00:00:03 thanksandbye 00:02:06 np 00:16:52 davb has changed the topic to: OpenACS: Free Web Toolkit, 4.5=Tcl and proud of it! 00:17:58 paje, tickle davb. 00:17:58 denshi: excuse me? 00:18:40 bye guys 00:18:42 denshi has quit () 00:21:27 heh 00:49:06 hazmat has left #openacs 00:56:47 talilee (~talli@188.muka.lasv.snfccafj.dsl.att.net) has joined #openacs 01:07:07 markd2 (~Snak@r-41.21.alltel.net) has joined #openacs 01:10:24 friday night at #openacs 01:10:30 wheee! 01:12:55 did you like the "weird group, excellent support and enthusiam" quote describing out little gang? 01:14:30 i resent the enthusiasm part 01:14:41 you are one of the most enthusiastic! 01:15:11 i guess 01:15:21 * talilee says this very unenthusiastically 01:15:48 * markd2 isn't weird enough 01:16:19 ok, guess I was wrong, sorry. 01:16:25 markd2, you've gotten decidely less weird since you had your antlers removed 01:16:48 i know it was hard going during hunting season, but still. 01:17:37 talli: hows progress on openacs.org 4.5? 01:17:39 * davb ducks 01:18:05 antlers of the damned! 01:18:43 davb, if you duck, i will just kick you 01:19:00 progress was slowed by an unexpected case of the maui's 01:19:08 ah, 01:19:12 understandable. 01:19:31 that must be really difficult to have to go back so often... 01:19:40 yeah 01:19:46 i cry alot about it 01:19:49 I mean to NY 01:41:39 denshi (~chatzilla@adsl-216-62-223-193.dsl.rcsntx.swbell.net) has joined #openacs 02:10:13 hey talilee, shagster, rbm, markd2, davb: what's the best way to stack trace vars being defined in an adp page? 02:10:46 err... being printed in said script. the question is to find where they were defined, and what happened to them on the way. 02:12:19 good question. I dunno 02:12:59 probably use tclpro 02:13:33 I heard you can use it to debug in aolserver. 02:14:06 i'm finding 4.5 to be quite byzantine. 02:15:03 quite. 02:15:08 uh oh. another voice registered ... 02:15:26 oacs is quickly moving towards the worst of "enterprise" applications 02:15:34 too complex to be worth using 02:15:54 ? 02:16:24 have you seen any sites built with OACS, ON TIME yet? 02:16:47 talli: what are you guys doing with it? 02:17:01 or are you referring to 4 specifically? 02:17:10 we really build most things using AOLserver and Pzg 02:17:11 PG 02:17:14 ah. 02:17:19 and use the OACS stuff pretty sparingly 02:17:23 i see. 02:17:34 I can't pass verdit yet. 02:17:42 s/verdit/verdict/ 02:17:56 I think most people just build sites and get on with it... 02:18:11 I'm trying to find the irreducible set of operations to employ on 4.5. I'm pretty lost so far. 02:18:20 davb is probably right. 02:18:39 denshi: what is a irreducible set of operations? should I look that up? 02:18:42 acs3 was great because the flow of control was so obvious and so little got in the way. 02:19:12 but each module did its own thing and became a mess of code. 02:19:35 if you understood databases, and could grasp text processing, then you could quickly build something. That's all there was. 02:19:44 yes. 02:19:48 still can :) 02:19:56 davb, not really 02:20:03 why not. code is still there. 02:20:11 don't you need to understand the object hierarchy? 02:20:18 that's ncie that the code is all there 02:20:20 yeah, I know, just that the set of core operations is much larger. I can't fathom the vector space yet. 02:20:22 does that mean we 02:20:32 I meant 3.5 is still there. 02:20:35 anyway. 02:20:35 can go start hacking the linux kernel easily? 02:20:45 4.5 is a huge complex mass. 02:20:55 ah, yeah, so we're talking abotu the same thing 02:20:58 4.5 02:21:49 problem is, aD never decided what the core set of ops was before abandoning the project. so we are left with half implemented and duplicated implementations. 02:22:14 has dotlrn cleaned anything up or did it add complexities? 02:22:16 but we need to get 4.5 out the door and very quickly decide wha tneeds to stay and what needs to go or be rewritten. 02:22:35 talilee. it makes at ALOT easier to setup and add pieces from pacakages. 02:22:45 with the new portal system, which is 100% dotlrn independent. 02:22:48 cool 02:22:51 ditto for the applets 02:23:11 so you can use the file storage, bboard, calendar etc applets with new portal and be dotlrn free. 02:23:18 but no postgresql yet. 02:23:52 we need to simply and better define the apis. the service contract stuff helps alot I think. 02:24:00 so packages can use services from others etc... 02:24:24 I have been looking at the code for over a year... part time on and off 02:24:51 the problem is the docs only explain the low-level functions of the system, not how to use them to build something. 02:25:06 we need a how to put the pieces together doc. 02:25:51 I need to write up a proposal for my user contributed cusomization cookbook/doc kinda of thing and ask people who have actually used it to contribute. 02:26:40 cool 02:27:17 should service-contract be replaced by wdsl? or does wdsl describe sites rather than pages? 02:29:39 I'm heading towards some strange conclusions working on 4.5 02:36:31 1. either relational databases need to advance, or 4.5 is a bad idea. 02:37:08 2. 4.5 is a monsterous piece of work. I doubt tcl has ever been pushed so far. I wonder if it can go any further. 02:37:48 what does 1 mena? 02:37:53 the whole object hierarchy thing? 02:38:31 GEM_ (~mcmullan@MULTICS.MIT.EDU) has joined #openacs 02:38:46 GEM_ is now known as GEM 02:40:44 hello, GEM, by what do I call you in meatspace? 02:41:28 talilee: what I mean is that 4.5 presents an abstraction layer that I instinctively feel should be natively be in the schema, or not there at all. 02:41:53 I just don't know enough relational calculus to describe a full proof of that instinct. 02:42:11 talilee is now known as talli 02:42:30 well, since many many people have said the same thing, it may be dangerously true 02:43:05 Hi denshi 02:43:34 service contract was inspired by wdsl, you probably should ask neophytos more about that. 02:43:53 denshi: I'm Greg 02:44:03 meaning this, talli: the relational database is probably the best weapon we developers have against complexity. 02:44:16 * GEM was pondering making a library package :_) 02:44:30 denshi: i agree 02:44:43 denshi: well alot of the code lives in the database. 02:45:10 vinod lives in the database 02:45:14 we can define data of arbitrary complexity that *knows* how it is validated; we can query it to find out it's structure, and we can mathematically prove that it will not be corrupted via normalization. 02:45:28 markd2: that explains that v_pants table. 02:45:30 ow, my head... 02:45:53 I think I sent talli an email about this...brb 02:46:33 oh, here it is: 02:46:43 from a comp.databases post.. 02:46:59 > Is there a standard way (using SQL) to query the structure of a database? 02:47:10 Has anyone ever paused for a moment, and considered how unbelievably odd 02:47:15 this question is? And then paused again, and reflected on the incredible fact 02:47:25 that this question can be answered at all, in any RDBMS? (although the details may vary) 02:47:47 Today I received a brochure in the mail for a software development conference. Delegates will be assailed on topics like "Software Patterns: Are they 02:47:58 Useless?" and a couple of talks that wail endlessly about how hard it is to reverse engineer/cross compile a Java/C++/C/COBOL/D-flat application to 02:48:21 figure out what it does. The conference features an endless parade of esteemed engineers explaining why their latest methodology solves this or that problem of software engineering. 02:48:40 heh 02:49:20 denshi: actually, no, I don't think that it is odd that most database implementations store their metadata in database tables 02:49:51 GEM: no, that's the cool part about databases. their structure is transparent and queryable. 02:49:51 Label GEM not found. 02:50:27 hmmm. I thought that was fixed. 02:50:30 you know, i think it was very interesting that someone posted about how the object hierarchy is supposed to store metadata about content that is in other tables 02:50:55 sorry, the CR 02:50:57 It seems perfectly logical to me that the designers will realize that the required information aboue tables can be very usefully kept in system-space tables 02:51:17 but hten other people posted that in implementation the CR holds content 02:51:21 denshi: *smile* THanks. I've been here y started this channel 02:51:22 heh, GEM whose nick is in all caps, has been acquainted with chump :) 02:51:23 the former seems reasonably 02:51:42 err, my keyboard is eating keystrokes lately 02:51:47 the latter seems, uh, dangerous 02:51:49 talli: no, wait. the object system only holds metatdata. 02:51:56 I have been here before.l 02:52:01 I started this channel 02:52:01 i meant the CR 02:52:05 the content repository is especially for content, keeping it in the content-item tables. 02:54:57 'scuse me - time to go to sleep 02:55:57 i see 02:55:57 'night. 02:57:25 think 5 GB is enough to squeeze oracle on? 02:57:40 yes 02:57:54 cool. 02:58:04 well, depending on the amount of data.... 02:58:20 just a test system, openacs. 03:00:32 so anyway, I have to think that there's a way that a relational DB could present the 4.5 object & permissions system in a more natural way. 03:00:48 ah. excellent. 03:01:00 I should go read Dodd's papers. Supposedly, modern RDBMSes only implement 6 of his 12 points. 03:01:18 sounds about right. 03:03:01 brb 03:03:08 talli has quit (Read error: 104 (Connection reset by peer)) 03:08:31 talli (~talli@188.muka.lasv.snfccafj.dsl.att.net) has joined #openacs 03:11:14 good night, thanks for the interesting questions. I would love to explore this more. 03:12:01 night dav 03:12:45 I'll just say this: the idea of reducing openacs to the core operations is a great idea. the opeancs-core should be lighter and simpler, but extremely well though out so that the reduced set of operations are the right ones. this is kinda what mccusker was saying yesterday at treedragon about language. 03:12:51 and I am really going now :) 03:12:53 davb has quit ("Client Exiting") 03:16:22 he's all about leaving on the last note. 03:27:42 markd2 is now known as markdzen 03:28:08 markdzen is now known as markd2 03:28:32 markd2 is now known as loko-zenno 03:29:05 loko-zenno is now known as markd2 03:30:47 denshi is now known as tzolf 03:30:57 tzolf is now known as tzodd 03:36:04 tzodd is now known as denshi 03:38:53 heh 04:51:47 markd2 has quit ("wheeeee") 05:03:13 hey denshi 05:03:35 ngh. 05:06:38 what's moving, talli? 05:09:33 not much 05:09:39 just got back from chinese with the folks 05:09:45 why are you online on a fri night? 05:10:10 deadline 05:10:19 ah, i see 05:10:36 oh, and there's nothing worth dating in Dallas. 05:11:28 haha 05:11:38 any idea how to get from kde2.2 to kde3 on redhat? 05:11:46 motherfriggers don't make it easier 05:11:47 easy 05:12:07 no idea, sorry. 05:12:43 have you looked at Gentoo Linux? 05:15:21 the name is familiar, but it has evaded my attention. 05:15:57 new distro, of course 05:16:05 with a BSD package management system 05:17:51 how does the bsd system work? 05:21:41 talli has quit (Read error: 104 (Connection reset by peer)) 05:22:44 talli (~talli@188.muka.lasv.snfccafj.dsl.att.net) has joined #openacs 05:22:54 dell hardware officially sucks 05:25:06 really, it's not that so much as that you're part of their beta-testing. 05:25:51 for what? when they implant us with commodity bio-chips? 05:26:00 this shit they sell for 2K is really crap 05:28:37 no, i mean right now. as a customer, you're doing the work that a beta-test and burn-in used to do. 05:29:06 nowadays it's cheaper to ship crap and do returns on the statistical rate of defects. 05:29:53 shit 05:29:55 neat 05:32:16 yeah, dell's real innovation is their inventory control. 05:32:28 well, that's it 05:32:28 IIRC, they have a negative inventory at all times. 05:32:42 they're actually making money off the interest. 05:32:48 i'm finding me a good hardware developer and stickin' with 'em 05:32:54 i think i'll try apple 05:34:08 I'm looking for a good small vendor, but I think they've all been wiped out. 05:48:16 so what is your prescription for improving the OACS? 05:48:48 i don't know yet. I'm still a 4.5 neophyte, comparatively. 05:51:25 what, is there another conversation running concurrently? 06:02:15 speaking of that, where are our ambiguous miata drivers? 06:02:40 pumping iron, probably 06:08:36 never saw that article 06:44:38 have a good one 06:44:46 l8r 06:44:55 denshi has quit () 08:04:37 pardini (~ricardo@206213.virtua.com.br) has joined #openacs 08:12:04 pardini has quit ("gtkBitchX Lite I said!") 12:40:22 davb (dave@alb-24-58-162-46.nycap.rr.com) has joined #openacs 14:24:01 morning, all 17:27:13 <_djg_> _djg_ (~dirk@212.84.246.68) has joined #openacs 18:39:57 [Global Notice] Hi all. The latest edition of the Occasional OPN News is up on http://openprojects.net/news.shtml .... thank you for using OPN and have a great day! :) 18:44:19 bbl installing oracle 18:44:21 davb has quit ("Client Exiting") 19:27:35 davb (oracle@alb-24-58-162-46.nycap.rr.com) has joined #openacs 19:27:42 * davb reinstalls oracle 19:29:34 davb: oracle sounds like a lot of fun 19:29:39 are you a registered DBA yet? 19:29:50 no way... 19:30:12 its not bad, the openacs docs are very good. 20:14:45 davb has quit ("Client Exiting") 21:35:14 markd2 (~Snak@r-41.113.alltel.net) has joined #openacs 22:08:03 1152 / 16 22:08:04 72 22:08:07 72 * 1 22:08:09 72 * 16 22:08:14 72.0 * 1 22:08:15 72 22:08:16 72.0 * 166 22:08:17 11952 22:08:20 grrr.. damn 6 key 22:08:23 72.0 * 16 22:08:23 1152 22:54:11 markd2: does cocoa make .net smell like alpaca piss for real? 22:54:17 yep 22:54:26 not that .net needs anything to make it smell like that 23:11:40 denshi (toddg@linux128.ma.utexas.edu) has joined #openacs 23:15:17 yo denshi 23:15:23 hey 23:15:33 i have a new invention 23:15:40 cool 23:15:53 sorry, i already have indoor plumbing 23:16:43 gee, I couldn't tell 23:16:58 1. uncap mushrooms, fry top in olive oil. Concurrent to this, gently fry spinach with garlic, rosemary, sage, & mint. 23:17:30 2. pull mushrooms off when tops are done. pull spinach off when it's thinking about becoming paste, but not reaclly comitted to it. 23:18:25 3. stuff mushrooms with spinach thingie, goat cheese, and oregano. cap 'em, toss them back in the oil; top with baby swiss. 23:18:32 4. serve on pasta/ 23:18:42 this is so much more rewarding that coding. 23:18:52 s/that/than/ I can't type this week. 23:20:44 so apache2 is final. everyone see that? 23:21:04 finally? 23:21:04 yeah, that's cool 23:21:08 what's it been? 3 years? 23:21:18 probably. 23:21:37 but mod_perl isn't done, which is kind of funny. 23:21:54 because no one can read anyone else's code 23:22:13 I'm not sure whether Jakarta or mod_perl has the largest installed base, but mod_perl is pretty critical in apache land. 23:22:28 yeah 23:23:07 who uses perl for web programming anymore, other than the slashdorks? 23:23:12 talli, I'm willing to bet it's because perl 5.6's threading implementation is so terrifying. 23:24:34 how does perl6 look? 23:24:42 big 23:25:21 whoah. debian woody is heading for release. 23:25:59 perl6; I don't know about the language definition, but I know a bit about the VM. 23:26:28 will perl6 be a big improvement? 23:26:59 Dan Sugalski runs that, and although he's pretty unexposed to modern language features, he's getting a lot of input from academics. so there's that. 23:29:13 perl6 will be an enormous improvement in the runtime, but I don't know about the language. Once again, the world will be suprised by Larry Wall's perverted monster brain. 23:30:19 is wall that big of a smartie or is he just a freak? 23:31:16 freak, mostly. I don't use perl anymore, but you have to concede that he changed all the rules in language design. 23:31:53 are you just using ruby now? 23:33:05 tcl, lisp, ruby, C, python. in that order, I think. 23:34:58 actually I'll shelve ruby if matz doesn't kick out 1.8 or 2.0 any day now. which isn't likely. 23:36:15 again, whoa. anyone fooled with Subterfuge? 23:36:54 markd2 has quit ("wheeeee") 23:38:24 http://subterfugue.org/ 23:38:24 A: http://subterfugue.org/ from denshi 23:38:34 A: Sysadmins Strike Back 23:38:34 added comment A1 23:42:05 python seems to be the language of the day 23:42:12 kinda like perl a few years ago 23:44:36 python is lisp without symbols, macros, or codewalking. 23:45:34 http://www.norvig.com/python-lisp.html 23:45:34 B: http://www.norvig.com/python-lisp.html from denshi 23:46:58 B: Python for LISP programmers 23:46:58 added comment B1 23:51:39 which on the whole bodes well for me. 23:56:27 that subterfuge stuff is some crazy shit. 23:57:15 arbitrary system call permutation filters written in a high-level language..