IRC log of dig on 2007-02-13
Timestamps are in UTC.
- 01:33:19 [djweitzner]
- djweitzner (n=djweitzn@pool-70-108-167-73.washdc.east.verizon.net) has joined #dig
- 02:42:39 [djweitzner]
- djweitzner has quit ()
- 02:54:42 [djweitzner]
- djweitzner (n=djweitzn@pool-70-108-167-73.washdc.east.verizon.net) has joined #dig
- 04:03:46 [djweitzner]
- djweitzner has quit ()
- 04:18:27 [timbl]
- timbl (n=timbl@212.49.149.114) has joined #dig
- 04:41:21 [timbl]
- timbl has quit (Read error: 113 (No route to host))
- 11:35:23 [djweitzner]
- djweitzner (n=djweitzn@pool-70-108-208-211.washdc.east.verizon.net) has joined #dig
- 12:49:21 [RalphS]
- RalphS (n=swick@30-6-113.wireless.csail.mit.edu) has joined #dig
- 12:54:56 [djweitzner]
- djweitzner has quit ()
- 13:42:31 [[DiG]Tratonis]
- [DiG]Tratonis (n=odintg@i577A90A6.versanet.de) has joined #dig
- 13:43:03 [[DiG]Tratonis]
- [DiG]Tratonis has left #dig
- 15:54:33 [djweitzner]
- djweitzner (n=djweitzn@gandalf.mindlab.umd.edu) has joined #dig
- 16:24:15 [sandro]
- sandro has quit ("Terminated with extreme prejudice - dircproxy 1.0.5")
- 16:24:29 [sandro]
- sandro (n=sandro@homer.w3.org) has joined #dig
- 18:15:39 [djweitzner]
- djweitzner has quit ()
- 18:20:49 [djweitzner]
- djweitzner (n=djweitzn@gandalf.mindlab.umd.edu) has joined #dig
- 18:29:31 [yosi_s]
- yosi_s (n=chatzill@w3cdhcp27.w3.org) has joined #dig
- 20:09:24 [djweitzner]
- djweitzner has quit ()
- 20:31:03 [yosi_s]
- DanC, I think I have cwm not splitting bNodes in proofs
- 20:31:07 [yosi_s]
- tests are running
- 20:32:29 [DanC]
- ooh!
- 20:32:32 [DanC]
- very timely
- 20:33:15 [DanC]
- I just checked in some xml/xpath related stuff. I wasn't as careful as perhaps I should have been about running tests before I checked it in
- 20:39:16 [yosi_s]
- the problem was that there was no preexissting way in cwm to get a formula containing a specific set of triples
- 20:39:30 [yosi_s]
- which is what was needed
- 20:39:37 [yosi_s]
- this has now been changed
- 20:39:46 [yosi_s]
- it is, perhaps, not perfect
- 20:40:01 [yosi_s]
- hmm...
- 20:40:49 [DanC]
- are you gonna be here for a little while?
- 20:41:04 [DanC]
- i.e. if I cvs update and find it's all broken, will you be here to help me fix it?
- 20:43:32 [yosi_s]
- [[
- 20:43:38 [yosi_s]
- @prefix : <gmpbnode#>.
- 20:43:40 [yosi_s]
- @keywords is, of, a.
- 20:43:42 [yosi_s]
- dan a Man; home [].
- 20:43:43 [yosi_s]
- { ?WHO home ?WHERE.
- 20:43:45 [yosi_s]
- ?WHERE in ?REGION } => { ?WHO homeRegion ?REGION }.
- 20:43:47 [yosi_s]
- { dan home ?WHERE} => {?WHERE in Texas} .
- 20:43:48 [yosi_s]
- ]]
- 20:44:01 [yosi_s]
- what should the proof look like?
- 20:45:42 [yosi_s]
- DanC?
- 20:48:22 [DanC]
- 1. parsing. 2. dan home [], by extraction from 1
- 20:48:55 [DanC]
- 3. { dan home ?WHERE} => {?WHERE in Texas} . by extraction. 4. dan home [ in Texas] by rule on 3 applied to 2
- 20:49:13 [DanC]
- 5. ... { ?WHO homeRegion ?REGION }. by extraction from 1
- 20:49:37 [DanC]
- 6. dan homeRegion Texas by rule from line 5 applied to 4
- 20:49:39 [DanC]
- QED.
- 20:49:59 [yosi_s]
- ok. Here we have a problem. Cwm will generate: 4. [ in Texas ]
- 20:50:54 [DanC]
- well, that's clearly bogus...
- 20:51:43 [DanC]
- ... does it have dan home [] in evidence at step 6?
- 20:53:23 [yosi_s]
- http://pastebin.ca/354204
- 20:53:52 [yosi_s]
- yes DanC, it does have that in evidence
- 20:56:07 [DanC]
- lines 2 and 6 are split, bogusly
- 20:56:31 [DanC]
- #
- 20:56:31 [DanC]
- 8: :dan :homeRegion :Texas .
- 20:56:31 [DanC]
- #
- 20:56:31 [DanC]
- [by rule from step 7 applied to steps [2, 6
- 20:57:39 [DanC]
- ew... is cwm's reasoning unsound altogether?
- 20:57:53 [DanC]
- lines 2 and 6 have different justifications, which suggests they _should_ be split.
- 20:58:22 [yosi_s]
- that was my point
- 20:58:26 [DanC]
- is this a case of one triple having 2 justifications, and cwm getting confused as a result?
- 20:58:34 [yosi_s]
- no
- 20:58:49 [yosi_s]
- every triple has one justification in this example
- 20:59:26 [yosi_s]
- but the fact that dan lives somewhere, and that that somewhere is in Texas, have different justifications
- 21:00:49 [DanC]
- the rule { dan home ?WHERE} => {?WHERE in Texas} can't conclude 2 triples on my line 4, can it.
- 21:03:33 [DanC]
- ouch. my head is exploding.
- 21:04:23 [yosi_s]
- welocome to my world
- 21:04:28 [yosi_s]
- welcome
- 21:04:41 [DanC]
- there's some wierd skolemization going on somewhere in cwm's reasoning... or something.
- 21:05:37 [yosi_s]
- the problem is not the reasoning
- 21:05:55 [yosi_s]
- the problem is the expressivity of the proof format
- 21:05:59 [DanC]
- I think this example goes beyond the proof structures we have so far. right.
- 21:06:27 [yosi_s]
- I don't see how to fix this --- n3 is not sufficiently introspective
- 21:06:37 [DanC]
- cwm should probably throw an exception when asked to generate a proof in this case, for now. file a bug or something.
- 21:07:41 [DanC]
- meanwhile, the code you've got is no worse than what we had before, and in some cases better, yes?
- 21:08:54 [DanC]
- agenda+ moving from cvs to hg
- 21:09:08 [DanC]
- agenda+ moving swap from cvs to hg
- 21:10:12 [yosi_s]
- yes
- 21:10:37 [yosi_s]
- I had introduced a bug
- 21:10:41 [yosi_s]
- rerunning tests
- 21:28:18 [DanC]
- how does cwm know/decide to unify the conclusion of { dan home ?WHERE} => {?WHERE in Texas} with dan home []? that has me puzzled.
- 21:29:18 [[DiG]Tratonis]
- [DiG]Tratonis (n=odintg@i577A94A0.versanet.de) has joined #dig
- 21:29:47 [DanC]
- I bet it has to re-open a formula to do that.
- 21:31:27 [yosi_s]
- that was not hard at all for cwm
- 21:31:37 [yosi_s]
- the bNode is a constant
- 21:31:45 [yosi_s]
- it simply does the right thing
- 21:32:28 [yosi_s]
- we only run into a problem because cwm is trying to represent a bNode with a bNode in the proof, and it's not working
- 21:34:27 [DanC]
- "the bNode is a constant" <- that's skolemization, in a nutshell.
- 21:35:31 [yosi_s]
- yes
- 21:35:50 [yosi_s]
- -- committed
- 21:41:01 [Tratonis]
- Tratonis (n=odintg@i577A94A0.versanet.de) has joined #dig
- 21:41:24 [Tratonis]
- Tratonis has left #dig
- 21:43:07 [DanC]
- yay; the original dan/texas case from the bug report works.
- 21:45:01 [[DiG]Tratonis]
- [DiG]Tratonis has quit (Read error: 104 (Connection reset by peer))
- 21:48:48 [DanC]
- Subject: toward cwm/N3/reason proof of httpRange-14 argument
- 21:48:48 [DanC]
- Date: Mon, 11 Sep 2006 09:46:52 -0500
- 21:48:51 [DanC]
- trying that one...
- 21:50:13 [DanC]
- definitely
- 21:51:08 [DanC]
- the httpRange-14 argument proof seems to work...
- 21:51:38 [DanC]
- it shows that I haven't implemented support for log:supports in the check.py --report option
- 21:51:40 [yosi_s]
- yosi_s has quit ("Chatzilla 0.9.77 [Firefox 1.5.0.9/2007012620]")
- 22:05:36 [timbl]
- timbl (n=timbl@146-115-112-112.c3-0.lex-ubr1.sbo-lex.ma.cable.rcn.com) has joined #dig
- 22:24:24 [DanC]
- yosi left at xx:51
- 22:24:51 [DanC]
- but he fixed that bnode splitting bug first!
- 22:25:06 [DanC]
- of course, he found a research problem while he was at it
- 22:36:56 [sandro]
- sandro has quit ("Getting off stoned server - dircproxy 1.0.5")
- 22:37:28 [sandro]
- sandro (n=nsandro@homer.w3.org) has joined #dig
- 22:53:11 [timbl]
- timbl has quit ()
- 23:30:45 [timbl]
- timbl (n=timbl@146-115-112-112.c3-0.lex-ubr1.sbo-lex.ma.cable.rcn.com) has joined #dig