alise> What; my processor doesn't do virtualisation.
19:15:38 <alise> Phantom_Hoover: So does banana-flavoured cheese.
19:15:49 <alise> So I can't do 64-bit with VirtualBox. Bet QEMU can do it.
19:16:18 <Phantom_Hoover> Using VB for writing bootloaders and such doesn't seem like such a good idea.
19:16:44 <alise> VirtualBox you mean?
19:16:47 <alise> Oh, I'm just trying Open Genera.
19:17:16 <alise> Phantom_Hoover: For actual OS development you usually want Bochs.
19:17:17 <Phantom_Hoover> Of course I mean VirtualBox. Writing bootloaders in Visual Basic is so stupid it's almost cool.
19:17:26 <alise> Slow as molasses, but you can do debug output and even some debugging (this is hellish).
19:17:43 <alise> So Bochs when shit is fucked up, QEMU when shit is cool.
19:22:22 <alise> Yeah. Wait ... are we coding /now/?
19:22:32 <alise> Hold up there boy, there's an awful lot of planning and procrastinating to do first! :D
19:23:31 -!- oerjan has quit (Quit: Later).
19:24:07 <alise> Of course to the first question or the second statement?
19:24:51 <alise> Phantom_Hoover: We should probably decide on the actual Lisp. Proper Common Lisp would take 50 years to implement all the subtleties of Roman numeral formatting and LOOP and whatnot, so I'd suggest that that's out. But Scheme is ... not ideal for this, obviously.
19:24:59 <alise> Therefore I propose the obviously superior solution, that of designing our own dialect.
19:25:15 <alise> Worked for the Symbolics guys!
19:25:19 <alise> Phantom_Hoover: ABSOLUTELY FUCKING NOT.
19:25:33 <alise> What say you to yasm?
19:25:42 <alise> It's just like nasm, except BSD-licensed. :P
19:26:44 <alise> You're like me when I even had the necessary amount of self-delusion to convince myself to start big projects! And I LIKE that. :P
19:26:46 <Phantom_Hoover> So long as I don't have to grapple with AT&T indirection, I'm good.
19:27:48 <alise> So, presumably you know assembly; that's handy, as I ... don't, not more than the absolute basics.
19:33:45 <alise> Phantom_Hoover: Eh?
19:38:02 <alise> Nothing much; immutable conses, there's always a popular one. :P
19:40:24 <Phantom_Hoover> Immutable conses seem a bit weird for a low-level system.
19:41:38 <alise> Phantom_Hoover: Perhaps, but actually mutating conses basically never comes up in code. Not in well-designed code, anyway.
19:41:44 <alise> And if you're iterating over that cons... /ouch/
19:41:50 <alise> Well, over the list with that cons in it.
19:42:02 <Phantom_Hoover> alise, true, but functions like nreverse can be useful.
19:42:26 <alise> Reversing a linked list is slow.
19:42:45 <alise> Perhaps you are correct.
19:43:11 <alise> Phantom_Hoover: Erm, that's pretty much what you have to do.
19:43:14 <alise> Consider the structure of the cons cells.
19:43:34 <alise> First you have to iterate through all of them, to find the last element; save the car of the first node, set the car of the first node to the last one.
19:43:42 <alise> You end up copying all the values and changing the cars of all of them.
19:43:51 <alise> This won't be significantly faster than just creating new conses.
19:44:10 <alise> Besides, if we prioritised performance over safety...
19:44:21 <alise> Phantom_Hoover: Ah, like that. That'd be even slower then, I think.
19:44:50 -!- zzo38 has joined.
19:44:55 <alise> Consing is fast, you know. :)
19:45:54 <alise> Maybe you are right.
19:46:12 <alise> But if we are not careful we could end up with "C++ -- the S-Expression edition".
19:46:19 <alise> Phantom_Hoover: Racket does, by default.
19:46:31 <alise> (Previously, PLT "I Can't Believe It's Not R5RS" Scheme.)
19:46:42 <alise> Apparently this broke very few programs in the wild.
19:46:48 <oklopol> umm, alise, reversing a list by reversing the cdr's basically means copying integers around
19:47:02 <alise> Of course, they just had to change it to m(cons|car|cdr|set-car!|set-cdr!).
19:47:12 <alise> oklopol: well, i never said it would be slow
19:47:15 <alise> just slower than the car method
19:47:19 <alise> unless i'm misunderstanding
19:47:24 <oklopol> "<alise> Phantom_Hoover: Ah, like that. That'd be even slower then, I think."
19:47:34 <alise> than the car method
19:47:53 <alise> Phantom_Hoover: anyway you can easily have a mutable cons type too
19:48:13 <oklopol> (more local to switch cdr's than car's)
19:48:14 <zzo38> Does yasm support 888ASM style syntax?
19:48:16 <Phantom_Hoover> alise, I have a vague recollection that that's the defined behaviour...
19:48:17 <alise> Phantom_Hoover: obviously we should have some CLOS-style object system underpinning everything, if we want every object (well, apart from things like integers) to have a persist flag (and because it's generally useful)
19:48:20 <alise> multi-dispatch, definitely.
19:48:41 <alise> zzo38: probably not; they're not psychic, so they can't instantly know of and implement your new projects.
19:48:45 <alise> unless you've told them?
19:49:04 <Phantom_Hoover> I thought '(1 2 3) was technically immutable, while (list 1 2 3) is guaranteed mutable.
19:49:12 <zzo38> alise: Of course I know they are not psychic
19:49:34 <zzo38> But I mean, if the way the program works is sufficient to make 888ASM style syntax work
19:49:48 <alise> zzo38: I imagine if you modified the code...
19:49:58 <alise> zzo38: Or: defined a bunch of macros.
19:50:23 <alise> Phantom_Hoover: In Scheme? I don't know.
19:50:31 <alise> Let's not worry about what They do, anyway. :P
19:50:43 <alise> Phantom_Hoover: A major blocker for this project will be the UI... because UIs suck >_>
19:51:36 <zzo38> Phantom_Hoover: Do you need a UI?
19:52:14 <Phantom_Hoover> zzo38, well, in a technical sense, yes. It's the "G" that sucks.
19:53:11 <AnMaster> hm, what was the relationship between quines and TCness. I seem to remember there was one.
19:53:12 <zzo38> Are you trying to make a Lisp machine?
19:53:23 <alise> Phantom_Hoover: We need /some/ interface.
19:53:36 <alise> AnMaster: all TC machines with sufficient output capabilities have quines (fixed-point theorem for turing machines)
19:53:45 <alise> Phantom_Hoover: A UI is an interface, not just a graphical one.
19:53:54 <alise> Phantom_Hoover: Anyway, I'm sure we can come up with something.
19:53:59 <alise> That's hardly a current problem, anyway.
19:54:14 <alise> Phantom_Hoover: I do suggest we use graphical capabilities, even for a linear interface like a command line. Because that lets us do things like images and stuff easily.
19:54:20 <alise> But that's no fun!
19:54:25 <alise> We want little persisted objects on the screen.
19:54:36 <zzo38> I don't know about Lisp machine, but for Forth it is easy to just make it display "ok" and allow any commands to be typed in at the "ok" prompt. For Lisp, I don't know, but I suppose you can do something
19:55:04 <alise> Phantom_Hoover: Everything is a Lisp object, dammit!
19:55:23 <alise> Even C has the screen as a C "object" (pointer).
19:55:42 <Phantom_Hoover> alise, I thought the screen was done through port I/O.
19:56:08 <alise> It's done via memory (controlled by the video card).
19:56:15 <alise> Okay, you can do fancy stuff by talking to the video card; but pixel blitting is memory.
19:56:44 <alise> Talking to the card?
19:56:50 <alise> Mostly unneeded unless you're doing complicated graphics.
19:56:58 <alise> Well, it's mapped from the card.
19:57:30 <alise> Phantom_Hoover: I was about to say "we don't even have to support USB", but the PS/2 emulation is done by the BIOS and you can't talk to the BIOS in long mode; plus it's so slow why would you want to?.
19:57:53 <alise> And VMs can do PS/2 mice and keyboards to start with :P
20:01:30 <alise> VGA text mode is done with mmapping.
20:01:34 <alise> So is video mode in everything.
20:01:39 <alise> (Simple blitting video mode.)
20:01:49 <alise> Here, delegate the UI work to me, I'll enjoy it :P
20:04:10 <zzo38> Text mode just uses put directly in memory, the ASCII in one byte and the next one is the colors
20:04:25 <zzo38> Usually it is CP437 ASCII
20:04:27 <Phantom_Hoover> For some reason, finding decent docs for low-level display code is difficult.
20:04:36 <alise> zzo38: You can change the font, though.
20:04:37 -!- CakeProphet has quit (Ping timeout: 260 seconds).
20:04:42 <alise> Phantom_Hoover: It's all in OSdev material.
20:05:04 <alise> But really, text mode is not worth bothering with after you get stuff running, apart from debug; actually doing stuff with text mode is a bitch with all the wrapping and scrolling and everything.
20:05:06 <zzo38> Phantom_Hoover: I know how the memory is laid out for text mode and for CGA modes
20:05:15 -!- Vegabondmx has joined.
20:05:18 <alise> Besides, you can do more fun, object-y Lisp UIs with graphical stuff. :P
20:05:29 <alise> Note: I do not like WIMP GUIs.
20:05:30 <Phantom_Hoover> This is probably because MediaWiki's search function sucks.
20:05:48 <alise> But an automatically tiling (but mouse controlled layout; like Plan 9's acme, if you've used it) object display?
20:05:53 <zzo38> You can also use the BIOS to write text to the screen
20:06:02 <alise> Object display: Literally, "windows" (frames) are objects with a view.
20:06:12 <alise> You can use an Emacs-style buffer switcher "on steroids", too.
20:06:21 <alise> ...also some floating windows, for displaying some objects.
20:06:33 -!- CakeProphet has joined.
20:06:57 <alise> Phantom_Hoover: Of course, my UI ideas are not very fleshed out because I'm a hopeless visionary.
20:07:13 <alise> But since this project has ... slightly lower ambitions ... than aliseOS, I should be able to cook up something awesome.
20:07:54 <alise> Phantom_Hoover: It's... uh...
20:08:03 <alise> Phantom_Hoover: You know Feather?
20:08:13 <alise> You know the stage where ais523 is with Feather, the bit he can explain?
20:08:17 <alise> Well, I haven't even got there yet.
20:08:19 <alise> Phantom_Hoover: Of.
20:08:34 <Phantom_Hoover> I know vaguely of it, but I don't know where he got stuck.
20:09:15 <zzo38> I also have some ideas and when I get a new computer and write Linux distribution I will write a window manager with my own ideas
20:09:18 <alise> Basically, aliseOS will make you toast, take over the internet so you have some more computing power, make your mind explode every time you use it, and sexually gratify your dog*.
20:09:29 <alise> *Advanced edition only.
20:09:47 <AnMaster> <alise> AnMaster: all TC machines with sufficient output capabilities have quines (fixed-point theorem for turing machines)
20:10:01 <alise> aliseOS is limit of OS as awesome goes to infinity
20:10:46 <zzo38> But I don't have a dog
20:10:57 <alise> zzo38: It will buy you a dog first.
20:11:28 <zzo38> But I don't need a dog
20:11:35 <AnMaster> alise, what about this one: The source code consists of the letter "b" followed by a bf program. The "b" is mandatory. When executed the implementation begins by outputting the letter c as the first char.
20:11:36 <alise> Yes, but the dog will be cuddly, you see.
20:11:48 <alise> AnMaster: again, this does not have the necessary output capabilities
20:11:59 <alise> which have been quantified i think but i don't think we've done more than discuss them
20:12:06 <alise> i'd like to quantify them at some point but haven't yet
20:12:11 <alise> david madore has written about this AnMaster
20:12:12 <AnMaster> alise, can you specify what exactly "sufficient output capabilities" is
20:12:21 <alise> see what i just said.
20:12:27 <zzo38> But I don't want a dog
20:12:32 <zzo38> What I want is the source-codes
20:12:44 <alise> http://www.madore.org/~david/computers/quine.html
20:12:51 <alise> AnMaster: this has a detailed explanation of the fixed point theorem
20:12:57 <alise> grep /The fixed-point theorem/
20:13:41 <alise> "Yow! I've just lost the SOURCE CODE for all my QUINE PROGRAMS! What will I DO NOW with just the BINARIES?" -- David Madore
20:13:54 <zzo38> alise: Yes I have seen that quotation
20:14:01 <zzo38> I copied it into FORTUNE file, even
20:14:33 <Phantom_Hoover> alise, how does one access persistent objects after a reboot?
20:15:05 <alise> Phantom_Hoover: Um. The system automatically restores the state of the previous session.
20:15:08 <alise> Because the objects are persistent.
20:15:16 <zzo38> Maybe you have to store all memory state in hard-drive
20:15:17 <alise> Like, the window manager will have the "open objects" persisted.
20:15:24 <zzo38> And then you can restore exactly the state of before
20:15:25 <alise> It will try to show them on screen, which will cause them to be loaded from disk.
20:15:29 <alise> etc., etc., etc. recursively.
20:15:40 <Phantom_Hoover> alise, so shutting down is the same as hibernating on lesser OSes?
20:15:55 <zzo38> Except that you might have to reinitialize some things of other hardware
20:16:09 <alise> Phantom_Hoover: Yes, except... the RAM is written every .5 seconds while running, not all at once.
20:16:24 <alise> Phantom_Hoover: Also, this means that if you trip over your power cord, you never miss more than N intervals (e.g. .5 seconds) of work.
20:16:26 <alise> Which is, you know, good.
20:17:04 <Phantom_Hoover> We will be bothering with multiple processes, won't we?
20:17:27 -!- CakeProphet has quit (Ping timeout: 260 seconds).
20:19:15 -!- CakeProphet has joined.
20:19:39 <alise> Well, there are no processes, really, if everything is just living objects.
20:20:17 <Phantom_Hoover> OK, but what if you want to interleave two things at once?
20:20:20 <alise> You look at objects on the screen, you modify them, pass them to other objects: where is the process? there are long-running background things, like say in "threads"; there are the object's paintings on the screen, which they are told to do by the UI...
20:20:28 <alise> there are their reactions to the buttons you press.
20:20:46 <alise> but "process" is a vague concept, both for merely object display/mutation /and/ long-running computation: and we have no need of it.
20:21:17 <Phantom_Hoover> alise, so your text editor can fiddle with the WM's internals?
20:21:27 <zzo38> My idea of the BIOS, is, to be compatible with IBM PC BIOS calls, and if there is no boot devices, or you push DELETE key while booting, or you just configured it to do so, or if the operating system attempts to switch to ROM BASIC, then it shall enter interactive Forth mode and display the "ok" on the screen.
20:21:28 <alise> No: this is why we have sandboxing.
20:21:42 <alise> In a way, note that the WM's internals would be /private/ to it; so there is no real worry there; but we can have proper sandboxing too.
20:23:24 <alise> Perhaps not threading as the pthreads and C folk know it; perhaps something more rigorous, without shared state.
20:23:29 <alise> Who knows? It's almost an active research topic.
20:23:32 <alise> I suggest we look at what Oberon does.
20:23:54 <alise> It has a similar objects-not-processes model, and it's well-designed; it will help.
20:24:05 <alise> Which /is/ the language, in a way.
20:24:34 <zzo38> Can you please tell me what is wrong with this program? http://sprunge.us/BgPV http://sprunge.us/UjTE
20:24:48 <alise> zzo38: What error does it result in?
20:25:09 <alise> \parindent=0.1pt? What, why the .1pt?
20:25:30 <alise> I do suggest that you use LaTeX.
20:25:55 <zzo38> Many errors, including "Missing } inserted", also there is a copy of one of the pages near the end of the document copied at the beginning of the printout, and the tables have no lines and won't wrap like HTML tables do.
20:26:14 <alise> Phantom_Hoover: "An efficient multitasking model is supported in a single-process by using short-running commands and cooperative background task handlers."
20:26:15 <zzo38> How do I make it wrap and stuff like HTML tables do?
20:26:21 <alise> Phantom_Hoover: This would need to be adjusted for SMP.
20:26:33 <alise> zzo38: missing } inserted -- you have a } missing somewhere
20:26:44 <alise> zzo38: and if you use LaTeX, it includes a table environment that does everything
20:26:50 <alise> writing it yourself will be hard
20:27:23 <Quadrescence> Output written on srcprint.pdf (1449 pages, 20374279 bytes). http://i.imgur.com/2eX2x.jpg
20:27:25 <zzo38> Does LaTeX do like HTML tables that can be wrapped?
20:27:53 <alise> What filesystem do floppies use again? Fat-12?
20:28:10 <alise> <Quadrescence> Output written on srcprint.pdf (1449 pages, 20374279 bytes). http://i.imgur.com/2eX2x.jpg
20:28:24 <alise> Phantom_Hoover: Well, true.
20:28:39 <Quadrescence> haha I AM PRINTING 1449 PAGES OF SOURCE CODE ON LEGAL-SIZED PAPER DO YOU HAVE A PROBLEM WITH THAT ALISE
20:28:47 <alise> Seems Oberon is not in any existing filesystem.
20:28:51 <alise> Quadrescence: Can you... afford that?
20:29:12 <alise> But the waste of paper.
20:29:28 <alise> 1449 pages is rather a lot for a pointless print-out.
20:29:41 <alise> no, just asking whether you really don't have anything better to do with it :p
20:30:12 <alise> the code looks far too small on that paper.
20:30:41 <Quadrescence> but I don't want to print Output written on srcprint.pdf (3628 pages, 32294410 bytes).
20:30:42 <zzo38> Also the page numbers in the table of contents are wrong, and there is a copy of page 28 before the first page of the table of contents, although Yap says it is page 26
20:31:01 <zzo38> The page it is a copy of is labeled 26 in the table of contents, though
20:31:49 <zzo38> No I passed it through Plain TeX once
20:32:01 <zzo38> Why do I need any twice passes?
20:32:22 <zzo38> I put everything in a macro and make it call the macro multiple times
20:32:30 <zzo38> So that is like multiple passes, a bit?
20:32:45 <Quadrescence> In latex usually you pass it through a few times to resolve references and counts and stuff
20:33:54 <zzo38> Like, see the http://sprunge.us/BgPV the part \def\PrintDocument that is the macro I am making it multiple times to resolve references and counts and stuff.
20:34:57 <zzo38> Also can you tell me if you found a mistake in the other file http://sprunge.us/UjTE
20:36:12 <zzo38> See if you can find the part I missed??
20:40:15 <zzo38> See the file http://zzo38computer.cjb.net/icosahedral/printout/main.dvi is the printout, you can see what is wrong with page-numbers and you can see some of the text is off of the page, and the other problems
20:41:58 <zzo38> Can you please tell me how to correct these problems?
20:44:29 <alise> zzo38: Why is it in landscape?
20:44:34 <alise> Why are the margins awful? Aieee, please use LaTeX.
20:44:46 <zzo38> It isn't in landscape?
20:45:23 <zzo38> Maybe your computer converted it to landscape because there is text off of the page
20:45:32 <zzo38> And how do I use LaTeX?
20:45:41 <alise> zzo38: Pretty easily.
20:45:41 <Phantom_Hoover> This seems similar to what alise had in mind, to a degree.
20:45:51 <alise> You can get rid of almost all your definitions; LaTeX includes those features.
20:45:55 <zzo38> The table of contents is correct except for the page numbers, as you should be able to see.
20:46:03 <alise> Phantom_Hoover: TUNES is the One True Idealist OS; 1995 to present, not a single bit of development.
20:46:07 <alise> It's as old as I am.
20:46:16 <alise> The founder, Fare, is a fan of Lisp machines.
20:46:25 <zzo38> Where is the document for how to use those features in LaTeX?
20:46:32 <alise> Quadrescence: Not even that. Well, in a few weeks, I suppose.
20:46:45 <alise> zzo38: You could read the LaTeX manual; that probably comes with your distribution; or you could use a tutorial, like Wikibooks'.
20:46:54 <alise> zzo38: http://en.wikibooks.org/wiki/LaTeX
20:47:02 <alise> You could just start at http://en.wikibooks.org/wiki/LaTeX/Document_Structure.
20:47:58 <zzo38> OK I can see the Wikibooks file now
20:48:10 -!- Flonk has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]).
20:49:36 <zzo38> I still want to learn how LaTeX works though, so that I can know how to write those kind of macros
20:49:42 <Phantom_Hoover> alise, this is going to be very different to a conventional OS, isn't it?
20:50:28 <alise> Phantom_Hoover: Extraordinarily.
20:50:47 <alise> No concept of binaries, filesystems, conventional UIs... other languages...
20:50:59 <alise> Unix-style OS design is extraordinarily boring, of course.
20:51:00 <zzo38> So, the things like "\subsection*" can be used to implement the "==Pragma TeX OmitFromContents" that I have?
20:51:13 <alise> zzo38: They're basically very complicated macros. Well, not that complicated.
20:52:38 <Phantom_Hoover> alise, I still don't understand how we do two things at once.
20:53:13 <zzo38> Icosahedral RPG is not like a scientific article/report, though.
20:53:37 <alise> Phantom_Hoover: Complicatedly.
20:53:42 <alise> zzo38: That is alright.
20:53:49 <alise> zzo38: the article or book class will suit your needs. Maybe memoir.
20:53:59 <zzo38> And what if I want to omit the \author part?
20:54:00 <alise> Memoir is like all the classes in one.
20:54:23 <zzo38> Probably book class should I use?
20:54:44 <alise> I've never used the book class. I'd use either article or memoir.
20:54:47 <zzo38> It is supposed to be like a book, that you can print out and bind it as a book and published
20:54:54 <alise> memoir is a very good class for books.
20:55:08 <zzo38> Except it says \chapter is only for books and reports
20:55:08 <alise> You can use the same commands as article class.
20:55:17 <alise> \chapter works fine in memoir.
20:55:29 <zzo38> OK. The description on Wikibooks does not mention it.
20:55:50 <zzo38> Is the memoir class good for publishing a book?
20:55:59 <alise> zzo38: http://www.tex.ac.uk/ctan/macros/latex/contrib/memoir/memman.pdf Very comprehensive documentation of memoir.
20:56:08 <alise> It does book typography and the like by default.
20:56:20 <alise> You can ignore most of the typographical, formatting stuff in the memoir manual; it's very excessive.
21:00:58 <alise> Phantom_Hoover: Complicatedly!
21:02:07 <zzo38> Are commands such as \def and \newcount and so on still usable in LaTeX?
21:03:37 <zzo38> The Icosahedral RPG rules is meant to be printed in black&white, except for the cover art which can be colored.
21:04:11 <alise> zzo38: Yes, though for commands you should use \defcommand.
21:04:52 <alise> http://www.emerson.emory.edu/services/latex/latex_19.html
21:05:22 -!- relet has quit (Read error: Connection reset by peer).
21:06:30 <zzo38> Does LaTeX still use Plain TeX or is stuff defined in plain.tex unavailable?
21:06:47 -!- relet has joined.
21:06:56 <alise> It doesn't use Plain TeX, but it has analogues for all the commands.
21:07:05 <alise> I guess Plain TeX /might/ work with LaTeX.
21:07:17 -!- tombom has joined.
21:09:31 <zzo38> I want to use Plain TeX for some things, because there are some things in LaTeX which I don't like
21:16:38 <zzo38> Also there is the problem of still not automatically wrapping text in columns
21:16:59 <zzo38> It doesn't do that with the icoruma.tex macros that I have, either, it also has the same problem
21:17:27 -!- cheater99 has quit (Ping timeout: 276 seconds).
21:18:02 <zzo38> How can you do it like HTML where it will automatically determine the proper widths for wrapping the text in the tables?
21:20:53 <alise> zzo38: latex can do columns
21:20:56 <alise> twocolumn environment or something
21:20:59 <alise> or was it multicol
21:21:04 <alise> it does do that with the tables
21:22:32 <AnMaster> why did I just think "D&D - flatland edition"
21:22:51 <AnMaster> that would be some interesting dice though
21:23:14 <AnMaster> zzo38, do you know what flatland refers to?
21:24:24 <AnMaster> zzo38, so this would be D&D for a 2D world, with some 2D replacement for dices
21:31:05 -!- ais523 has joined.
21:31:25 <zzo38> I can see however, that LaTeX does have a lot of stuff useful for Icoruma, such as \begin{enumerate} for <+> and \begin{itemize} for <->
21:31:32 -!- Flonk has joined.
21:33:05 <AnMaster> Phantom_Hoover, how would you toss one, you wouldn't have gravity?
21:33:25 <AnMaster> well, if you did, then people would have to climb over each other
21:34:11 <zzo38> Also the \footnote command works differently?
21:34:33 <AnMaster> Phantom_Hoover, you seem to forget the 2D nature of this
21:34:51 <AnMaster> Phantom_Hoover, wouldn't you only have one, the one you are in?
21:35:22 -!- cheater99 has joined.
21:40:35 <alise> zzo38: \footnote{text}
21:40:42 <alise> or \footnotemark[number]
21:40:45 <alise> then \footnotetext[num]{...}
21:41:28 <Warrigal> Our world seems pretty easy to two-dimensionalise. Just squeeze it between two plates that are really close together.
21:44:24 <zzo38> alise: The LaTeX \footnote{text} works fine for Icoruma, but it is different than Plain TeX?
21:44:36 <zzo38> And what is the command to activate LaTeX?
21:44:46 <alise> With LaTeX, run the "latex" binary instead of "tex".
21:45:36 <zzo38> When entering "latex" it says it is pdfTeX
21:46:03 <zzo38> But what if I want DVI?
21:46:50 <alise> pdfTeX outputs dvi too by default
21:46:55 <alise> only if you enter pdflatex will it output pdf
21:47:05 <alise> pdfTeX has non-PDF features too like typography and stuff that LaTeX utilises
21:47:52 <zzo38> O, that is why it is pdfTeX
21:52:39 <alise> Since you are doing a book, I suggest using document class memoir.
21:52:56 <alise> You can also use 10, 11 or 12pt text size (and it will scale the rest accordingly) with \documentclass[NNpt]{memoir}
21:52:57 * Sgeo wonders what alise's opinion of things like Wt and Blitzen is
21:53:18 <alise> Sgeo: wt is strange. i don't like it much -- link to blitzen?
21:53:38 <Sgeo> http://blitzen.sourceforge.net/new/index.php
21:53:46 <Sgeo> Something similar, except for C/GObject
21:54:29 <Sgeo> No, but I plan to become decent quickly
21:54:38 <Sgeo> I'm planning on making a UI toolkit
21:56:02 <alise> Phantom_Hoover: So, the OS! Which is now my favourite someone-else's-project.
21:56:06 <alise> Also problem, SEP both ways.
21:56:31 <alise> Make sure you manage my crazy ideas which I am likely to suggest >_>
21:56:39 <alise> Hey that's my job :D
21:56:59 <alise> Phantom_Hoover: sgeo invented PSOX
21:57:07 <alise> he's insane in the "Oh, jesus." way :|
21:58:36 <Sgeo> I meant, I'm making a GUI toolkit
21:59:26 <Sgeo> You're saying e still has time to save emself?
21:59:57 <alise> Flonk: Hypergeometric spatial disk organisation for deduplicative Lisp morphism storage theory.
22:01:23 <Sgeo> So, need layout control, widgets (there will be 2), and events (clicking)
22:01:27 <alise> See, he /is/ sufficiently crazy.
22:01:28 <Sgeo> Anything else?
22:01:34 <alise> Sgeo: Why are you doing that?
22:01:46 <Sgeo> alise, to make writing a HUD for Active Worlds sane
22:01:52 <Flonk> the only word in this sentence I really understood was 'for'
22:01:57 <alise> How do you plan to handle updating the UI without user interaction?
22:02:32 <Sgeo> clicking is a user interaction, right? But it needs to be updated separately from that, so
22:03:07 <Flonk> Phantom_Hoover: I've never done anything with it
22:03:32 <Phantom_Hoover> Flonk, what alise is talking about is abstracting away the disk completely.
22:03:38 <alise> no, i was talking about babble
22:03:43 <alise> only half of those words were meaningful
22:08:23 <Phantom_Hoover> fungot, what do you think of the ideas so far presented?
22:08:24 <fungot> Phantom_Hoover: hence my comment saying that they're partly mocking it, i'd kick them in the least... not working with the standard ( which it does regularly) they all roll on their backs and wave their paws in the air and point out that i have
22:12:40 <alise> Quadrescence: interestingly nobody ever prescribes the things i say to age until they know what my age is.
22:12:47 <oklopol> "<AnMaster> well, if you did, then people would have to climb over each other" <<< or you could have a few layers
22:12:48 <alise> everyone else in here has known for years; get over it
22:13:22 <alise> hmm, late 2008 i think.
22:13:49 <oklopol> alise isn't exactly all that young anymore
22:17:17 <oklopol> maybe he meant if you've checked from DSM you can't be sane
22:18:14 <oklopol> Quadrescence: ben. of doubt or something
22:18:55 <Flonk> Phantom_Hoover: why do you ask everyone it they know Lisp? :D
22:19:16 <Phantom_Hoover> Because we need someone sane to vet our Lisp OS ideas.
22:20:08 <Sgeo> I .. read a bit about Scheme some time ago
22:20:14 <Sgeo> Does that count for anything?
22:20:36 <Flonk> well all I know is (f o o), so i guess I'm useless there
22:20:42 <ais523> I know bits of several different incompatible bits of lisp
22:20:46 <ais523> *incompatible dialects
22:20:49 -!- Darkfull has joined.
22:21:11 <ais523> Phantom_Hoover: allegedly
22:21:14 <Sgeo> Phantom_Hoover, ais523 teaches Java [or teacher assistant?]
22:21:25 <oklopol> Quadrescence: to me it looks like it could've been a bad attempt at jokingness, in case there's something insane about checking if you're sane (although this is probably something you are not that interested in discussing)
22:21:59 <Darkfull> Help http://www.4security.com.br/2010/07/26/artigo-apache-com-mysql-e-monitoramento-com-bijk-no-debian-lenny/
22:22:45 <Sgeo> Anyone feel like opening elinks or similar?
22:23:27 -!- Darkfull has left (?).
22:23:39 <oklopol> "<Quadrescence> alise: stupidity is confused with nativity" <<< i'm not sure i get this, can you elaborate
22:24:55 <oklopol> hey Darkfull brought us a link
22:34:06 <Phantom_Hoover> alise, I still think we should cannibalise an existing CL implementation, BTW.
22:37:27 <alise> Phantom_Hoover: no
22:37:36 <alise> one, CL lacks a lot of things we'd like, like low-level hardware access;
22:37:49 <alise> two, CL implementations will be VERY tied to existing OSes and very hard to make work for our purposes -- trust me there --
22:38:25 <Phantom_Hoover> OK, so we write all of the frills that Lisp needs from scratch?
22:39:41 <alise> Sure. We're writing all the frills an OS needs from scratch
22:39:49 <alise> Besides, C would interfere with stuff, with memory and the like.
22:43:06 <zzo38> Unfortunately 888ASM is x86 only, with limited support for 32-bit and no support for floating point. But other than that I like the syntax best and it supports macros and things like that, too.
22:43:11 <alise> We'll probably need to write the JIT compiler in assembly, for speed and managing hardware, but we can code the frills in Lisp itself.
22:43:54 <alise> Phantom_Hoover: I mean, these things really blur the OS-language distinction.
22:44:12 <alise> Phantom_Hoover: Or we can just build it up with macros and the like.
22:44:43 <Phantom_Hoover> Memory access would need to be included at a low level, wouldn't it?
22:44:55 <zzo38> alise: Is it x86 assembly? Is it protected mode? Now if you do it to other computer you need C instead of assembly, a bit
22:45:08 <alise> for the address space
22:45:40 <zzo38> O, so it won't work if you don't have a 64-bit computer
22:45:41 <Flonk> good night people.
22:45:51 -!- Flonk has left (?).
22:46:13 <alise> zzo38: You could use QEMU.
22:46:16 <alise> It can emulate 64-bit.
22:46:26 <alise> But yeah, the 64-bit address space is important for our plans.
22:46:30 <zzo38> I have Bochs, I think it can emulate a 64-bit computer as well??
22:46:40 <alise> Bochs is slower though.
22:46:41 <zzo38> But that is OK, you can make it 64-bit program only, if you want to
22:46:45 <alise> You'd get much less than 2 GiB of space in total on your computer.
22:47:14 -!- Flonk has joined.
22:48:37 <Flonk> :D no its just that my clock tells the wrong time
22:53:04 <Flonk> so you guys know everything, right? :D
22:53:37 <Flonk> I'm just looking at lambda calculus (and havent come very far yet).
22:54:09 <Flonk> would (λx.xx)2 be 2+2, or 2*2, or something else?
22:54:31 <alise> lambda calculus doesn't have integers by default, so do you mean the church numeral?
22:54:58 <alise> then yes, it's 2 squared
22:55:04 <alise> You can verify this yourself
22:55:28 <Flonk> alright, thanks :)
22:55:57 -!- nooga has joined.
22:56:53 <alise> what don't you get about \b e. e b?
22:57:37 <zzo38> How to I configure my Wikibooks account to display the newest draft by default instead of the checked versions?
22:59:34 <alise> Phantom_Hoover: try and expand it manually
22:59:41 <alise> (\f x. f (f x)) (\f x. f (f x))
22:59:44 <zzo38> Actually I didn't find it
22:59:49 <alise> (\x. (\f x. f (f x)) ((\f x. f (f x)) x))
23:00:37 <alise> alpha renaming, you mean?
23:01:56 <alise> "A new way of analyzing grids of numbers known as matrices" --physorg.com
23:02:54 <oklopol> "in our new formalism matrices are written upside down, which makes some equations look much more natural"
23:04:02 <oklopol> mine was, that's how professional what alise said sounded to me
23:04:24 <alise> i just thought it was hilarious that a physics news website, supposedly
23:04:27 <alise> had to explain what a matrix was
23:06:09 <alise> http://www.physorg.com/news199631037.html
23:08:11 <alise> "Yehuda Eliezar" -- someone trying to say Eliezer Yudkowsky
23:08:17 <Sgeo> I'm uninstalling VS2010
23:10:09 <Sgeo> Actually, I'm not
23:10:37 <oklopol> so it answers the long asked question of when does a banded matrix have a banded inverse
23:11:03 <oklopol> i wrote the end of the sentence first so now it's all wrong and blah
23:11:59 -!- zzo38 has quit (Quit: YES CARRIER).
23:14:01 <oklopol> a b f x = (b ... (b f)) x, b occurs a times, now the outermost b will apply (b ... (b f)) to x b times, where the depth of (b ... (b f)) is now a-1, by induction each of these applies f to its argument b^(a-1) times, so all in all f gets applied b^a times to x
23:15:04 <oklopol> maybe that makes some sense, i've always just sort of trusted my feeling that composition looks exponentiationy.
23:16:17 <oklopol> note that (b f) applies f to its argument b times, so the induction base works
23:16:30 <oklopol> one b so f applied b^1 times
23:17:14 <ais523> composition's more like multiplication
23:17:18 <ais523> it's application that's exopnential-like
23:18:11 <oklopol> i meant application as you can probably see
23:20:49 -!- calamari has joined.
23:21:12 <Gregor> SQUIDS ARE ATTACKING, PLEASE MAN YOUR BATTLE STATIONS.
23:21:36 <Gregor> calamari: Yup. It has a Debian chroot too 8-D
23:21:46 <Gregor> I got frustrated looking for a half-decent WM for the screen size though.
23:22:08 <calamari> just ported dpkg to android/bionic
23:22:26 <Gregor> Get aptitude next plzkthx?
23:22:27 <calamari> haven't actually tested it yet lol
23:22:34 <Gregor> Testing is for the weak!
23:22:58 <calamari> but I was thinking the first package should definitely be egobf
23:23:02 <oklopol> so okay what does a . b do... (a . b) f = a (b f), so * x = (bf ... (bf x) nested a times, car-bf applies f to the cdr b times, and it follows by induction that f gets applied exactly b*a times
23:23:31 <oklopol> because application wasn't exponentiation but exponentiation-like
23:23:57 <oklopol> so addition, that was more work or?
23:24:41 <Gregor> cal153: Why is your name the first tab completion for "cal<tab>"
23:24:52 <Gregor> calamari: EgoBF is of vital importance. As is cplof.
23:26:34 <ais523> I'll update it if I ever go back to WIkipedia
23:26:48 <Gregor> What is a "metapedian"?
23:27:01 <ais523> oklopol: well, I haven't gone back there ever
23:27:06 <ais523> Phantom_Hoover: that's not what the word normally means
23:27:11 <ais523> at least within Wikipedia
23:27:37 <Gregor> This site makes my nose hurt.
23:27:45 <ais523> and I have no reason to believe everyone who redefines random words on the Internet
23:27:50 <Gregor> Phantom_Hoover: Yesh :P
23:28:19 <calamari> does dpkg absolutely require perl?
23:29:45 <calamari> bash: /data/local/bin/dpkg: No such file or directory
23:29:52 <calamari> yeah my son pushed enter for me lol
23:29:53 <oklopol> can you somehow browse the content of metapedia
23:30:15 <calamari> so I'm assuming dpkg is looking for something but not telling
23:30:27 <Flonk> well, good night again :D
23:30:29 -!- Flonk has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]).
23:31:09 <oklopol> oh main_page, no link from metapedia.org except the one in latest news
23:31:21 <calamari> hmm I guess it doesn't think it's executable
23:31:38 <Gregor> calamari: That's probably the dynamic loader.
23:32:04 <Gregor> objdump it to see what dynamic loader it wants, then objdump another binary to see what dynamic loader it wants.
23:34:31 <calamari> file format elf32-littlearm for both...
23:35:08 -!- Quadrescence has quit (Read error: Connection reset by peer).
23:36:05 -!- Quadrescence has joined.
23:36:09 <Gregor> calamari: Hrm, objdump isn't telling me what I thought objdump would tell me :P
23:36:22 <calamari> the start address is different
23:36:47 <Gregor> readelf -l /bin/bash | grep interpreter
23:36:53 <Gregor> That's what you actually want
23:38:03 <Gregor> Uhhh, are they different then? :P
23:38:12 <Gregor> Yup, I've seen that problem a lot.
23:38:20 <Gregor> -dynamic-linker flag to ld
23:38:48 <Gregor> Ideally that shouldn't be necessary, if it is that suggests your GCC configuration is wrong ...
23:38:56 <Gregor> Phantom_Hoover: ... huh?
23:39:38 <Gregor> Phantom_Hoover: I just poked around 'til I found the command that gave me what I wanted :P
23:39:58 <calamari> Gregor: android ndk doesn't really want you to create native binaries.. maybe that's why?
23:40:13 <Gregor> calamari: Ahhhh, that'd do it, yeah.
23:40:37 <Gregor> calamari: So just provide the requisite -dynamic-linker flag manually through GCC like so: -Wl,-dynamic-linker,/whatever/foo.so
23:40:51 <Gregor> Why have I munged with ELFs so much >_<
23:41:10 <Gregor> No, although that'd work too.
23:41:32 <Gregor> And since that is how it's documented, probably wise to use it that way X-D
23:50:57 <Gregor> Phantom_Hoover: Not since I watched http://www.youtube.com/watch?v=-0Xa4bHcJu8
23:54:18 -!- oerjan has joined.
23:54:59 <calamari> well it ran it but gave me segfaults.. trying my build on a hello world
23:56:02 <calamari> oh sorry, that came across wrong
23:56:14 <calamari> I am trying to compile a hello world to see if it segfaults too
23:56:38 <calamari> I have one built with the gloogle tools but it isn't so good for configure scripts