00:25:07 [QUIT] ink quit: Ping timeout for ink[1Cust5.tnt22.sfo3.da.uu.net] 00:42:02 billh joined #tunes 00:42:09 [QUIT] billh quit: billh 00:58:49 td left #tunes 01:54:33 red joined #tunes 01:54:54 [QUIT] red quit: [x]chat 01:55:54 [QUIT] smoke quit: Ping timeout for smoke[15dyn30.delft.casema.net] 02:14:25 smoke joined #tunes 03:05:31 [NICK] mark4 changed nick to: I440r 05:23:18 mort_ joined #tunes 05:24:13 [QUIT] mort_ quit: . 05:54:06 Alex[lex|yacc] joined #tunes 05:56:13 hi 06:52:47 [QUIT] gREMLiNs quit: [BX] Reserve your copy of BitchX-1.0c16 for the BeOS today! 06:58:55 [QUIT] smoke quit: Ping timeout for smoke[15dyn245.delft.casema.net] 07:28:22 smoke joined #tunes 08:20:55 smklsmkl joined #tunes 08:33:14 Alex[] joined #tunes 08:35:13 hey Alex[] 08:37:12 pwet smklsmkl 09:08:41 XeF4 joined #tunes 09:08:45 abi, slate? 09:08:45 slate is a unifying/hybridizing of self/beta/lisp at http://www.tunes.org/~water/slate-home.html 09:09:06 what is the url for the most recent slate pages? 09:15:22 http://diktuon.arrow.cx/list.php?ns=tutorial 09:16:26 Warning: Unable to connect to PostgresSQL server: connectDBStart() -- 09:16:27 connect() failed: No such file or directory Is the postmaster running 09:16:27 at 'localhost' and accepting connections on Unix socket '5432'? in 09:16:30 0 nodes in tutorial 09:18:47 bah. their server is fubar 09:46:54 morton joined #tunes 09:59:46 [QUIT] morton quit: Read error to morton[AC948AE1.ipt.aol.com]: Connection reset by peer 10:08:34 [QUIT] smklsmkl quit: BitchX-1.0c13: just when you thought it couldn't get any better 10:17:07 smklsmkl joined #tunes 10:22:56 [QUIT] XeF4 quit: Read error to XeF4[supernova.sevenheavens.com]: EOF from client 10:29:06 Alex[] left #tunes 10:37:02 tcn joined #tunes 10:40:05 [QUIT] smklsmkl quit: Ping timeout for smklsmkl[ppp35.dial-in.verkkotieto.fi] 10:44:58 morton joined #tunes 10:57:55 [QUIT] nate37 quit: bbl 11:04:24 smklsmkl joined #tunes 11:08:46 [QUIT] downix quit: downix has no reason 11:15:21 Alex[] joined #tunes 11:36:39 [QUIT] Alex[] quit: Mangez un castor, sauvez un arbre. 11:38:35 morton is away. (irc.N) 11:49:50 tcn left #tunes 11:53:58 washort joined #tunes 12:25:32 [QUIT] smklsmkl quit: Ping timeout for smklsmkl[ppp35.dial-in.verkkotieto.fi] 12:28:53 smklsmkl joined #tunes 12:44:58 [QUIT] morton quit: Ping timeout for morton[AC872FF9.ipt.aol.com] 13:43:17 XeF4 joined #tunes 13:57:31 your mom's server is fubar 13:58:47 smklsmkl joined #tunes 13:58:58 (btw, that db error is fixed) 13:59:22 ((my fault)) 14:17:15 eihrul joined #tunes 14:22:16 [QUIT] Alex[] quit: Mangez un castor, sauvez un arbre. 14:23:31 gREMLiNs joined #tunes 14:26:32 [NICK] gREMLiNs changed nick to: zzzkittie 15:16:33 [QUIT] washort quit: Ping timeout for washort[d136.narrowgate.net] 15:32:30 washort joined #tunes 16:51:52 [QUIT] smklsmkl quit: back tomorrow 17:18:59 washort joined #tunes 17:19:52 re 17:25:36 [QUIT] washort quit: Ping timeout for washort[d145.narrowgate.net] 17:26:49 washort` joined #tunes 17:27:54 washort`: that's what you get for dissing Canada Dry :P 17:29:19 Gah, Vernors is _much_ stronger then Canada Dry! :P 17:30:03 yeah 17:30:07 vernors is definitely better 17:35:36 [QUIT] washort` quit: ircII EPIC4pre2.508 -- Accept no limitations 18:03:43 tcn joined #tunes 18:31:09 [QUIT] rares quit: [x]chat 18:43:54 [QUIT] I440r quit: Ping timeout for I440r[purplecoder.com] 18:52:09 water joined #tunes 18:52:56 [TOPIC] water: TUNES: Free Reflective Computing System - http://www.tunes.org/ 18:52:58 water: url for slate docs that does not return a PostgreSQL error? 18:53:14 pgsql errors now, is it? 18:54:07 water looks 18:54:56 where did you get an error 18:54:57 ? 18:55:42 in the url smklsmkl gave me 18:55:50 which was? 18:55:52 I have since forgotten it, but I believe it was a .cx 18:56:08 http://diktuon.arrow.cx seems to work fine 18:56:32 aha.. it was a php script 18:56:38 with a somethingsomething=tutorial 18:56:49 ok check the namespaces listing 18:57:03 tutorial is in there, and you can browse in that 18:57:41 I440r joined #tunes 18:57:42 http://diktuon.arrow.cx/show.php?ns=Slate&name=intro fails 18:57:45 tutorial works now 18:58:02 yeah that's not the right url 18:58:14 although there *was* something there earlier, iirc 18:58:23 so something *is* wronf 18:58:25 wrong even 19:03:12 water: hm.. I've been puzzling all day over how to handle parentheses as a language extension 19:03:40 xef4: forth has had some experience with that 19:04:02 water: I know.. 19:04:11 which afaik falls under their meta-compiler work 19:04:16 ok 19:04:45 water: but most approaches I have thought up are too CPU-intensive 19:05:00 such as? 19:05:45 such as making ( return a function which retuns a list consisting of itself and the next token recursively until it recieves a message from ) 19:06:11 hm 19:11:25 which somehow seems like a rube goldberg approach 19:11:55 not necessarily 19:12:14 does proper tail-recursion sound right for this? 19:13:22 yes, but still CPU intensive 19:13:36 maybe 19:14:04 seems like a small price to pay for flexibility 19:18:50 unless I plan to implement an OS in the same language 19:19:20 heh 19:19:24 a good optimizer could optimize the recusion away 19:23:25 what good is a language if you can't write your OS in it? 19:23:47 what language *can't* an os be written in? 19:24:11 efficiently 19:24:11 (assuming proper environment support) 19:24:17 heh 19:24:33 say, I heard about squeak-os :) 19:24:48 yeah, but it ain't self-hosted 19:25:06 although it does fit on a floppy :) 19:25:19 oh.. it's a C-compiled VM? 19:25:45 that and the boot code is OSKit-derived 19:25:54 figures :) 19:26:06 what did you expect? 19:26:29 until some real money gets invested, it won't get much support 19:26:40 it's promising 19:26:55 true 19:27:10 and the windowing system and gui code *are* squeak native :) 19:27:40 it reminds me of the OLD mac and amiga OSes 19:27:59 how so? 19:28:08 size-wise 19:28:12 oh 19:29:04 if they get their act together on tightening squeak's belt-line properly, then it will be a lot smaller 19:29:31 at least it's getting away from the idea of the OS hogging the machine :) 19:29:32 despite the features-to-size ratio, there's still a lot of bloat in it 19:29:48 yep 19:29:57 yup.. the Forth equivalent would be maybe 30k :) 19:30:00 and squeak has threading and such :) 19:30:16 yeah but you can't really comment 30k of forth code :) 19:30:33 or browse it in a forth-native gui app :) 19:30:37 30k compiled, mabye 100k source 19:30:48 oh ok 19:31:21 and in Forth the 'browser' could be a block editor 19:31:36 hm 19:31:54 hey have we discussed forth-style slate? 19:31:58 I read Thinking Forth.. saw some nice tricks in there, i'm getting convinced that blocks are the way to go 19:32:10 forth-style slate, no.. 19:32:21 it's not rocket-science.... 19:32:40 basically identify namespace access with word concatenation 19:33:18 and each namespace's contents relate to the syntax of the forth system involved 19:33:32 sort of like a meta-compiler 19:35:12 does that sound right to you? 19:35:24 how's that different from the old 'vocabulary' system? 19:35:52 i don't know... i'm not familiar enough with the fine details of forth 19:37:11 say you have a ROOT namespace, which contains the words FORTH and ASSEMBLER.. 19:37:52 typing FORTH makes that namespace accessible 19:38:15 btw this should also apply to simple forth words like push pop etc and combinatory words 19:39:11 what's this about 'each namespace's contents relate to the syntax of the forht system involved'? 19:39:33 are you talking different varieties of forth? 19:39:45 sure 19:39:53 or just different forth environments 19:39:57 like 16-bit and 32-bit? 19:40:07 water nods 19:40:10 among other things 19:40:53 the idea is that by having a dynamic namespace system (which it already has), you can simulate language syntaxes and grammars 19:41:44 like having an ASSEMBLER vocabulary.. 19:41:46 it doesn't respect the namespace system metaphor very well, but that's a small price to pay 19:41:53 dunn 19:41:55 dunno 19:42:04 that's common to have 19:42:40 you type ASSEMBLER and you're in assembler mode, basically 19:43:03 ok 19:44:38 it's easier to make the 'sub-language' look like forth, but you can write your own parser and just use Forth's dictionary 19:45:07 that's how the list-processing extensions work 19:45:18 ok 19:47:11 forthers write 'little languages' all the time 19:47:32 of course 19:47:58 this would probably be oriented a little differently, you see 19:48:14 since the stacks are implicit in forth 19:48:15 ok... 19:50:04 with slate, the main mechanism to specify syntactic differences is implicit delegation 19:53:17 which is..? 19:53:35 oh just a fancy term for inheritance 19:53:42 plus mixins 19:54:53 basically either the namespaces inherit from other (usually previous) namespaces, or their meta-objects run algorithms that generate the slots on demand 19:56:54 ok.. in regular Forth you have a search-order... say + isn't found in the ASSEMBLER namespace, the interpreter looks in the next namespace in the search order, finds it in FORTH, executes it.. 19:57:16 but you can always define a different version of + in ASSEMBLER 19:57:44 well this is just about getting the namespace idea to allow extensible syntax, forth-style 20:00:42 i'm looking at this from the perspective of implementing SLATE in FORTH.. you're thinking the other way round right? 20:00:54 kind of 20:01:05 just seeing how to do forth-y things in slate 20:02:07 keep in mind i can make a pure stack language in one object in slate :) 20:03:44 well.. the postfix syntax seems pretty important. It allows the compiler/interpreter to take things one step at a time 20:04:00 sure 20:04:18 like something a zen would say :) 20:04:31 but i can make an elaborate stack object :) 20:05:41 so :) 20:06:27 you want to hear what I dislike about forth? 20:06:37 sure 20:07:03 hey put it into the zope review for forth :) 20:08:21 gotta think about this one :) 20:09:52 the need to put a space after every 'word' can be annoying, when doing things like comments 20:09:54 how about "it's only a little more readible than apl" ;) 20:10:04 hm 20:10:53 color forth would alleviate that annoyance.. 20:11:10 yeah i noticed that discussion on CLF 20:11:57 i'm gonna have a cforth mode in Retro.. not much to it 20:13:06 postfix syntax, I can't complain about that. it becomes an advantage when you're doing complex formulae 20:14:23 lack of popularity, there's a big problem :) 20:15:34 lack of a squeak-a-like? 20:16:24 such as self-hosted gui and environment 20:16:38 s/environment/application set/ 20:20:00 well.. there's just not much example code to look at 20:21:38 no books in the stores.. 20:22:07 nothing to compete with all the java hype, or whatever it is this month 20:22:13 (xml?) 20:23:03 directory services? :) 20:23:07 haah 20:23:41 lack of employment opportunities :( 20:24:24 the same could be said of inventing 20:24:36 i'm using it a little at work, to test things out.. but it doesn't mesh well with things like Access 20:24:58 yeah i do database work and design myself 20:25:12 i'd kill to use smalltalk 20:25:14 hey, I finally get to use pgsql at work :) 20:25:43 bah one small step 20:26:33 there are smalltalk systems designed for database implementation and admin, design, etc 20:27:09 navy guys want their data stored in a central location, instead of copies sent out.. pretty cool :) 20:27:40 P 20:28:14 and we're having so much trouble with NT4 and W2K as servers, we may be going over to unix 20:28:47 i'm currently in charge of unifying our *entire* admin system to the nth degree 20:28:51 NT just makes simple things like backup/restore friggin impossible 20:29:01 yeah i know 20:29:11 there is no 'superuser' 20:29:17 retarded 20:29:21 it seems we have the same job 20:29:29 roughly 20:30:02 who's "us" for you? 20:30:18 i have my bosses convinced that we can make a constellation of databases for security and privacy of info, yet keep it unified 20:30:28 eh? 20:30:33 i meant you and i 20:30:50 I mean, your group you're in charge of 20:31:01 heh what group? :) 20:31:31 a ship, or what? 20:31:45 there are two officers and 3 enlisted guys who i consult with 20:31:53 oh yes 20:32:01 an aircraft carrier 20:33:23 heh.. i've heard stories.. a guy who left my company a couple months ago, he used to do security auditing in the navy.. they found one carrier that was wide open to the internet :) 20:33:59 yeah the administrators are nin-com-poops 20:34:30 but since i'm not "trained", i can't advise them formally 20:35:39 yeah, damn bureaucrats :) 20:35:53 aka stupid people 20:36:34 and i'm not exaggerating in describing these people as less-than-bright 20:36:43 oh well 20:36:48 big pain in the ass with the 1388-2B standard is that they specified that it had to use a relational db, with foreign keys.. makes it difficult to achieve the same goal using a better design 20:36:50 water snaps out of it 20:37:28 we work with a lot of people at what was an IBM plant.. "I Believe in Meetings".. it's all they do!! 20:37:41 heh 20:37:55 never get any work done 20:38:49 there's another project we're doing tech manuals for, and the actual equipment is still in breadboard 20:38:56 doesn't even work 20:39:10 no one i work for knows enough buzzwords to be a threat 20:39:33 geez 20:39:38 that explains a lot :) 20:40:32 so your boss doesn't talk about web-enabling the administrative functions? that's good :) 20:40:46 yep 20:41:17 in fact, i do most of that kind of thing myself and just demo and use it when they like it :) 20:41:43 pretty much the only bottleneck is having to use Access 20:41:52 (which is a really big one of course) 20:41:57 hah 20:42:30 have you got a C compiler or something that can use the DAO .dll without Access? 20:42:57 no they don't want me writing in C 20:43:08 ehh? 20:43:22 in fact, they want me to avoid VB wherever possible :) 20:43:22 not enough C programmers around in case you die? 20:43:30 try 0 20:43:47 what are you supposed to do if you can't program? :) 20:43:48 0 for VB programmers as well 20:44:03 heh 20:44:28 they let me use VB if it can't be done any other way 20:44:47 i'm trying to train some of my coworkers to do a little programming.. the ones who know Access pretty well.. 20:44:57 the idea is that we're going to hand over the software with a small manual for 1 admin and many users to other ships 20:45:31 besides, i'd much sooner code in Smalltalk than C when it comes to databases 20:45:39 or VB for that matter 20:46:16 you use queries much? 20:46:23 yep 20:47:03 SQL.. worst language ever 20:47:11 oh well 20:47:18 it's poweful at least 20:47:48 well it gets annoying when you need recursion 20:48:05 you have this slow ugly hybrid of vb and sql 20:48:06 can't hold a candle to smalltalk, though, in it's own intended category 20:48:20 heh yep 20:48:43 everything to learn in access is just a trick 20:49:01 nothing is systematically-designed 20:49:02 yeah, "where'd they bury this..." 20:49:11 that's why i like smalltalk 20:49:34 it takes me 5 minutes to code something in smalltalk that takes days to do properly in access 20:49:37 did you ever figure out how to access the 'hidden' and 'description' attributes of objects in the database windows? 20:50:02 yeah 20:50:19 writable? 20:50:26 hm 20:50:39 i think so, although i didn't try 20:51:14 i've about given up on that 20:51:21 seen access 2000? 20:51:21 heh 20:51:30 nope 20:51:49 does it use xml properly? 20:51:52 well they just moved a buncha shit around on the screen 20:52:11 i haven't installed it 20:52:11 oh wonderful 20:52:26 just helped out another guy with a query 20:52:41 he say's it's crap :) 20:53:02 how much better is oracle? 20:53:22 never used it.. our biggest customer does.. 20:53:39 a 5-fridge size thing.. 20:53:56 [QUIT] clintonr quit: oops. 20:54:10 hm 20:54:18 best I can figure, it's like pgsql 20:54:22 i'll just plug smalltalk then 20:55:09 like pgsql, but designed for really fuckin huge things 20:55:29 yeah that's what i thought 20:55:57 this is too much talk about work for me 20:56:01 haha 20:56:08 yeah, me too.. i'm on vacation 20:56:55 man, you'd think it's july 4th here 21:11:49 water: i found a paper in Henry Baker's archive on instruction scheduling by trying a few permutations generated from a crude machine model versus a more precise machine model :) 21:12:05 ok 21:12:34 don't let lmaxson know about it... >:) 21:12:51 hm? why not? 21:13:06 (kidding) 21:13:15 so, was the crude model good enough? 21:14:35 well, he's inferring that its possible to get better instruction schedules 21:14:39 by just running the various permutations 21:14:42 and seeing which one is faster 21:14:51 (of course, he admits that this only works within single basic blocks) 21:15:04 and you still need a crude model to weed out most permutations 21:23:08 the same could be said of neural nets, you know 21:23:20 or genetic algorithms 21:25:10 yes :) 21:25:16 but, as a local heuristic 21:25:29 if you know the previous instruction in the schedule 21:25:51 and have multiple choices for following, it would be simpler than creating an accurate machine model :) 21:27:45 yeah but choosing instructions seems to naturally presume an accurate machine model 21:28:25 whether in one's head or in software... you're just not using it at code generation time 21:29:30 [QUIT] tcn quit: Ping timeout for tcn[cci-209150250059.clarityconnect.net] 21:40:13 water: Oracle runs on small machines too and runs 10-100 times faster than Access. 21:41:47 unfortunately, my issue is with complexity of use, but obviously this comment falls on deaf ears 22:34:04 td joined #tunes 00:31:55 billh joined #tunes