IRC log of dig on 2010-08-20

Timestamps are in UTC.

11:32:22 [DIGlogger]
DIGlogger (~dig-logge@groups.csail.mit.edu) has joined #dig
11:32:22 [bartol.freenode.net]
topic is: Decentralized Information Group @ MIT http://dig.csail.mit.edu/
11:32:22 [bartol.freenode.net]
Users on #dig: DIGlogger RalphS mhausenblas melvster webr3 presbrey ernestchiang_ Yudai Pipian gbot9 sandro nunnun_away jube drrho ericP
11:40:00 [bblfish]
bblfish (~bblfish@ALagny-753-1-57-216.w90-35.abo.wanadoo.fr) has joined #dig
11:47:18 [melvster]
melvster has quit (Ping timeout: 240 seconds)
12:14:18 [bblfish]
bblfish has quit (Quit: Leaving.)
12:30:17 [melvster]
melvster (~melvster@p579F9F77.dip.t-dialin.net) has joined #dig
12:32:34 [bblfish]
bblfish (~bblfish@216.219.70.137) has joined #dig
12:43:22 [lkagal]
lkagal (~lkagal@pool-96-237-240-136.bstnma.fios.verizon.net) has joined #dig
12:51:29 [Simpson_2]
Simpson_2 (~bart@office.netage.nl) has joined #dig
12:51:32 [Simpson_2]
hi all
12:51:51 [Simpson_2]
i've found some issues in the rdflib.js code, how do I report them ?
12:53:22 [webr3]
http://dig.csail.mit.edu/issues/tabulator/
12:53:26 [webr3]
but...
12:53:46 [webr3]
there's an issue w/ the issues
12:56:07 [Simpson_2]
is there a way to report them ? :)
12:57:52 [webr3]
yeah normally there, just at the minute it's not working - could always drop it in here too
13:16:38 [Simpson_2]
okay I have some very wierd reproducable issue with tabulator and rdflib.js
13:17:05 [Simpson_2]
if I you take: http://brandweeraa.resc.info/data/messages/18723c471e0f48dc6a8e7466758997d7030e281a
13:17:29 [Simpson_2]
and excute: select ?unit ?label WHERE { ?msg <http://vocab.resc.info/communication#unit> ?unit . ?unit <http://www.w3.org/1999/02/22-rdf-syntax-ns#label> ?label . }
13:17:37 [Simpson_2]
on it, it doesn't return anything
13:17:51 [Simpson_2]
if you take: http://brandweeraa.resc.info/data/messages/c02f36867087b67a3b005c8844ecccb074f61303
13:18:04 [Simpson_2]
excecute the same query you get the results.
13:18:29 [Simpson_2]
as expected, so the precense of more then 2 units in the graph seem to confuse the query matcher
13:19:16 [Simpson_2]
if I comment out 'patern.sort(easiestQuery)
13:20:31 [Simpson_2]
in function $rdf.IndexedFormula.query.match2 everything works as expected...
13:21:09 [Simpson_2]
in the logs you can see that if this function is stil enabled, it suddenly flips around the ?unit and ?msg match and doesn't fetch the external graphs.
13:21:19 [Simpson_2]
hm stil no flood kick ;)
13:41:10 [oshani]
oshani (~oshani@31-34-137.wireless.csail.mit.edu) has joined #dig
13:51:12 [mhausenblas]
mhausenblas has quit (Quit: brb)
14:00:05 [oshani]
oshani has quit (Quit: Mama nidi!)
14:07:10 [oshani]
oshani (~oshani@31-34-137.wireless.csail.mit.edu) has joined #dig
14:39:06 [kennyluck]
kennyluck (~kennyluck@114-25-209-129.dynamic.hinet.net) has joined #dig
14:39:07 [kennyluck]
kennyluck has quit (Excess Flood)
14:40:18 [jube]
jube is now known as nori
14:40:38 [nori]
webr3: was there an export of issues/ ?
14:42:55 [kennyluck]
kennyluck (~kennyluck@114-25-209-129.dynamic.hinet.net) has joined #dig
15:10:14 [webr3]
webr3 has quit (Ping timeout: 276 seconds)
15:14:28 [webr3]
webr3 (~nathan@host86-140-58-83.range86-140.btcentralplus.com) has joined #dig
15:17:14 [nori]
http://blog.whats-your.name/public/tabIssues.n3
15:26:54 [lkagal]
lkagal has quit (Quit: lkagal)
15:27:31 [lkagal]
lkagal (~lkagal@pool-96-237-240-136.bstnma.fios.verizon.net) has joined #dig
15:41:05 [oshani]
oshani has quit (Quit: Mama nidi!)
15:46:43 [oshani]
oshani (~oshani@31-34-137.wireless.csail.mit.edu) has joined #dig
16:22:36 [wlfsbrg]
wlfsbrg (~wlfsbrg@unaffiliated/wlfsbrg) has joined #dig
16:22:45 [wlfsbrg]
wlfsbrg has left #dig
16:43:05 [othree]
othree (~othree@admin39.ct.ntust.edu.tw) has joined #dig
17:37:20 [bblfish]
bblfish has quit (Ping timeout: 260 seconds)
17:39:09 [bblfish]
bblfish (~bblfish@216.219.69.211) has joined #dig
17:56:47 [bblfish]
bblfish has quit (Ping timeout: 276 seconds)
17:58:14 [knappy]
knappy (~haoqi@18.111.72.249) has joined #dig
18:01:50 [knappy]
knappy has quit (Client Quit)
18:04:35 [knappy]
knappy (~haoqi@dhcp-18-111-36-78.dyn.mit.edu) has joined #dig
18:06:35 [timbl]
timbl (~timbl@31-35-252.wireless.csail.mit.edu) has joined #dig
18:07:22 [timbl]
FIOS fibre eaten "by a squirrel"
18:07:44 [timbl]
DIGlogger, pointer?
18:07:44 [timbl]
See http://dig.csail.mit.edu/irc/dig/2010-08-20#T18-07-44
18:09:00 [timbl]
So now, because of the relative URI problem, now I have a mixture of items in the same file some in tabulator.org and some in dig.xvm domains
18:10:54 [marisol]
marisol (~marisol@pool-141-154-35-71.bos.east.verizon.net) has joined #dig
18:15:21 [timbl]
Which should i go for, I wonder?
18:16:35 [timbl]
presbrey, ericp figured it was your fault for calling his code with a base URI when it accepts data incoming?
18:17:21 [sandro]
/me wonders who owns tabulator.org
18:17:39 [ericP]
which is to say, it's possibe that if you call it with a baseURI of "", it will do what we need
18:17:50 [oshani]
sandro, I think melvster owns it
18:19:13 [timbl]
Ok, ericp -- so if I start using this onlu as tabulator.org and not dig.xml and change all my pointers to remove any xvm then you think it will be consistent with itself?
18:22:17 [timbl]
(I still don't think yo cantotally blame the way you are called for writing all those absolute URIs.)
18:31:38 [mcherian_]
mcherian_ (~mathew@123.123.253.31) has joined #dig
18:47:40 [timbl]
Sign .. So I fixed the absolute URIs to make them relative in the file http://tabulator.org/wiki/semanticweb/roadmap/state and that i snow error 500
18:49:48 [timbl]
Sigh .. So EricP I fixed the absolute URIs to make them relative in the file http://tabulator.org/wiki/semanticweb/roadmap/state and that s now error 500
18:52:48 [ericP]
sounds like i need to come in and spend some quality time
18:53:00 [ericP]
also that i need test cases
18:53:07 [timbl]
I don't know how much you can do from there
18:53:23 [timbl]
There is a test case
18:54:14 [ericP]
i need to come in to pick up vegetables anyways (CSA dole-out on fridays)
18:54:40 [ericP]
ideally, a piece of python that makes some SWObjects API calls and gives back a demonstrably wrong result
18:56:36 [timbl]
It would great also if you could talk about what the apache module is you are proposing to Alexandre too and what it does (sparql update?) and what it doesn't (save the file?)
18:56:40 [mcherian_]
mcherian_ has quit (Ping timeout: 260 seconds)
18:57:20 [timbl]
"OK
18:57:21 [timbl]
The server encountered an internal error or misconfiguration and was unable to complete your request.
18:57:21 [timbl]
Please contact the server administrator, root@localhost and inform them of the time the error occurred, and anything you might have done that may have caused the error.
18:57:21 [timbl]
More information about this error may be available in the server error log."
18:58:12 [timbl]
Wheras the tabulator says it got for the same URI at 14:46:44 a 500 error
19:04:13 [timbl]
Now it gets 200 and 283 statements
19:21:18 [timbl]
___________________________
19:21:24 [timbl]
Another of theseo 200 error messages
19:21:58 [timbl]
datawiki dead?
19:22:10 [timbl]
http://tabulator.org/wiki/semanticweb/roadmap/state#Iss1281736045706
19:22:30 [timbl]
back up
19:22:40 [timbl]
I wonder whither it wanders.
19:22:49 [melvster]
yes i got tabulator.org ... when it became free, happy to sign it over anytime, but for now it points to csail ...
19:24:11 [timbl]
Thank, yes, in due course that would be a good idea
19:24:38 [timbl]
Thank you for grabbing it!
19:54:27 [bblfish]
bblfish (~bblfish@ALagny-753-1-57-216.w90-35.abo.wanadoo.fr) has joined #dig
19:54:52 [presbrey]
a newly proposed apache module?
19:55:15 [presbrey]
sparql update?
19:55:36 [timbl]
ask ericp
19:56:26 [presbrey]
ericP: with swobjects? C++-linked apache modules are sooo cool
20:21:14 [RalphS]
RalphS has quit (Quit: leaving ...)
20:21:48 [timbl]
This is presbrey's mod_swobjects ?
20:22:24 [timbl]
It has the write back to disk commented out?
20:22:51 [timbl]
presbrey, compare the mod_swobjects and the python rig you have now?
20:25:43 [ericP]
ears here
20:30:21 [ericP]
presbrey, not a newly proposed apache module, just a re-use of the one you wrote
20:31:26 [ericP]
it's my impression that uncommenting http://swobjects.svn.sourceforge.net/viewvc/swobjects/branches/sparql11/apache/mod_sparul.cpp?revision=1193&view=markup#l77 will do the job
20:32:09 [presbrey]
maybe
20:32:40 [presbrey]
the python edition is far beyond in features
20:32:54 [ericP]
gotcha. which would you find cooler?
20:33:21 [ericP]
also, i'd like to try SWObjectDriver("", F) to see if working just in relative URIs solves timbl's serialization prob
20:33:25 [presbrey]
in terms of what, language, features?
20:33:36 [presbrey]
other people are going to be able to contribute to python much easier
20:33:37 [ericP]
features
20:33:52 [presbrey]
ha python by a landslide?
20:34:05 [presbrey]
impl... fastcgi vs. loading libstdc++ into apache
20:34:43 [presbrey]
if you can get python/swig/g++ working, you can use this on your own host without root modifications
20:34:58 [timbl]
By the way, for the state.n3 fiels I edited to make the URIS relative
20:35:07 [presbrey]
I think thats way cooler for wide deployment
20:36:27 [ericP]
it's funny; seems so contrived to have IPC to other languages, but yes, we'd get more community contributions to a python implementation
20:36:42 [timbl]
Alexandre suggests running it on data.w3.org (space.w3.org?) rather than in www.w3.org
20:38:34 [ericP]
want to take a moemnet to look at SWObjectDriver("", F) and see if we can put this issue to bed easily?
20:38:46 [timbl]
cooler even for normal sysadmins who don't necessarilioty know/like python?
20:39:29 [timbl]
To what extent is c[+] just a standard they all trust and understand?
20:39:32 [presbrey]
even in the midst of MIT, scripts.mit.edu, widely deployed fastcgi
20:39:46 [presbrey]
I will need it to be python/fcgi to run a data space at http://presbrey.mit.edu/foaf
20:40:07 [presbrey]
you can run python/fcgi on groups.csail.mit.edu as well
20:40:23 [ericP]
good point, IPC is actually a feature there
20:40:58 [presbrey]
we can get much further, faster with an implementation that involves the least # of sysadmins
20:41:45 [timbl]
so what -- is it all done with the .htaccess file locally in the file system?
20:41:54 [presbrey]
or httpd.conf
20:42:04 [timbl]
s/what/what is the hook/
20:42:09 [presbrey]
for the data wiki?
20:42:14 [timbl]
if you are a sysadmin
20:42:22 [marisol]
marisol is now known as mari-away
20:42:31 [ericP]
i think it's sort of reversed at w3; sysadmins are reluctant to add things like python to the requirements for a service. module-only is easier to deploy
20:42:33 [timbl]
I suppose for two sorts of use.
20:42:55 [timbl]
1) A straight datawiki space, a uniform space of editable stuff
20:43:21 [timbl]
2) within n existing website, to be able to save files with a specific extension to make the editable.
20:43:37 [ericP]
true
20:44:03 [timbl]
would be nice to be able to turn on either at will using htaccess
20:44:25 [timbl]
or even using the server directory index pages
20:44:59 [ericP]
presbrey, did you ever test mod_sparul with l77-79 uncommented?
20:48:25 [webr3]
developing rdflib.js question - ar people taking the aproach of modifying rdflib.js directly, or modifying each class individualy then regenerating rdf lib?
20:48:52 [timbl]
always the latter
20:49:07 [timbl]
Not really classes.
20:49:27 [timbl]
Different files add diffeernt functionality to the same class, in the ay yo can with js
20:51:31 [presbrey]
no I didn't finish mod_sparul to a point where tests were needed
20:56:09 [presbrey]
I'm surprised w3 would rather load libstdc++ and mod_sparul into apache than mod_fcgid
20:59:24 [presbrey]
s/surprised/interested why/
21:00:15 [webr3]
timbl, thanks
21:01:00 [ericP]
presbrey, fewer dependancies on the server
21:01:27 [ericP]
and a bit cheaper (which would only matter if it got popular)
21:01:31 [betehess]
betehess (~betehess@betehess.w3.org) has joined #dig
21:02:37 [timbl]
Alexandre, EricP and Joe Presbrey were discussing how best to mplement this
21:03:06 [timbl]
wonder what w3.org woul dprefer -- python under fastcgi or c+
21:03:21 [timbl]
presbrey: I'm surprised w3 would rather load libstdc++ and mod_sparul into apache than mod_fcgid
21:03:21 [timbl]
[16:59] presbrey: s/surprised/interested why/
21:03:42 [betehess]
or a Scala implementation based on Jena
21:04:05 [ericP]
i argued that we generally favor fewer dependancies
21:04:29 [timbl]
betehess, you'd clearly like to it from scratch -- everyone hs their fav languages
21:05:25 [betehess]
timbl, I would first prefer to have a perfect overview of what we want. Especially an area perfectly delimited
21:05:42 [betehess]
including the ACLs stuff
21:05:54 [betehess]
who will use that
21:05:58 [betehess]
and to do what
21:06:15 [timbl]
Well, I know Ted wanted the same W3C ACLS to be used on w3.org but maybe on space.w3.org one could use new ACLs too
21:07:24 [timbl]
So there is an overview, imperfect of course, in http://www.w3.org/DesignIssues/ReadWriteLinkedData.html and http://www.w3.org/DesignIssues/CloudStorage.html
21:07:52 [timbl]
They could be striped down to a smaller spec
21:08:16 [timbl]
just referencing the other specs like sparql, http, etc
21:09:51 [presbrey]
I would like to see the data wiki in scala
21:10:57 [timbl]
But given that there is a working* version of the thing running now, why not run with that?
21:11:38 [ericP]
because working code will never be as pure and efficient as wish-code
21:11:54 [timbl]
I don't think you can really leave a scala-based system for W3C systeam to maintain and repond to calls on.,
21:12:24 [betehess]
timbl, I made sure it was already the case :-) (with ted's evaluation and approval)
21:12:26 [timbl]
Mind you you could argue that Jena is an established codebase.
21:12:39 [betehess]
Jena is a widely used API
21:12:49 [betehess]
in the Java world
21:12:49 [timbl]
Ted said you coudl put scala on w3.org?
21:13:10 [betehess]
yes if it is motivated
21:13:41 [mari-away]
mari-away is now known as marisol
21:13:43 [betehess]
one of the problems was jigsaw compatibility, but we first managed that and then decided to not use it anymore :-)
21:14:23 [betehess]
the whole systeam decided that Scala was worth trying it during the F2F in Spain
21:14:55 [betehess]
* ted has sparql update in his local agenda file for monday. if there is an irc log from #dig to point to please have pointer for monday
21:15:05 [oshani]
oshani has quit (Quit: Mama nidi!)
21:15:22 [timbl]
DIGlogger, pointer?
21:15:22 [timbl]
See http://dig.csail.mit.edu/irc/dig/2010-08-20#T21-15-22
21:15:43 [betehess]
anyway, before speaking of the technical solution, we have to be sure of what we want exactly
21:16:07 [oshani]
oshani (~oshani@31-34-137.wireless.csail.mit.edu) has joined #dig
21:16:19 [timbl]
That is described in those 2 design issues articules.
21:16:40 [timbl]
There is a lot of motivation in those too
21:16:47 [betehess]
where is joe's code?
21:17:19 [betehess]
and also the documentation to know how to deploy and use the data wiki?
21:18:28 [betehess]
question: is the use case of doing SPARQL/UPDATE on rdf files in addressed in any of our specs?
21:18:34 [betehess]
and if not, will it be?
21:21:11 [timbl]
Joe, does http://presbrey.mit.edu/ have pointers to all the latest code?
21:22:16 [timbl]
re: the use case of doing SPARQL/UPDATE on rdf files in addressed in any of our specs -- yes, there is work in the SPARQL working group in a document we just reviewed which I hope will end up speccing this scenario.
21:25:23 [presbrey]
no the latest code is on dig.xvm
21:25:46 [timbl]
you don't have a repo it all goes back to?
21:26:42 [timbl]
If Alex wanted to clone it , are there enough breadcrumbs?
21:37:26 [betehess]
does the current code uses the ACL ontology?
21:37:36 [betehess]
s/uses/use/
21:38:50 [timbl]
yes
21:40:25 [presbrey]
oops I meant dig.csail
21:40:40 [presbrey]
there are plenty of breadcrumbs
21:40:53 [presbrey]
it runs on dig.xvm, it can run on dig.csail with a little elbow grease
21:41:03 [presbrey]
its stored in the dig SVN
21:41:15 [presbrey]
pretty standard 120 lines of python
21:41:20 [presbrey]
will work in python >2.4
21:41:46 [presbrey]
with apache mod_fcgid
21:42:40 [presbrey]
Apache claims mod_fcgid delivers 'very similar' performance to in-process
21:43:13 [presbrey]
and that it is now preferred to mod_php by php developers, etc.
21:43:22 [presbrey]
http://httpd.apache.org/mod_fcgid/
21:46:18 [presbrey]
ericP: are you in today?
21:46:32 [presbrey]
I can come over and look at the relativizing stuff with you
21:46:47 [presbrey]
I heard timbl crying earlier
21:48:29 [presbrey]
so betehess: how do I get scala inside apache?
21:49:48 [Simpson_2]
re
21:53:03 [Simpson_2]
timbl, I've ran into some more issues with the tabulator code see: http://dig.csail.mit.edu/irc/dig/2010-08-20#T12-51-51
22:06:29 [Simpson_2]
night all
22:06:32 [Simpson_2]
Simpson_2 has quit (Quit: Ex-Chat)
22:13:14 [betehess]
presbrey, i (ericP) am at the 'tute
22:13:20 [betehess]
wanna drop by?
22:14:52 [oshani]
oshani has quit (Quit: Mama nidi!)
22:23:10 [presbrey]
I just came by and the doors were all locked
22:23:21 [presbrey]
so I'm in central now
22:25:05 [presbrey]
I've got my laptop now to try stuff 1st hand
22:25:50 [presbrey]
I don't really want to come back down unless you're staying for a while?
22:27:57 [ericP]
where's central?
22:28:11 [ericP]
re locked doors, sorry, was in betehess's office
22:28:23 [presbrey]
mass ave and prospect
22:29:10 [ericP]
i can come by on my motorcycle
22:29:16 [ericP]
timbl,
22:29:28 [ericP]
ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAQEA3GhU+vv4vwT3LAemGPe0uRIc7op42B3+8nFbBa9/p0EuyebSyrQpKij2gh2TtB0ZdF/kOo7rSi+OTZ9+3uLFNsJIhgWQHIy2DvFa6zZye08PlegpY7EZG4e3DXeqlYiQy24lSztLNE8JEiOLlqe7P7THMLMtPjODlv+RM2s+tgaLOPP1Wo58CFwjlYikQ7biDXWfwW5GrlcZ4iSovNeh4p6tDW4sqxGNLlIGZsSgVcheafnYVyisp60doxwfS5Fx0zO9V3vUkGCk6BBl0zeI8foaqQhHy0zGXjfNSErv+OxAfXmMH+kCuxgK14dI48L/zut/fvlBMcmPeFnj81ggyw== 25052004-eric@w3.org
22:29:46 [presbrey]
sure
22:29:48 [timbl]
tx
22:29:49 [presbrey]
you use that key for everything
22:30:11 [ericP]
yup
22:30:37 [Pipian]
Pipian has quit (Remote host closed the connection)
22:31:06 [ericP]
presbrey, where about in central?
22:34:03 [timbl]
What is the screen sharing command on a mac ... open vnc://localhost:8888/ kinda fails silently
22:34:23 [presbrey]
its probably port 5900
22:34:29 [presbrey]
I never heard of 8888 for vnc
22:34:50 [timbl]
I have it locally on 5901
22:35:16 [timbl]
ssh tunneld throug h to 5900 so i actually did open vnc://localhost/5901/
22:35:51 [timbl]
Clicking on that ^ I get "A computer can not connect to itself and share its screen."
22:40:28 [knappy]
knappy has quit (Quit: knappy)
22:42:54 [timbl]
timbl has quit (Quit: timbl)
22:59:45 [oshani]
oshani (~oshani@31-34-137.wireless.csail.mit.edu) has joined #dig
23:06:14 [oshani]
oshani has quit (Quit: Mama nidi!)
23:24:37 [oshani]
oshani (~oshani@31-34-137.wireless.csail.mit.edu) has joined #dig
23:48:47 [betehess]
betehess has quit (Quit: Leaving)
23:48:57 [oshani]
oshani has quit (Quit: Mama nidi!)