cwm-dev/paw

Convene, take roll, review agenda

<DanC> <Zakim> On the phone I see Li, Steve, DanC, [UMD], MIT515, TimBL

<DanC> <Zakim> [UMD] has DanK, Vlad

<DanC> lkagal1, do you intend to dial in?

<DanC> <Zakim> On the phone I see Li, Steve, DanC, [UMD], YosiS, TimBL

<DanC> <Zakim> [UMD] has DanK, Vlad

<DanC> PROPOSED: to meet 23 March at 11aET

<DanC> YS: is that MIT spring break?

<DanC> I'm available 23 March

<DanC> timbl some risks, but ok

<DanC> so RESOLVED.

<DanC> scribe: everybody

<DanC> scribe: everybody/somebody/nobody

Vlad's progress integrating pychinko and cwm

<vlatko> http://svn.mindswap.org/pychinko/pychinko-cwm

<timbl> Vlad: The last time we tried to run it, we had a clash with rdflib, but now we have a version forked to be decoupled frpom rdflib.

<timbl> ...You have to install this form is you are going to work with cwm.

<timbl> ... --pythink is now --rete

<vlatko> http://svn.mindswap.org/pychinko/pychinko-cwm/pychinko

<DanC> eikeon, are you able to reproduce vlad's results?

<timbl> ... So you have to change your python path

<vlatko> swap\test\rete

<timbl> swap/test/rete

<DanC> connolly@dirk:~/w3ccvs/WWW/2000/10/swap/test$ cvs update -d rete

<DanC> cvs update: Updating rete

<DanC> U rete/anon.n3

<DanC> U rete/anon.ref

<DanC> ...

<DanC> would vlad and/or yosi pls send mail about this to public-cwm-talk@w3.org ?

<DanC> (we've spent about the amount of time I wanted to spend; I'd like to take this to public email)

<timbl> retest rete/detailed.tests

<timbl> ImportError: No module named pychinko

<timbl> PYTHONPATH=$PYTHONPATH:/devel/pychinko/phychinko-cwm/pychinko retest rete/detailed.tests

<timbl> doesn't work either

<DanC> VK: yes, I'll send mail to the public list

<DanC> .

<DanC> .

<DanC> .

PAW demo item? How is it going? Debugging cwm --why and proof checking

<eikeon_> /usr/local/bin/python2.4 pfproxy.py --test-rein

<DanC> . /home/connolly/paw/www/2005/demo/

<DanC> $ md5sum pfproxy.py

<DanC> 49cd535fe4402590ea5eefd654a20f1a pfproxy.py

<eikeon_> md5 pfproxy.py

<eikeon_> MD5 (pfproxy.py) = 49cd535fe4402590ea5eefd654a20f1a

<yosi_s> my phone won't operate plugged in

<DanC> connolly@dirk:~/paw/www/2005/demo$ PYTHONPATH=~/w3ccvs/WWW/2000/10 python2.4 pfproxy.py --test-rein

<DanC> loading rein/engine.n3

<DanC> made an engine

<DanC> rules with request: {5}. ready to think...

<DanC> think done: {7}

<DanC> filtering...

<DanC> filter done: {5}

<DanC> Expressing proof as a graph...

<DanC> Done, 162 top-level statements.

<DanC> Found poof. Serializing...Done.

<DanC> Now checking proof...

<DanC> Now checking proof... Starting valid on None

<DanC> Traceback (most recent call last):

<DanC> File "pfproxy.py", line 452, in ?

<DanC> _testRein()

<DanC> File "pfproxy.py", line 442, in _testRein

<DanC> result = check.valid(proof)

<DanC> File "/home/connolly/w3ccvs/WWW/2000/10/swap/check.py", line 216, in valid

<DanC> raise RuntimeError

<DanC> RuntimeError

<DanC> in my book RuntimeError is a way of confessing you haven't coded up a case that you intend(ed) to

<yosi_s> I've done things to my copy of pfproxy

<DanC> md5sum of check.py in my swap is a76d50812a107952388bd970e5f74ad2

<yosi_s> it will not be the same

<DanC> mail your pfproxy to paw-team@lists.mindswap.org ?

<yosi_s> syosi@mr-burns:~/CVS-local/WWW/2000/10/swap$ md5sum check.py

<yosi_s> 81fc808a35546ece3ed3bc1f7c8c73b4 check.py

<yosi_s> ok

<yosi_s> that runtimeError at line 216

<yosi_s> appears to be me breaking pfproxy

<DanC> yosi, are you subscribed to paw-team@lists.mindswap.org? do you have bandwidth to read problem reports sent by eikeon ?

<DanC> YK: yes, I'm subscribed

<DanC> yes, send your pfproxy to paw-team or public-cwm-talk, please, yosi

<DanC> YS: I changed the interface to valid... it takes a [premises] arg

<DanC> home/connolly/paw/www/2005/demo/

<lkagal1> DanC, just saw your note. Sorry I can't dial in, I'm in the TAMI meeting.

<DanC> no problem, lkagal1 ; I wasn't expecting you.

<vlatko> svn+ssh://mindswap.org/home/svn/paw/www/2005/demo

<DanC> TimBL: this meeting clashes with a TAMI meeting.

<eikeon_> http://www.policyawareweb.org/2005/demo/pfproxy.py

<DanC> PROPOSED: to meet

<eikeon_> With clue to what the URL is at the bottom of: http://www.policyawareweb.org/2005site/admin.html

<DanC> PROPOSED: to meet Thu 23 Mar ...

<DanC> PROPOSED: to meet Thu 23 Mar 1pET

<yosi_s> http://lists.w3.org/Archives/Public/public-cwm-talk/2006JanMar/0029.html my mail about failing to run --rete

<DanC> RESOLVED: to meet Thu 23 Mar 1pET, to be confirmed a day in advance

<eikeon_> Instructions here: http://www.policyawareweb.org/2005site/dev.html

<yosi_s> ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAQEAn3Ofd6+4axA3jpdCeTFVOHM+WZ2wxOuyZaniUFrz4fACQa0hdf7xgSJ6XUxwFIssWMQUWJVcVks3/0fXhFhB5mUyJM7W0lQcBDkh8d9QVmGjnm5F/hHm2wGheq1wLKfM8h6g5I9a2rHItQ3Eh4sUup7Ut9qLngtSJSjehpCvQS+CdcURItpLeFbsmBobvR/udrHmMXfsr3tP6/K6Qc3U9dhQ6ryq6gYBasJnWTAL0pdLPWaVOdu8Z5G9FQUWGSnuVKFdRTjJbnwP7aFwc7fLvRWQeG3VAicURmCHDdoK4k5f3EHexoBgLlwXXtRQNoSS0ByMheCQq5BvppmP573+hw== syosi@mr-burns.w3.o

<yosi_s> rg

<DanC> ACTION: eikeon to get yosi's key in the right place for access to paw demo

<vlatko> yosi, i hadn't checked in the latest version of pycwmko - that was the reason for th error.

<vlatko> please try again

<DanC> .

<DanC> .

<DanC> .

cwm/Inferenceweb integration

<DanC> http://iw4.stanford.edu/iwvalidator/

<yosi_s> vlatko, a `svn update` updated no files

<DanC> http://www.w3.org/2000/10/swap/test/reason/

<vlatko> pycwmko is in the swap directory

<vlatko> do a cvs update on cwm

<DanC> in particular, http://www.w3.org/2000/10/swap/test/reason/to-pml.n3

<DanC> http://www.w3.org/2000/10/swap/test/reason/

<yosi_s> ok vlatko, that simple test worked

<vlatko> great

<vlatko> i

<vlatko> i'll try to optimize the conversion of terms b/t cwm and the rete

<DanC> http://www.w3.org/2000/10/swap/test/reason/t1.pml

<vlatko> actually, i can even try to use the cwm terms (term.Symbol, term.LabelledNode) in the rete itself

<vlatko> to integrate it more

<yosi_s> vlatko, so I noticed with my second version

<DanC> [[

<DanC> <b>PML Validation Result</b><p>

<DanC> <b>[Step 1] RDF validation ...</b><br>

<DanC> Loading ontology: http://inferenceweb.stanford.edu/2004/07/iw.owl<br>

<DanC> Time used : (41 ms) <br>

<DanC> ...

<DanC> ]]

<timbl> <b>PML Validation Result</b><p>

<timbl> <b>[Step 1] RDF validation ...</b><br>

<yosi_s> having it never generate any rdflib terms is harder

<vlatko> hm ok. the term conversion is a mess - i will clean it up

<vlatko> it's working but i'm not happy with it

<timbl> Need [http://www.w3.org/2000/01/rdf-schema#Literal] but found a literal<br>

<vlatko> yosi, if it's difficult to use the same terms, maybe i can subclass them then

<DanC> Need [http://inferenceweb.stanford.edu/2004/07/iw.owl#InferenceStep] but found [http://www.w3.org/2000/10/swap/reason#Parsing]<br>

<DanC> Concern: [http://www.w3.org/2000/10/swap/test/reason/t1.pml#aux__g47, http://inferenceweb.stanford.edu/2004/07/iw.owl#isConsequentOf, 688954:109dddf58fe:450a]<br>

<timbl> "This URL does not contain an IW document."

<DanC> the tool I want to use is http://iw3.stanford.edu:8080/iwbrowser/

<timbl> http://iw4.stanford.edu/iwbrowser/NodeSetBrowser?url=http%3A%2F%2Fwww.w3.org%2F2000%2F10%2Fswap%2Ftest%2Freason%2Ft1.pml

<DanC> "This URL does not contain an IW document."

<eikeon_> yosi_s: Ron is asking what your preferred username

<DanC> ACTION: Li to follow up on errors from the IW proof browser

<DanC> ACTION: DanC to mail Li about the public-cwm-talk mailing list.

<yosi_s> EST

<DanC> yes, yosi; the agenda says "9 Mar at 11:00ET for 90 mins". but we've done the critical business for today. we could adjourn. or we could excuse you.

<li_> hi

<timbl> grep url= .svn/entries

<DanC> .

<DanC> .

<DanC> .

N3 changes and n3.n3

<eikeon_> Where did we get to re: decimals

<DanC> , asked timbl

<DanC> YS: not well tested.

<DanC> TimBL: when unicode stuff from n3.n3 is converted to .html, it screws up

<DanC> ... in the regular expressions. \uxxxxx

<DanC> YS: there's cwm option that might fix it

<timbl> Neeeded: please cwm print strings using unicode escapes and ascii.

<yosi_s> [\U00010000-\U000EFFFF] turns to 4 chars [\uxxxx\uxxxx-\uxxxx\uxxxx]

<yosi_s> the range is the middle two

<DanC> Vlad's paper writing action continues.

<DanC> VK: yosi was able to run the rete stuff, after I checked in a missing file

<DanC> ADJOURN.

<DanC> ah, I see now, yosi

<DanC> <Zakim> SW_PAW(CWM*)11:00AM has ended

<DanC> <Zakim> Attendees were +1.410.455.aaaa, +1.301.390.aabb, DanC, Li, Steve, DanK, Vlad, TimBL, YosiS, MIT515

<DanC> does anybody else know how to use scribe.perl to produce something like http://www.policyawareweb.org/2005/pf-dev/02-15-paw-minutes from today's IRC log?

<DanC> or... is anybody else willing to just mail the raw IRC log to public-cwm-talk?

<DanC> eikeon_, maybe?

<eikeon_> Yes?

<eikeon_> Ah...

<eikeon_> I could give scribe.perl a go

<DanC> I think the directions are linked fromt he bottom of the 15 Feb minutes

<DanC> what I usually do is: start with the raw IRC log, run it thru scribe.perl, look at the output, then edit the IRC log and repeat

<DanC> Scribe: DanK

<DanC> (those will get noticed by scribe.perl )

<DanC> Attendees were +1.410.455.aaaa, +1.301.390.aabb, DanC, Li, Steve, DanK, Vlad, TimBL, YosiS, MIT515

<DanC> Attendees were DanC, Li, Steve, DanK, Vlad, TimBL, YosiS

<eikeon_> File->Save As gives me an xml file

<DanC> scribe.perl eats lots of log formats, but perhaps not that one

<eikeon_> Copy / paste looks more like raw bits.

<eikeon_> I'll try running that through

<eikeon_> DanC: Can you email me the raw log this time... and I'll make sure to run an irc client that let's me get at one that scribe.perl is happy with next time ;)

<DanC> ok...