IRC log of dig on 2012-01-26

Timestamps are in UTC.

00:11:43 [melvster]
melvster has quit (Ping timeout: 255 seconds)
00:30:58 [cheater]
cheater (~cheater@dslb-084-057-001-153.pools.arcor-ip.net) has joined #dig
01:20:47 [scor]
scor (~scor@drupal.org/user/52142/view) has joined #dig
01:25:39 [Pipian]
Pipian (~pipian@18.111.116.24) has joined #dig
01:36:52 [scor]
scor has quit (Quit: scor)
02:20:49 [scor]
scor (~scor@drupal.org/user/52142/view) has joined #dig
02:22:25 [scor]
scor has quit (Client Quit)
02:26:20 [nunnun]
nunnun is now known as nunnun_away
02:29:07 [nunnun_away]
nunnun_away is now known as nunnun
02:29:37 [nunnun]
nunnun is now known as nunnun_away
02:33:23 [scor]
scor (~scor@drupal.org/user/52142/view) has joined #dig
02:44:45 [scor]
scor has quit (Quit: scor)
03:22:38 [Pipian]
Pipian has quit (Quit: Pipian)
04:42:11 [rszeno]
rszeno has quit (Quit: Leaving.)
06:00:40 [Yudai]
Yudai has quit (Ping timeout: 276 seconds)
06:01:07 [Yudai]
Yudai (~Yudai@p656629.tokynt01.ap.so-net.ne.jp) has joined #dig
06:23:09 [timbl]
timbl (~timbl@178-82-173-100.dynamic.hispeed.ch) has joined #dig
06:24:14 [mhausenblas]
mhausenblas (~mhausenbl@188.141.67.15) has joined #dig
06:33:31 [presbrey]
timbl, you will have much less merging trouble if use git pull --rebase
06:34:27 [presbrey]
git pull --rebase origin master
06:34:56 [presbrey]
history will be cleaner too
06:45:00 [presbrey]
"By rebasing your branch to master, you put all of your changes on top of the master branch. This allows your commits to be grouped (not interwoven by date), reduces the ubiquitous (merge) commit to something that never needs to be explored, pushes all requirement to fix merge conflicts onto the branch, and always results in a clean merge to master"
06:46:18 [presbrey]
http://mettadore.com/analysis/the-ever-deployable-github-workflow/
06:46:25 [presbrey]
http://darwinweb.net/articles/the-case-for-git-rebase
06:48:48 [mhausenblas]
mhausenblas has quit (Quit: mhausenblas)
06:52:44 [presbrey]
http://gitready.com/advanced/2009/02/11/pull-with-rebase.html
06:57:02 [timbl]
So anything I pull under any event should always be git pull --rebase origin master ?
06:58:08 [timbl]
Certianly the merge commits are a pain
07:01:51 [timbl]
timbl has quit (Quit: timbl)
08:19:50 [tlr]
tlr (~tlr@188.188.84.198) has joined #dig
08:48:39 [mhausenblas]
mhausenblas (~mhausenbl@wlan-nat.fwgal01.deri.ie) has joined #dig
09:43:51 [melvster]
melvster (~melvin@p4FF97B51.dip.t-dialin.net) has joined #dig
09:54:31 [tlr]
tlr has quit (Ping timeout: 260 seconds)
10:53:29 [danbri_]
danbri_ has quit (Ping timeout: 245 seconds)
11:45:32 [mhausenblas]
mhausenblas has quit (Quit: mhausenblas)
11:48:07 [mhausenblas]
mhausenblas (~mhausenbl@wlan-nat.fwgal01.deri.ie) has joined #dig
12:25:35 [scor]
scor (~scor@drupal.org/user/52142/view) has joined #dig
12:41:17 [Ralph]
Ralph (Ralph@30-7-118.wireless.csail.mit.edu) has joined #dig
12:41:24 [Ralph]
Ralph is now known as Ralh
12:41:27 [Ralh]
Ralh is now known as RalphS
13:38:07 [manu-db]
manu-db has quit (Ping timeout: 255 seconds)
13:38:33 [manu-db]
manu-db (~msporny@digitalbazaar.com) has joined #dig
13:40:22 [tlr]
tlr (~tlr@188.189.76.122) has joined #dig
13:43:53 [betehess]
betehess has quit (Quit: Leaving)
13:48:02 [kjetilkWork]
kjetilkWork has quit (Remote host closed the connection)
14:03:27 [scor]
scor has quit (Quit: scor)
14:06:18 [betehess]
betehess (~betehess@betehess.w3.org) has joined #dig
14:08:46 [danbri]
danbri (~danbri@cable-146-255-156-245.dynamic.telemach.ba) has joined #dig
14:57:30 [Pipian]
Pipian (~pipian@31-33-105.wireless.csail.mit.edu) has joined #dig
15:04:46 [scor]
scor (~scor@drupal.org/user/52142/view) has joined #dig
15:42:24 [presbrey]
Pipian, did you have anything about cert expiry in your original foaf-ssl doc?
15:42:36 [Pipian]
I do not believe so.
15:52:21 [presbrey]
hey bblfish, you around?
15:52:26 [nunnun_away]
nunnun_away is now known as nunnun
15:52:40 [bblfish]
yes
15:52:46 [presbrey]
I'm a little confused on your stance on x509
15:53:22 [presbrey]
because I know we both agree about extending the trust network via URIs, LD, resolvers, issues and signers
15:53:30 [bblfish]
yes
15:53:32 [presbrey]
cosigners, liberty boxes :), server clients
15:53:34 [presbrey]
all day long
15:54:07 [presbrey]
I guess maybe my problem is with proposing extended development thru/via further x509 'features'
15:54:34 [bblfish]
well i am just saying we should be faithful to what the x509 cert says
15:54:39 [bblfish]
as far as possible
15:54:57 [bblfish]
Mo points out if you want to make an infinitely long lasting certificate it is easy
15:54:58 [presbrey]
oh so I guess I'm really more of a JSON/LD day
15:55:05 [presbrey]
guy*
15:55:28 [bblfish]
well when we have the JSON certificates with time stamps the same will hold
15:55:46 [presbrey]
we can already have that
15:55:55 [bblfish]
not working in browsers
15:55:57 [presbrey]
we have linked data and we have json-ld
15:56:03 [bblfish]
in a relatively secure way
15:56:16 [bblfish]
I am speaking of the certificate
15:56:28 [bblfish]
not the linked data side of things
15:56:33 [presbrey]
anyone can give you the first hint
15:56:40 [bblfish]
?
15:56:40 [presbrey]
all you need is one URI to start your discovery
15:57:00 [presbrey]
we already get one from the SAN, isn't that enough?
15:57:23 [presbrey]
"not working in browsers", what?
15:57:53 [bblfish]
ok, let's start from scratch
15:57:59 [bblfish]
what is the problem?
15:58:01 [presbrey]
...also...
15:58:18 [presbrey]
its a really bad idea to give the idea that when their x509 expires, its null and void, and they can 'throw it away'
15:58:21 [bblfish]
why don't you Skype me, I am a bit tired of typing
15:58:44 [bblfish]
or call me
15:58:51 [presbrey]
if a hacker finds an expired webID, they can reactivateit
15:59:15 [melvster]
i think it's not that hard ... x.509 is a set of claims and LD is a set of claims
15:59:15 [presbrey]
expired *according to the cert
15:59:33 [melvster]
some will be orthogonal, some may be coincident, some may conflict
15:59:43 [melvster]
LD oriented consumers should have a primary accent on LD
15:59:46 [bblfish]
if a hacker finds a self signed expired cert they can reactivate it
15:59:49 [presbrey]
unless you sign it
15:59:58 [bblfish]
*self-signed* is important
15:59:59 [melvster]
X.509 consumers should have a primary accent on X.50
16:00:01 [melvster]
9
16:00:47 [presbrey]
we can protect self-signed certs by moving the not{Before,After} check from the x509 into the LD
16:01:56 [bblfish]
yes, we can also do that. But why bother right now since we can also put it in the cert
16:01:59 [presbrey]
otherwise you have
16:02:17 [presbrey]
2 suggested SPARQLs to implementors
16:02:25 [presbrey]
doing 5 different checks
16:02:27 [presbrey]
building a trust score
16:02:55 [bblfish]
but you were the one suggesting linked signature no?
16:03:00 [presbrey]
1 sparql for signed, 1 other sparql for self-signed
16:03:04 [melvster]
but im surprised the browsers allow an expired certificate to work
16:03:30 [presbrey]
Firefox prompts you, Chrome is silent allow
16:04:22 [melvster]
ahh
16:04:28 [bblfish]
well do you want to suggest that dates be part of the key in the profile>
16:04:34 [bblfish]
?
16:04:45 [presbrey]
melvster already did, and yes, I agree
16:04:56 [bblfish]
in any case having the x509 write a date is not a problem either
16:05:02 [presbrey]
we should play nice as x509 as possible
16:05:10 [presbrey]
eg. use their terms for our schema
16:05:24 [presbrey]
-startdate - notBefore field
16:05:24 [presbrey]
-enddate - notAfter field
16:05:56 [presbrey]
LD will be easier for adoption than extending SSL routines
16:06:03 [presbrey]
ultimately we can't do all the work ourselves :)
16:06:20 [presbrey]
and we want to protect all webIDs equally
16:06:26 [bblfish]
this is going to make the webid test a bit more complex
16:06:36 [bblfish]
and the rdf to be written out more advanced
16:06:37 [presbrey]
not just ones written by browser vendors
16:07:12 [presbrey]
people will think if it like
16:07:20 [presbrey]
a calendar appointment with start and end date
16:07:26 [presbrey]
very common model to implement
16:07:52 [presbrey]
we plan to extend profiles additionally soon anyway
16:08:03 [bblfish]
well it's ok if you want to extend the ontology to allow validity dates on keys
16:08:12 [presbrey]
I think more granular role setup in LD will be needed soon
16:09:14 [bblfish]
ok. well you can propose that if you want
16:09:28 [bblfish]
I am ok, if everyone is willing to implement it
16:09:34 [melvster]
dont forget there's other ways to authenticate that yield a user URI that dont involve X.509
16:10:03 [bblfish]
yes. true
16:10:17 [presbrey]
yes and anything will be build in LD will auto-apply to all containers/transports :)
16:10:23 [bblfish]
still se should still be true to the cert
16:10:25 [presbrey]
future proof
16:10:30 [bblfish]
but this would add a double check
16:10:59 [melvster]
maybe there should be things like expiredKey and revokedKey ... henry suggested oldKey
16:11:21 [melvster]
im told revocation is considered a hard problem
16:11:43 [bblfish]
it is for services that don't have webid melvster
16:12:01 [bblfish]
because they either have to go to the big CA in the cloud and ask them for a list of all revoked certs
16:12:11 [bblfish]
which can become soon a denial of service attack
16:12:41 [presbrey]
:oldKey, :notBefore, :notAfter, all sound good to me
16:13:01 [bblfish]
well old key means you have to change the cert
16:13:07 [bblfish]
I mean the
16:13:10 [presbrey]
eg. specifing when an :oldKey "became old"
16:13:13 [bblfish]
profile
16:13:20 [presbrey]
_when_
16:13:41 [presbrey]
old is ambiguous but makes checking for activeKey easy :)
16:13:44 [bblfish]
no I suppose one has to say the following: a key is valid unless oe has info that it is expired
16:14:04 [presbrey]
when you move it to oldKey, just mark it with the date
16:14:25 [bblfish]
yes. But what presbrey wants is for someone when they make the key to be able to publish the key expiration date
16:14:30 [bblfish]
too
16:14:37 [bblfish]
that would be equivalent to having the info in the cert
16:14:44 [presbrey]
have both
16:15:07 [bblfish]
yes, so you see things soon get more complicated. that's why I'll leave it to you to propose something
16:15:55 [presbrey]
arg but I just like to code...
16:17:40 [bblfish]
me too
16:17:48 [bblfish]
that's why WebId is so simple
16:17:54 [bblfish]
;-)
16:18:17 [bblfish]
because every silly decision one has to make people can argue on it for ever
16:18:31 [bblfish]
which is why Faebook is so big and our networks are so smal
16:19:06 [melvster]
rfc 4158 says 'A valid certification path cannot be built if T falls outside of the certificate validity period' whatever that means, i guess the rfc should be respected where it's mandatory
16:21:27 [presbrey]
but there is no path with self-signed?
16:22:22 [presbrey]
unless the null path is still a path?
16:22:29 [rszeno]
rszeno (~rszeno@79.114.79.130) has joined #dig
16:25:39 [melvster]
hmm not too sure ... it's quite a big spec
16:26:33 [bblfish]
it's probably the null path
16:27:19 [bblfish]
it kind of makes for a good reason to have a non-self signed cert
16:27:54 [melvster]
FOSS guys like self signed
16:28:26 [presbrey]
yes we do
16:28:32 [melvster]
lol :D
16:28:39 [presbrey]
;)
16:29:46 [presbrey]
do you know if www.cacert.org allows you to specify a SAN?
16:31:03 [melvster]
presbrey: btw slightly different topic, I'm now using commands like : for (( ;; )) ; do inotifywait test.html ; b ; curl -T test.html http://util.data.fm/ ; done
16:31:29 [melvster]
b is my own little script that monitors my productivity
16:31:47 [melvster]
but what it does is automatically deploy a file as soon as i change it
16:32:19 [melvster]
you could imagine using the same technique to send messages etc.
16:32:32 [bblfish]
ok, well the server cert is self signed, so that's ok. :-)
16:33:22 [bblfish]
still the idea of adding validity dates to the key is not a bad one
16:34:58 [bblfish]
but also it is really the certificate that has validity dates
16:35:03 [bblfish]
mhh
16:35:28 [melvster]
i think in this way you could possibly get some real time LD working
16:35:43 [bblfish]
what has validity dates could be the WebID-->key relation. Ie the claim that the owner will be the only one knowing that relation for such and such a time
16:35:48 [melvster]
similar to your wait idea
16:35:56 [bblfish]
wait?
16:36:07 [melvster]
on data.fm
16:36:18 [bblfish]
i forgot
16:36:28 [melvster]
is a way to do semi-realtime LD
16:36:29 [presbrey]
inotify is awesome
16:36:44 [melvster]
you use it already?
16:37:07 [presbrey]
yes hook it to templates/code to auto-recopmile
16:37:09 [melvster]
i think it could work really well desktop to desktop with data.fm + some scripting
16:37:13 [melvster]
ahh ok
16:37:16 [presbrey]
there is a ?wait already, its sorta like a fake basic inotify
16:37:27 [melvster]
yes i know
16:37:42 [presbrey]
ah cool, r u thinking of using it?
16:37:49 [melvster]
soon yes! :)
16:38:11 [melvster]
realtime stuff makes impressive demos
16:38:15 [melvster]
(and tools)
16:38:17 [presbrey]
yes great!
16:38:55 [presbrey]
I'm sure you will let me know when you find probs
16:39:16 [presbrey]
haven't tested it much yet
16:39:21 [melvster]
more a case of putting the pieces together
16:39:49 [melvster]
we have awesome infrastructure now
16:40:34 [tlr]
tlr has quit (Ping timeout: 245 seconds)
16:41:42 [melvster]
presbrey: you use inotify on a mac?
16:49:15 [presbrey]
no on linux
16:51:55 [melvster]
oh cool
17:00:24 [mattl]
mattl has quit (Remote host closed the connection)
17:27:01 [timbl]
timbl (~timbl@178-82-173-100.dynamic.hispeed.ch) has joined #dig
17:33:58 [presbrey]
the python-inotify module is pretty efficient
17:53:55 [RalphS]
RalphS has quit (Ping timeout: 252 seconds)
17:54:31 [Ralh]
Ralh (Ralh@30-7-118.wireless.csail.mit.edu) has joined #dig
18:06:12 [mhausenblas_]
mhausenblas_ (~mhausenbl@wlan-nat.fwgal01.deri.ie) has joined #dig
18:07:46 [mhausenblas]
mhausenblas has quit (Read error: Connection reset by peer)
18:07:47 [mhausenblas_]
mhausenblas_ is now known as mhausenblas
18:20:48 [mattl]
mattl (mattl@fsf/staff/mattl) has joined #dig
18:22:30 [bblfish]
here's one for the group https://www.destroyallsoftware.com/talks/wat
18:22:48 [bblfish]
fun
18:40:34 [amy]
amy has quit (Read error: Connection reset by peer)
18:42:13 [amy]
amy (~amy@30-6-207.wireless.csail.mit.edu) has joined #dig
19:37:32 [mhausenblas]
mhausenblas has quit (Quit: mhausenblas)
19:42:54 [mhausenblas]
mhausenblas (~mhausenbl@wlan-nat.fwgal01.deri.ie) has joined #dig
19:45:10 [mhausenblas]
mhausenblas has quit (Client Quit)
20:30:18 [presbrey]
melvster, are you typically using turtle or json with data.fm?
20:30:45 [melvster]
presbrey: I used to use turtle, but im switching now to json-ld
20:32:15 [presbrey]
what do you do for storage / querying?
20:32:50 [presbrey]
wasn't aware of the json parser for tabulator ;)
20:33:13 [melvster]
storage on json-ld currently now just a POST / PUT
20:33:21 [melvster]
and i send in the json
20:33:30 [melvster]
GET to get it back
20:33:39 [presbrey]
cool so no libs or wrappers
20:33:42 [melvster]
Content-Type: application/json
20:33:45 [presbrey]
I'll just do that too
20:34:33 [melvster]
tho i guess it may become application/ld+json ; form=compacted
20:34:43 [melvster]
or whatever is in the latest spec
20:34:46 [presbrey]
jquery can send Accept header easily
20:35:00 [melvster]
sure tho i use native XHR
20:35:16 [presbrey]
I wonder how much a json parser for tabulator would speed it up
20:35:21 [melvster]
xhr = new XMLHttpRequest();
20:35:21 [melvster]
xhr.open('DELETE', file, false);
20:35:21 [melvster]
xhr.setRequestHeader('Content-Type', 'appliction/json; charset=UTF-8');
20:35:21 [melvster]
xhr.send(body);
20:35:24 [melvster]
etc
20:35:32 [presbrey]
since they have them built-in in C now JSON.parse/stringify
20:35:43 [presbrey]
err C++ :)
20:35:45 [melvster]
im finding it pretty fast already
20:35:55 [melvster]
the calendar now has colors!
20:36:28 [melvster]
http://calendar.data.fm/
20:36:33 [melvster]
still a bit buggy
20:36:36 [mattl]
colors? on my internet?
20:36:38 [presbrey]
great
20:36:52 [melvster]
so the nice thing is you can do things like the following:
20:36:57 [melvster]
appointments in blue
20:37:03 [melvster]
little todo items in red
20:37:05 [presbrey]
hm the ?next= login points to appbuilder
20:37:12 [presbrey]
for webid
20:37:25 [melvster]
then clear out all the red items periodically with a script etc.
20:37:56 [melvster]
re: appbuilder sorry, fixed
20:39:10 [melvster]
i use the http://www.w3.org/ns/ui#color property
20:40:00 [melvster]
yes ive been having really positive experiences with json ld so far
20:41:54 [melvster]
ill do some calendar sharing too
20:46:41 [melvster]
hmmm i guess i should do delete event
21:04:16 [presbrey]
I want to make a CORS schema too
21:04:26 [presbrey]
for at least Origin, Methods
21:04:51 [melvster]
there's always vocab.data.fm until you put them on purl.org :)
21:04:53 [presbrey]
so we can control that for URIs in ACL from .meta
21:05:15 [presbrey]
no its going to w3
21:05:28 [melvster]
oh cool
21:06:06 [presbrey]
I started there eg. http://ns.data.fm/posix/stat
21:06:32 [presbrey]
but I hope the w3.org will long outlast data.fm :)
21:06:42 [presbrey]
shorter URL too!
21:08:26 [presbrey]
where can we put CORS schema? /ns/http/cors ?
21:09:09 [presbrey]
http://data.fm/proxy?uri=http://www.w3.org/2006/http
21:09:20 [presbrey]
some good stuff there for setting up redirects
21:11:01 [presbrey]
http://data.fm/proxy?uri=http://www.w3.org/2011/http newer
21:11:52 [presbrey]
looking quite polished already
21:14:30 [melvster]
yes
21:15:57 [presbrey]
what apps are you looking at for real-time?
21:18:54 [presbrey]
desktop messaging?
21:19:55 [presbrey]
this ui might be free: http://roundcube.net/screens
21:21:30 [melvster]
presbrey: havent really anything concrete in mind just would be a nice thing to have
21:21:38 [melvster]
maybe some gaming stuff
21:21:42 [melvster]
maybe messaging / chat
21:21:45 [melvster]
maybe markets
21:22:03 [melvster]
or just tabulating things together
21:23:39 [presbrey]
whats the windowing on your calendar?
21:23:58 [presbrey]
just jquery ui?
21:28:10 [melvster]
presbrey: you mean the dialog modal ... thats jquery ui
21:28:25 [kennyluck]
kennyluck has quit (Ping timeout: 252 seconds)
21:28:28 [melvster]
quick and easy
21:28:31 [melvster]
just .dialog()
21:29:29 [melvster]
Huffington Post: Why 2012 will be 'The Year of the Semantic Web'
21:29:30 [melvster]
lol
21:30:51 [kennyluck]
kennyluck (~kennyluck@114-43-125-60.dynamic.hinet.net) has joined #dig
22:08:20 [Pipian]
Pipian has quit (Ping timeout: 260 seconds)
22:09:41 [Pipian]
Pipian (~pipian@30-9-3.wireless.csail.mit.edu) has joined #dig
22:38:46 [tlr]
tlr (~tlr@91.183.112.171) has joined #dig
22:50:06 [Ralh]
Ralh has quit ()
22:55:49 [melvster]
it's tempting to do a social net clone on data.fm
23:13:53 [scor]
scor has quit (Quit: scor)
23:29:46 [tlr]
tlr has quit (Quit: tlr)
23:42:38 [melvster]
melvster has quit (Ping timeout: 252 seconds)
23:47:01 [timbl]
timbl has quit (Quit: timbl)
23:58:49 [scor]
scor (~scor@drupal.org/user/52142/view) has joined #dig
23:59:30 [bblfish]
bblfish has quit (Quit: bblfish)