IRC log of dig on 2006-08-07
Timestamps are in UTC.
- 08:00:27 [tlr]
- tlr (n=roessler@ip-83-99-49-179.dyn.luxdsl.pt.lu) has joined #dig
- 11:53:17 [RalphS]
- RalphS (n=swick@30-7-207.wireless.csail.mit.edu) has joined #dig
- 12:50:35 [DIGlogger]
- DIGlogger (n=dig-logg@pink-panther.csail.mit.edu) has joined #dig
- 12:50:35 [leguin.freenode.net]
- topic is: Decentralized Information Group at MIT CSAIL (http://dig.csail.mit.edu/ )
- 12:50:35 [leguin.freenode.net]
- Users on #dig: DIGlogger RalphS tlr lkagal sandro DanC eikeon evoli crowell_work
- 13:43:38 [ryanlee]
- ryanlee (n=ryanlee@c-65-96-190-159.hsd1.ma.comcast.net) has joined #dig
- 13:58:36 [evoli]
- 'lo
- 14:02:59 [DanC]
- hmm... dig meeting today?
- 14:05:27 [DanC]
- I see a propsed agenda from Ralph, so I guess so.
- 14:05:43 [DanC]
- (subject seems to have a typo: "2006-07-07 DIG proposed agenda")
- 14:07:59 [RalphS]
- RalphS has changed the topic to: 07 Aug DIG agenda https://lists.csail.mit.edu/mailman/private/diggers/2006-August/000978.html
- 14:08:11 [RalphS]
- oops, indeed, a typo in msg subject
- 14:08:27 [DanC]
- DanC has changed the topic to: DIG (http://dig.csail.mit.edu/ ) meets today at 11:30aET (Mon, 7 Aug)
- 14:09:54 [DanC]
- Ralph, I got Zakim to read the agenda from http://www.w3.org/2001/sw/grddl-wg/weekly-agenda
- 14:10:36 [DanC]
- it reminded me that the telcon calendar doesn't use the RDF calendar vocab. I wonder if that's worth DIG time, or if that's just a W3C thing.
- 14:16:47 [evoli]
- DIGlogger: point
- 14:16:47 [evoli]
- I'm logging. I don't understand 'point', evoli. Try /msg DIGlogger help
- 14:30:55 [RalphS]
- nice, Dan, re: Zakim reading from http://www.w3.org/2001/sw/grddl-wg/weekly-agenda
- 14:33:53 [yosi_s]
- yosi_s (n=chatzill@w3cdhcp27.w3.org) has joined #dig
- 14:41:40 [lkagal]
- lkagal has quit ()
- 14:58:38 [jambo]
- jambo (i=jambo@30-6-112.wireless.csail.mit.edu) has joined #dig
- 15:00:25 [sheets]
- sheets (n=dsheets@31-33-99.wireless.csail.mit.edu) has joined #dig
- 15:29:49 [Zakim]
- Zakim (n=rrs-brid@homer.w3.org) has joined #dig
- 15:30:14 [RalphS]
- zakim, this is dig
- 15:30:15 [Zakim]
- ok, RalphS; that matches DIG_weekly()11:30AM
- 15:30:21 [Zakim]
- +MIT531
- 15:30:28 [yosi_lap]
- yosi_lap (n=syosi@30-5-64.wireless.csail.mit.edu) has joined #dig
- 15:30:34 [RalphS]
- zakim, mit531 has Yosi, Ralph
- 15:30:34 [Zakim]
- +Yosi, Ralph; got it
- 15:30:52 [RalphS]
- Meeting: DIG Weekly
- 15:31:10 [claudiof]
- claudiof (i=claudiof@31-35-17.wireless.csail.mit.edu) has joined #dig
- 15:31:14 [RalphS]
- Agenda: https://lists.csail.mit.edu/mailman/private/diggers/2006-August/000978.html
- 15:31:30 [RalphS]
- agenda+ Old action followup
- 15:31:46 [RalphS]
- agenda+ Followup to Fresnel discussion
- 15:31:52 [RalphS]
- agenda+ Around the table
- 15:32:05 [RalphS]
- agenda+ Tabulator work progress/plans
- 15:32:13 [RalphS]
- abenda+ dbview work progress/plans
- 15:32:24 [RalphS]
- agenda+ dbview work progress/plans
- 15:32:28 [RalphS]
- agenda+ tracker work status
- 15:33:15 [RalphS]
- zakim, mit531 also has Harvey, Rob, Sheets, Jim, Claudio, Jesus, Adam
- 15:33:16 [Zakim]
- +Harvey, Rob, Sheets, Jim, Claudio, Jesus, Adam; got it
- 15:33:50 [RalphS]
- zakim, take up agendum 1
- 15:33:50 [Zakim]
- agendum 1. "Old action followup" taken up [from RalphS]
- 15:33:51 [alerer]
- alerer (i=markus@30-5-79.wireless.csail.mit.edu) has joined #dig
- 15:34:01 [Zakim]
- +DanC
- 15:34:08 [crowell]
- crowell (n=crowell@31-35-95.wireless.csail.mit.edu) has joined #dig
- 15:34:24 [harveyj]
- harveyj (n=harveyj@30-5-253.wireless.csail.mit.edu) has joined #dig
- 15:34:29 [crowell_work]
- crowell_work has quit (Remote closed the connection)
- 15:34:40 [RalphS]
- [DONE] ACTION: Ryan identify the important parts of breadcrumbs on swada to be backed-up
- 15:34:48 [RalphS]
- [DONE] ACTION: Ralph request the parts of swada identified by Ryan be added to T!Gs backup plan
- 15:34:56 [RalphS]
- ACTION: Ralph invite an Alloy person to demo in this meeting
- 15:34:57 [RalphS]
- -- continues
- 15:35:03 [RalphS]
- [PENDING] ACTION: Sandro to discuss fresnel/UI/etc with Ryan and TimBL and perhaps others; ETA: ~3 weeks
- 15:35:06 [RalphS]
- -- continues
- 15:35:29 [RalphS]
- ACTION: Sandro to discuss fresnel/UI/etc with Ryan and TimBL and perhaps others; ETA: ~3 weeks
- 15:36:18 [RalphS]
- zakim, take up agendum 3
- 15:36:18 [Zakim]
- agendum 3. "Around the table" taken up [from RalphS]
- 15:36:44 [DanC]
- fyi, tabulator issues by developer: http://dig.csail.mit.edu/issues/tabulator/issue?@columns=title,topic,id,activity,status,assignedto&@sort=priority&@group=assignedto&@filter=status&@pagesize=50&@startwith=0&status=-1,1,2,3,4,5,6,7&@dispname=Developer%20Status
- 15:37:55 [DanC]
- (I mostly failed to get what I needed via IRC last week; I needed phone contact.)
- 15:38:04 [RalphS]
- Adam: working on UI stuff, putting icons into different fields in the outliner
- 15:38:28 [Zakim]
- +RyanL
- 15:39:03 [sheets]
- (So maybe Thursday phone contact is important)
- 15:39:50 [RalphS]
- DanC: see issue 88
- 15:41:39 [RalphS]
- zakim, Lalana just arrived in mit531
- 15:41:39 [Zakim]
- +Lalana; got it
- 15:42:04 [RalphS]
- Jesus: work on hashing RDF graphs, talked with Yosi about this
- 15:42:22 [lkagal]
- lkagal (n=lkagal1@30-5-242.wireless.csail.mit.edu) has joined #dig
- 15:42:27 [RalphS]
- ... objective is to be able to quickly compare two graphs
- 15:42:50 [DanC]
- (Jesus's work is intended to contribute to performance issues in PAW, fyi, ralph. I learned this via Yosi)
- 15:43:14 [RalphS]
- Claudio: having problems fixing a bug when creating actions
- 15:43:43 [RalphS]
- zakim, Adam has left the room
- 15:43:43 [Zakim]
- sorry, RalphS, I do not recognize a party named 'Adam'
- 15:43:50 [RalphS]
- zakim, Adam has left mit531
- 15:43:50 [Zakim]
- -Adam; got it
- 15:44:04 [RalphS]
- Jim: working on UI update stuff, view area
- 15:44:16 [RalphS]
- ... finished putting in a button to expand and collapse query selection area
- 15:44:20 [RalphS]
- ... some bug squashing
- 15:44:26 [RalphS]
- ... expect to finish this today
- 15:44:38 [DanC]
- (http://dig.csail.mit.edu/issues/tabulator/issue21 is some part of Jim's work lately)
- 15:44:42 [RalphS]
- ... syncing Adam's sparql stuff with new changes
- 15:45:01 [RalphS]
- David: embroiled in source retrieval and how to best do this
- 15:45:21 [RalphS]
- ... moving from procedural architecture to event-driven
- 15:46:14 [RalphS]
- DanC: may be necessary to commit bits soon even if broken
- 15:46:37 [RalphS]
- Rob: trying to work postgres into dbview to export Roundup data
- 15:46:54 [RalphS]
- ... talked with Tim a bit about a dbview release soon, so working on code cleanup
- 15:46:55 [DanC]
- (dbview and postgress; I'm interested to know if that's a 10-minute-hack like one would hope)
- 15:47:06 [RalphS]
- ... will work on test cases this week, hope to borrow code from cwm
- 15:47:40 [RalphS]
- ... mysql and postgres are largely similar
- 15:48:07 [RalphS]
- ... reworking code to make it mroe modular
- 15:48:17 [RalphS]
- Harvey: working on TAMI interface for another day or two
- 15:48:30 [RalphS]
- ... then will look at Chris' scheme code and try to export it in ntriples
- 15:48:44 [DanC]
- agenda + adding scheme to the python/perl/java/prolog list consciously
- 15:49:32 [RalphS]
- Yosi: working on cwm bug interaction between canonicalization and regression testing
- 15:49:45 [RalphS]
- ... and another cwm bug that is elusive to reproduce
- 15:49:51 [RalphS]
- ... hoping to make a cwm release tarball
- 15:50:01 [RalphS]
- zakim, Joyce just arrived in mit531
- 15:50:01 [Zakim]
- +Joyce; got it
- 15:50:16 [RalphS]
- Yosi: starting to look into getting my new backward chainer integrated
- 15:51:49 [RalphS]
- Lalana: next steps for PAW not yet decided
- 15:52:13 [harveyj]
- RalphS, http://swiss.csail.mit.edu/users/harveyj/tami-demo/view.ssp?node-uri=s3:arrest-1 is a snapshot of the interface as of Friday
- 15:52:15 [RalphS]
- ... working on ISWC paper joint with Ora and Uni Bologna on context-aware and policy-aware middleware
- 15:52:30 [RalphS]
- ... hoping to finish Tabulator paper this week, due 11 Aug
- 15:52:31 [DanC]
- agenda + AAAI debrief (how did lk think the talk went? lessons learned?)
- 15:52:47 [RalphS]
- ... will send draft on 8 or 9th for comments
- 15:53:47 [RalphS]
- Joyce: working on calendar view and issue 69
- 15:54:04 [jambo]
- "calendar view of connolly's travel schedule gives lots of "No value ..." alert dialogs"
- 15:54:07 [RalphS]
- ... also working on tests and documentation
- 15:54:18 [DanC]
- (lkagal, I can't see any updates from you on the tabulator paper. I trust that's by design.)
- 15:54:33 [DanC]
- (last commit seems to be r1811 | connolly | 2006-07-31 08:11:37 -0500 )
- 15:54:34 [RalphS]
- ... am not sure what to do next on retrieving collections
- 15:54:37 [harveyj]
- and http://dig.csail.mit.edu/TAMI/cph/v2/deduce.scm has a log of a command-line interaction with the system in the comments on the bottom
- 15:54:48 [RalphS]
- ... would like help on understanding how sources retrieve collections
- 15:55:03 [RalphS]
- David: sources appear as arrays
- 15:55:09 [RalphS]
- ... collections are arrays of nodes
- 15:55:30 [lkagal]
- DanC, I'm going to work on the tabulator paper this week.
- 15:55:34 [RalphS]
- ... but you shouldn't need to worry about the representation in the store
- 15:55:46 [eikeon]
- eikeon has quit ()
- 15:55:55 [eikeon]
- eikeon (n=eikeon@dsl092-168-155.wdc2.dsl.speakeasy.net) has joined #dig
- 15:56:08 [RalphS]
- DanC: coaching tabulator work while Tim is out
- 15:56:57 [RalphS]
- ... challenging because lots of what Tim wants is "generic RDF browser" stuff
- 15:57:13 [RalphS]
- ... would like one phone call/week with each Tabulator developer
- 15:57:21 [RalphS]
- ... W3C GRDDL Working Group starting up
- 15:57:51 [RalphS]
- ... W3C GEO group starting up, poked around some French data, found lots of 404s, triggered a discussion on URI use
- 15:58:22 [RalphS]
- ... discussion thread in W3C Healthcare and Life Sciences interest group also on URI use
- 15:58:31 [DanC]
- The Research Libraray in the 21st Century event at U.T. Austin 11-12 Sep
- 15:58:32 [RalphS]
- ... getting frustrated with too many passwords
- 15:58:41 [RalphS]
- ... thinking about installing OpenID
- 15:59:12 [RalphS]
- Ryan: working on integrating openID with Longwell
- 15:59:25 [RalphS]
- ... thinking about running an openID server on simile.mit.edu
- 15:59:41 [RalphS]
- ... will be on vacation from Wed for 1.5 wks
- 16:00:13 [DanC]
- (sandro's action is done to my satisfaction, though perhaps not to sandro's)
- 16:01:05 [RalphS]
- Ralph: mostly administrative stuff but hope to talk about SwapMe plans with Ora and Deepali this week
- 16:02:11 [RalphS]
- close this agendum
- 16:02:20 [RalphS]
- zakim, take up agendum 2
- 16:02:20 [Zakim]
- agendum 2. "Followup to Fresnel discussion" taken up [from RalphS]
- 16:02:45 [RalphS]
- close agendum 1
- 16:02:58 [RalphS]
- Ryan: talked more with the IBM folk, not sure what their next steps are
- 16:03:17 [DanC]
- (which IBM folk? Elias?)O
- 16:03:31 [RalphS]
- ... Elias interested in Fresnel lenses
- 16:04:24 [alerer]
- alerer has quit (Read error: 110 (Connection timed out))
- 16:04:33 [RalphS]
- move to next agendum
- 16:04:33 [Zakim]
- agendum 4. "Tabulator work progress/plans" taken up [from RalphS]
- 16:05:00 [DanC]
- rel tabulator/4, timbl sent a "my understandings" msg that has a good summary
- 16:05:17 [DanC]
- From: Tim Berners-Lee <timbl@w3.org>
- 16:05:17 [DanC]
- To: tabulator@csail.mit.edu
- 16:05:17 [DanC]
- Subject: [Tabulator] My understandings
- 16:05:17 [DanC]
- Date: Tue, 1 Aug 2006 12:57:02 +0800 (Mon, 23:57 CDT)
- 16:05:34 [RalphS]
- DanC: TimBL would like another Tabulator release before the term starts
- 16:06:28 [RalphS]
- Sheets: yeah, release is on my plate
- 16:06:33 [RalphS]
- move to next agendum
- 16:06:33 [Zakim]
- agendum 5. "dbview work progress/plans" taken up [from RalphS]
- 16:07:20 [RalphS]
- Rob: tabulator examination of dbview results is reasonably good
- 16:07:24 [RalphS]
- ... that's the way I test
- 16:07:36 [RalphS]
- DanC: what about a chapter describing dbview in tabulator
- 16:07:53 [RalphS]
- Rob: I'd like to run dbview on dig/issues -- i.e. Roundup data
- 16:08:00 [RalphS]
- ... this would give people two views of the same data
- 16:08:12 [RalphS]
- move to next agendum
- 16:08:12 [Zakim]
- agendum 6. "tracker work status" taken up [from RalphS]
- 16:08:59 [DanC]
- (where does the tracker work live? hmm... msg from Ralph of 6 Jul refers to http://dig.csail.mit.edu/2006/tracker/rif/api/rdfdump ...)
- 16:10:24 [DanC]
- http://dig.csail.mit.edu/2006/tracker/rif/tracker.php
- 16:12:37 [RalphS]
- Claudio: I start working as an intern at IBM T.J. Watson in September
- 16:13:22 [DanC]
- hmm... SVN history of tracker.php seems to be contaminated by unix/dos line ending foo.
- 16:13:37 [RalphS]
- close this agendum
- 16:13:46 [RalphS]
- move to next agendum
- 16:13:46 [Zakim]
- agendum 7. "adding scheme to the python/perl/java/prolog list consciously" taken up [from DanC]
- 16:14:22 [RalphS]
- DanC: we've talked about integrating more programming languages
- 16:14:36 [RalphS]
- ... we're adding scheme to the mix
- 16:15:15 [RalphS]
- Harvey: the hairy scheme stuff is Chris and Jerry's code
- 16:15:39 [RalphS]
- DanC: it's best if people don't code alone
- 16:15:45 [RalphS]
- Harvey: Chris looks at my code
- 16:16:44 [RalphS]
- ... my main task for the next week or two is to get Chris and Jerry's reasoner exporting in ntriples
- 16:17:36 [RalphS]
- ... most of their code is part of their scheme distribution
- 16:17:49 [RalphS]
- move to next agendum
- 16:17:49 [Zakim]
- agendum 8. "AAAI debrief (how did lk think the talk went? lessons learned?)" taken up [from DanC]
- 16:18:44 [RalphS]
- Lalana: my talk went well, even though I planned for a 30-minute talk in 20 minutes
- 16:19:02 [RalphS]
- ... but I got no questions about why I used Semantic Web techniques
- 16:19:17 [RalphS]
- ... a company in Rhode Island expressed interest in applying our work
- 16:19:29 [RalphS]
- ... they want an access plug-in for a distributed group
- 16:19:43 [RalphS]
- ... they're visiting me tomorrow
- 16:20:09 [RalphS]
- ... I declined to sign a non-disclosure but they're coming anyway
- 16:20:34 [RalphS]
- DanC: I particularly liked your slide style
- 16:20:52 [RalphS]
- Lalana: I had a detailed example of representing policies in several ways but had to skip over it
- 16:21:12 [RalphS]
- Dan: I ran into Ben Kypers again
- 16:21:25 [RalphS]
- ... he didn't get to ask Tim his question
- 16:21:31 [DanC]
- Kuipers
- 16:21:37 [RalphS]
- s/Kypers/Kuipers/
- 16:21:56 [RalphS]
- ... Kuipers' algernon work lets you reason over data without having all of it available
- 16:22:19 [DanC]
- -> http://www.cs.utexas.edu/users/qr/algernon.html Algernon and Access-Limited Logic
- 16:22:23 [RalphS]
- ... algernon work not currently active but Kuipers noodling on possibility of reactivating
- 16:24:09 [RalphS]
- ... Larry Lessig said good things about Semantic Web at Wikimania
- 16:24:17 [ryanlee]
- Denny V.?
- 16:25:39 [DanC]
- ooh... looking for the wikimania blog entry I saw, I see tabulator screenshots from somebody in http://www.wshoy.sidar.org/index.php?2006/08/05/36-tabulator-un-navegador-para-la-web-semantica-parte-ii
- 16:25:45 [RalphS]
- next meeting: 21 August
- 16:26:35 [RalphS]
- DanC: DIG f2f?
- 16:26:52 [RalphS]
- Ralph: I have 31 August in my calendar
- 16:27:24 [DanC]
- I'm holding 28 Aug; I'm supposed to make flight arrangements 3 weeks in advance, which is today.
- 16:28:43 [Zakim]
- -DanC
- 16:28:48 [Zakim]
- -RyanL
- 16:28:52 [Zakim]
- -MIT531
- 16:28:53 [Zakim]
- DIG_weekly()11:30AM has ended
- 16:28:55 [Zakim]
- Attendees were Yosi, Ralph, Harvey, Rob, Sheets, Jim, Claudio, Jesus, Adam, DanC, RyanL, Lalana, Joyce
- 16:29:41 [DanC]
- ah... here's the Wikimania item that made me so jealous; and it's from the very same Denny V http://semantic.nodix.net/2006/08/maybe-hottest-conference-ever.html
- 16:30:48 [DanC]
- is Jesus on IRC? "Si para muestra hace falta un bot?n, aqu? el esperado ejemplo de uso de Tabulator.". that uses either vocabulary or verb conjugations that I can't work out
- 16:31:09 [DanC]
- muestra is something like "teach" or "show"
- 16:31:20 [DanC]
- "falta" is something like "lack"
- 16:31:29 [harveyj]
- harveyj has quit ()
- 16:31:42 [DanC]
- "Si" is either "yes" or "if".
- 16:31:49 [yosi_lap]
- yosi_lap has quit ()
- 16:32:07 [DanC]
- "If to show I lack a button..." perhaps? hmm.
- 16:32:26 [crowell]
- crowell has quit ()
- 16:32:48 [yosi_lap]
- yosi_lap (n=syosi@30-5-64.wireless.csail.mit.edu) has joined #dig
- 16:32:59 [DanC]
- hmm... "Si para muestra hace falta un bot?n, aqu? el esperado ejemplo de uso de Tabulator."
- 16:33:08 [DanC]
- ah... accents were getting mangled.
- 16:33:55 [DanC]
- I think "probablemente el ejemplo que m?s me ha gustado de momento" means "probably my favorite example"
- 16:56:38 [yosi_lap]
- yosi_lap has quit ()
- 16:57:29 [Zakim]
- Zakim has quit (Remote closed the connection)
- 17:02:02 [ryanlee]
- ryanlee has quit ()
- 17:02:13 [RalphS]
- Regrets: Danny, TimBL
- 17:24:40 [yosi_lap]
- yosi_lap (n=syosi@30-5-64.wireless.csail.mit.edu) has joined #dig
- 17:30:40 [ryanlee]
- ryanlee (n=ryanlee@30-6-75.wireless.csail.mit.edu) has joined #dig
- 17:36:57 [tlr]
- tlr has quit ("I")
- 17:37:32 [jambo]
- DanC, muestra hace falta un boton is really just kind of "to show a necessary example", i think, but I havent spoken spanish in about 3 years now :(
- 17:39:39 [DanC]
- hmm
- 17:40:45 [jambo]
- or rather, an unseen example
- 17:43:34 [jambo]
- But yeah, that says "probably my favorite example thus far"
- 17:50:43 [jambo]
- hrmm, with that example, the map zooms in further than google maps has data for... that is sort of aggravating, seeing as I make a call to their own function to see how far I can zoom in
- 17:51:49 [ryanlee]
- ryanlee has quit ()
- 18:02:21 [ryanlee]
- ryanlee (n=ryanlee@18.42.7.202) has joined #dig
- 18:12:51 [sheets]
- Re: content negotiation, 'q' seems to mean both quality of content and preference in negotiation -- which is correct?
- 18:49:09 [ryanlee]
- ryanlee has quit ()
- 18:52:20 [DanC]
- both, I think
- 18:52:58 [DanC]
- i.e. when the client sends, it's expressing a preference; when the server responds, it's expressing quality
- 18:53:58 [sheets]
- I am still interested
- 18:54:43 [sheets]
- http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html
- 18:54:59 [sheets]
- in 14.1 it states " The example
- 18:54:59 [sheets]
- Accept: audio/*; q=0.2, audio/basic
- 18:54:59 [sheets]
- SHOULD be interpreted as "I prefer audio/basic, but send me any audio type if it is the best available after an 80% mark-down in quality." "
- 18:55:25 [sheets]
- But this is at odds with what I've read elsewhere about q= being used for client preferences
- 18:56:21 [DanC]
- how is it at odds? it seems consistent, to me
- 18:57:34 [sheets]
- q= can't mean both "Send me content with a quality markdown" and "Send me this content after this other content"
- 18:57:58 [sheets]
- What if I prefer low quality audio over high quality audio?
- 18:58:19 [DanC]
- that's considered irrational
- 18:58:32 [DanC]
- i.e. the protocol can't express it
- 18:58:41 [sheets]
- Why?! If I am on a slow connection then it's totally rational...
- 18:58:47 [sheets]
- hmm
- 18:59:08 [sheets]
- It seems the protocol can express it if more than one attribute were used instead of just 'q'
- 18:59:15 [DanC]
- there was discussion of ways to express bandwidth limitations, but I think they didn't get standardized
- 18:59:23 [sheets]
- ok
- 18:59:45 [ryanlee]
- ryanlee (n=ryanlee@30-6-75.wireless.csail.mit.edu) has joined #dig
- 18:59:46 [sheets]
- It still seems wrong that 'q' has two different meanings
- 18:59:54 [DanC]
- I think the discussion is in the www-talk archives from '92 or '93 if you really want to see all the gory details. (or maybe in an http-wg archive)
- 19:00:28 [sheets]
- If I prefer one type of audio over another but I still want them at high quality "audio/foo; q=1.0, audio/bar; q=0.1" may not do what I expect
- 19:00:50 [DanC]
- er... audio quality is not what the protocol expresses..
- 19:01:08 [DanC]
- "quality" in the http spec refers to loss of integrity due to conversion from one MIME type to another.
- 19:01:51 [sheets]
- fine, but apparently it is also used for content-type preference ordering
- 19:02:05 [DanC]
- also? it's *only* used for preference ordering.
- 19:02:32 [sheets]
- "SHOULD be interpreted as "I prefer audio/basic, but send me any audio type if it is the best available after an 80% mark-down in quality.""
- 19:02:51 [sheets]
- Doesn't that mean that '0.8' is used for quality information too?
- 19:02:54 [DanC]
- that's not saying "after adding noise to the audio signal"
- 19:04:06 [DanC]
- that's saying "if you have audio/mp3 at q=1 and audio/basic at q=0.5 , send me the audio/basic, but if the audio/basic has q=0.1, then send me the mp3"
- 19:04:54 [DanC]
- the 0.5 vs 0.1 is at the discretion of the server/publisher. it's not a measurable thing.
- 19:05:21 [sheets]
- But the client is just using arbitrary 'q' values to rank their preferred content-types
- 19:05:38 [DanC]
- yes, and the server is using arbitrary 'q' values to rank their preferred content types.
- 19:05:56 [sheets]
- but these are not the same set of arbitrary values
- 19:06:03 [DanC]
- no?
- 19:06:24 [DanC]
- by virtue of the fact that both parties agree to the HTTP spec, they are.
- 19:06:53 [sheets]
- if the server says "audio/mp3 is 0.8 and audio.basic is 0.1"
- 19:06:58 [sheets]
- but the client doesn't know this
- 19:07:07 [sheets]
- and just uses arbitrary values for ranking purposes only
- 19:07:19 [sheets]
- then whatever semantics the server has assigned will be unknown to the client
- 19:07:25 [DanC]
- yup.
- 19:07:32 [sheets]
- and will be invoked with possibly incorrect results
- 19:07:47 [sheets]
- because 'q' expresses both ranking and quality
- 19:07:55 [DanC]
- well, results that are sub-optimal by some other measure, but they'll be correct per http
- 19:08:12 [sheets]
- so then it seems like the spec is brain-damaged here
- 19:08:35 [sheets]
- unless there is some way to untangle quality and ranking that I don't know about
- 19:08:35 [DanC]
- well, it has limitations. but you don't see that it works for most cases?
- 19:09:02 [sheets]
- It seems like the quality aspect of 'q' can and should be handled by other content-type attributes
- 19:09:05 [DanC]
- the spec treats preference and quality as the same thing. it's a simplifying assumption that works in most cases, no?
- 19:09:10 [sheets]
- and that most people just use it for ranking
- 19:09:21 [DanC]
- most people never use this part of HTTP at all.
- 19:09:36 [sheets]
- Sure... that doesn't really excuse it, though
- 19:10:12 [DanC]
- well, yes, it does: why should the spec be even _more_ complicated when, to first (or 2nd or 3rd..) order, noone will use the extra complexity
- 19:10:13 [DanC]
- ?
- 19:10:51 [DanC]
- it's always possible for to use more round-trips; the server can always respond with 406 "well, here's what I've got; you decide".
- 19:10:59 [sheets]
- I would argue it would actually be simpler: 'q' would mean one thing and one thing only
- 19:11:38 [DanC]
- it does mean one thing and one thing only: it tells you which variant to pick.
- 19:11:48 [DanC]
- you just don't like what it means. ;-)
- 19:12:17 [sheets]
- You're right, I don't
- 19:12:20 [sheets]
- because it's wrong
- 19:12:21 [sheets]
- ;-)
- 19:12:51 [DanC]
- well, hop back in your time machine and get involved in the design discussion. ;-) or change a billion web server installations ;-)
- 19:13:29 [DanC]
- don't hold your breath ;-)
- 19:13:49 [DanC]
- actually, there is some TCN/transparent-content-negotiation thing out there...
- 19:14:01 [DanC]
- ... it's complicated that I never looked at it closely. But maybe it appeals to you.
- 19:14:30 [DanC]
- http://www.ietf.org/rfc/rfc2295.txt
- 19:48:06 [yosi_lap]
- yosi_lap has quit ()
- 20:29:16 [RalphS]
- RalphS has quit ("bye for today")
- 20:59:20 [lkagal]
- lkagal has quit ()
- 21:14:21 [yosi_s]
- yosi_s has quit (Remote closed the connection)
- 22:04:12 [timbl]
- timbl (n=timbl@m615e36d0.tmodns.net) has joined #dig
- 22:05:13 [claudiof]
- claudiof has quit ()
- 22:14:14 [timbl]
- sourcefetch iasynchronicity makes it trickeir to debug than moset?
- 22:14:18 [timbl]
- ,ost?
- 22:18:06 [timbl]
- We have Leef as a user of the library ... cool.
- 22:25:44 [sheets]
- No, firefox crashes with a segfault. And yeah, LeeF seems to be a user. :-)
- 22:27:33 [timbl]
- segfault -- ouch ... where to they hang out on IRC, I wonder
- 22:28:30 [timbl]
- A segfaul should get you some moral support if not actual help and fixes!
- 22:41:08 [lkagal]
- lkagal (n=lkagal1@98.68.171.66.subscriber.vzavenue.net) has joined #dig
- 22:44:02 [timbl]
- Hi Lalana.
- 22:44:12 [timbl]
- remind me when the paper dedline is?
- 22:44:19 [timbl]
- kkaga
- 22:44:23 [timbl]
- lkagal
- 22:46:59 [ryanlee]
- ryanlee has quit ()
- 22:47:22 [sheets]
- timbl, apparently not irc.mozilla.org#firefox -- I got absolutely no response
- 23:02:50 [lkagal]
- timbl, the tabulator paper is due Friday, Aug 11th.
- 23:22:34 [jambo]
- jambo has quit (Read error: 110 (Connection timed out))