IRC log of dig on 2011-12-13

Timestamps are in UTC.

12:22:30 [DIGlogger]
DIGlogger (~dig-logge@groups.csail.mit.edu) has joined #dig
12:22:30 [hubbard.freenode.net]
topic is: Decentralized Information Group @ MIT http://dig.csail.mit.edu/
12:22:30 [hubbard.freenode.net]
Users on #dig: DIGlogger RalphS cheater melvster bblfish rszeno timbl presbrey amy kennyluck nunnun_away ericP sandro mattl betehess manu-db manu1 Yudai
12:40:52 [danbri]
danbri (~danbri@ip176-48-210-87.adsl2.static.versatel.nl) has joined #dig
12:44:04 [cheater]
cheater has quit (Ping timeout: 240 seconds)
12:53:59 [timbl_]
timbl_ (~timbl@c-24-62-225-11.hsd1.ma.comcast.net) has joined #dig
12:54:00 [timbl_]
timbl_ has quit (Client Quit)
12:57:27 [timbl]
timbl has quit (Ping timeout: 240 seconds)
14:01:21 [timbl]
timbl (~timbl@31-33-140.wireless.csail.mit.edu) has joined #dig
15:07:35 [scor]
scor (~scor@drupal.org/user/52142/view) has joined #dig
15:24:28 [nunnun_away]
nunnun_away is now known as nunnun
15:31:15 [presbrey]
timbl, time to meet about tabulator + ACL spec today?
15:33:22 [timbl]
ummm
15:34:14 [timbl]
Sure, any time from now through lunch
15:38:09 [timbl]
Alex could meet over lunch
15:39:47 [tlr]
tlr (~tlr@195-86-126-98.dsl.easynet.nl) has joined #dig
15:48:02 [timbl]
I'm afraid the afteroon is full
16:07:25 [presbrey]
k I'll come down now timbl
16:18:14 [presbrey]
arg bike tire flat
16:19:52 [presbrey]
fixing, cya soon
16:22:46 [timbl]
good luck with the tire
16:38:21 [nunnun]
nunnun is now known as nunnun_away
16:40:37 [Pipian_]
Pipian_ (~pipian@31-33-77.wireless.csail.mit.edu) has joined #dig
16:53:46 [rszeno]
rszeno has quit (Ping timeout: 252 seconds)
17:00:56 [nunnun_away]
nunnun_away is now known as nunnun
17:01:08 [tlr]
tlr has quit (Quit: tlr)
17:06:29 [rszeno]
rszeno (~rszeno@79.114.79.149) has joined #dig
17:48:33 [bblfish]
Is betehess around? I have not seen him in a while
17:54:38 [melvster1]
melvster1 (~melvin@p5797F115.dip.t-dialin.net) has joined #dig
17:55:44 [melvster]
melvster has quit (Ping timeout: 252 seconds)
18:52:56 [Pipian__]
Pipian__ (~pipian@30-9-3.wireless.csail.mit.edu) has joined #dig
18:54:45 [Pipian_]
Pipian_ has quit (Ping timeout: 248 seconds)
19:01:40 [bblfish]
timbl in the WebID spec we currently make something of the distinction between terms on the web: those that are defined in a document and the others
19:02:19 [bblfish]
see section https://dvcs.w3.org/hg/WebID/raw-file/tip/spec/index-respec.html#verifying-the-webids
19:09:50 [timbl]
http://www.chem.helsinki.fi/~jonas/git_guides/HTML/CVS2git/
19:09:59 [timbl]
CVS to git Transition Guide
19:10:09 [timbl]
Wondering about it for CWM
19:28:39 [bblfish]
having cwm in git or hg would I am sure be useful
19:29:06 [bblfish]
it's a pitty cwm does not do rdfa
19:31:09 [rszeno]
imo cwm need first a cmd line option to pass custom accept header
19:34:23 [timbl]
yes? what for in particular?
19:34:34 [timbl]
(maybe should add some curl command line options)
19:35:16 [bblfish]
it's a pitty DanC did not continue work on swap scala http://code.google.com/p/swap-scala/updates/list - I want to look at that in more detail at some point now that I can write scaal. But I think betehess is doing something in that space too
19:38:46 [rszeno]
most of the sites now respond with text/html even if they have rdf, setting in webacces the accept will make then to send rdf
19:41:02 [rszeno]
i'm also looking to DanC scala-web, i started one month ago but i'm not too confortable with scala
19:42:53 [bblfish]
It takes a bit of time.
19:43:06 [bblfish]
make sure to use intellij it helps a lot
19:43:12 [bblfish]
latest version 11 is out now
19:44:04 [bblfish]
for me Scala takes a lot of time, mostly because there is so much to learn from the functional programming side
19:44:15 [bblfish]
so I am learning a bit about Monads and things like that
19:44:45 [rszeno]
i'm lisper, :) scala is odd for me, :)
19:45:03 [bblfish]
ah yes. I am a javaite
19:45:22 [bblfish]
though I did program in lisp when I was 13
19:45:24 [rszeno]
intellij is a ide? or only a ide?
19:45:31 [bblfish]
it's an IDE
19:45:38 [bblfish]
like emacs
19:45:42 [rszeno]
i use emacs, :)
19:45:55 [rszeno]
scala have a mode for emacs
19:45:58 [bblfish]
ah ok. Well So did James Gosling
19:46:09 [bblfish]
who wrote emacs
19:46:42 [bblfish]
but you should check out IntelliJ anyway. They have done the best job, so if you want to compare emacs with the best of IDEs
19:48:18 [rszeno]
thank you for info and advice, yes variablity means usualy progress, so i'm for trying different things, :)
19:50:05 [bblfish]
iI think RDF is the new lisp
19:50:06 [bblfish]
:-)
19:50:24 [bblfish]
it's lisp pushed down even simpler: down to relations :-)
19:50:38 [bblfish]
emacs needs to be rewritten in rdf
19:52:40 [bblfish]
perhaps they'll put together rdf machines, like they once had lisp machines
19:53:03 [bblfish]
Ah when will we get RDF support in the CPU...
19:59:42 [rszeno]
i'm against droping things which already work to make new ones to achive same thing, :) imo, is better to build glue for what already exists
20:00:59 [bblfish]
I was kidding of course
20:01:18 [bblfish]
but it's odd that nobody has posed those questions with respect to rdf
20:01:29 [bblfish]
any other language people think of writing specialised cpus
20:01:33 [bblfish]
they even did for java
20:02:00 [bblfish]
:-) just a thought
20:02:17 [rszeno]
kpu - knowledge processing unit, concept was invented around '70 i guess
20:02:57 [rszeno]
tokio, their program who fail, based on prolog
20:04:14 [rszeno]
i hope i'm not wrong, i don't remeber details, years mainly, :)
20:05:26 [rszeno]
wrong, 1992, http://www.informatik.uni-trier.de/~ley/db/conf/fgcs/index.html
20:09:54 [bblfish]
anyway it's pretty cool that DanC got RDFa working in that scala version!
20:11:58 [rszeno]
i guess DanC focus is more on proof cheking then parsing in cwm-scala
20:13:50 [bblfish]
no, I think he has parsers for every rdf serialisation in there
20:14:01 [bblfish]
which is amazing because he wrote them in so little code
20:15:05 [rszeno]
:)
20:16:16 [rszeno]
brb
20:16:22 [rszeno]
rszeno has quit (Quit: Leaving.)
20:25:41 [timbl]
bblfish, re ": perhaps they'll put together rdf machines, like they once had lisp machines", haystack may have in fact been a rdf machine IIRC - R-code
20:26:02 [Pipian__]
Pipian__ has quit (Quit: Pipian__)
20:30:20 [bblfish]
ah yes. Though I think haystack had no notion of graph so putting all the triples into one container was bound to lead to trouble
20:31:12 [bblfish]
but cine to think of it I did come across someone who write something like executable rdf, ...
20:31:49 [rszeno]
rszeno (~rszeno@79.114.93.227) has joined #dig
20:43:26 [timbl]
----------
20:43:27 [timbl]
https://docs.google.com/View?id=dd9g3qjp_103fdxjk3qt
20:43:40 [timbl]
Attribution, Open IP and Highly Factual, Highly Granular Data
20:46:05 [timbl]
Bon App!
20:50:00 [timbl]
Datr
20:53:58 [Pipian_]
Pipian_ (~pipian@18.111.8.83) has joined #dig
20:56:37 [timbl]
timbl has quit (Quit: timbl)
21:02:38 [timbl]
timbl (~timbl@31-33-140.wireless.csail.mit.edu) has joined #dig
21:12:24 [scor]
scor has quit (Quit: scor)
21:12:49 [lkagal]
lkagal (~lkagal@30-7-251.wireless.csail.mit.edu) has joined #dig
21:12:53 [lkagal]
timbl, http://web.mit.edu/~wstyke/Public/cv.rdf
21:13:15 [timbl]
http://web.mit.edu/~wstuke/Public/cv.rdf
21:13:23 [timbl]
tx
21:44:10 [lkagal]
timbl, http://web.mit.edu/~frango/Public/6.932/mayank.rdf
21:45:15 [lkagal]
http://web.mit.edu/frango/Public/6.932/social_schema
21:55:07 [lkagal]
lkagal has quit (Ping timeout: 244 seconds)
21:59:07 [Pipian_]
Pipian_ has quit (Quit: Pipian_)
22:00:16 [Pipian_]
Pipian_ (~pipian@18.111.8.83) has joined #dig
22:00:16 [Pipian_]
Pipian_ has quit (Client Quit)
22:04:10 [presbrey]
ok so I grabbed https://github.com/linkeddata at betehess's suggestion, shall we stash tabulator there? timbl, make last commit please :)
22:04:12 [rszeno]
rszeno has quit (Quit: Leaving.)
22:04:56 [betehess]
good job re: linkeddata
22:06:09 [presbrey]
yes I think you are also right re. gh:/linkeddata/tabulator, gh:/linkeddata/tabulator-firefox, gh:/linkeddata/tabulator-chrome, etc.
22:06:43 [presbrey]
optionally splitting rdflib.js into its own as well
22:07:13 [presbrey]
or calling it tabulator to avoid naming conflict with C's librdf
22:07:21 [RalphS]
RalphS has quit (Ping timeout: 252 seconds)
22:08:10 [timbl]
and linkeddata/rdflibjs
22:08:19 [rszeno]
rszeno (~rszeno@79.114.98.172) has joined #dig
22:11:50 [lkagal]
lkagal (~lkagal@30-7-251.wireless.csail.mit.edu) has joined #dig
22:17:30 [timbl]
The library isn't Tabulator -- Tabulator is a project or an app
22:17:50 [timbl]
Yes, problem with things called rdflib librdf etc
22:21:01 [melvster1]
... and rdfquery ... is jenit's js library
22:26:38 [presbrey]
I like rdflib.js, seems like only our stuff comes up when googling it
22:28:47 [bblfish]
are you moving away from hg ? Or is it just to grow visibility ?
22:28:58 [presbrey]
hey bblfish
22:29:01 [presbrey]
both
22:29:10 [bblfish]
ah ok
22:29:12 [presbrey]
tiny thing we need on WebID/browsers:
22:29:23 [presbrey]
Warning: Your WebID will expire in 4 days on Sun Dec 18 18:11:48 2011
22:29:35 [presbrey]
when using mine today... :)
22:29:38 [melvster1]
oh noes
22:29:53 [melvster1]
i set mine to 100 years for both GPG / WebID :)
22:29:59 [bblfish]
ah yes. Something to add to the browsers bug issues, when they understand this
22:31:26 [bblfish]
we published new version of http://webid.info/spec yesterday btw.
22:33:08 [presbrey]
why is there a note on renegotiation?
22:33:37 [bblfish]
let me look
22:33:56 [bblfish]
which note?
22:34:25 [presbrey]
sec 3.2.3
22:34:45 [bblfish]
because it is important that people know that that issue has been fixed
22:35:11 [bblfish]
and also at this stage that people know that we know this
22:35:21 [presbrey]
the renegotiation topic is unrelated to WebID imo
22:35:39 [bblfish]
ah it is very important if you want to have a good user experience
22:35:42 [presbrey]
I can add a note that I am also familiar with the BEAST attack
22:36:03 [bblfish]
the current spec tries to make renegotiation central
22:36:16 [bblfish]
for 100% https web
22:36:31 [bblfish]
if you don't have renege, you have horrible user experience
22:36:37 [bblfish]
on a 100% https web
22:36:38 [lkagal]
lkagal has quit (Ping timeout: 252 seconds)
22:37:08 [bblfish]
because the browser will ask the user for his certificate even when there is no need for it
22:37:17 [bblfish]
i.e. page has not need for security
22:37:45 [bblfish]
the wiki of the webid group shows how one can do this in apache
22:37:51 [presbrey]
how is browser behavior / ui tied to SSL renegotiation in the protocol?
22:38:02 [bblfish]
well it's a major use case
22:38:12 [bblfish]
look at the sequence diagram
22:39:41 [presbrey]
most servers I know of now are disabling renegotiation, using server-preference for the cipher, and setting the allowed to a fixed high set
22:40:17 [presbrey]
cheap TLS frontends like stud and stunnel give good performance by alleviating SSL burden from the webserver
22:40:22 [bblfish]
well so they will not be able to use that feature
22:40:34 [bblfish]
it is not required
22:40:37 [presbrey]
anyway, all of this is 0% related to webID imo
22:40:48 [bblfish]
I disagree
22:41:04 [bblfish]
WebID is about bringing lots of different technologies together
22:41:26 [bblfish]
it's an explanation of how you can put things together that nobody thought of putting together
22:41:45 [presbrey]
the word "renegotiation" only appears 1 time in the spec in a <p class="note">
22:42:12 [bblfish]
oh ok. Well the spec is not final
22:42:14 [bblfish]
:-)
22:42:19 [presbrey]
it seems really weird to bring that word/issue/bugfix into webid spec at that level and frequency
22:42:50 [timbl]
timbl has quit (Ping timeout: 252 seconds)
22:43:35 [bblfish]
it can be improved as I said. This is a lot better than what we had before
22:44:03 [bblfish]
it's readable at least. Anyway, I seem to be doing most of the work there. I am happy to have people help out
22:44:10 [presbrey]
cool
22:44:20 [presbrey]
thanks! I didn't mean to criticize
22:44:38 [Pipian_]
Pipian_ (~pipian@31-34-37.wireless.csail.mit.edu) has joined #dig
22:44:52 [bblfish]
essentially this spec is about getting people from security and linked data to meet
22:45:19 [bblfish]
so we need to explain some things that seem obvious for each side without getting overly complicated, but without simplifying too much
22:45:36 [bblfish]
if you simplify one side is going to get upset
22:45:40 [presbrey]
I didn't get to read it all yet, I can form some more complete notes or do some editing by hand this holidays, let me know how...
22:45:57 [bblfish]
that would be really useful
22:46:11 [bblfish]
I think we will have a couple of pieces of feedback in the coming week
22:46:35 [bblfish]
just check the editors draft linked from the spec in hg for the latest version
22:48:27 [presbrey]
dvcs hg or?
22:49:13 [bblfish]
yes
22:49:19 [bblfish]
it's linked to from the spec at the top
22:49:23 [bblfish]
hg
22:50:57 [timbl]
timbl (~timbl@31-33-140.wireless.csail.mit.edu) has joined #dig
22:51:58 [bblfish]
oh and the names on the authors list is a bit fictional at present
22:52:04 [bblfish]
some have not contributed in a year
22:52:09 [bblfish]
others not at all
22:52:17 [bblfish]
so there is space to put new names up there
22:54:36 [presbrey]
cool when are you trying for next draft?
22:56:24 [timbl]
timbl has quit (Quit: timbl)
22:56:32 [bblfish]
January we should have something new out
22:56:33 [bblfish]
it's an ongoing process
22:56:36 [melvster1]
seems to me one group of identity folks want to use email style identifiers and one group want to use http:// uri's
22:57:00 [presbrey]
emails are uris too ;)
22:57:05 [bblfish]
yes
22:57:13 [presbrey]
can we keep to just URI or do we have to say http?
22:57:28 [presbrey]
eg. https URLs are even better
22:57:31 [bblfish]
the webid spec currently concentrates on https
22:57:46 [bblfish]
we could then have another one that shows how to tie claims together
22:57:52 [bblfish]
in the end
22:58:03 [bblfish]
webid foaf:mbox <mailto:..>
22:58:44 [bblfish]
the problem is that if we add every URI then the spec becomes long, and difficult to implement
22:58:58 [melvster1]
i think the spec says that all webids are http urls ... or at least that was implied from my understanding of the 'profile page'
22:59:16 [bblfish]
well sounds ok to me
22:59:21 [bblfish]
it's a WEB Id
22:59:26 [bblfish]
not an SMTP id
23:00:07 [melvster1]
sure but as i say there's a whole bunch of people that want to use email like identifiers on the web
23:00:18 [bblfish]
yes, but they are not on our team
23:00:32 [bblfish]
i.e., those people don't usually try to interact with us
23:00:48 [bblfish]
never saw them come up with serious suggestions
23:01:04 [bblfish]
it's just us trying to be open to them, where frankly they don't care
23:01:17 [bblfish]
so we can add these things, then they'll just say we're complicated
23:01:18 [presbrey]
hehe ok
23:01:23 [presbrey]
well we could just say
23:01:27 [presbrey]
"WebID URIs are resolved to public key using the mechanism designated by its URI scheme"
23:01:38 [bblfish]
yes,
23:01:48 [bblfish]
and add. This document focuses on https uris
23:01:50 [presbrey]
but if that means webFinger for email, why is there rdf/foaf for HTTP?
23:01:51 [Pipian_]
Pipian_ has quit (Read error: Operation timed out)
23:02:32 [bblfish]
well we have GRDDL
23:02:42 [bblfish]
so we would have to then develop the GRDDL section
23:02:55 [bblfish]
you see things get quite complex, as zoon as you do that
23:03:07 [melvster1]
Personally i think you cover all bases by saying a WebID is a URI that refers to an agent, and that the WebID Protocol is a way to link an X.509 certificate to the public key on that http(s) URI
23:03:10 [bblfish]
but I am for it
23:03:51 [bblfish]
Well I think we can be explicit and say that we focus on https uris
23:03:59 [bblfish]
in this spec
23:04:54 [bblfish]
Well at the same time a WebID that would use say an XRI is not really a Web id anymore
23:05:01 [bblfish]
it's kind of
23:05:05 [bblfish]
but you know
23:05:36 [melvster1]
you mean XRD / Webfinger?
23:05:40 [bblfish]
it is extremely easy to make things exceedingly complicated
23:05:55 [bblfish]
yes
23:06:05 [melvster1]
that's different from XRI
23:06:18 [bblfish]
ah ok. Well
23:06:32 [bblfish]
As I see it our base are the linked data folk
23:06:55 [bblfish]
the openid people and others don't really come to us. And I'd be happy to add support for XRD
23:06:59 [git]
git (~git@sh1-ext.rs.github.com) has joined #dig
23:06:59 [git]
[tabulator] presbrey created master (+1148 new commits): http://git.io/Nb47hA
23:06:59 [git]
[tabulator/master] Adding the skeleton tabulator extension structure - jambo
23:06:59 [git]
[tabulator/master] importing 2005 tabulator repository to chrome content - presbrey
23:06:59 [git]
[tabulator/master] fixed for directory restructuring - jambo
23:06:59 [git]
git has left #dig
23:07:01 [bblfish]
but it would need some serious commitment
23:07:04 [bblfish]
from them
23:07:26 [bblfish]
in the end we want to create the social web. not spend our life working out identity systems
23:07:44 [bblfish]
I am going to try to show that we can get WebID to work with OAUth in read-write-web
23:08:00 [melvster1]
nice
23:08:13 [bblfish]
then when a bunch of us have experience with that we can document it
23:08:18 [bblfish]
and make a spec out of it
23:11:41 [lkagal]
lkagal (~lkagal@30-7-251.wireless.csail.mit.edu) has joined #dig
23:11:54 [bblfish]
it's just trying to build form what we have that works together rather than try to mathematically cover all ground
23:12:23 [melvster1]
presbrey: great job on the github
23:12:54 [timbl]
timbl (~timbl@31-33-140.wireless.csail.mit.edu) has joined #dig
23:16:08 [git]
git (~git@sh1-ext.rs.github.com) has joined #dig
23:16:08 [git]
[tabulator] presbrey pushed 6 new commits to master: http://git.io/q6rlCw
23:16:08 [git]
[tabulator/master] Moved mashup here from dig.csail.mit.edu - Tim Berners-Lee
23:16:08 [git]
[tabulator/master] Working on CORS proxy system still - Tim Berners-Lee
23:16:08 [git]
[tabulator/master] Moved fn from rdf/sparql*.js - Tim Berners-Lee
23:16:08 [git]
git has left #dig
23:16:16 [presbrey]
thx melvster1 hopefully better for everyone :)
23:17:40 [presbrey]
timbl, https://github.com/linkeddata/tabulator should have your latest
23:20:27 [presbrey]
(thats really tabulator-firefox)
23:23:19 [presbrey]
timbl, do you prefer to call it rdflib2.js or rdflib.js?
23:28:35 [presbrey]
I put a redirect in at http://dig.csail.mit.edu/hg/tabulator/
23:28:50 [timbl]
rdflib.js
23:29:27 [timbl]
I had a tem rdflib2 which is now i think called rdfa-rdf-lib.js which has rdfa and needs jq
23:31:39 [timbl]
it may get just folde into rdflib
23:32:15 [timbl]
Ok, if we have google karma for rdflib.js then the project should be called justrdflib
23:32:57 [timbl]
do lineddata/rdflib
23:33:02 [timbl]
s/do/so/
23:34:31 [presbrey]
rdflib is python
23:34:36 [presbrey]
we have google karma for rdflib.js