IRC log of dig on 2012-10-11

Timestamps are in UTC.

02:30:21 [danbri]
danbri has quit (Remote host closed the connection)
02:49:03 [rszeno]
rszeno has quit (Quit: Leaving.)
03:20:41 [kennyluck]
kennyluck (~kennyluck@119.161.158.96) has joined #dig
03:40:33 [scor]
scor has quit (Quit: scor)
04:41:28 [melvster]
melvster (~melvin@p4FF966F1.dip.t-dialin.net) has joined #dig
06:08:46 [melvster1]
melvster1 (~melvin@p4FF9634B.dip.t-dialin.net) has joined #dig
06:11:12 [melvster]
melvster has quit (Ping timeout: 265 seconds)
06:16:33 [melvster1]
the new firefox developer toolbar is quite nice ... sort of like a command line for the browser
06:16:48 [mhausenblas]
wow, melvster1 early bird
06:16:53 [mhausenblas]
morninh
06:16:57 [melvster1]
lol ... morning
06:17:04 [mhausenblas]
s/morninh/morning
06:17:29 [mhausenblas]
yup, indeed, re FF
06:18:16 [melvster1]
nice!
06:18:23 [melvster1]
but why launch and not search?
06:18:35 [mhausenblas]
well, look at the available actions ;)
06:18:45 [mhausenblas]
you don't only search, you do things there
06:19:01 [mhausenblas]
btw, melvster1 I think I might have something for you
06:19:21 [melvster1]
mhausenblas: ooh great ... O.o
06:19:43 [mhausenblas]
oh, well, more like, you might be interested to check out https://github.com/mhausenblas/ld-in-couch
06:19:51 [mhausenblas]
feedback etc welcome
06:20:08 [melvster1]
wow
06:20:24 [mhausenblas]
(part of a NoSQL/cloud benchmarking initiative I'm doing with some others)
06:20:47 [mhausenblas]
we'd like to better understand what stores and with what layout are 'RDF friendly'
06:20:57 [melvster1]
mhausenblas: how do you deal with _id and _rev fields?
06:21:10 [mhausenblas]
(background: http://webofdata.wordpress.com/2011/05/02/nosql-linked-data-processing/)
06:21:16 [mhausenblas]
how do you mean?
06:21:42 [melvster1]
doesnt couch supply _id and _rev when you do a GET?
06:21:53 [mhausenblas]
the _id so far is auto-generated, as is the _rev
06:21:58 [mhausenblas]
yes Couch does
06:22:01 [melvster1]
and you need the right _rev when doing an update ...
06:22:07 [mhausenblas]
do you think one could leverage this somehow?
06:22:10 [mhausenblas]
yup
06:22:13 [melvster1]
so not 100% RESTful (tho most things aren't)
06:22:37 [mhausenblas]
what exactly is not RESTful?
06:22:47 [melvster1]
do you think couch should have it's own content type?
06:22:50 [mhausenblas]
you mean that the payload doesn't contain URIs?
06:22:57 [mhausenblas]
for JSON?
06:23:02 [melvster1]
yes
06:23:05 [mhausenblas]
hmm
06:23:15 [mhausenblas]
TBH, didn't think about this one
06:23:34 [mhausenblas]
it's more like finding an optimal way to squeeze the graph into docs
06:23:51 [mhausenblas]
not so much about interfacing
06:24:09 [mhausenblas]
optimal wrt to requirements I've listed there ...
06:24:11 [melvster1]
looks awesome tho!
06:24:18 [mhausenblas]
tx
06:24:33 [mhausenblas]
way from production ready but good enough for doing some experiements
06:25:05 [mhausenblas]
however, even more exciting from my POV is HDT
06:25:07 [mhausenblas]
as seen in https://github.com/mhausenblas/shodan
06:25:41 [mhausenblas]
acts somewhere between a static RDF dump and a full SPARQL endpoint
06:26:03 [mhausenblas]
very efficient, compressed, flexible and rich interfacing
06:27:16 [melvster1]
looks amazing
06:27:31 [melvster1]
http://www.rdfhdt.org/ seems to be down ...
06:27:34 [mhausenblas]
tx, it (HDT) is really cool, should give it a try
06:27:36 [mhausenblas]
oh
06:27:42 [mhausenblas]
damn
06:28:02 [mhausenblas]
gotta tell Mario (he's a colleague here at DERI)
06:28:05 [mhausenblas]
tx
06:28:24 [melvster1]
wonderful!
06:28:37 [mhausenblas]
they might roll out a new version or so
06:28:52 [mhausenblas]
dunno why the heck they take down the site for that, though :(
06:31:42 [melvster1]
mhausenblas: interesting serialization ... were you not tempted with one form of JSON-LD or RDF/JSON?
06:32:29 [mhausenblas]
there, I said it
06:32:59 [melvster1]
lol
06:33:07 [melvster1]
other question does couch allow multiple objects in an array [ {...},{...} ... ] this is how LD models multiple objects
06:33:20 [mhausenblas]
I reckon the only useful serialisation would be along the line of http://www.w3.org/wiki/JTriples
06:33:30 [mhausenblas]
(not because I proposed it, though)
06:33:52 [mhausenblas]
yes, melvster1, one can nest as deeply as one wants
06:34:07 [melvster1]
but do you have to put everything in { }
06:34:14 [melvster1]
or can you start with [ ]
06:34:29 [mhausenblas]
no, whatever you like - Couch is agnostic to the payload
06:34:34 [mhausenblas]
as long as it is valid JSON
06:34:40 [melvster1]
sweet
06:34:59 [mhausenblas]
well, you can have attachments, as well
06:35:05 [mhausenblas]
but not relevant for our use case
06:35:17 [mhausenblas]
the main question now is to index stuff
06:35:18 [melvster1]
my vague thought was that if each serialization has its own content-type then we can have ways to read them and change between the ... of course couch itself probably doesnt let you do that
06:35:26 [mhausenblas]
hence I looked into elasticsearch connector
06:35:33 [melvster1]
s/the/them
06:35:35 [mhausenblas]
hmmm
06:35:39 [mhausenblas]
good question
06:35:49 [mhausenblas]
I think I can do this in Couch
06:36:46 [mhausenblas]
yeah, via shows … see http://guide.couchdb.org/editions/1/en/show.html
06:37:33 [mhausenblas]
ok, gotta dash off, grab some more coffee
06:37:39 [mhausenblas]
TTYL, melvster1
06:37:47 [melvster1]
cool!
06:37:52 [melvster1]
thx for the pointers! :)
06:38:01 [mhausenblas]
yw ;)
07:06:07 [trueg]
trueg (~trueg@HSI-KBW-46-237-239-121.hsi.kabel-badenwuerttemberg.de) has joined #dig
08:32:06 [deiu]
deiu (~andrei@unaffiliated/deiu) has joined #dig
08:49:12 [rszeno]
rszeno (~rszeno@79.114.63.31) has joined #dig
09:34:13 [danbri]
danbri (~danbri@80.71.30.113) has joined #dig
09:48:30 [danbri]
danbri has quit (Remote host closed the connection)
09:48:56 [danbri]
danbri (~danbri@80.71.30.113) has joined #dig
09:50:50 [mhausenblas]
mhausenblas has quit (Quit: http://mhausenblas.info/#i says TTYL)
09:53:22 [danbri]
danbri has quit (Ping timeout: 245 seconds)
10:10:16 [trueg]
trueg is now known as trueg_away
10:44:34 [trueg_away]
trueg_away is now known as trueg
11:00:43 [trueg]
trueg is now known as trueg_away
11:14:34 [RalphS]
RalphS (RSwick@30-7-118.wireless.csail.mit.edu) has joined #dig
11:45:48 [kennyluck]
kennyluck has quit (Quit: kennyluck)
12:01:56 [trueg_away]
trueg_away is now known as trueg
12:06:11 [betehess_laptop]
betehess_laptop (~betehess@31-33-151.wireless.csail.mit.edu) has joined #dig
13:20:43 [scor]
scor (~scor@drupal.org/user/52142/view) has joined #dig
13:24:09 [bblfish]
betehess: betehess_laptop: are you on rdf working group? (Or anyone else?)
13:24:46 [bblfish]
mhh let me bring that to #swig
13:26:25 [bblfish]
just noticed that one needs full URL context to correctly interpret SPARQL betehess
13:26:31 [bblfish]
in Jena I think
13:49:20 [betehess_laptop]
yes, SPARQL is defined against RDF, hence absolute URIs
13:49:31 [betehess_laptop]
I'm not part of the RDF WG
13:49:55 [bblfish]
ah ok: just added a new bug report https://github.com/w3c/banana-rdf/issues/35
13:50:40 [bblfish]
betehess_laptop: SPARQL is defined againts absolute URIs, but you can use relative uris there too ( to name graphs for example )
13:51:08 [betehess_laptop]
I don't think this is sound in SPARQL
13:51:44 [bblfish]
both Jena and Sesame do it
13:51:54 [betehess_laptop]
I don't think so
13:52:02 [bblfish]
they do I checked
13:52:12 [betehess_laptop]
more precisely: you may not have what you expect
13:52:19 [betehess_laptop]
well, I think I had issues with that
13:52:31 [betehess_laptop]
because of the way they are resolving the uris
13:52:37 [betehess_laptop]
you should not rely on that
13:52:47 [bblfish]
well I can imaine they get it wrong, as usual with anything webby
13:52:49 [betehess_laptop]
it's clearly a gray area
13:53:56 [bblfish]
use case: POST a SPARQL query to a collection: to query across a number of graphs in the collection with relative URLs
13:55:14 [bblfish]
http://www.w3.org/TR/sparql11-query/#relIRIs
13:56:17 [betehess_laptop]
yeah, you define BASE
13:56:29 [betehess_laptop]
and the API lets you define BASE from the outside
13:57:27 [bblfish]
argh god. The work one has to do here to get really basic things working!
13:58:02 [bblfish]
speaking about debate on ldp list!
14:14:08 [jnpato_]
jnpato_ (jnpato@30-7-93.wireless.csail.mit.edu) has joined #dig
14:15:00 [jnpato]
jnpato (jnpato@30-7-93.wireless.csail.mit.edu) has joined #dig
14:16:19 [jnpato_]
jnpato_ has quit (Client Quit)
14:16:42 [jnpato]
jnpato has quit ()
14:17:04 [jnpato]
jnpato (jnpato@30-7-93.wireless.csail.mit.edu) has joined #dig
14:19:45 [betehess_laptop]
betehess_laptop has quit (Remote host closed the connection)
14:29:12 [presbrey]
presbrey has quit (Excess Flood)
14:29:35 [presbrey]
presbrey (~presbrey@2001:4830:2446:b5:aede:48ff:fe00:6001) has joined #dig
14:58:01 [bblfish]
ah good progress on the list!
15:00:57 [trueg]
trueg is now known as trueg_away
15:15:03 [deiu]
deiu has quit (Ping timeout: 260 seconds)
15:15:50 [Pipian-Work]
Pipian-Work (~Pipian@2001:470:8b2d:7d8:d094:718c:a908:8cc0) has joined #dig
15:56:38 [jnpato]
jnpato has quit (Ping timeout: 245 seconds)
17:38:26 [RalphS]
RalphS has quit (Ping timeout: 246 seconds)
17:38:48 [RalphS]
RalphS (RSwick@w3cdhcp61.w3.org) has joined #dig
17:43:32 [trueg_away]
trueg_away is now known as trueg
18:24:13 [bblfish]
bblfish has quit (Remote host closed the connection)
18:27:42 [bblfish]
bblfish (~bblfish@AAubervilliers-652-1-23-200.w90-35.abo.wanadoo.fr) has joined #dig
18:29:36 [trueg]
trueg is now known as trueg_away
18:32:27 [bblfish]
bblfish has quit (Ping timeout: 260 seconds)
18:56:48 [melvster]
melvster (~melvin@p4FF9674D.dip.t-dialin.net) has joined #dig
18:57:24 [melvster1]
melvster1 has quit (Ping timeout: 260 seconds)
19:33:34 [melvster]
melvster has quit (Ping timeout: 260 seconds)
19:52:58 [bblfish]
bblfish (~bblfish@AAubervilliers-651-1-329-35.w83-114.abo.wanadoo.fr) has joined #dig
20:09:02 [melvster]
melvster (~melvin@p4FF9674D.dip.t-dialin.net) has joined #dig
20:12:02 [Pipian-Work]
Pipian-Work has quit (Quit: Pipian-Work)
20:19:44 [melvster]
melvster has quit (Ping timeout: 260 seconds)
20:20:57 [RalphS]
RalphS has quit ()
20:45:11 [scor]
scor has quit (Quit: scor)
20:50:29 [Pipian-Work]
Pipian-Work (~Pipian@30-9-71.wireless.csail.mit.edu) has joined #dig
23:03:02 [rszeno]
rszeno has quit (Quit: Leaving.)
23:31:01 [Pipian-Work]
Pipian-Work has quit (Quit: Pipian-Work)
23:34:20 [danbri]
danbri (~danbri@80.71.30.113) has joined #dig