IRC log of dig on 2012-11-18

Timestamps are in UTC.

00:29:46 [kennyluck]
kennyluck has quit (Quit: kennyluck)
01:08:47 [kennyluck]
kennyluck (~kennyluck@203.117.93.98) has joined #dig
01:09:45 [kennyluck]
kennyluck has quit (Remote host closed the connection)
01:10:05 [kennyluck]
kennyluck (~kennyluck@203.117.93.98) has joined #dig
01:27:41 [kennyluck]
kennyluck has quit (Read error: Connection reset by peer)
01:27:43 [kennyluck_]
kennyluck_ (~kennyluck@203.117.93.98) has joined #dig
01:36:10 [kennyluck]
kennyluck (~kennyluck@203.117.93.98) has joined #dig
01:36:22 [kennyluck_]
kennyluck_ has quit (Read error: Connection reset by peer)
01:45:11 [kennyluck]
kennyluck has quit (Read error: Connection reset by peer)
01:45:30 [kennyluck]
kennyluck (~kennyluck@203.117.93.98) has joined #dig
02:31:15 [kennyluck]
kennyluck has quit (Read error: Connection reset by peer)
02:31:25 [kennyluck]
kennyluck (~kennyluck@203.117.93.98) has joined #dig
03:27:40 [timbl]
cwm can sign things
03:27:42 [timbl]
Timestamp: 2012/11/17 22:26:55
03:27:43 [timbl]
Error: WebSocket is not defined
03:27:44 [timbl]
Source File: chrome://tabulator/content/js/rdf/dist/rdflib.js
03:27:45 [timbl]
Line: 6630
04:12:22 [kennyluck]
kennyluck has quit (Read error: Connection reset by peer)
04:12:26 [kennyluck_]
kennyluck_ (~kennyluck@203.117.93.98) has joined #dig
04:43:11 [kennyluck_]
kennyluck_ has quit (Read error: Connection reset by peer)
04:43:16 [kennyluck]
kennyluck (~kennyluck@203.117.93.98) has joined #dig
05:10:12 [presbrey]
which browser doesnt have WebSocket timbl?
05:12:30 [presbrey]
my firefox 16.0.2 and chrome 24 work
05:35:25 [kennyluck]
Safari perhaps? cause it's prefixed?
05:47:53 [presbrey]
the Source File looks like tabulator extension
05:48:23 [presbrey]
of course it shouldnt even be triggered unless the data-source sent an 'Updates-Via' header
05:49:38 [presbrey]
some (older?) firefoxes need MozWebSocket
05:55:03 [Snuupy]
woah, you guys are alive!
06:04:30 [melvster]
melvster (~melvin@p5797FB12.dip.t-dialin.net) has joined #dig
06:39:00 [melvster]
melvster has quit (Read error: Connection reset by peer)
06:39:45 [melvster]
melvster (~melvin@p5797FB12.dip.t-dialin.net) has joined #dig
06:45:23 [melvster]
melvster has quit (Quit: Leaving.)
06:46:32 [melvster]
melvster (~melvin@87.151.251.18) has joined #dig
06:51:22 [cheater_1]
cheater_1 (~cheater@p4FD0FFC3.dip.t-dialin.net) has joined #dig
06:54:43 [cheater__]
cheater__ has quit (Ping timeout: 260 seconds)
07:50:53 [rszeno]
rszeno has quit (Quit: Leaving.)
09:16:31 [bblfish]
bblfish (~bblfish@78.251.73.39) has joined #dig
09:46:14 [Snuupy]
Snuupy has quit (Excess Flood)
09:51:56 [Snuupy]
Snuupy (~Snuupy@unaffiliated/snuupy) has joined #dig
11:01:52 [trueg_away]
trueg_away is now known as trueg
11:23:38 [trueg]
trueg is now known as trueg_away
11:25:40 [kennyluck]
kennyluck has quit (Quit: kennyluck)
11:56:44 [timbl]
Firefox extension
12:37:30 [kennyluck]
kennyluck (~kennyluck@bb115-66-236-93.singnet.com.sg) has joined #dig
13:49:29 [trueg_away]
trueg_away is now known as trueg
14:06:35 [trueg]
trueg is now known as trueg_away
14:25:58 [rszeno]
rszeno (~rszeno@79.114.118.96) has joined #dig
15:42:16 [kennyluck]
kennyluck has quit (Ping timeout: 246 seconds)
15:43:09 [kennyluck]
kennyluck (~kennyluck@bb115-66-236-93.singnet.com.sg) has joined #dig
17:22:36 [GitHub32]
[13tabulator] 15timbl pushed 1 new commit to 06master: 02http://git.io/lcS-fw
17:22:36 [GitHub32]
13tabulator/06master 14b45d4dc 15Tim Berners-Lee: Attach pane basically works, right hand panel (view attached thing) is a bit of a kludge. Re-added the image pane, as the atach pane uses it.
17:24:11 [GitHub193]
[13tabulator] 15timbl pushed 1 new commit to 06master: 02http://git.io/Ue9Bdw
17:24:11 [GitHub193]
13tabulator/06master 14f240f96 15Tim Berners-Lee: Ooops forgot the changes to the style file
17:41:39 [deiu]
deiu (~andrei@2001:470:8b2d:7d4:474:c1a8:d174:fda6) has joined #dig
17:41:39 [deiu]
deiu has quit (Changing host)
17:41:39 [deiu]
deiu (~andrei@unaffiliated/deiu) has joined #dig
17:47:36 [betehess]
betehess (~betehess@c-24-61-44-145.hsd1.ma.comcast.net) has joined #dig
17:48:50 [betehess]
bblfish, you actually *can* query a Sesame Graph without going through a Repository
17:48:54 [betehess]
see https://github.com/w3c/banana-rdf/blob/077854b7bc51246484743743875a162deacacca3/plantain/src/main/scala/plantain/PlantainGraphSparqlEngine.scala#L19 :-)
17:58:59 [kennyluck]
kennyluck has quit (Read error: Connection reset by peer)
17:59:11 [kennyluck]
kennyluck (~kennyluck@bb115-66-236-93.singnet.com.sg) has joined #dig
17:59:39 [kennyluck]
kennyluck has quit (Read error: Connection reset by peer)
18:02:02 [kennyluck]
kennyluck (~kennyluck@bb115-66-236-93.singnet.com.sg) has joined #dig
18:32:36 [deiu]
bblfish, do you have the WebID sequence diagram as CSV?
19:13:19 [trueg_away]
trueg_away is now known as trueg
19:44:55 [trueg]
trueg is now known as trueg_away
19:54:35 [bblfish]
bblfish has quit (Remote host closed the connection)
19:55:07 [bblfish]
bblfish (~bblfish@78.251.73.39) has joined #dig
20:07:19 [bblfish]
bblfish has quit (Remote host closed the connection)
20:07:20 [trueg_away]
trueg_away is now known as trueg
20:22:20 [trueg]
trueg is now known as trueg_away
20:25:02 [bblfish]
bblfish (~bblfish@78.251.92.218) has joined #dig
20:49:51 [presbrey]
bblfish, if you want to have regex we should support simple globbing too
20:50:26 [presbrey]
most users do not write /admin/.*, they write /admin/*
20:51:56 [presbrey]
also do we really want to incorporate blank nodes? this is the first proposal to do so
20:54:32 [presbrey]
such a pattern also seems to duplicate eg.
20:54:41 [presbrey]
acl:defaultForNew </admin/>
20:57:00 [presbrey]
also in this particular scenario, it costs more to compile the regex pattern than to evaluate it
20:58:26 [presbrey]
in more complex examples, the server now needs a resident regex cache
20:59:23 [melvster]
perhaps arbitrary regex could be an attack surface too depending on who has accesss
21:23:51 [presbrey]
betehess, do you know how I can parse ant style in python or php?
21:24:19 [presbrey]
and javascript? :)
21:24:22 [betehess]
shouldn't be difficult
21:24:33 [betehess]
we'll need to define the regex grammar anyway
21:25:00 [betehess]
at the end, any language should be able to compile them down to their own native regex style
21:26:09 [presbrey]
at the end?
21:26:14 [betehess]
http://trac.mach-ii.com/machii/wiki/ANTPatternMatcher
21:26:29 [betehess]
just three wildcards
21:26:46 [betehess]
having both ** and * is pretty cool
21:32:42 [bblfish]
bblfish has quit (Remote host closed the connection)
21:35:56 [presbrey]
hm that does seem cool
21:36:22 [presbrey]
now I wonder if he meant posix regex, perl regex, or one of the others
21:39:16 [presbrey]
presbrey has quit (Remote host closed the connection)
21:41:23 [deiu]
betehess, what was your argument for avoiding 303 redirects in LDP?
21:41:41 [deiu]
can you put it into one sentence? (I need it for the new spec)
21:43:59 [betehess]
let me think
21:45:15 [presbrey]
presbrey (~presbrey@2001:4830:2446:b5:aede:48ff:fe00:6001) has joined #dig
22:01:47 [betehess]
deiu, what about something about these lines:
22:01:48 [betehess]
[[
22:01:51 [betehess]
1. HTTP 303 tells an agent that the content for the requested URI is somewhere else. When combined with LDP -- where the Write operation is defined -- it makes difficult to know where to actually write.
22:01:51 [betehess]
2. hash-URIs (with http 200 for the hashless URI) makes very explicit the relationship between the RDF resource and its LDPR.
22:01:51 [betehess]
Because of 1. and 2., HTTP 303 should be avoided in favor of hash-URIs.
22:01:52 [betehess]
]]
22:02:26 [betehess]
my main issue is clearly "where should I write back?"
22:02:45 [betehess]
and saving a round-trip
22:05:37 [betehess]
makes me think that we may have a terminology issue: how do we call http://example.com/foo#bar ?
22:05:55 [betehess]
.. if we say that http://example.com/foo is an LDPR
22:06:06 [betehess]
http://example.com/foo#bar is an RDF resource
22:06:24 [betehess]
but with RDF resource, you loose the fact that HTTP/LDP is involved
22:06:45 [betehess]
or is it implied because of the http scheme for the URI?
22:07:47 [deiu]
hmm
22:08:28 [deiu]
Also, if you allow 303 redirects, you will have to also specify the max number of redirects at some point.
22:09:26 [betehess]
right
22:09:46 [deiu]
So far I think it's important to say that by using a hash URI, there is no longer the need to make the difference between the identifier (hash URI) and the document it points to.
22:09:50 [betehess]
if we could avoid have to specify that :-)
22:09:56 [betehess]
s/have/having/
22:10:16 [deiu]
yeah, I've put this part in red (as a current ISSUE)
22:10:31 [betehess]
you mean: there is a difference, but the relationship is obvious?
22:10:35 [deiu]
yeah
22:10:41 [betehess]
yes
22:10:53 [betehess]
it just makes everything easy
22:11:09 [deiu]
I'll leave it like that for now then
22:11:40 [betehess]
wow
22:12:11 [deiu]
it's _way_ shorter than the initial one
22:12:15 [deiu]
which is good
22:12:54 [betehess]
by implementing my new LDP back-end, I now strongly feel the need to say that an LDPC *is* a SPARQL endpoint, where the graphs denote its LDPR
22:13:20 [betehess]
(limited to SELECT, CONSTRUCT, ASK... no UPDATE)
22:13:55 [deiu]
until you implement ACL
22:13:57 [betehess]
that will be the best way to query the collection
22:14:06 [betehess]
what would ACL change?
22:14:11 [deiu]
once ACL is up and running, you also allow PUT/POST?
22:14:14 [deiu]
and DELETE
22:14:24 [betehess]
well, not sure how UPDATE fits there yet
22:14:34 [betehess]
I'd still use my client-side API
22:14:39 [betehess]
instead of SAPRQL
22:15:14 [betehess]
I have DELETE, but this works at the LDPR level
22:15:22 [deiu]
hmm
22:15:37 [betehess]
I mean, I have all the udpate operations on LDPR, no need to go through the LDPC
22:15:51 [deiu]
for me UPDATE/DELETE implies having a basic ACL running
22:15:55 [betehess]
sure
22:16:08 [betehess]
but for now, I'm just focused on the bacj-end
22:16:20 [betehess]
ACLs will come later
22:17:47 [deiu]
well, if you could store the identity of the user requesting the creation the LDPC/R (through a WebID-TLS auth), you can at least support UPDATE/DELETE for that user
22:18:02 [deiu]
that's why I'm trying to push the idea of ldp:Owner
22:18:05 [betehess]
I'd need a little ACL engine
22:18:35 [deiu]
yeah, but it's a basic match for the LDPC/R owner :-)
22:19:53 [betehess]
to be honest, I don't need ACLs in my app for now
22:21:39 [deiu]
I need a minimal ACL engine, at least supporting ownership
22:22:08 [deiu]
to separate applications storing stuff on the LDP back-end
23:09:19 [deiu]
https://dvcs.w3.org/hg/WebID/raw-file/tip/spec/identity-respec.html
23:09:31 [deiu]
this is the first draft
23:10:40 [deiu]
time to go home
23:10:42 [melvster]
melvster has quit (Ping timeout: 252 seconds)
23:12:18 [deiu]
deiu has quit (Quit: Leaving)
23:45:25 [betehess]
betehess has quit (Quit: Leaving)