00:07:14 hazmat has left #openacs 01:00:14 denshi has quit () 02:49:55 abbaJ (~jabba@adsl-65-65-96-146.dsl.austtx.swbell.net) has joined #openacs 03:26:39 davb (dave@alb-24-58-162-46.nycap.rr.com) has joined #openacs 03:27:17 davb has quit (Client Quit) 05:07:58 rbm has quit (Read error: 113 (No route to host)) 05:16:29 til has quit (Read error: 60 (Operation timed out)) 05:19:27 Psychephylax has quit (Read error: 110 (Connection timed out)) 06:24:00 rbm (rmello@fslc.usu.edu) has joined #openacs 06:44:04 andyn has quit ("leaving") 06:47:22 hazmat (~ender@adsl-66-123-57-58.dsl.lsan03.pacbell.net) has joined #openacs 07:24:22 dlk (dlk@walter.ita.chalmers.se) has joined #openacs 08:03:36 dlk has quit (Read error: 104 (Connection reset by peer)) 08:03:46 dlk (dlk@walter.ita.chalmers.se) has joined #openacs 08:04:08 dlk has quit (Read error: 104 (Connection reset by peer)) 08:04:34 dlk (dlk@walter.ita.chalmers.se) has joined #openacs 08:05:04 dlk has left #openacs 08:19:57 dullk (dlk@walter.ita.chalmers.se) has joined #openacs 08:20:51 dullk is now known as dlk 08:28:04 hazmat has left #openacs 10:14:42 dlk is now known as dlk-lunch 11:57:21 til (til@port-212-202-128-197.reverse.qsc.de) has joined #openacs 12:06:44 dlk-lunch is now known as dlk 12:44:49 dlk has left #openacs 13:37:32 davb (~dave@rrcs-nys-24-97-22-204.biz.rr.com) has joined #openacs 13:41:10 Psychephylax (~nickb@pool197.eng.cv.net) has joined #openacs 13:48:43 hi paje 13:48:43 salut, davb 13:48:47 hi Psychephylax 13:51:24 Hey all 13:51:28 hi paje :P 13:51:37 Dumb bot :-/ 14:12:45 Psychephylax: what's up? 14:26:01 Nothing much...just writing documentation 14:26:49 ooohhh fun. 14:37:34 Actually, according to my boss I am good at it 14:39:23 great 14:39:25 brb 14:39:28 davb has quit () 14:40:56 davb (~dave@rrcs-nys-24-97-22-204.biz.rr.com) has joined #openacs 14:41:03 jim (~jim@12-233-187-5.client.attbi.com) has joined #openacs 14:41:10 hi jim 14:41:13 hiya 14:41:37 Wow, google is getting really personal. It keeps telling me I am brilliant! 14:41:58 in a googly sort of way? :) 14:42:32 no, it says right at the bottom, You are brilliant, we are hiring :) 14:43:25 ok, seems my machine is more stable now... traced some instability to a bad 256mb ram card 14:44:06 that is no fun. can you send it back for a new one? 14:44:13 maybe 14:44:19 good luck. 14:44:25 I would hate to lose 256mb of ram. 14:44:57 I have 2 others, I swapped with the one that was also in the machine (pulled it out to try testing stability) 14:45:11 ah. 14:45:18 one is untried/unused... might put that one in too 14:45:53 there's not really a need tho, I'm not running oracle anymore 14:46:19 right. 14:48:03 you think we should still wait for k2pts? I'm still thinking the ideas are far enough apart to have different contracts 14:50:33 which ideas... please refresh me :) 14:52:04 he has this idea of having more than just delete-package-instance, so you can more finely granulate packages for the purpose of letting users (as opposed to coders) do more wrt composing packages... 14:52:28 so there would be a package manipulation sc? 14:53:04 he wrote acs-sc so I think is overlaps something he planned on doing with it :) 14:53:08 my idea is to get delete-instance by itself in a separate contract, so we can add stable deletes to oacs 14:53:40 right. What other sort of operations was he thinking of? 14:55:39 on the one hand, delete is an operation that would operate on package instances, so to be more truely oop, one should include ops like mount-instance, unmount-instance, new-instance, set-instance-properties, get-instance-property 14:56:39 on the other hand, delete-instance is needed -quickly- in order to ensure stability of installations 14:57:43 right. Don didn't seem to concerned about it as a problem that would hold up a release. 14:58:27 I keep vascelating on whether I agree with that... 14:58:39 sure, but you are not in charge :) 14:58:50 yes, I know that :) 14:59:13 Could we get away with a check in the delete package code, that says, sorry can't do that because this instance contains objects. 14:59:24 I really wish stability were a higher priority 14:59:33 define stable... 14:59:35 :) 14:59:43 every link works would be nice. 14:59:47 existing links always work 14:59:57 have you seen the testing server? that is one of the tests. 15:00:19 no, sure haven't... 15:00:21 Problem is, we inherited a half-baked toolkit and it is over a year past due. 15:00:28 let me find the link. 15:00:53 http://213.107.207.131:8000 15:01:18 you can register to look at the tests. you can volunteer (send email) to get write access to add tests or enter test data. 15:02:08 does k2pts know all the operations he wants in this sc? We could set them up, but not worry about implementations until later. 15:03:28 ie. if an implementation doesn't exist, use the existing code. that would be like a default implementation. 15:03:32 is it possible to ask if part of a contract is implemented? 15:03:56 yes, you can see if a certain implementation of an operation exists. 15:03:56 is it possible to say, this contract includes that contract by reference? 15:04:14 I am not sure what that means, so I can't say :) 15:10:08 abbaJ has quit (Remote closed the connection) 15:10:20 looking at java, an interface can extend another interface... 15:12:08 but I don't completely understand all the connections and terms used in sc, partly because many of them don't have equivalents in oacs 15:13:15 or if they do, either they're not named the same or I didn't know about them up to now... 15:14:07 I've read the docs a few times... btw, they say there are some examples... but I haven't seen them in the package's fiels 15:15:57 I think until I have a stronger understanding of it, some of the things I say could be misinformed comments... 15:16:25 k2pts (~nkd@213.149.180.151) has joined #openacs 15:16:29 hey jim 15:16:32 heya 15:16:33 seen the chatlogs 15:16:33 I haven't seen 'the', k2pts 15:16:58 I didn't have time to look at it very deeply, but here's what I suggest: 15:17:22 I wanna know.... have you evah seen the rain? 15:17:28 seen the-rain 15:17:28 I haven't seen 'the-rain', jim 15:17:44 implement the delete instance as you planned to and then we can incrementally improve it by specifying a more 15:17:59 general contract that makes use of the "old" contract (delete_instance) 15:18:08 part of the reason I suggest you do that is that 15:18:18 dotLRN has something that's called an applet... 15:18:25 so a contract can include another by reference? 15:18:32 in that case, we can start 15:18:35 you can do something that's similar to applets with package instances 15:18:47 no (re: reference) but you can have 15:19:02 a delete_instance operation that can be implemented for all contracts using the old one 15:19:06 does that make sense? 15:19:25 davb: are you around? 15:19:36 so, you're saying have 2 contracts that are separate? 15:20:07 they could easily merge. one of the greatest advantages with acs-sc is that you can incrementally improve the contracts without lost of generality 15:20:57 you can also include more operations in the future 15:21:03 in a contract 15:21:17 does that make sense? 15:21:19 so it's possible to extend a contract? 15:21:24 I think so... 15:21:26 "yes" 15:21:52 lemme get some basic info since you're here... 15:21:59 ok go ahead 15:22:12 the docs say there are a pair of examples... where can I find those? 15:22:31 check the search package and the openfts-driver 15:22:41 the search package specifies two contracts, namely 15:22:41 ok 15:22:51 FtsEngineDriver, FtsContentProvider 15:23:13 so openfts-driver includes an implementation for FtsEngineDriver 15:23:31 also, the acs-content-repository package has an implementation for FtsContentProvider 15:23:52 basically, a contract is a list of operations 15:23:59 ok 15:24:01 while an implementation is a binding of the operations of 15:24:09 a contract with specific/actual functions 15:24:29 the major advantage is that instead of calling openfts__search (for example) you call 15:24:32 how do arguments get passed to contract operations? 15:24:56 table that question... 15:24:57 call FtsEngineDriver "openfts" .... 15:25:26 jim: didn't understand the question 15:26:25 I'm trying to get my head around how sc fits in generally 15:26:59 it's a way to provide orthogonal support for integration...let me explain 15:27:10 ok 15:27:12 take a search engine as an example 15:27:50 Google! 15:27:57 :-D 15:28:06 the search package could have been tightly integrated with openfts, instead the search package is implemented using calls to FtsEngineDriver and FtsContentProvider 15:28:38 that way you can change your full-text search engine (htdig, swish, openfts) according to your needs...you need not implement a new search package for each one of these 15:28:46 jim: does that make sense? 15:28:53 yes... 15:28:58 let me ask this: 15:29:21 is it possible for more than one package to include an implementation of a particular contract? 15:29:33 sure...that's the point exactly 15:29:45 for example, dotlrn has something that's called an applet 15:29:53 an applet is a contract specification 15:30:27 there are several implementations of an applet (a portlet for instance is an applet) 15:30:52 and not only do you implement the operations specified in the contract, you make a claim that the entire contract is implemented? 15:31:02 in turn a portlet could be a contract in itself and then have bboard-portlet, file-storage-portlet, calendar-portlet, events-portlet...etc 15:31:18 hi k2pts! 15:31:19 yes that's one of the things that need be improved 15:31:21 hey davb 15:31:33 davb: I have a question for you. give me a moment 15:32:29 jim: by improvements we could loosen the contstraing of having to implement the whole contract...you could have optional operations... 15:33:04 davb: I think I missed one stylesheet when I send you the package and it's very important 15:33:28 just as in java, if a class implements an interface, it does so not only by including all the methods in the interface, but also says "implements name" in the class header 15:33:42 it's called to-10.xsl and it will automatically transform feeds from 0.9 or 0.91 (also tried it with 0.92) to 1.0 15:33:46 jim: more or less yes 15:34:40 k2pts: look at the instance_delete thing for a moment... 15:35:12 but note that java isn't a message redirector (this is a powerfull pattern in a language) while acs-sc is a message redirector, you could implement everything on top of acs-sc and still have something that is as fast as before but more flexible...(this is what was done with dotlrn)...I mean you can change any part that is implemented with acs-sc at any point without any side-effects 15:35:33 davb: still there? 15:35:33 there are none. 15:35:35 k2pts: ok. I need that stylesheet. 15:35:56 ok, let me email it to you...I had to change something to work 15:36:01 there would be an operation, instance_delete(package_instance_id) 15:36:26 davb: one problem is that it works with xsltproc but not with ns_xml...I have to explain this for a minute 15:36:58 and a contract that encloses it, let's call it InstanceOps 15:38:26 with the idea being that eventually every package would implement this contract if it wanted to be able to delete a package instance, and specify how that would be done in a manner local to the package's data model... 15:39:08 davb: sent...note that it will work if you specify an http address for the two stylesheets inside the import tags...I just needed to make it work with "file://" instead...when I use "file://" instead of "http://" it works with xsltproc but not with ns_xml...I'm probably missing something, so I could use a pair of eyeballs on this one :) 15:40:11 then in the sitemap, you'd have the unmounted applications link, which would lead you to the list of unmounted packages... 15:40:15 k2pts: i think you are right. I had this problem with the stylesheets for chump. I think it is the way ns_xml implements the import. 15:40:52 jim: right, you can have a contract named Package with only one operation initially, InstanceDelete. Then you can add more... 15:41:20 when you need to delete a package instance...instead of calling bboard_delete_instance...you would call: 15:42:06 Package bboard InstanceDelete 15:42:07 that is 15:42:27 acs_sc_call contract_name implementation_name operation arguments 15:43:31 jim: still there? 15:43:31 there are none. 15:43:42 it would be useful if the type of instance to delete could be queried from the acs_object.object_type field... 15:44:21 yes you 're right but... 15:44:24 then you could say delete_instance(1234), it would look up 1234 to see if it's an instance, see what kind of instance, and dispatch 15:44:29 packages doesn't use object_types 15:45:19 so when your contract is package-specific (i.e., it uses package keys to select the implementation) you use the package keys...otherwise you use the object_type 15:45:35 the two examples I have offered earlier demonstrate this... 15:45:38 i.e. 15:46:02 FtsEngineDriver distinguishes between its implementations using the package-key (e.g. openfts-driver) 15:46:03 while 15:46:26 FtsContentProvider is based on the object type (e.g. bboard_message, note, etc) 15:46:49 can the package key be always gotten from the id of a package instance, according to the present data model? 15:47:13 yes, this is what you should do... 15:47:47 you will call the InstanceDelete(instance_id) but in order to select which implementation is the right one you should also use the package key 15:48:02 for example, you have two packages 15:48:03 bboard 15:48:06 and notes 15:48:10 ok 15:48:16 you have the following instances for bboard: 15:48:18 1234 15:48:22 1235 15:48:24 1236 15:48:30 and the following for notes: 15:48:34 9876 15:48:36 9875 15:48:38 9874 15:49:04 you have a page that deletes an instance called iamthepage-for-deleting-an-instance.tcl 15:49:22 inside that page, you'll have one call that looks like: 15:49:51 find the package_key for this instance (using the instance_id) 15:49:56 and then call: 15:50:29 acs_sc_call "ApmPackage" $package_key InstanceDelete $instance_id 15:50:32 more or less that's it 15:50:54 does that make sense? 15:50:59 yes... 15:51:02 ok 15:51:06 I have to head out 15:51:14 ok... 15:51:14 I have a cold these days... 15:51:19 thanks; 15:51:38 feel free to email me any questions...I'm a bit late answering mail these days but I hope I have your sympathy 15:51:43 one sec... 15:51:44 davb: we talk about rss soon... 15:51:46 sure 15:52:19 in your call, what is "ApmPackage"? is that something that exists now? 15:52:21 k2pts:ok. I hope you feel better. 15:52:38 yeah, meetoo 15:53:19 jim: ApmPackage is the name of your contract (I assumed that that would be the name that you give to your contract) 15:53:41 then you use $package_key to identify which implementation to use 15:53:53 I might have two, since a Package is different from a PackageInstance 15:54:00 and then call the implementation for InstanceDelete with argument $instance_id 15:54:15 I don't think it's a good idea 15:54:24 it's still one package 15:54:41 you could have operations: InstanceDelete, TheWholePackageDelete 15:55:16 does that make sense? 15:55:21 PackageDelete would then iterate InstanceDelete, then take itself 15:55:37 that's one way to do it 15:55:54 or you could have an operation that needs to be implemented 15:56:15 I mean write specific code as you do for instance delete 15:56:39 inside your implementations 15:57:09 ok, have to head out...we talk later guys 15:57:25 my life is a mess past couple of months...I'm still trying to sort things out :) 15:57:41 best wishes, l8r 15:57:47 k2pts has left #openacs 15:58:14 davb: so is that good enough to start planning? 16:00:45 jim: package_key really is the name of the package. see apm_packages table. I have 6 different instances of edit-this-page. the package-key is always edit-this-page, but the package_name and package_id are different. 16:01:17 I see... 16:01:35 so that the example makes a little mroe sense. 16:02:44 what would cement this together, is if I had the id of an instance, can I always get the package key, even if the package is not mounted anywhere? 16:03:24 jim: I think for a wholePackageDelete you would directly call the procs you created for InstanceDelete instead of calling a service contract. it is perfectly ok for a package to have knowledge of its own internal representation. 16:03:27 jim: yes. 16:04:07 so we have the pieces... 16:04:33 site_nodes is the table tha tkeeps track of what is mounted where. 16:07:06 brb 16:07:34 ok, would you think we should have two operations in the contract initially, one for packages, one for package instances? 16:07:39 ok 16:12:46 doesn't the drop script take care of totally removing a package? 16:13:06 that works. if it doesn't we need to fix the drop scripts, so I think we can hold off on that. 16:15:10 heh. over the the bboard, simon, the testing organizer suggests we just kill/drop all the _really_ bad packages. 16:15:28 I pretty much agree. we can add them in one at a time as they are fixed or rewritten. 16:15:52 the whole point of the package system is to allow independent development of packages to enhace the mostly stable core. 16:17:51 which ones are bad/ 16:17:52 ? 16:18:30 not sure. I haven't used most of them. 16:18:33 (wp-slim is much better with the patches for allowing html, for example) 16:18:54 right. wp slim is poorly implemented, but actually does what it is supposed to do. 16:19:31 it would be nice if its fixed to use the CR properly. 16:19:45 maybe there aren't that many really bad ones. 16:19:47 that's what your patches do tho, yes? 16:20:09 well, I'm with you, I haven't used many of the packages.... 16:20:21 they just make it minimally work. the content_types, etc are all wrong if I remember correctly. or it could have been another package :) 16:21:39 for example, wp0slim worked great by itself..until you install another package. 16:22:04 * davb can't wait for dotlrn 16:22:25 we worked to separate instances tho, yes? 16:22:29 now I just need to sneak a linux box into work here. 16:22:36 jim: yeah, it works fine now. 16:22:36 heh :) 16:22:49 not too hard, the IT guy has no problem with it. 16:23:16 I figure I will get a demo running on my home machine and show them what it can do before asking for a linux server. 16:24:47 * rbm eyes davb's post to the aolserver list 16:25:04 morning btw 16:25:08 hi rbm 16:25:20 hey davb 16:25:50 heya rbm 16:25:53 I dreamt with my dad last night, now I'm all homesick 16:25:57 we got out first virus last night! (on the new win2k server) 16:26:13 maybe a visit is in order 16:26:21 jim: to my dad? 16:26:24 yeah 16:26:46 He's a bit far, in Manaus, Brazil :) 16:26:51 ahh 16:26:59 I would take the bus but it'd take a while to get there 16:59:29 Hey Roberto 17:01:17 * rbm reads about the security breach 17:01:19 hey Psyche 17:05:29 has anyone tried webmail? 17:05:42 (well, lately? :) 17:06:20 brb...need to reboot 17:06:28 Psychephylax has quit (Read error: 104 (Connection reset by peer)) 17:18:44 jim: no 17:21:06 I am waiting for shagster to write a new one :) 17:21:32 Ouch! I typed "/list" accidentally 17:21:37 paje: are you alive? 17:21:37 no idea, rbm 17:21:43 paje: :) 17:21:43 rbm: huh? 17:21:47 paje: botsnack 17:21:47 :) 17:33:18 wow. i didn't know this: "For Steve Ball, the "most significant advantage" of Tcl compared to other programming languages remains internationalization: "Tcl rocks when it comes to Unicode and handling/manipulating character encodings." 17:38:12 Who's steve ball? 17:38:46 big XML in TCL guy. 17:39:37 he wrote TclXML 17:39:39 and TclDOM 17:39:39 okay 17:39:55 I was trying to get him involved with OpenACS :) 17:40:00 and? 17:40:15 he said it was in the future, but I haven't heard anything in awhile. 17:40:32 apparently he has paying clients, so is too busy to take on any more projects. 17:40:54 I should learn C. 17:41:13 grab the white bible 17:42:01 The C Programming Language? 17:42:10 its on my list :) 17:43:49 also grab a primer and "The C Companion" by Holub 17:44:12 the latter is an advanced adjunct to a primer 17:44:19 I was going to get the C Primer Plus. highly rated at amazon. 17:44:26 I have a lot of stuff in my TOLEARN list. Too bad school doesn't leave me much time to actually learn useful things 17:44:35 that's a good primer 17:45:27 hopefully I will start school learning useful things RSN. 17:45:31 cool. 17:45:36 davb: Got a URL? I have a $25 gift certificate at Amazon I could use 17:45:42 davb: RSN? 17:45:43 sure... 17:45:46 Real Soon Now 17:45:56 * jim played a little with oacs-4.5's cms... nothing crashed! 17:46:24 http://www.amazon.com/exec/obidos/ASIN/0672322226/qid=1017337572/sr=8-3/ref=sr_8_67_3/103-0202080-3978231 17:46:24 A: http://www.amazon.com/exec/obidos/ASIN/0672322226/qid=1017337572/sr=8-3/ref=sr_8_67_3/103-0202080-3978231 from davb 17:46:30 A:|C Primer Plus 17:46:30 titled item A 17:46:35 A: there is also a C++ version 17:46:35 commented item A 17:46:43 jim: excellent. 17:47:16 and your head didn't explode? that is an overly complex interface. it exposes all the nasty bits underneath. 17:48:16 jim: do you have an author on the C Companion? I can't seem to find it. 17:48:24 Why are people saying "an easy to read introduction to C++" for a C bok? 17:48:27 s/bok/book/ 17:48:49 Allan I. Holub 17:49:41 I think amazon is broken. those reviews belong to the C++ version. it incorrectly links those as different editions of the same book. 17:49:51 you can preview it over at safari.oreilly.com 17:50:06 how can a 400 page book, be a "pocket" book? 17:50:09 preview what? 17:50:14 http://www.holub.com/aihbooks.html 17:50:14 B: http://www.holub.com/aihbooks.html from jim 17:50:35 B:|Books written by Allen Holub 17:50:35 titled item B 17:51:02 damn, its out of print. 17:51:29 I can get it new or used from amazon "marketplace" though 17:51:41 how much is it? 17:51:56 $30 new, $20 or so used. 17:52:39 I bought his "Compiler Design in C"... 17:52:51 960 pages@?!?!?!1 17:52:57 * rbm goes out for a bit 17:53:31 I like this title better: http://www.amazon.com/exec/obidos/ASIN/0070296898/qid=1017337841/sr=1-3/ref=sr_1_3/103-0202080-3978231 17:53:57 ouch, the C primer is HUGE 18:01:32 bbl- lunch 18:21:47 ok i'm back. 18:34:07 hazmat (~ender@adsl-66-123-57-58.dsl.lsan03.pacbell.net) has joined #openacs 18:41:36 hi hazmat 18:47:36 ola 20:00:42 denshi (~chatzilla@adsl-216-62-223-193.dsl.rcsntx.swbell.net) has joined #openacs 20:01:08 hello, evil peoples. 20:01:24 evil? 20:01:40 evil is a dignified profession, rbm. 20:02:13 hi denshi 20:02:34 hello mr b. 20:02:53 denshi: oh 20:05:12 so is anyone actually doing something productive today? 20:06:48 nope. 20:07:18 * davb checks out the competition 20:08:28 I've given up hope on finding employ in the tech industry again. 20:08:38 now I'm looking to go into professional evil. 20:09:00 concentrated or unconentrated? 20:09:22 dunno... it seems to be a 'who you know' kind of field. 20:09:40 you know, the Z*PE web site has lot of links to stuff people are working on. 20:10:01 we need some pages where people can enter what kind of projects they are working on and get organized to build new packages etc... 20:10:09 hopefully openacs.org 4.5... 20:10:42 actually mostly talking, not alot of actual code... 20:11:52 ie. online education. links to a couple of implemented sites, but not any actual code projects I can see. 20:11:52 you know, that would take 30 seconds of an admin's time: just a static page with general-comments enabled. 20:12:16 yes. 20:12:21 or ETP 20:12:26 but don't tell talli. 20:15:19 davb: I wholeheartedly agree 20:19:35 markd2 (~markd2@166.102.41.109) has joined #openacs 20:19:45 hi paje 20:23:09 rbm: cool 20:23:16 I am just impatient :) 20:23:49 I was looking at Zope in case I can't sneak linux in here to see what me other options were that will run on windows. 20:29:02 can openacs-4.5 run under win? 20:29:20 howbout under apache+mod_aolserver? 20:31:24 I don't think it iwll run under apache+mod_aolserver 20:31:34 it doesn't do enough. 20:31:48 i think it would be easier to run AOLserver under cygwin 20:36:42 davb: I need to try compiling AOLserver under cygwin 20:37:17 I need to wait until I get a win2k box I can play with. 20:46:27 http://www.lrc.state.ky.us/2002rsrecord/hr256.htm 20:46:27 C: http://www.lrc.state.ky.us/2002rsrecord/hr256.htm from denshi 20:47:05 C: Kentucky HR256: "Encourage the purchase of a submarine to patrol the waters of the Commonwealth and search and destroy all casino riverboats." 20:47:05 commented item C 20:48:32 jim: are there some nsd modules that are used by openacs4.5 that don't compile natively on windows? libxml, notably? 20:49:04 denshi: they should compile under cygwin. 20:49:19 yeah, but cygwin's a performance hit. 20:49:28 I think libxml is part of cygwin isn't it? 20:49:35 denshi: Have you measured? 20:49:35 I don't know for sure... 20:50:08 I don't know if the performance hit is really significant. 20:50:09 I haven't tried libxml. just traditionally, cygwin binaries aren't as fast as native ones. 20:50:18 denshi: Again, have you measured? 20:51:41 Maintaining a windows version of an application may be much more expensive than the alledged performance hit. 20:51:44 if doable at all 20:54:44 take a gander at all those #ifdefs in the nsd source. it's not impossible. 20:55:12 denshi: maybe the modules you mentioned will compile/run under windows... but I'm still curious if oacs-4.5 would work using apache and mod_aolserver, even under linux 20:56:13 i thought petru was going to maintain mod_aolserver? 20:56:33 that was a while ago though... 21:00:19 jim: I think it is a safe bet that no, mod_aolserver doesn't have the new stuff that oacs4.5 needs. 21:00:49 I think ns_xml 2 has stuff so it will compile under windows... 21:01:16 yep, it seems to. 21:01:49 I have to see if the guy who wrote it will release the code so we can finish it up. ns_xml 2 has alot of nice new commands for xml. 21:02:49 jim: http://openacs.org/doc/openacs-4/win2k-installation.html 21:02:49 also aolserver 4.0 has removed all win32 code. 21:02:51 give it a whirl. 21:03:03 cool. I will try that. 21:03:23 meaning it won't run under win32? 21:04:08 jim: right. although I think someone was working on fixing it so it will work... 21:04:32 aha 21:04:34 http://acs-misc.sourceforge.net/ns_rel_beta2.html 21:04:51 binaries of aolserver 4 early beta code for windows 21:05:03 cool. comes with nsxml. 21:05:14 and all the other modules needed for aolserver. 21:05:57 so it must be possible to make it work. 21:08:44 cool its running! 21:09:23 spiffy. in win98 no less. 21:13:50 license the same? 21:14:33 sure, its just aolserver compiled. 21:14:41 he can't unGPL it :) 21:15:18 well, meaning for the new version 21:15:55 oh, afaik its the same. 21:16:34 darn, it crashed. 21:16:43 have to get a little bit better OS anyway... 21:20:27 bbl 21:20:30 davb has quit () 21:41:58 markd2 has quit ("Wheeeeee") 22:31:58 lalal 22:32:41 shalalalala 22:33:20 beeppie do-wha! 22:34:50 denshi: I wasn't talking especifically about AOLserver when I said "if doable at all". I've looked at the nsd sources and all those horrible #ifdef's are the reason why win32 support was dropped in AOLserver 4 22:37:46 Hey, anybody know if there is a current aolserver debian maintainer? 22:38:21 shagster: it's probably orphaned by now 22:38:52 shagster: I plan to take over that package 22:39:06 shagster: and make new packages for AOLserver modules (nsxml, etc.) 22:39:21 In fact, I plan to package everything necessary to run OpenACS 4 22:39:24 Ahh, okay. I just sent a message to the old maintainer ask if anybody was and if not, I would :) 22:39:44 rbm: I was going to start looking at that this weekend 22:39:59 Once I learn how to make a .deb package :) 22:40:27 So I guess I'm saying, need some help? 22:40:28 jim: BTW, I made my first ITP today 22:40:38 shagster: Help's always welcome 22:40:40 you did? :) 22:40:55 I'm not on debian-devel mailing list right now 22:40:58 Anybody know of a good gui builder for Python? Any toolkit 22:41:12 doesn't matter much if it's non-free. I just need something that works well. 22:41:24 shagster: you know where the readings are? 22:41:36 the readings? nope 22:41:40 jim: Yes, let me grab the url for you 22:42:22 shagster: www.debian.org/devel/ has lots of links 22:42:43 Okay, thanks. 22:43:02 rbm: the url of your package page? 22:43:26 yeah, I'll take a look at that :) 22:43:32 I'm going to start building (or linking) to AOLserver/OpenACS packages this weekend and needed the deb info 22:43:53 shagster: ?? 22:44:14 that is me 22:44:31 denshi has quit () 22:44:59 jim: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=140371 22:46:17 shagster: I don't understand what you meant by your last phrase 22:46:45 shagster: it might be nice to comb out all the modules into each their own package 22:46:54 Sorry, I was just trying to find out information about debian packages for AOLserver 22:47:11 Brent Fulgham is still listed as maint... 22:47:13 I wanted to add links at http://uptime.openacs.org/aolserver-openacs 22:48:25 ok, gotta go... 22:48:32 (idling( 22:48:36 )) 22:48:56 This weekend I'm going to start seeing what different 'package' formats need to be created (rpms, debs, solaris .pkg, etc) or linking to current packages 22:50:47 jim has quit (Remote closed the connection) 22:53:00 I'm working with Jonathan Marsden to make OpenACS 4 rpms and debs. He has been taking care of the rpms. 23:28:34 select timestamp('2002-03-23 04:00'); works under 7.1, not under 7.2 23:29:33 * til looks for postgres changelog 23:39:30 * The timestamp() function is no longer available. Use timestamp 23:39:31 'string' instead, or CAST. 23:40:38 and the only package that uses this is calendar, apparently. 23:45:09 * til walks over to the sdm