Wednesday, May 08, 2013
xsf@muc.xmpp.org
May
Mon Tue Wed Thu Fri Sat Sun
    1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
   
             
XSF Discussion | Logs: http://logs.xmpp.org/xsf/ | Agenda https://trello.com/b/Dn6IQOu0/board-meetings

[00:22:53] *** Jef shows as "away" and his status message is " (Away as a result of being idle more than 5 min)"
[00:23:17] *** Jef shows as "online"
[00:28:40] *** Lance has joined the room
[00:28:40] *** Lance shows as "online"
[01:03:58] *** stpeter has left the room
[01:37:36] *** Lance shows as "away"
[01:56:30] *** Lance shows as "online"
[02:28:39] *** Jef has left the room
[04:08:41] *** kevin has joined the room
[05:06:04] *** Lance has joined the room
[05:06:05] *** Lance shows as "online"
[05:16:14] *** jabberjocke has left the room
[05:40:13] *** Lance has left the room
[05:51:01] *** jabberjocke has joined the room
[05:51:01] *** jabberjocke shows as "online"
[05:52:27] *** Alex has joined the room
[05:52:58] *** Alex shows as "online"
[05:53:31] *** Alex has left the room
[06:26:59] *** Kev shows as "online"
[06:32:45] *** jabberjocke has left the room
[06:43:23] *** Alex has joined the room
[06:52:40] *** Neustradamus has joined the room
[07:30:39] *** Lloyd has joined the room
[08:58:30] *** kevin has left the room
[09:03:11] *** kevin has joined the room
[09:32:24] *** jabberjocke has joined the room
[09:32:24] *** jabberjocke shows as "online"
[09:39:51] *** Alex shows as "away" and his status message is "Auto-Status (untätig)"
[09:39:54] *** Alex shows as "online"
[10:20:23] *** Alex shows as "away" and his status message is "Auto-Status (untätig)"
[10:40:23] *** Alex shows as "xa" and his status message is "Auto-Status (untätig)"
[10:41:49] *** Alex shows as "online"
[10:53:45] *** Lloyd shows as "online"
[11:30:19] *** Neustradamus shows as "away"
[11:56:28] *** Lloyd shows as "online"
[12:15:07] *** Alex shows as "away" and his status message is "Auto-Status (untätig)"
[12:15:07] *** Alex shows as "online"
[13:10:12] *** jabberjocke has left the room
[13:43:16] *** kevin has left the room
[13:55:13] *** Kev shows as "away"
[14:01:51] *** Kev shows as "online"
[14:14:25] *** Kev shows as "away"
[14:53:58] *** kevin has joined the room
[14:57:27] *** stpeter has joined the room
[15:01:49] *** Kev shows as "online"
[15:05:40] *** Alex has left the room
[15:06:47] *** Florob has joined the room
[15:49:08] *** jabberjocke has joined the room
[15:49:08] *** jabberjocke shows as "online"
[15:57:01] *** Florob has left the room
[16:17:13] *** Lance has joined the room
[16:17:13] *** Lance shows as "online"
[16:17:27] <stpeter> hi Lance!
[16:17:34] <stpeter> any further thoughts on our SDP problem?
[16:18:19] <Lance> just that I wish there were easily available SDP parsers
[16:18:30] <stpeter> in what languages?
[16:18:37] <Lance> JS and Python
[16:18:57] <stpeter> I'm kind of surprised we don't have one in JS given all the WebRTC work happening
[16:19:15] <Lance> I've seen some partial works for that
[16:19:33] <Lance> I think Phono might have a good implementation, not sure on licensing though
[16:19:56] <Lance> but the SOX idea is looking reasonable
[16:20:12] <stpeter> Lance: ugly but practical, yes
[16:20:28] <stpeter> Lance: I didn't write it up yet, but I think I will before the end of this week
[16:20:36] <Lance> right, just one more thing in the disco stack to check
[16:20:40] <stpeter> :)
[16:20:45] <Lance> support sox? use that. support jingle? use that
[16:20:52] <stpeter> right
[16:21:11] <stpeter> anyway I will run it up the flagpole and see who salutes :-)
[16:25:14] <Kev> Am I being dense that I don't immediately know what sox is?
[16:25:24] <Lance> sip/sdp over xmpp
[16:26:37] <Kev> I've completely missed this.
[16:26:44] <Kev> Wherewhowhywhatnow?
[16:26:46] <Lance> it was mentioned at the summit
[16:26:55] <Lance> the main issue is with the direction that webrtc has gone
[16:27:01] <stpeter> Kev: Lance and I had a chat about it here the other day
[16:27:04] <Lance> which has standardized on using SDP
[16:27:06] *stpeter fishes out the logs
[16:27:07] <Kev> I must have not been in that track.
[16:27:14] <Kev> Unless you mean cusax (sp?)
[16:27:34] <stpeter> http://logs.xmpp.org/xsf/130506/#20:27:18
[16:27:40] <stpeter> Kev: not CUSAX
[16:27:47] <Lance> so browser apps right now just pass SDP blobs around to set up video/audio sessions
[16:27:54] <stpeter> CUSAX is Combined Use of SIP and XMPP = dual-stack clients
[16:28:13] <Lance> and browsers dont provide a way to parse or get the raw data in the SDP blob
[16:28:34] <stpeter> SoX is for XMPP-only endpoints to send plain SDP
[16:28:54] <Lance> so just lots of error prone parsing/translating back and forth to jingle
[16:30:11] <stpeter> wow, I haven't logged into stpeter@jabber.org for a while -- I have 100+ buddy requests
[16:34:05] <Kev> This sox stuff doesn't sound very appealing!
[16:34:58] <Lance> Kev: yes, very ugly and blobby
[16:35:45] <Lance> but also simpler for certain useful cases
[16:35:49] <Kev> I hope we're putting it inside the message body!
[16:36:01] <Lance> and JSON encoding it, for good measure
[16:39:40] <stpeter> :)
[16:40:32] <stpeter> nice! ".stpeter@jabber.org would like to subscribe to your presence"
[16:47:31] <stpeter> Kev: in fact, SoX is just about the ugliest thing you've ever seen, but it works -- it's really http://xmpp.org/extensions/inbox/zoep.html revisited
[16:52:44] *** Kev shows as "away"
[17:03:27] *** Kev shows as "online"
[17:05:56] <Kev> So you mean I might expect to see things as beautiful as http://xmpp.org/extensions/inbox/zoep.html#examples-call?
[17:06:08] <Kev> Because that's what I've always thought has been missing from life.
[17:06:13] *** stpeter shows as "dnd" and his status message is "in a WebEx meeting"
[17:06:49] <Kev> That example's also illegal, isn't it?
[17:07:50] <Lance> yep, with the < in the headers
[17:08:00] <stpeter> I haven't looked at the ZOEP thing in a long time
[17:08:09] <Kev> From what I can tell, you're missing out.
[17:18:37] *** Lloyd shows as "online"
[17:21:37] <stpeter> :P
[17:23:09] *** Lloyd shows as "online"
[17:24:13] *** Lloyd has left the room
[18:11:32] *** Alex has joined the room
[18:34:37] *** Kev shows as "away"
[18:42:36] *** Kev shows as "online"
[18:59:03] *** Kev shows as "away"
[19:04:07] *** stpeter shows as "online"
[19:27:01] *** kevin has left the room
[19:27:01] *** kevin has joined the room
[19:53:46] *** Lance has joined the room
[19:53:51] *** Lance shows as "online"
[20:17:34] *** Lance has left the room
[20:28:55] *** Alex shows as "away" and his status message is "Auto-Status (untätig)"
[20:38:24] *** kevin shows as "away" and his status message is "(Away as a result of being idle more than 5 min)"
[20:48:54] *** Alex shows as "xa" and his status message is "Auto-Status (untätig)"
[21:02:13] *** Kev shows as "online"
[21:04:44] *** Lance has joined the room
[21:04:46] *** Lance shows as "away"
[21:05:15] *** Lance shows as "online"
[21:11:54] *** stpeter shows as "away" and his status message is "wandered off..."
[21:23:57] *** stpeter shows as "online"
[21:28:15] <Lance> stpeter: I also started on a xep for letting a server prep jids a client sends it: http://legastero.github.io/customxeps/extensions/jidprep.html
[21:38:39] <stpeter> Lance: ah, good, I've been thinking a lot in the last 24 hours about some Python code for internationalization, but only to output the codepoint tables for IDNA2008 and PRECIS
[21:39:48] <stpeter> of course, Matthew Wild would say I should do it in Lua ;-)
[21:39:49] *** MattJ has joined the room
[21:39:57] <stpeter> speaking of whom! ;-)
[21:40:04] <MattJ> Hello!
[21:40:09] <stpeter> but multiple implementations are a good thing
[21:40:23] <stpeter> MattJ: the topic is a server-side service that preps JIDs for you
[21:40:30] *MattJ bookmarks the room, because one can never have enough bookmarks
[21:40:40] <stpeter> heehee
[21:40:47] <MattJ> Want one? I can code it in about 5 minutes :)
[21:41:00] <stpeter> MattJ: for some reason, Lance and I and a few others seem to use this room for technical discussions at times
[21:41:18] <stpeter> MattJ: you can code most everything in 5 minutes :P
[21:41:20] <MattJ> Trying to move jdev out of prosody@?
[21:41:33] <stpeter> jdev is dead, long live jdev!
[21:41:47] <MattJ> Heh
[21:41:52] <stpeter> besides, muc.xmpp.org isn't hammered like conference.jabber.org
[21:42:12] <stpeter> and I'd assume that XSF IPR rules apply here, for whatever that is worth
[21:42:56] <MattJ> Rules, pft
[21:43:04] <stpeter> that's the spirit!
[21:46:17] <MattJ> Do you also want the server to split it perhaps? or only prepping?
[21:46:53] <stpeter> prepping seems like enough
[21:47:07] <stpeter> the client could send whole JIDs or only JID-parts
[21:47:57] <stpeter> as Jack pointed out in the related email thread a while back, that doesn't help you until you're connected to the XMPP server, so one might want to offer an HTTP service as well, but IMHO the XMPP service would be a fine start
[21:48:34] *** Kev shows as "away"
[21:55:03] <Lance> yeah, an HTTP service would be nice, but at that point, a whole heck of a lot of XMPP data would be nice to have exposed over standardized REST calls
[21:55:16] <stpeter> heh
[21:56:06] <stpeter> it's an open question how much really needs to go over XMPP -- many IQ things could happen over HTTP (oh, say, vCard get)
[21:57:48] <MattJ> HTTP? You're moving the goalposts :P
[21:57:59] *MattJ adds, because it's only another couple of lines...
[21:58:02] *stpeter glances at PEP-0008
[21:59:02] <MattJ> Lance, GET or POST?
[21:59:15] <Lance> MattJ: I'd say just prepping the JID as given is enough. Clients can split them pretty easily
[21:59:35] <Lance> GET should work
[21:59:37] <Lance> cacheable
[22:00:30] <Lance> stpeter: If you figure out getting precis to work in Python, i'll bundle it in to sleek
[22:00:41] <stpeter> Lance: it's a deal!
[22:01:07] <stpeter> Lance: but step one is to at least figure out how to output the tables I need -- that might lead to a building block we can use
[22:01:53] <MattJ> Lance, http://q.zash.se/d1302228.txt
[22:02:18] <MattJ> GET localhost:5280/prep?jid=USER@HOST
[22:02:21] <Lance> \o/
[22:02:27] <MattJ> (prep assumes you named it mod_prep)
[22:02:39] <MattJ> (/prep assumes you named it mod_prep)
[22:02:59] <MattJ> Message correction, I'm growing to like it :)
[22:03:04] <stpeter> heh
[22:03:18] <Lance> it is handy
[22:06:07] <Lance> ok, so prosody supports this now, and that's all that counts. so time to tidy up this protoxep and send it to the editor
[22:06:27] <stpeter> yay!
[22:06:30] <MattJ> Ha, I just saw your XEP in the backlog
[22:06:35] <MattJ> I can make the plugin support that protocol
[22:07:02] <Lance> Yeah, it's pretty much identical already, just a namespace change
[22:07:20] <Lance> But specifying jid-malformed for the error response is a good idea
[22:07:24] <MattJ> I chose not to nest the <jid>, just makes the code simpler, and I didn't see a need for it
[22:07:54] <Lance> yeah, makes sense
[22:08:42] <MattJ> I'll commit this to prosody-modules
[22:08:51] <Lance> i think i was modelling how bind works, but that's not needed if we're not splitting
[22:08:54] <MattJ> along with a couple of other plugins I've been working on...
[22:10:00] <stpeter> sheesh, my inbox is back at 666 messages -- why don't I seem able to shake that?
[22:10:08] <stpeter> I'll need to delete a whole lot of messages, I think ;-)
[22:10:15] <MattJ> Good luck :)
[22:10:31] <stpeter> I got it down from ~3000 to this number and just got stuck
[22:10:54] *** Jef has joined the room
[22:13:15] <MattJ> Lance, I'll change the namespace to urn:xmpp:jidprep:tmp
[22:14:10] <MattJ> and wondering if perhaps the JID could just go in the path... GET /prep/user@host
[22:15:21] <Lance> that would work pretty nicely
[22:16:40] <MattJ> Makes the code smaller too :)
[22:17:16] <MattJ> and it now returns HTTP 400 on an invalid JID, unless you have a better code
[22:17:45] <Lance> i can't think of one off the top of my head
[22:17:56] <Lance> except for 418
[22:18:42] <MattJ> 422 Unprocessable Entity (WebDAV; RFC 4918)
[22:19:06] <MattJ> I do like 451 :)
[22:19:28] <Lance> :)
[22:19:35] <stpeter> heh
[22:19:43] <stpeter> I don't think we ever got that one registered
[22:20:01] *MattJ is a Bradbury fan
[22:20:35] <Lance> ok, new version pushed. should match implementation
[22:21:35] <Lance> bah, forgot schema change
[22:21:36] <MattJ> Looks perfect
[22:21:44] <MattJ> Bah, schemas :)
[22:21:58] *stpeter shakes his fist at the XML gods!
[22:22:10] <Lance> its ok they're not normative anyway :p
[22:22:21] <stpeter> the gods or the schemas? ;-)
[22:22:34] <Lance> haha
[22:22:46] <MattJ> :P
[22:23:30] <Lance> ok, now that's updated
[22:23:38] <Lance> do I just email the xml file to you stpeter?
[22:24:00] <stpeter> Lance: I can pull it from a URL or whatever
[22:24:20] *stpeter finds http://legastero.github.io/customxeps/extensions/jidprep.xml
[22:24:25] <MattJ> Formalities? This isn't the IETF :)
[22:24:40] <stpeter> yeah thankfully :P
[22:25:33] <stpeter> I'm going to experiment with literate programming in my i18n program, we'll see how it goes :-)
[22:26:06] <Lance> the only problem I ever had with literate programming was that none of the systems worked like I wanted it to
[22:26:20] <Lance> i think its a rite of passage, making your own literate programming framework
[22:26:23] <stpeter> who needs a system?
[22:26:26] <stpeter> heh
[22:31:37] <stpeter> and it's in the inbox
[22:31:42] <Lance> \o/
[22:32:21] <Lance> and now I don't have to feel bad about not bundling Mbs of character data to stringprep in the browser
[22:32:27] <stpeter> hehe
[22:32:28] <MattJ> :)
[22:32:31] <stpeter> indeed
[22:32:38] <stpeter> make the server do all that work
[22:33:18] <MattJ> Long live simple clients
[22:33:27] <stpeter> :)
[22:34:19] <stpeter> that brings me back to Jingle and SDP and SoX -- what I had always hoped for was a way for the client to tell the server "hey I want to call Alice" and the server would work out all the details -- the whole offer/answer model is so complex...
[22:38:09] <Lance> yeah, that would be nice
[22:38:21] <stpeter> probably just a dream :-)
[22:40:01] <Lance> maybe, but sox is pretty close
[22:40:08] <stpeter> really? I don't think so
[22:40:23] <stpeter> SoX is just an ugly workaround for offer-answer madness
[22:40:28] <Lance> especially since most clients just use a web view anyway which will eventually have webrtc support
[22:40:30] <stpeter> but it's within that same paradigm
[22:40:37] <stpeter> hmm, maybe
[22:40:38] <Lance> just grap the sdp blob, send, be happy
[22:40:47] <stpeter> you're more of an optimist than I am :P
[22:41:22] <Lance> probably
[22:41:52] <Lance> but its looking like webrtc will hopefully solve a lot of these problems and pave the road
[22:42:01] <stpeter> nod
[22:42:15] <stpeter> I still worry about silos
[22:42:18] <stpeter> but I suppose I always will
[22:42:43] <Lance> i dont think sox would be any more siloable than things already are
[22:42:51] <stpeter> oh, no, I didn't mean that
[22:42:52] <Lance> its the exact same data in a jingle stanza
[22:42:57] <stpeter> yepper
[22:50:18] <stpeter> bbl
[22:50:22] *** stpeter has left the room
[22:53:48] *** Alex shows as "online"
[22:59:28] *** kevin has left the room
[23:00:06] *** kevin has joined the room
[23:04:25] *** Alex shows as "away" and his status message is "Auto-Status (untätig)"
[23:07:27] *** Alex shows as "online"
[23:07:27] *** Alex has left the room
[23:08:15] *** jabberjocke has left the room
[23:16:54] *** Neustradamus has left the room
[23:19:08] *** Neustradamus has joined the room
[23:56:09] *** Lance has left the room
[23:56:15] *** Lance has joined the room
[23:56:15] *** Lance shows as "online"
[23:57:08] *** Lance has left the room
[23:57:15] *** Lance has joined the room
[23:57:15] *** Lance shows as "online"