Channel: prev | nextHour: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23Date: prev | next | index
lispTuesday, 18 July, 2006
Style: cnoolb | irssi | clog | f00f | sexp
Filter: Quit-off | Join-off | Part-off | Nick-off | Mode-off | Topic-off | Ignore-filter-words-off | Url-only-off
Highlight: From-me | To-me | My-highlights
 
<starseeker>pjb: Heh - guess so.00:00:17
eriderQuit with message ("I don't sleep because sleep is the cousin of death!").00:00:55
ThomasIQuit with message ("bye").00:01:46
ThomasIQuit with message ("bye").00:01:46
<starseeker>Sigh. Thousands of interesting subjects, only one lifetime00:01:48
<Drakeson>yes, that's the problem. so the developers (mainly game developers now) have to confine themselves to opengl/directx or go hunt for special features in new cards00:01:50
<starseeker>Drakeson: Ah. No wonder the OpenGraphics project is still active.00:02:23
<starseeker>Drakeson: Open source software doesn't get to take advantage of all the advanced features, so garden variety OpenGL which is well supported will do very well.00:02:56
<Drakeson>yes, but the sort of programming needed to take advantage of today's graphic cards needs engineers to write them (it's no fun). so it has to be done by big players (i.e. companies who can afford it)00:05:21
<starseeker>Drakeson: The advanced features you mean, or just to make them work period?00:05:50
estEntered.00:06:10
<Drakeson>i know people in my profession (electrical engineering) do not embrace open-source as much as people in CS do00:06:31
<Drakeson>starseeker: we do not have a mature xgl/aiglx/... yet00:07:03
<pjb>AFA programmers are concerned, if you just provide the documentation of your hardware, we're happy. Anyway, we know you can't program :-)00:07:27
<starseeker>Ah.00:07:36
<starseeker>So that's what xgl is - an attempt to make X work off of just OpenGL?00:08:18
<starseeker>pjb: Heh - true enough.00:08:36
<starseeker>Has anybody ever used the CADR emulator? http://www.unlambda.com/cadr/index.html00:09:31
estQuit with message (Client Quit).00:09:35
<Drakeson>pjb: no, that doesn't work. ATi and nVidia wouldn't endanger their precious!!!! technology by opening up all their documentations (not that everything is documentated, of course)00:10:33
<pjb>starseeker: I booted it up a long time ago. IIRC, it has changed since.00:10:33
<Drakeson>what a pity!00:10:36
<pjb>That's why I don't own an ATi nora a nVidia and probably never will.00:11:01
tritcheyMiniEntered.00:12:13
<Drakeson>I believe that ATi who has better hardware and worse drivers really needs to rely on the poeple outside, by providing special features, drivers, and docs for open-source00:12:31
<starseeker>How do I start up chaosd?00:12:45
est_Quit with message (Read error: 145 (Connection timed out)).00:13:28
<Drakeson>back to my perl-to-lisp lesson ...00:13:36
<starseeker>pjb: That's why I wish they would stick everything propritary on the darn board - I don't really need to know the details, but a working driver would be nice00:13:52
lemonodorEntered.00:14:03
AdamantEntered.00:14:24
<starseeker>Grr. See chaosd directory - what chaosd directory?00:15:19
<starseeker>Oh, nevermind - multiple tarballs.00:16:19
FufieQuit with message ("Leaving").00:16:58
FufieQuit with message ("Leaving").00:16:58
<starseeker>Anybody around who knows how the design of the old MIT CADR machines compared to the more modern stuff made at Symbolics?00:17:54
<starseeker>Was it a major leap ahead?00:18:00
<pjb>starseeker: working drivers cannot be provided by the hardware vendor, as Xerox prooved. That's because of a non working printer drivers, that RMS started the freedom software movement.00:18:24
<pjb>Any programmer can debug a bad driver, given the sources and the hardware documentation.00:18:45
<Dawid[Programmer>So that's what xgl is - an attempt to make X work off of just OpenGL?--> You could look at it that way, basically each window is rendered to its own target (Texture) and then the scene put together much like a 3d game.. This provides for great graphically effects adn speed ups when it comes to transparency00:18:54
<starseeker>pjb: Hehehe00:19:34
mbishopQuit with message (Remote closed the connection).00:20:23
<starseeker>pjb: I never thought about it that way. Still, what are the odds of another RMS appearing? That's gotta be a million to one shot...00:20:32
<pjb>We are all another RMSes!00:20:46
<pjb>Just give us the hardware doc!00:20:55
<starseeker>pjb: Point.00:21:30
Dawid[ProgrammerQuit with message ("( www.nnscript.de :: NoNameScript 4.01 :: www.XLhost.de )").00:21:58
JaskoQuit with message (Read error: 110 (Connection timed out)).00:22:08
<starseeker>Normally I'm a tightwad, but I would buy an opengraphics card if they actually produce one - I really hope they succeed.00:22:15
<starseeker>Dawid[Programmer: Does that make for any inefficiencies, or is it pretty much a wash?00:23:33
rtoy_Quit with message (Remote closed the connection).00:24:01
<starseeker>Grr - anybody know how to make one's own text appear as something other than dark grey in xchat?00:25:09
<spiaggia>what, you are not using beirc?00:25:31
<starseeker>Unfortunately not - haven't had time to try getting CLIM Desktop up yet00:25:54
mejjaEntered.00:26:04
<spiaggia>actually, you don't need the entire desktop to run beirc00:26:53
<starseeker>Is the CLIM desktop setup still current?00:26:54
<starseeker>Hmm. Well, I'm feeling like wandering through mud tonight...00:27:16
<spiaggia>I am using it on my laptop at home, but not at work.00:27:22
*starseeker starts by trying to pull a recent McCLIM...00:28:15
<xarq>starseeker: yeah, it's in the preferences in xchat00:29:05
<starseeker>I wish someone could answer the question as to what (if anything) the copyright status of the McCLIM spec is00:29:10
<starseeker>xarq: I've been looking - I don't see a label.00:29:20
mc__Quit with message (Remote closed the connection).00:29:28
<starseeker>Foreground isn't it00:29:36
<spiaggia>starseeker: McCLIM spec?00:29:39
<xarq>starseeker: I wouldn't know, I haven't used xchat in about a year00:29:49
<starseeker>spiaggia: Well, the CLIM spec00:29:50
<spiaggia>starseeker: It's included in source form in the McCLIM distribution, and nobody has complained so far.00:30:16
<starseeker>Ah that was it - extra colors #3000:30:19
<starseeker>spiaggia: Right, but I don't think silence is necessarily concent in that respect.00:30:38
<spiaggia>that might be true, but I think it is pretty safe. It is unclear who owns it and so it would be hard for anyone in particular to complain.00:31:12
<spiaggia>why do you care?00:31:31
HatchBack176Quit with message ().00:31:40
<starseeker>spiaggia: I've been thinking it would be an interesting exercise to combine the CLIM spec with the source code to make a literate document00:32:02
<starseeker>clarifying the spec as needed00:32:14
<starseeker>but that would effectively make the code and the documentation one and the same, which sort of flies in the face of a lot of licensing ideas.00:33:19
<starseeker>the Garnet documentation, for example is very definitely NOT public domain.00:33:30
<starseeker>That caught me by surprise00:33:44
<starseeker>That's why I had started that Gardener's idea about the ANSI draft spec - if we could get all the major players to say it's OK, we could start to use that to do some interesting things with documentation, literate programming, etc.00:34:58
<starseeker>but it would be a big waste if we put in the effort and then someone objects00:35:12
<spiaggia>starseeker: both for the CLIM spec and the ANSI CL standard, we actually need new documents that are more pedagogical in nature. And I think it would be a good thing to start a community project to write such documents. It wouldn't bee too hard.00:36:41
*starseeker looks up pedagogical00:36:58
syamajalaQuit with message (Connection timed out).00:37:25
tanukiEntered.00:37:45
<dankna_>actually I think the CL standard is nicely pedagogical00:38:13
<dankna_>I wholeheartedly agree with you about the CLIM spec00:38:19
<starseeker>spiaggia: You mean teaching documents as opposed to detailed specs?00:38:22
<starseeker>or more precise?00:38:27
<starseeker>I've heard the CLIM spec leaves a fair bit to be desired, in a lot of ways00:38:49
<dankna_>I read most of it but that was a while ago00:39:39
tavon_Entered.00:39:47
*starseeker finds dict.org less than useful for pedagogical, time for the dead tree...00:39:52
<dankna_>what I can never get any beginner to believe is that the best way to learn is by looking things up as you go... I'm not sure it's true for everyone, but it seems like it ought to be00:40:08
<dankna_>but documents that give conceptual overview are very important also00:40:19
<dankna_>he means teaching documents00:40:31
<starseeker>OK, got it.00:40:56
<starseeker>:-)00:40:58
<starseeker>Made me resort to the printed dictionary for the first time in years ;-)00:41:09
tavon_Quit with message (Client Quit).00:41:46
<dankna_>hehe00:41:54
<dankna_>well, it's always a nice experience to learn a new word :)00:42:07
<spiaggia>starseeker: I am not thinking about user manuals, but more like reference documentation with lots of examples and cross references.00:42:27
<spiaggia>starseeker: user manuals would be very valuable as well, but they are much harder to write.00:42:48
<dankna_>yeah... examples are important even when they don't illustrate anything subtle00:42:54
thsQuit with message (Read error: 110 (Connection timed out)).00:42:55
<starseeker>What I would like to see happen is the draft spec being combined with one or more lisp implementations (probably SBCL and CMUCL primarily for licensing reasons), and incorporating examples as unit tests throughout00:43:07
<starseeker>spiaggia: I agree.00:43:17
<dankna_>which is pretty much what the CL-standard examples are there to do, show subtleties. which of course CL has a lot of.00:43:24
<dankna_>hmmm, combined with implementations, that's an interesting approach00:43:39
<spiaggia>dankna_: yeah, and sometimes the HyperSpec is a little terse wrt examples00:43:51
<starseeker>Blame Tim Daly - he made a believer out of me ;-)00:43:54
<housel>starseeker: it all sounds like a great idea, modulo the copyright problems00:44:34
<starseeker>housel: Exactly, hence this page: http://wiki.alu.org/Project_FreeSpec00:45:04
<dankna_>right. it needs examples for even the simple things, so you can be sure you understand the context00:45:12
<starseeker>dankna_: Exactly. You don't ever assume it's so simple it doesn't need context, and you never get caught later00:45:47
maxmLeft.00:45:48
<dankna_>what you really need are one to five really good writers who are willing to put large chunks of time into it... that's basically what it takes for that sort of project00:45:51
<dankna_>right00:45:54
<starseeker>dankna_: Which we might be able to scare up if we could get some coherent organization and clear permissions lined up (sigh)00:46:23
<starseeker>dankna_: One of my long term dreams (and the reason I was curious about Lisp Machines) is a system which is solidly founded in formal proof logic and design, from the hardware up00:46:58
<starseeker>a truly trustworthy computer00:47:04
<dankna_>hm, are you familiar with Andrew Appel's work on proof-carrying code?00:47:20
<starseeker>and naturally lisp comes to mind as the ideal language to have as a foundation :-)00:47:23
<pjb>starseeker: you could put ACL2 to use, then.00:48:06
<starseeker>dankna_: No, actually. I'm aware of the bitc design work going on in Coyotos but not this effort - thanks!00:48:13
<starseeker>pjb: Right :-). ACL2, the HOL system, a couple of others could all be used as targets for proofs generated by the "compiler"00:48:44
<michaelw>dankna_: not george necula?00:48:47
<starseeker>ACL2 could be built right in00:48:52
<dankna_>there's certainly something very conceptually pure about having Lisp all the way down, but I actually gave up on that a while ago00:48:57
<dankna_>the scope of what people want to do is already too large for any one language00:49:03
tavonEntered.00:49:22
<dankna_>which is why one of my interests is cross-language infrastructure that can dissolve the boundaries00:49:25
<starseeker>dankna_: that's no problem - if you think about it, we write lisp compilers in C. Why not C compilers in Lisp?00:49:28
<starseeker>it all comes down to putting it in a form the machine can understand00:49:40
<dankna_>I agree that it's possible technically, just not in manpower00:49:44
<starseeker>true :-(00:49:50
<michaelw>starseeker: see Zeta-C00:49:58
<dankna_>because for it to work, Lisp would have to control a greater portion of the market than C does presently00:50:03
syamajalaEntered.00:50:13
<dankna_>(because C doesn't do anywhere close to everything anymore, either)00:50:18
<starseeker>michaelw: Cool!00:50:33
<dankna_>George Necula? *googles*00:50:37
thsEntered.00:50:49
tanukiQuit with message (Remote closed the connection).00:52:06
<starseeker>dankna_: My one hope that this might be changed is that such a system could satisfy the two great demands that current systems cannot provided - reliability and security which can be guaranteed, provided the hardware doesn't fail00:52:07
<dankna_>michaelw: cool to that, too :)00:52:27
estEntered.00:52:37
<dankna_>thanks for pointing that out, I apparently didn't do enough background reading00:52:48
tanukiEntered.00:53:03
<dankna_>starseeker: oh, it could, yet you still have an adoption problem. basically I want tools to treat any language like it's Lisp, y'know?00:53:35
<starseeker>:-)00:53:43
<starseeker>Well, the only thing which can fight the adoption problem is time, and merit. After all, I think the post office still used Win3.1 (shudder)00:54:17
<dankna_>there are actually movements in that direction from a number of places. AspectJ is a reimplemented Java which adds a language feature; since the compiler was already designed with extensibility in mind, it could in principle be reworked and expanded to allow CL-macro-like processing of something that resembles Java code and can link with Java.00:54:51
<dankna_>haha, yes00:54:58
<starseeker>What was that quote from somebody at sun - "we managed to drag half the C++ programmers halfway to common lisp" or something like that00:55:51
<dankna_>at some point there'll be a profound convergence among languages sort of like the hardware convergence we're seeing today with Macs and PCs: people will realize that the differences have been getting smaller for a long time00:56:01
<dankna_>Paul Graham, wasn't it?00:56:05
djmQuit with message (Client Quit).00:56:18
<starseeker>"We were not out to win over the Lisp programmers; we were after the C++ programmers. We managed to drag a lot of them about halfway to Lisp."00:56:37
<starseeker>- Guy Steele, Java spec co-author00:56:37
<starseeker>found it00:56:41
<xarq>dankna_: it's been smooth sailing for hardware ever since openfirmware00:57:03
<dankna_>xarq - well, yes, very true00:57:27
<starseeker>Hardware is shaking out - I think the ability to completely virtualize Intel and AMD cpus and use stuff like Xen will be one of the final steps to a really acceptable PC platform, technologically speaking. After that, it's a question of doing it Right00:58:24
<dankna_>agreed, absolutely00:59:03
<starseeker>Is it true that originally Symbolics was started to make good computers for hackers, and got taken over by the suit types?00:59:22
<dankna_>that said, just because Xen is the wave of the future and we know it's going to take over at some point, does not mean that now is the right moment to sink a lot of money into it00:59:25
DrakesonLeft.00:59:28
<starseeker>dankna_ agreed00:59:36
<starseeker>it's too soon00:59:39
<dankna_>I'm not old enough to have firsthand knowledge of that, sorry00:59:40
<starseeker>letsee, where did I see that...00:59:51
<starseeker>I always thought that was a neat idea - have a pseudo nonprofit company that made quality hardware for the geek niche market01:01:23
tanukiQuit with message ("Rising Above Mediocrity: http://bgeiger.net/weblog/").01:02:03
<dankna_>there are companies like that in certain niches. soekris.com sold me my router.01:02:05
<dankna_>it's a neat idea, I just don't see it as a big world-changing idea. but it's nice when it happens.01:02:30
<starseeker>dankna_: Oh, I agree - it's not world changing. But it might serve as a "testbed" for ideas which could then be licensed for larger production, if they happen to work.01:03:26
<dankna_>yeah01:03:33
<dankna_>but I'll stick to the software, it's the area where I know enough to hopefully make a difference :)01:04:19
<starseeker>Oh, agreed. I'm not much of a hardware guy, actually.01:04:48
<starseeker>But perhaps the existance of things like this http://www.mosis.org/ could make it easier for a non-profit Symbolics to function01:05:14
<dankna_>yeah, seen that01:05:34
<starseeker>I should get the Freespec site active again - it would be worthwhile if we really could clear that up.01:06:49
<slyrus>is this the lisp freespec that was discussed on the gardeners list?01:07:16
<starseeker>Yes01:07:20
<starseeker>slyrus: I don't know if it's possible, but I don't know what else to do about that particular situation.01:08:04
<slyrus>that would be useful. I'm sure this has been discussed far too many times here and elsewhere, but I think the best way to free it up is to publish derivative works and see who comes after you01:08:11
<starseeker>slyrus: Heh - that's a bit like going tiger hunting by wearing big boots and stomping around in the brush as hard as you can.01:08:48
|klette|Quit with message ("Leaving").01:09:12
|klette|Entered.01:09:25
<starseeker>slyrus: It might be workable, except for two things: If I understand copyright law, there's no "silence gives consent" provision - they could wait until it was a raging success and then kill it with a lawsuit.01:10:00
<starseeker>and if they did sue, there would be no money for a defense01:10:13
<xarq>true, if there are 15 copyright violations it doesn't give you extra safety - you can all be sued at once just as easily if it's a large company01:10:19
<dankna_>correct01:10:51
<dankna_>I think you're better off writing material from scratch01:11:00
mkhlEntered.01:11:16
<dankna_>Linux didn't start by modifying commercial Unix.01:11:25
<starseeker>The common lisp spec, unfortunately, is far beyond my abilities to recreate, even starting with the ISO lisp as a reference01:11:44
<dankna_>yeah, but that's what you're left with01:12:09
<starseeker>It's almost a thousand pages put together by a lot of smart people over a period of years, working full time.01:12:16
<starseeker>I agree - if nothing else, ISO is a start01:12:24
<starseeker>it's explicitly released under a usable license.01:12:34
<xarq>a sequel to PCL would be a good step01:13:01
<starseeker>but I'm also faced with the problem - how do I create a spec that is equalivent to ANSI common lisp without infringing on the document which DEFINES ANSi common lisp?01:13:04
<dankna_>well01:13:42
<starseeker>I suppose I could try for "better than ANSI Common Lisp" but as a newbie I can tell you how far I'd get with that.01:14:20
<xarq>isn't copyright law that if 70% of the work is different, it's not in violation?01:14:31
<dankna_>you want to make sure to cover all the same subjects but without borrowing phrasing or the organization of individual pages. borrowing the list of chapters ought to be fine.01:14:35
<dankna_>xarq: I don't think so, but I'm not a lawyer01:14:43
<starseeker>That's a tricky one.01:14:51
<xarq>there are hundreds of books on C which are very similar in nature, and have to be01:15:05
<xarq>well, dozens maybe01:15:17
<starseeker>True, but a spec is a bit different from a regular book - it must be precise to be useful.01:15:26
<dankna_>a standard technique is to read the stuff, take notes - but not in sentences, just words that remind you what it was about - and then, working from your notes without the original material open, summarize it01:15:31
<dankna_>this advice is maybe too low-level to be helpful01:15:39
<dankna_>but it's all I've got01:15:43
<starseeker>No problem - glad to get it01:15:50
<starseeker>The gardner's list pretty much ignored it :-(01:15:59
<dankna_>cool01:16:07
<foom>why does the common lisp spec need to be rewritten?01:16:44
<starseeker>I would like to make one try at the main document - if all of the major organizations listed as having contributed, and the key individuals all say no problem, that should be OK.01:16:47
<starseeker>foom: It doesn't - it's just we can't get cleary copyright info on it at present.01:17:04
<foom>oh, nobody knows who owns it?01:17:24
<starseeker>Well, I have a list of about 180 contributing individuals and organizations, who participated in the original process of creating it01:17:51
<dankna_>see the archives of comp.lang.lisp, there's been a lot of discussion about this01:18:16
<starseeker>presumably many of them own small pieces of it, but even they probably don't know what any more.01:18:17
<dankna_>in particular KMP has an opinion on it01:18:23
<foom>huh.01:18:33
<foom>the CLHS on lisp.org says its copyright 1996, harlequin.01:18:59
<foom>odd. :P01:19:01
<dankna_>which I don't really remember the details of01:19:03
<dankna_>the CLHS is a derivative work and has a separate copyright from the spec it's based on01:19:16
<housel>there's been plenty of discussion of the status of the ISO spec on c.l.l01:19:33
<starseeker>This would be less frustrating if one of two things happened 1) the ANSI folks let us have an "unofficial" copy of the text, which we can't claim as ANSI but can use, or b) the draft spec is no longer available sitting there in tempting tex format ;-)01:19:44
<starseeker>the ISO spec is free and clear, isn't it?01:19:54
<xarq>the CLHS on lispworks says it's copyright lispworks LTD 1996-200501:20:03
<housel>ISO standards are almost never free and clear01:20:17
<foom>"Parts of this work incorporate material taken from American National Standard X3.226, copyright 1994, and is used with permission of the X3 Secretariat, ITI, 1250 Eye St., NW., Suite 200, Washington, DC 20005 and of the copyright holder, American National Standards Institute. ANSI/X3.226 was developed by Technical Committee X3J13, Common Lisp."01:20:23
<foom>if harlequin could get permission from them to use it, why can't everyone else?01:20:40
<starseeker>the CLHS was arrived at through some process years ago between the hyperspec authors and ANSI, IIRC.01:20:42
<starseeker>harlequin was involved with the creation of the spec01:20:56
<starseeker>they had special clout ;-)01:21:03
<xarq>you could still ask (:01:21:14
<starseeker>I think Camm Maguire was talking to ANSI some time back, but he didn't (or couldn't) say much about it and I don't think anything ever came of it.01:21:41
<starseeker>I should check with him.01:21:58
<starseeker>what's wrong with the ISO Lisp? I could have sworn I saw a "this isn't TECHNICALLY ISO Lisp, but it's the same and it's free" message on that site.01:22:32
*starseeker goes googling01:22:36
<xarq>I asked a copyright holder if I could use their work, and they said that 49% of my work could be their document I wanted to use01:22:59
<xarq>might give you some hope01:23:12
<starseeker>http://islisp.info/01:23:37
<starseeker>xarq: That's encouraging.01:23:45
<starseeker>http://islisp.info/legalese.html - is this in dispute?01:24:23
<housel>not for the ISLISP standard, that I know of01:25:41
<housel>(not (eql +ISLISP+ +CL+))01:26:34
bmpQuit with message ().01:26:51
<starseeker>Right. Sorry, I was confused - I thought ISO Lisp was ISLisp01:26:57
tanukiEntered.01:27:11
<starseeker>Even if someone were to start a "new, clean" spec, I doubt it would get much interest or support.01:27:39
<starseeker>the Lisp community as a whole does not seem too interested in such matters.01:27:56
<housel>people interested in doing that sort of work are few and far between01:28:27
<starseeker>It's a shame - I can't help thinking Lisp is a bit "diluted" right now.01:31:06
ths_Entered.01:31:15
<sigma957>starseeker: then what's wrong with scheme?01:31:35
<starseeker>sigma957: It doesn't seem to have the "industrial software implementation" chops that Lisp does, for whatever reason01:32:26
*sigma957 nods01:32:57
<starseeker>sigma957: If you look, ACL2, Maxima and Axiom are all commonly run on GCL, which is probably the least common one I seem mentioned in comp.lang.lisp01:33:34
rr--Quit with message (Read error: 110 (Connection timed out)).01:33:45
<starseeker>To be fair, that's partially because (except for Maxima now, not sure about ACL2) there are ANSI issues on both sides of the mirror in those programs.01:34:16
dankna-lapEntered.01:35:07
tritcheyMiniLeft.01:35:12
<starseeker>I guess if I had to put it in one line it would be "The Lisp community should strive to avoid user visible implementation differences that do not serve to significantly enhance functionality."01:35:32
<starseeker>Ah, well.01:37:09
<starseeker>I think once GCL and Axiom are both on the ANSI bandwagon that will improve my outlook on the issue ;-)01:37:51
thsQuit with message (Read error: 113 (No route to host)).01:38:34
ths_Changed name to ths.01:40:48
<housel>GCL isn't?01:41:38
PripQuit with message ("Sto andando via").01:41:49
bmpEntered.01:43:02
<starseeker>Well, it can be compiled either as an "almost ANSI compliant" flavor or the older flavor.01:43:40
<starseeker>Maxima uses the almost ANSI one, Axiom the older one (IIRC).01:43:53
<starseeker>I don't know how close the newer GCL is to ANSI compliance - I should ask Camm.01:44:15
<starseeker>The thing is, a lot of the more useful CL.net stuff doesn't run on GCL yet (CFFI being an obvious example.) So for Axiom, we can't take advantage of any of that work yet.01:46:05
<starseeker>It should come with time though.01:46:13
<starseeker>Well, thanks all. Interesting discussion.01:48:20
<slyrus>starseeker: sorry, was/am away for dinner. more later. you're right, but I think they would have little other recourse besides shutting it down, BICBW.01:49:31
<xarq>they wouldn't have much to gain by shutting it down - more people using common lisp means more software sales for lispworks01:51:08
<foom>yeah, until SCO buys lispworks01:51:51
<dankna-lap>they have to protect their rights or lose them, when they become aware of a violation01:51:59
<dankna-lap>at least trademarks work that way and I think copyrights have something similar, but IANAL01:52:14
<foom>dankna-lap: no. that's not true.01:52:15
<xarq>unless they granted permission01:52:17
<dankna-lap>oh01:52:25
<housel>copyrights are not similar to trademarks01:52:34
<dankna-lap>I am pleased to be corrected, then01:52:54
<foom>depending upon something of questionable legality, even if the current owner doesn't seem to mind/notice isn't necessarily a good idea.01:53:24
<housel>indeed01:53:43
<xarq>otoh if they think it will generate more sales they will readily grant permission01:53:51
hikozaemonEntered.01:54:07
KompottkinQuit with message ("Und tschüß!").01:54:09
<xarq>and if they fail to grant permission, you can expect a good possibility of getting a cease and desist01:54:24
tanukiQuit with message ("Rising Above Mediocrity: http://bgeiger.net/weblog/").01:54:29
<DrForr>SCO is sniffing around lispworks? Goodie.01:56:32
<foom>this is the way rumors get started. =)01:56:52
<xarq>no, SCO is too busy getting fried in court01:57:19
<foom>DrForr: yes, please submit the story to slashdot!01:57:22
<foom>breaking news! :)01:57:29
<xarq>now that would be the ultimate in bad publicity01:58:04
<xarq>common lisp sued for 6 billion dollars01:58:26
Pupeno_Entered.01:59:10
redblueQuit with message (Client Quit).01:59:43
<DrForr>Heh, obviously my sarcasm detector needs to be retuned. Given all of the stupid moves SCO's made buying out LispWorks seemed perfectly rational, if not curtains for the entire language.02:00:05
mulliganQuit with message (Remote closed the connection).02:03:18
ajaEntered.02:04:03
redblueEntered.02:05:39
tanukiEntered.02:10:11
mulliganEntered.02:10:15
PupenoQuit with message (Connection timed out).02:12:10
aquanautQuit with message ().02:17:07
tavonQuit with message ().02:23:35
mkennedyEntered.02:24:20
voidengineerEntered.02:24:39
rtoymEntered.02:26:36
<voidengineer>hello02:27:46
<spiaggia>hello voidengineer02:28:34
arbscht_Changed name to arbscht.02:28:39
<arbscht>good afternoon spiaggia02:29:43
<voidengineer>spiaggia hello02:31:21
<spiaggia>hello arbscht02:31:29
<spiaggia>arbscht: I understand classes have started again, right?02:31:57
aquanautEntered.02:32:12
<arbscht>that's right, I'm on campus now02:32:18
<arbscht>and a little annoyed; they seem to have removed emacs at the labs02:32:32
<spiaggia>oh?02:36:06
dolmansEntered.02:36:19
<spiaggia>I think they must have reinstalled the machines and did not install Emacs again, nor Lisp probably. I'll talk to them.02:36:38
<spiaggia>we are talking Linux, right?02:36:56
<slyrus>starseeker: still around? it would be interesting to explore what the worst-case scenario of using the ANSI spec would be.02:37:50
<spiaggia>arbscht: no, wait. That can't be right. Linux without Emacs is unthinkable.02:38:05
karstensrage_altLeft.02:38:13
<slyrus>not to mention unusable (by me at least)02:38:20
<spiaggia>hello slyrus02:38:32
<slyrus>hey beach02:38:35
<arbscht>yep, it is linux :(02:38:52
<arbscht>they seem to have reorganised and upgraded a lot of things. I'm searching for traces of emacs in case they've simply moved it, but am having no luck02:40:23
karstensrage_altEntered.02:40:31
karstensrage_altLeft.02:40:37
<starseeker>slyrus: back02:41:07
<starseeker>slyrus: still here?02:41:17
<slyrus>hey starseeker. it would be interesting to consider what the downside would be. a C&D order about what? the entire work? parts? etc...02:41:38
<slyrus>what remedies might one be obliged to make?02:41:47
<spiaggia>arbscht: let me know if you find it. I'll try to find out what is going on.02:41:59
<arbscht>spiaggia: will do, thanks02:42:16
<starseeker>slyrus: The worst case would be a suit for damages, with financial compensation on the part of the violator, I think02:42:48
<starseeker>plus killing the project and voiding all the work put into it02:43:00
<slyrus>yes, and I wonder what damages one could reasonably claim in this case02:43:06
<starseeker>dunno. With the draft freely downloadable for ten years, I imagine that would be a bit tricky...02:43:30
<slyrus>lispworks might be able to claim some damages, others would probably have a harder time02:43:37
<starseeker>the company doing the hyperspec, too02:43:53
<starseeker>Harlequin(sp?)02:44:06
<starseeker>Symbolics Inc.02:44:10
<starseeker>Allegro02:44:26
<slyrus>who owns harlequin and symbolics? yes, allegro would be interesting to approach. I could imagine a more freely avaiable spec being something in their best interests.02:44:57
<starseeker>Guy Steele I think wrote much of one of the books the spec was based off of02:45:00
<starseeker>harlequin - not sure.02:45:10
<starseeker>Symbolics is a strange case - their assets were bought and they're back online, but it's some sort of weird holding company.02:45:30
<spiaggia>I think we could handle writing a document called (say) The Common Lisp Reference Manual.02:45:37
<rtoym>Hmm. The weak hash table tests no longer crash, which is good. But the weak pointers tend to be held on for longer on x86 than non-x86. Not sure why, but I'll blame the conservativeness.02:45:42
<starseeker>We went round and round about that back when we were thinking about the Macsyma source02:45:45
<slyrus>spiaggia: I like the title02:45:56
madmonky1Quit with message (Read error: 145 (Connection timed out)).02:46:10
<starseeker>We could, but that wouldn't be a spec in the sense of a language spec, correct?02:46:15
<spiaggia>starseeker: it would be as good (provided we take care to keep it correct)02:46:52
<starseeker>Well, then it would be a spec :-)02:47:09
<starseeker>My original idea was to have a new spec which was "very similar" to ANSI Common Lisp, but instead call it Community Lisp02:47:43
<spiaggia>starseeker: sort of, yes, but it would also be a much more complete document in terms of examples and explanations02:47:51
<starseeker>spaiggia: Right, but to my way of thinking that's just a spec done Right ;-)02:48:13
<starseeker>We agree.02:48:25
<spiaggia>yes, I think we do.02:48:33
LeafwQuit with message (Read error: 110 (Connection timed out)).02:48:35
<spiaggia>my estimate is that such a thing would be around 1k pages02:48:43
<starseeker>Yes02:48:47
<starseeker>at least02:48:49
<starseeker>I would like to see a unit test framework incorporated from the beginning, sort of like Paul Dietz's test suite02:49:25
mejjaQuit with message (Remote closed the connection).02:49:26
<spiaggia>most of the work could be distributed, like writing the spec and examples of each function. One person (or a few people) would have to work on the general organization and conventions and such.02:49:30
<starseeker>Yes02:49:37
<slyrus>perhaps we can eliminate the need for a spec with a detailed enough test suite02:49:47
gimbalQuit with message (Connection reset by peer).02:50:08
<starseeker>Hmm - that a VERY intersting idea. Anythhing which can run the test suite without error would be, by definition, compliant02:50:23
<starseeker>what an elegant way to test compliance02:50:32
<slyrus>the normative part would be the docstrings :)02:50:46
<starseeker>Well, I'd suggest using LaTeX for the text part, and including the code in noweb or similar style - docstrings are a bit limiting ;-)02:51:21
gimbalEntered.02:51:35
<starseeker>Is there some kind of "specification notation" for things like argument order, type limitations, etc?02:51:50
<spiaggia>I suggest texinfo, actually02:52:04
<starseeker>That would work too.02:52:11
<slyrus>smarkup!02:52:16
*starseeker googles smarkup02:52:28
*slyrus is surprised to see that google found something02:52:50
<arbscht>spiaggia: still nothing, but I've made a post on the faculty forums which may yield something. I'll let you know how that goes02:52:53
<spiaggia>arbscht: OK.02:53:03
<starseeker>slyrus: Probably the least informative page I've seen for a project ;-)02:53:16
<slyrus>thanks! :)02:53:21
<starseeker>There is also cl-typesetting02:53:21
<starseeker>Oh, is it yours? Sorry!02:53:32
<slyrus>it will be more informative after I finish the thesis02:53:35
*starseeker has a rather red face02:53:46
<slyrus>no need. the lack of documentation is appalling, considering that it's purpose is writing documentation and other text!02:54:16
<starseeker>Is it based on the cl-typesetting engine or is it something all its own?02:54:38
<slyrus>s/it's/its/02:54:45
<slyrus>it spits out latex02:54:52
<rr->isn't docbook the fashionable/politically correct way nowadays, as opposed to texinfo02:54:55
<starseeker>Cool.02:54:57
<slyrus>and xhtml02:55:02
<slyrus>and bibtex02:55:05
<spiaggia>rr-: does docbook have macros?02:56:03
<slyrus>this used to be extensions to gigamonkey's markup, but I removed all of that stuff02:56:11
<slyrus>and made it its own package02:56:11
<slyrus>so it used to be able to do cl-tyepsetting02:56:11
<starseeker>I would suggest cl-typesetting for obvious reasons, but it has almost as little documentation as smarkup :-(02:56:28
maxmEntered.02:56:29
<rr->i assume that is a rhetorical question02:56:32
<spiaggia>no, I don't know02:56:39
JaskoEntered.02:56:44
<maxm>hmm wtf is this some kind of freenode thing? autoleave after timeout?02:56:59
<starseeker>maxm - is your client set up to do that maybe?02:57:23
*maxm leaves this window alone, only to find out that client left the channel on its own02:57:29
<spiaggia>slyrus: the dogfood argument is in favor of something like smarkup, but it would have to be able to do at least what Texinfo does, i.e. produce indexes, cross references, and such.02:57:36
<maxm>well its ERC which I just started using02:57:37
<maxm>why would it have something like that by default02:57:50
<maxm>plus it does not happen with #emacs, only with #lisp02:57:59
<starseeker>maxm : dunno02:58:00
<maxm>very strange02:58:05
<spiaggia>maxm: I am using ERC right now and I have been connected for weeks02:58:22
<slyrus>spiaggia: yes, that's reasonable. for me things like TOC, lists of figures, etc... are all just passed off to the layer below (latex) and ignored in xhtml. not ideal for something like this.02:58:42
<maxm>maybe I'm hitting some kind of leave key by mistake... But I swear I left work and it was here, and I came home, just reattached my screen and it says I left02:58:50
rr-Quit with message ("Leaving").02:59:09
<maxm>ok lets leave it and see what happens02:59:17
<starseeker>I would suggest that the probable best way to proceed would be to pick texinfo or something else that has already the mature features we need, and a little awk magic can later convert it to whatever we ultimately decide on02:59:45
<spiaggia>slyrus: well, that doesn't work, because we might want to produce all of those things in an online version of the manual in the form of text02:59:46
<spiaggia>sounds very reasonable to me02:59:59
<slyrus>spiagga: yes, I agree.03:00:14
xarqQuit with message ("ERC Version 5.1.2 (IRC client for Emacs)").03:00:41
<starseeker>Obviously if a lisp solution matures we can hope on, probably providing a nice big debugging document for whatever system we switch to later ;-)03:00:47
BreezeQuit with message (Connection timed out).03:00:58
<starseeker>er s/hope/hop03:00:58
<slyrus>starseeker: or you can do what I'm doing which is using smarkup to generate latex (or texinfo or what-have-you) with the idea that working with sexp's is going to be easier for the folks writing the lisp spec03:01:00
<spiaggia>BBL03:01:14
<starseeker>That's a point.03:01:17
<starseeker>Is there an example smarkup document somewhere?03:01:31
<slyrus>it doesn't solve the problems spiaggia mentioned, but doesn't make things worse03:01:33
<slyrus>yes. or maybe there will be in a minute :)03:01:49
<starseeker>Can't texinfo produce text? If you can export to that that should work...03:01:53
<spiaggia>slyrus: well, there is a good chance that we would have to extend Texinfo anyway, so we should weigh that against adding features to something like smarkup03:02:32
<slyrus>sure, after losing all the markup :)03:02:36
<starseeker>slyrus: How hard would it be to add back in the cl-typesetting targeting?03:03:11
<lemonodor>rydis: you around?03:03:37
<slyrus>spiaggia: yeah, but my point is that having sexps that are used to generate the texinfo is better than native texinfo. it's a debatable point, at best :)03:04:24
<slyrus>hmm... I don't know. I never targeted cl-typesetting, but gigamonkey's stuff did and I was using that as my format for a while.03:04:33
<slyrus>I imagine it wouldn't be too bad03:04:40
<starseeker>I don't know how ready for prime time cl-typesetting is (it certainly isn't in terms of math) but we could at least have an "in house" environment03:05:23
<slyrus>no worse than what gigamonkey had to do for his markup, which didn't seem too horrible03:05:35
<slyrus>I'm not sure what that buys us for online stuff though. I would prefer to target xhtml in lisp and latex for dead trees, personally03:05:58
<starseeker>Hmm... true...03:06:18
<starseeker>OK, let's start there.03:06:33
<starseeker>I downloaded smarkup - which lisp do you recommend?03:06:48
<slyrus>at least until cl-tyepsetting has the math and bibliography handling of latex/bibtex03:06:50
<slyrus>which I don't see happening anytime soon, BICBW03:06:50
<starseeker>true :-(03:07:02
<slyrus>sbcl, of course03:07:20
<starseeker>you need - ch-asdf, anything else?03:07:37
<slyrus>and, as usually happens whenever someone downloads, I'm about to roll a new release :)03:08:07
<slyrus>ch-util03:08:07
<slyrus>cl-bibtex and puri too03:08:19
<starseeker>Hmm. OK, this will take a minute...03:09:16
<spiaggia>slyrus: I see, though Texinfo is probably not that complicated a program if you exclude the parsing bit. An sexp based notation would eliminate that problem.03:10:47
<slyrus>yes03:12:00
<slyrus>and would give us reader macros03:12:00
<starseeker>does ch-asdf require regular asdf?03:12:39
<slyrus>yes03:12:45
<starseeker>Um. Is there a big asdf file that will load all of this? ;-)03:13:07
<slyrus>as long as the stuff is in site-systems, just asdf:oos 'asdf:load-op :smarkup should do the trick03:13:58
<starseeker>hmm.03:14:15
<starseeker>I must not be set up right. arrgh03:14:24
<starseeker>how do I tell where site-systems is?03:15:24
<starseeker>nevermind03:15:44
<starseeker>/usr/lib/sbcl/site-systems/03:15:56
<slyrus>what's the error?03:17:53
<slyrus>where is sbcl installed?03:17:53
<starseeker>usr - it's a gentoo ebuild install03:18:14
<starseeker>Do the asd files go in site-systems, or how does it work? (I never did get the hang of this - I just load everything and then save an image ;-)03:18:59
<slyrus>yes03:19:33
mogunusQuit with message ("BitchX: made with real honey.").03:19:35
<slyrus>links to them, anyway03:20:01
<starseeker>how do you load cl-bibtex - it seems to use defsystem03:20:03
<slyrus>there's a bibtex.asd03:20:12
<starseeker>doesn't seem to be in the tarball - is it online?03:21:57
<starseeker>google doesn't know about it either :-(03:22:26
nocloudsQuit with message (Connection timed out).03:23:33
<slyrus>great. the anoncvs for cl-bibtex from subversions.gnu.org seems to be down.03:24:26
nocloudsEntered.03:24:32
sigma957Left.03:24:44
flounderingEntered.03:24:46
<starseeker>just curious - how come uffi instead of cffi?03:25:09
<Xach>cffi is pretty young03:26:11
<Xach>any project older than a year would probably not use cffi03:26:21
<starseeker>ah03:26:26
<starseeker>crud - there goes bibtex03:27:49
<slyrus>it predates cffi03:28:41
<<