- 10:16
- *** Log started on 2001-05-24 UTC-0400 ***
- 10:16 [RRSAgent]
- *** RRSAgent (swick@seahorse.w3.org) has joined channel #dspace
- 10:17 [mick]
- I'm also logging to a file on my machine, but web log is great.
- 10:18 [RalphS]
- there will be delays in updating mirrored copies of the irc log, but it will eventually catch up.
- 10:21 [RalphS]
- (a real-time log should be available via slow1.w3.org )
- 10:25 [dstuve]
- *** dstuve (dstuve1@cv279394-a.rslnma1.ma.home.com) has joined channel #dspace
- 10:25 [mick]
- sync
- 10:27 [RalphS]
- RRSAgent intentionally does not record out-of-band comments
- 10:27 [mick]
- just checking :)
- 10:27 [ArtB]
- Mick - if you XChat as your IRC client, it will log all of the traffic on a channel.
- 10:29 [mick]
- mIRC does this as well.
- 10:29 [RobTan]
- *** RobTan (robert_tan@hplb.hpl.hp.com) has joined channel #dspace
- 10:29 [RobTan]
- Good morning/afternoon
- 10:29 [mick]
- OK, standing by for EricM, pbreton, etc.
- 10:29 [RobTan]
- We are assembling the Bristol team
- 10:30 [RalphS]
- hmm. not accepting meeting id
- 10:30 [mick]
- I just got on no prob
- 10:30 [mick]
- I'm the only one on voice, so fare
- 10:30 [mick]
- so far
- 10:30 [ArtB]
- +Mick, Ralph, Art
- 10:31 [bwm]
- *** bwm (brian_mcbr@phobos.hpl.hp.com) has joined channel #dspace
- 10:32 [bwm]
- who/what is RRSAgent?
- 10:32 [ArtB]
- +EricP
- 10:32 [ArtB]
- +DaveS
- 10:32 [mick]
- RRSAgent - realtime weblog http://www.w3.org/2001/05/24-dspace-irc
- 10:32 [ArtB]
- +DaveR, Brian, Rob
- 10:33 [bwm]
- Ah
- 10:34 [em]
- *** em (em@dhcp065-024-049-027.columbus.rr.com) has joined channel #dspace
- 10:34 [mick]
- Hi eric, ralph attempting to raise you by phone
- 10:34 [pbreton]
- *** pbreton (pbreton@M10-423-4.MIT.EDU) has joined channel #dspace
- 10:34 [pbreton]
- Hi all
- 10:34 [mick]
- Hi Peter.
- 10:35 [ericP]
- *** ericP (eric@24.218.164.136) has joined channel #dspace
- 10:35 [ArtB]
- +Peter, EricM
- 10:35 [bwm_]
- *** bwm_ (brian_mcbr@phobos.hpl.hp.com) has joined channel #dspace
- 10:35 [pbreton]
- Hi Brian
- 10:35 [ArtB]
- +Gabe
- 10:36 [bwm_]
- Hi Peter
- 10:37 [RalphS]
- -> http://lists.w3.org/Archives/Public/www-rdf-dspace/2001May/0019.html pre-reading
- 10:39 [ArtB]
- EricM: what toolkits have you checke?
- 10:40 [ArtB]
- Peter: Jena, EricP's perl, EOR, Sergey's RDF db mapping, SiRPAC
- 10:40 [ArtB]
- Peter: Sergey's db info is too low level
- 10:41 [ArtB]
- RalphS: wrt when the data store gets created - virtual or permanent store - for a couple of reasons I'm nervous about the virtural approach
- 10:42 [ArtB]
- ... its difficult to carry through the orig semantics to the triple store
- 10:42 [ArtB]
- ... recovering semantics is more work
- 10:43 [ArtB]
- ... approach 1a - doing things in paralle may be easier
- 10:44 [ArtB]
- ... the add-on will facilitate adding new properties and extensibility without having to change the main DSpace tables
- 10:45 [ArtB]
- EricM: pretty much agree with Ralph.
- 10:45 [ArtB]
- ... I applaud your approach.
- 10:46 [ArtB]
- ... value adding new relationships will be easier if you use a triple store approach
- 10:46 [bwm]
- *** bwm (brian_mcbr@phobos.hpl.hp.com) has joined channel #dspace
- 10:47 [ArtB]
- DaveR: the value of the triple store will be necessary to give more value
- 10:49 [ArtB]
- ericP: the value of a triple store is store extraneous data
- 10:50 [ArtB]
- Approach 1c = a hybrid of 1a and 1b
- 10:51 [ArtB]
- Brian: I'm concerned about storing parallel data - sync issues
- 10:51 [ArtB]
- mick: seems like 1a and 1c is the general direction.
- 10:51 [ericP]
- 1c: := generic triple store used as a complement to the set of relational databases
- 10:53 [ArtB]
- Peter: I'm concerned about tool duplicity and tool availability
- 10:53 [gabe_hp]
- *** gabe_hp (gabe_beged@correl1.hp.com) has joined channel #dspace
- 10:54 [em]
- - jena, eor, forth group
- 10:55 [ArtB]
- EricM: I understand the need to have language consistency.
- 10:55 [ArtB]
- Redland?
- 10:55 [RalphS]
- -> http://139.91.183.30:9090/RDF/ ICS-Forth RDFSuite
- 10:56 [bwm]
- Redland does not have Java support or relational database
- 10:56 [RalphS]
- -> http://www.redland.opensource.ac.uk/ Redland
- 10:57 [ArtB]
- Does Jena have db support?
- 10:57 [bwm]
- Not yet
- 10:58 [ArtB]
- Brian: I did a relational db mapping with SiRPAC about one year ago.
- 10:58 [ArtB]
- Peter: I've received some email about someone that did some db support for Jena.
- 10:58 [bwm]
- DAve wants to speak
- 10:59 [ArtB]
- Peter: which have a RDBMS backend?
- 10:59 [ArtB]
- em: EOR and ICS-Forth and some Jena work
- 11:00 [ArtB]
- em: EOR uses SiRPAC
- 11:01 [ArtB]
- DaveR: Ian Dickinson's DoJo work [HPL-Bristol]
- 11:01 [ArtB]
- ... DoJo - gives some DAML support and some query support
- 11:01 [em]
- url reference to dojo?
- 11:01 [em]
- sounds very interesting
- 11:02 [bwm]
- not yet
- 11:02 [ArtB]
- DaveR: need to think about schema support when making your decision
- 11:02 [bwm]
- thats understood
- 11:04 [R-Tansley]
- *** R-Tansley (robert_tan@hplb.hpl.hp.com) has joined channel #dspace
- 11:04 [ArtB]
- mick: so 5 Java possibilities: Jena, EOR, Forth, persistent SiRPAC, DoJo
- 11:04 [RalphS]
- -> http://nestroy.wi-inf.uni-essen.de/rdf/sum_rdf_api/ [undated] Summary of discussions on RDF API
- 11:05 [ArtB]
- Brian: I would remove persistent SiRPAC and replace that with EOR since my persistent SiRPAC work is not being worked on.
- 11:05 [RobTan]
- *** RobTan (robert_tan@hplb.hpl.hp.com) has joined channel #dspace
- 11:06 [R-Tansley]
- *** R-Tansley (robert_tan@hplb.hpl.hp.com) has joined channel #dspace
- 11:07 [pbreton]
- hi rob
- 11:07 [ArtB]
- === strategies for generating RDF ====
- 11:13 [ArtB]
- Beans == fielded records
- 11:14 [ArtB]
- GINF: http://www-diglib.stanford.edu/diglib/ginf/
- 11:14 [em]
- :) thanks art
- 11:15 [em]
- also notes: http://www.interdataworking.com/
- 11:16 [ArtB]
- Peter, after this meeting I would appreciate if you would summarize your view [both polls] and send it to rdf-dspace.
- 11:18 [ArtB]
- em, do you know if folks have used EOR to solve a similar problem that Peter is thinking about
- 11:19 [em]
- not eor specific... but what I think i'm hearing is what is helpful is a very general rdf toolset with a set of applications (in this case focusing on data creation) that benifit from a set of known predicates
- 11:19 [em]
- this in essense is why we focused on building eor
- 11:20 [ArtB]
- ericP: a viewpoint of the polls - at one end you've got a 1:1 mapping of the fields of a form to a RDBMS table
- 11:20 [em]
- but the the set of machine rules (as opposed to procedural) is not somethign that eor dealt with
- 11:22 [ArtB]
- ericP, what's your viewpoint of the other end of the poll?
- 11:23 [ArtB]
- em: may want to record information about what you don't know - it can be very useful
- 11:27 [ArtB]
- DaveR: want a followup on storage API
- 11:28 [ArtB]
- DaveR: how to model changes to a document
- 11:28 [RalphS]
- indecs
- 11:28 [em]
- www.indecs.org
- 11:28 [em]
- also the harmony model
- 11:28 [bwm]
- www.indecs.org
- 11:28 [em]
- may be of interest... event modelling
- 11:29 [ArtB]
- em: followup - this was a good start; willing to have more such meetings
- 11:29 [ArtB]
- gabe_hp: once loading is figured, how to merge the two separate dbs
- 11:30 [ArtB]
- ... how to query both without caller knowing about it.
- 11:30 [bwm]
- A good ref for Index event model is http://www.dlib.org/dlib/january99/bearman/01bearman.html, see fig 5
- 11:30 [ArtB]
- ericP: that where my research has been
- 11:31 [RalphS]
- ArtB: could Peter write a summary of this discussion and post to the mailing list?
- 11:31 [RalphS]
- ... would like EricP to post his thoughts on querying across different databases
- 11:32 [ArtB]
- KUTGW Dspace guys!
- 11:32 [ArtB]
- Adjorned ...
- 11:32 [mick]
- THANK YOU!
- 11:33 [ArtB]
- Gabe - how's life in MagicDisk land?
- 11:33
- *** Log ended on 2001-05-24 ***