#topicmaps@irc.freenode.net log for 2004-03-24

This log is automatically generated by an IRC bot from the traffic on the #topicmaps IRC channel on the irc.freenode.net IRC server. This file has the traffic for 2004-03-24. If you have questions regarding this log, please contact larsga@ontopia.net.

00:48:50 grove grove has quit None (Remote closed the connection)
08:31:39 grove grove has joined #topicmaps
09:02:51 larsbot larsbot has joined #topicmaps
09:08:25 larsbot tmbot: url: http://www.ontopia.net/topicmaps/materials/tm-vs-thesauri.html
09:08:38 larsbot tmbot: title: Metadata? Thesauri? Taxonomies? Topic maps!
09:09:08 larsbot tmbot: comment: larsbot's paper for XML Europe 2004, about how topic maps compare to traditional information retrieval aids like metadata, taxonomies, and thesauri
10:34:55 arnarl morning
10:35:48 larsbot morning!
11:33:59 drrho drrho has joined #topicmaps
11:34:34 larsbot hi there, drrho
11:36:55 drrho hi lars!
11:37:45 drrho * drrho preparing again slides...:-|
11:38:09 larsbot not much fun, is it? :-)
11:38:15 drrho No.
11:38:16 larsbot * larsbot has to finish Extreme tutorial slides soon, too
11:38:24 larsbot plus 1000 words describing tutorial :-(
11:57:32 larsbot drrho: will you be in Amsterdam from the 14th? (sorry if Steve's already asked)
11:59:34 larsbot * larsbot -> squash
12:16:23 drrho drrho has quit None ("Client Exiting")
14:04:50 tonus tonus has joined #topicmaps
14:05:00 tonus %GabeW
14:05:04 tonus %arnarl
14:06:04 larsbot hi there
14:06:33 tonus hi larsbot
14:07:18 larsbot I don't think I
14:07:23 larsbot 've seen you before. are you new?
14:07:59 tonus not just new to this but totally new to irc, and just wandered in by accident
14:09:32 tonus i'm killing time before going to a conference where i may be supposed to know something about using the internet for conferencing
14:10:09 tonus so i thought i'd better see what sort of irc clients i have access to, and just launched konversation
14:15:36 tonus tonus has left #topicmaps ("Konversation terminated!")
14:16:46 lischke lischke has joined #topicmaps
14:16:50 lischke hi
14:17:19 lischke lars: what you think about my TMQL proposal with XPath and XQuery shortcuts?
14:18:07 lischke nobody answer except steve p.
14:19:34 larsbot * larsbot in meeting, sorry
14:19:47 lischke till then and much fun
14:20:16 larsbot I'll try to get back to you as soon as possible
14:32:18 eleusis eleusis has joined #topicmaps
14:48:56 lischke lischke has quit None ()
16:11:49 lischke lischke has joined #topicmaps
16:15:55 lischke hi
16:17:25 lischke lars: you have some minutes telling me what you think about my TMQL prop. ?
16:30:13 larsbot hi lischke
16:30:27 larsbot hang on. I'll take 30 secs to actually read it first :)
16:31:52 lischke :-)
16:33:00 larsbot ok. what you are proposing is what Jonathan Robie did in his "The Syntactic Web" paper
16:33:12 larsbot see http://www.idealliance.org/papers/xml2001/papers/html/03-01-04.html
16:33:49 lischke does he also use Xpath and XQuery shortcuts
16:33:52 larsbot yep
16:34:02 lischke which can be expanded to real XPath and XQuery?
16:34:18 larsbot it's XQuery functions
16:35:00 lischke with XPath its a little more complicated
16:35:17 larsbot XPath isn't powerful enough to do this
16:35:31 larsbot we've really considered this path quite a while ago and decided against it
16:35:41 lischke why
16:35:48 larsbot if you really want to convince people, what you should do is implement the TMQL use cases using this approach
16:36:11 larsbot see http://www.isotopicmaps.org/tmql/uc-solutions.html
16:36:37 larsbot my feeling is that doing a complex topic map query with XQuery, even with convenience functions, is just going to be too painful
16:37:09 larsbot from the point of view of the applications I've actually done, 5.2.2.18 is a really simple query
16:37:15 larsbot http://www.isotopicmaps.org/tmql/uc-solutions.html#q522.18
16:38:03 larsbot once you really have to do the full thing of traversing 7-8 associations recursively while tracking two different trees etc things get really hard
16:38:15 larsbot XQuery is well suited for XML, but it's not as well suited for a graph model
16:38:27 larsbot RDF is not using XQuery, for precisely the same reasons
16:38:42 larsbot even drrho, who loves XQuery, did a QL inspired by XQuery, instead of the approach you advocate
16:38:56 larsbot even Jonathan Robie was grumbling that his approach was too hard :-)
16:39:06 lischke ok...... just an idea
16:39:21 lischke did someone implement it?
16:39:58 larsbot Jonathan Robie actually did what his paper describes
16:40:04 larsbot he worked with real data and wrote real queris
16:40:20 larsbot I also know that Software AG built a kind of Omnigator on top of Tamino, using XQuery
16:40:49 larsbot however, one thing you should consider is whether you can take a different approach
16:40:53 larsbot take TMQL and compile it to XQuery
16:41:25 larsbot the user can write nice queries, and you can implement them without having to build an efficient query processor on top of a persistent backend you can't look inside
16:41:34 larsbot we do the same with tolog in the sense that we compile it to SQL
16:41:50 larsbot the important thing is to give the user something that's easy to work with
16:42:08 lischke ok im convinced
16:42:32 lischke when there will be a TMQL i think i could write the fastest query engine ontop of xtm4xmldb :-)
16:42:45 larsbot that's a challenge I'll happily accept :-)
16:42:51 larsbot in fact, any experiments you could do in this direction, especially based on the TMQL use cases would be very interesting
16:42:55 larsbot and they would definitely be looked at
16:43:11 lischke i didnt read all use cases, but whats with updating?
16:43:19 larsbot we're not doing that yet
16:43:21 lischke is there an update usecase
16:43:24 larsbot not yet
16:43:26 lischke thats really importand
16:43:38 larsbot it definitely is, but for now everybody's on their own, basically
16:43:47 larsbot nobody's done a QL for TMs with update capability yet
16:44:06 larsbot we are going to add it to tolog, and I expect we'll learn stuff from that that will eventually go into TMQL (whether it's based on tolog or not)
16:44:21 lischke than i should propose this, cause XQuery in eXist is able to update
16:44:22 larsbot experimentation in this direction would definitely be valuable
16:45:23 lischke steve has invited me to come to amsterdam, you think thats a good idea?
16:46:55 lischke how many ppl will be there?
16:47:08 larsbot I'm not sure how many there will be
16:47:26 larsbot * larsbot counts in head
16:47:32 larsbot anywhere between 10 and 20, I guess
16:48:16 lischke oh its a great honor to be invited...... i'm not sure i have to think about it
16:48:38 larsbot we're going to work on TMQL and TMCL for three days, and the existing ISO 13250 for two days
16:49:00 larsbot this is probably where the direction on TMQL and TMCL will be decided, so this is a crucial meeting
16:49:12 larsbot and probably having been there is going to give you a much clearer picture of what's happening
16:49:33 lischke is it also possible just hear what others say?
16:49:37 larsbot oh yes
16:49:42 larsbot you don't have to say anything
16:49:50 lischke but maybe i want :-)
16:49:51 larsbot in fact, you'll probably have to struggle a bit in order to get a word in :-)
16:50:08 larsbot you will be able to, but it may require a little effort :)
16:50:11 larsbot it's pretty informal
16:50:16 lischke when is deadline to say i will come?
16:50:25 larsbot you don't have to say anything before it starts
16:50:33 larsbot if you show up that's great, and if not then, well, you don't :)
16:50:43 larsbot it would be useful to know, but it's not necessary
16:51:00 lischke maybe its a good idea, we could drink one or two beers :-)
16:51:25 larsbot yep. there would be that, too :)
16:51:33 larsbot I certainly would be glad to get a chance to meet you
16:51:37 larsbot and amsterdam is a great place for beer
16:51:53 lischke ok thanx for answering my TMQL question, lets get back to work
16:51:56 lischke :-)
16:52:08 larsbot no problem
16:52:11 larsbot see you around :)
17:00:57 eleusis aloha
17:01:18 eleusis lischke dude, i'm not happy with the hypergraph source code >:|
17:01:50 eleusis the directory structure's a mess and the build.xml file doesn't make sense >_<
17:13:08 larsbot * larsbot has to go
17:13:11 larsbot see you all tomorrow!
17:13:14 larsbot larsbot has quit None ("Leaving")
17:19:37 lischke u have to use a programming environment like netbeans
17:19:53 lischke *its not my code :-)*
17:20:27 eleusis i know
17:20:50 eleusis but the build file doesn't make sense.. why the hell does it need to update from cvs every time it builds?
17:21:12 eleusis i was stuck with that problem last night, especially as the adsl connection died for half an hour
17:32:10 grove grove has quit None (Remote closed the connection)
17:32:23 lischke i dont use the build.xml, cause its not mine :(
17:32:30 eleusis heh
17:32:31 eleusis well
17:32:37 eleusis i did manage to get stuff to build
17:32:43 lischke with netbeans?
17:32:46 eleusis but then i had no idea how to use whatever was built
17:32:55 eleusis narh, with the build file
17:33:12 lischke sory i cannot help you with that, i just use it with netbeans
17:33:13 eleusis btw, i've dropped my TM editor shell idea for now :P
17:33:22 lischke why?
17:33:27 eleusis i've installed jython... python within java :D
17:33:51 eleusis jython allows me to have an interactive shell where i can import and interact with java classes :)
17:34:09 eleusis "from org.tmap import *"
17:34:10 eleusis :D
17:34:12 eleusis err
17:34:17 eleusis org.tmapi
17:43:20 larsbot larsbot has joined #topicmaps
17:44:52 eleusis wb lars
17:45:06 larsbot re :)
17:46:02 lischke eleusis : import org.tmapi.core.*
17:46:19 eleusis hmm
17:46:31 lischke or does it recursive import?
17:46:47 eleusis >>> from org.tmapi.core import *
17:46:48 eleusis >>>
17:46:51 eleusis that worked..
17:46:54 eleusis erm.. i dunno
17:46:58 eleusis maybe :P
17:47:36 larsbot try writing "locals()"
17:47:43 eleusis k
17:47:44 larsbot that will show you what variables you now have
17:47:56 eleusis >>> locals
17:47:56 eleusis <java function locals at 30195191>
17:48:03 eleusis whoa
17:48:10 eleusis way too much to paste
17:48:13 eleusis btw
17:48:32 eleusis how'd you know about this locals() thing, lars.. i wanted to know how to do that, earlier :D
17:48:50 larsbot have worked with Python for ~7 years :-)
17:48:59 eleusis hahah, no kidding
17:49:02 larsbot it's in the modules documentation, in the builtin functions section
17:49:03 eleusis excellent
17:49:12 eleusis * eleusis bookmarks lars as python guru
17:49:15 eleusis ;)
17:49:21 larsbot nah, not any more
17:49:30 larsbot haven't followed the 2.1/2.2/2.3/2.4 stuff closely enough
17:49:31 eleusis aww
17:49:37 larsbot I'm a python 1.5.2 guru :-)
17:49:40 eleusis ok
17:49:46 eleusis * eleusis re-bookmarks as python old-timer
17:49:49 eleusis :]
17:49:53 larsbot * larsbot now happy
17:50:00 larsbot [larsbot : python-old-timer]
17:50:05 eleusis lol
18:36:00 larsbot * larsbot got to run, this time for real
18:36:03 larsbot larsbot has quit None ("Leaving")
20:30:31 grove grove has joined #topicmaps
20:47:40 eleusis eleusis has quit None ("Leaving")
20:52:28 lischke lischke has quit None ()
23:27:29 grove grove has quit None (Remote closed the connection)