XMPP Council - 2013-06-19

  1. m&m

    T - 10 minutes

  2. stpeter

    sigh, I broke my Prosody installation at stpeter.im last night, I suppose I'll use jabber.org :-)

  3. m&m


  4. m&m

    how did you break it? I find that difficult to do (-:

  5. stpeter


  6. stpeter

    m&m dependency madness in debian

  7. m&m


  8. stpeter

    trying to upgrade to Prosody 0.9rc3

  9. m&m

    ok, we're two minutes late

  10. m&m

    I just pinged MattJ and Ralph

  11. MattJ wonders why he seems m&m as offline

  12. m&m


  13. MattJ wonders why he sees m&m as offline

  14. m&m

    I have no idea

  15. m&m


  16. m&m bangs gavel to start

  17. m&m

    1) Roll call

  18. MattJ


  19. m&m


  20. m&m


  21. Tobias


  22. m&m

    well, that's a quorum

  23. m&m

    2) Proposed XEP: SOX < http://xmpp.org/extensions/inbox/sox.html >   Accept as Experimental?

  24. fippo would love to see this accepted

  25. m&m

    I think it's fine as a starting point

  26. m&m

    no objections from me

  27. stpeter

    it's worth debating, anyway -- although discussion on the rtcweb@ietf.org indicates that even the use of SDP is in flux at in the WebRTC world

  28. fippo

    a starting point for a discussion about jingle and sox, but we can/should discuss that under AOB

  29. Tobias

    i have no objections either

  30. m&m


  31. MattJ

    Grr, sorry, power cut

  32. MattJ

    I'm +1

  33. m&m

    Kev and Ralph have a fortnight to raise objections

  34. m&m

    3) Proposed XEP: Chat Markers < http://xmpp.org/extensions/inbox/chat-markers.html >   Accept (revision) as Experimental?

  35. MattJ


  36. Tobias


  37. m&m

    no objections from me

  38. m&m

    Kev and Ralph again have a fortnight to object

  39. m&m

    4) Date of Next Meeting

  40. Peter Waher


  41. m&m


  42. m&m

    can it wait for AOB?

  43. Peter Waher

    I have made revisions to 2 proto-XEPs

  44. Peter Waher

    and would like you to consider advancing them to Experimental

  45. m&m

    Peter Waher: we can discuss in AOB

  46. Peter Waher

    the color field type, and HTTP over XMPP

  47. m&m

    right after this

  48. Peter Waher

    what is AOB?

  49. m&m

    just confirming some/most plan to make it

  50. m&m

    Any Other Business

  51. Peter Waher


  52. m&m

    I'll take silence as acceptance

  53. m&m

    5) Any other Business

  54. fippo

    sox has raised some questions on how we proceed with jingle...

  55. m&m

    Peter Waher: did you submit your revisions to the XEP editor?

  56. Tobias

    m&m, so same time next week?

  57. m&m

    Tobias: yes

  58. Peter Waher


  59. Peter Waher

    and to the standards list

  60. m&m

    Tobias: treat the "Date of Next Meeting" as a "un-rsvp" (-:

  61. stpeter

    I'm wondering if we need a new XEP editor

  62. m&m cracks whip @ XEP editor

  63. Tobias

    stpeter, or split the work up

  64. m&m

    I didn't add HTTP/XMPP to the agenda as I hadn't seen an update to the officail inbox

  65. m&m


  66. stpeter

    m&m yes that my bad

  67. Lance

    Tobias: or make an editorbot

  68. MattJ


  69. m&m


  70. fippo

    stpeter: your skill is wasted as editor. you need to get a secretary

  71. stpeter


  72. ralphm


  73. m&m

    well, this seems a fitting discussion for council

  74. ralphm

    sorry for being late

  75. MattJ

    My internet is going on/off :/

  76. m&m

    so, stpeter, how would you want the XEP editor role filled?

  77. m&m

    MattJ: ugh

  78. ralphm

    no objections to the protocol-xeps

  79. stpeter

    m&m: probably it makes sense to write a job description and ask for volunteers

  80. m&m

    ralphm: noted

  81. m&m

    stpeter: ok, since you're kind of overloaded now, maybe we can work that out between us (-:

  82. ralphm

    hi Dave

  83. fippo

    i think it's about offloading some of the work... stpeter isn't the only one who can judge things like typo-patches

  84. stpeter

    we do have http://xmpp.org/extensions/xep-README.html

  85. m&m

    Peter Waher: as to your request…once we get those updated XEPs into the inbox, I'm sure we can get them moved forward

  86. stpeter

    so that's a kind of job desciption to some extent

  87. ralphm

    There was talk about pull requests earlier. stpeter: would that make things better?

  88. m&m

    stpeter: that works for me

  89. Peter Waher


  90. Dave Cridland

    stpeter, Any idea which parts of your XEP Editor role are causing the most effort, and which you'd be most uncomfortable delegating?

  91. stpeter

    Dave Cridland: right, I shall give some thought to that and report back to the Council/Board

  92. ralphm


  93. Tobias


  94. m&m

    sounds like a plan

  95. ralphm

    I noted that fippo brought up a point about Jingle

  96. m&m

    I'll note that for next week's agenda, for the real chair (-:

  97. stpeter

    also, it might help to give more people git access (or it might not)

  98. m&m

    fippo: your turn!

  99. ralphm

    stpeter: (or you just merge in commits, pull requests)

  100. fippo

    ok... we've all heard how google (or at least a spokesperson) thinks that audio and video is not well integrated into xmpp.

  101. fippo

    and sox is, to some extend, running in parallel with jingle

  102. stpeter

    ralphm: yes, I already do that, my problem is that I have too much to do, people at Cisco expect me to work on internal projects these days, I have 20 Internet-Drafts to get finished, and I'm even writing some Python code these days related to internationalization ;-)

  103. fippo

    and with all the webrtc stuff happening we ought to make sure that webrtc and jingle work together really great

  104. Dave Cridland

    This seems like a sensible direction, yes.

  105. m&m

    fippo: ok, so what do you want the council to do?

  106. stpeter

    fippo: I totally agree -- although webrtc is rather a moving target right now :-)

  107. fippo

    m&m: until yesterday it was a question of "webrtc uses sdp. jingle doesn't. do we continue to do our own effort"

  108. fippo

    but then people over at rtcweb wanted to reopen that discussion

  109. Dave Cridland

    Right, but there's now an endless thread on whether or not WebRTC should use SDP, or at least expose it at all.

  110. stpeter

    yes, it's quite interesting reading

  111. ralphm


  112. fippo

    especially that one guy working on a large system built around jingle ;-)

  113. Dave Cridland

    Right. :-)

  114. ralphm

    and it really affects what we would like to push for

  115. ralphm

    apparently that large system doesn't (want to) use SDP at all

  116. stpeter

    ralphm: yes!

  117. fippo

    there are some current problems in jingle and sox even addresses some of them

  118. ralphm

    we have been discussing translating to/from SDP to 'pure' jingle

  119. fippo

    and of course there is the translation issue

  120. ralphm

    it seems some people want to ditch SDP and/or offer/answer from the API, leaving it up to signalling protocols entirely

  121. fippo

    (and i've had a really great day. worked around a crasher bug in chrome and had to reintroduce proprietary elements to the mapping)

  122. stpeter

    fippo: ouch

  123. ralphm

    I can see how having browsers generating SDP can be bad for forward compatibility

  124. Tobias

    m&m, is this still official meeting part over?

  125. ralphm

    Tobias: not sure

  126. m&m

    I hadn't banged the gavel yet

  127. ralphm

    I wonder if we should steer this process from our point of view

  128. fippo

    ralphm: and that is what i want the council to do ;-)

  129. ralphm

    'our' being the XMPP community in general and the council in particular

  130. stpeter

    yes, the Council is there to provide technical leadership, eh? ;-)

  131. MattJ nods

  132. ralphm

    stpeter: yeah

  133. ralphm

    (I have been discussing stuff with fippo earlier today)

  134. m&m

    who on the council wants to drive this?

  135. Dave Cridland

    Assign it to Kev. :-)

  136. m&m

    ralphm is tentatively volunteered? (-:

  137. Tobias


  138. stpeter


  139. ralphm

    I am interested in the space, but haven't been involved in Jingle at all

  140. m&m

    Dave Cridland: I had considered it (-:

  141. MattJ

    I know nothing about SDP, WebRTC, or any of stuff... and I fear I don't have the time to spend figuring it out

  142. MattJ

    I think this applies to many in the XMPP community (and council...)

  143. ralphm

    so I'd really need backing from people in our community that actually build stuff or at least know how to balance the issues involved

  144. Dave Cridland

    I can make a suggestion - the Council's well within its rights to delegate this to a small design/leadership team.

  145. fippo

    right... the problem of the xmpp community has always been that people didn't want to touch this media stuff

  146. m&m

    I was about to suggest that, Dave

  147. ralphm

    but I do feel this is a rare change to make a difference

  148. MattJ

    Dave Cridland, i.e. fippo? :)

  149. m&m

    starting with you and fippo!

  150. fippo

    and hopefully emil ivov!

  151. Dave Cridland

    I'm no more qualified than anyone else here.

  152. Tobias

    right...were nearly getting to the point where jingle ft works and is deployed :)

  153. m&m

    Dave Cridland: every group needs a chair

  154. ralphm

    I also note that we have Cisco people in our ranks

  155. ralphm

    that we should draft

  156. Dave Cridland

    Actually that's not a bad idea, I can chair it if you like. I'm effectively neutral on the issues.

  157. fippo

    the council even has a carbons expert and i think we'll need carbons (-:

  158. m&m

    I don't remember how official something like this needs to be

  159. stpeter

    true on carbons

  160. m&m

    fippo: sure

  161. stpeter

    maybe this is a good topic for discussion at the meetup in Berlin?

  162. ralphm

    I'd join in, too

  163. stpeter

    the webrtc situation should be clearer by then

  164. ralphm

    it has been a while since we created a SIG

  165. stpeter

    not that we need to wait

  166. m&m

    I've not been involved in the A/V club, though

  167. stpeter


  168. fippo

    stpeter: no bet on the situation

  169. stpeter

    and Emil will be in Berlin, too

  170. stpeter

    and other folks

  171. fippo

    m&m: this is about signalling

  172. Dave Cridland

    stpeter, In as much as I've followed the WebRTC stuff, I think the earliest it'll resolve is after Berlin. Probably not until after the next IETF.

  173. m&m

    fippo: it's been awhile since I've been involved in signalling (-:

  174. stpeter

    Tobias: about Jingle FT, yeah I agree

  175. m&m

    I can try to find some time to help, but I'm almost to a point of saturation as it is

  176. fippo

    you don't necessarily need experience with the media stuff, shuffling around some payload and deliver it in the right order is the bigger issue

  177. ralphm

    I do believe it is good to start forming opinions before Berlin, though

  178. stpeter

    there are other things Jingle can do beyond a/v

  179. stpeter

    ralphm: agreed

  180. m&m

    stpeter: true

  181. stpeter

    another approach, which we discussed a bit in Brussels, is to have Jingle-SDP

  182. m&m

    ok, give me a moment to try and focus here

  183. ralphm

    actually, one of the first things people want after audio/video is side channels

  184. ralphm

    like 'chat'

  185. stpeter

    ralphm: yes, very much so

  186. stpeter

    and the RTT stuff

  187. fippo

    ralphm: and that is the point where a full protocol stack kicks any of the webrtc samples

  188. ralphm


  189. m&m

    so, we have a SIG

  190. m&m

    Dave Cridland to chair

  191. fippo

    other than the cube slam stuff probably

  192. m&m

    and we have a number of interested parties

  193. fippo

    can we reuse the jingle@ list?

  194. ralphm


  195. m&m

    I was about to ask

  196. ralphm


  197. m&m

    ralphm: proto-SIG (-:

  198. m&m

    exploratory committee?

  199. m&m


  200. stpeter

    ah, XEP-0002 :-)

  201. stpeter

    the early days

  202. MattJ


  203. ralphm

    well, that's where we derive this particular super power from

  204. Tobias

    the...more bureaucratic days....

  205. Lance

    i have to run, but I volunteer for the SIG as well

  206. ralphm


  207. m&m

    focus people! I want to end this!

  208. stpeter

    such enthusiasm, such hope, such naïveté

  209. stpeter


  210. ralphm

    stpeter: thanks

  211. stpeter

    ok, shall we take this to the jingle@ list?

  212. Tobias


  213. Dave Cridland

    Sounds good.

  214. m&m

    let's just call it a committee, and use jingle@ to move forward

  215. ralphm


  216. Tobias

    m&m, +1

  217. stpeter

    m&m: +1

  218. MattJ


  219. m&m

    the official SIG-forming XEP can come sometime later

  220. stpeter


  221. stpeter


  222. ralphm


  223. Dave Cridland

    I don't think you need to make a XEP to make a SIG. We're not the IETF, you know.

  224. m&m

    Dave Cridland: "Anyone (not limited to members of the XMPP Standards Foundation) may propose the formation of a SIG by completing a XMPP Extension Protocol outlining the need for the SIG and its proposed focus."

  225. ralphm


  226. m&m

    so, in some aspects, we are

  227. ralphm

    I propose we form a SIG for WebRTC/Jingle

  228. fippo

    m&m: shall I write one?

  229. ralphm


  230. ralphm

    Guess we voted above

  231. m&m


  232. ralphm

    Yay, SIG in place

  233. stpeter


  234. m&m

    fippo: "the official SIG-forming XEP can come sometime later" (-:

  235. Dave Cridland

    m&m, That says that people may propose a SIG by a XEP, not that a XEP is required.

  236. fippo

    i even had http://hancke.name/jabber/jingle-hill prepared...

  237. m&m

    Dave Cridland: eh, true

  238. Tobias

    any other things?

  239. fippo

    but that assumed i would have to defend the jingle hill against the sox guys :-)

  240. m&m

    ok, so we have a SIG, and we've all voted ok (I suppose Kev gets a fortnight to object)

  241. ralphm

    fippo: I guess you get to do that on the list

  242. Dave Cridland

    m&m, Peter was sloppier in his standardese back then, we can take full advantage. :-)

  243. m&m

    oh, yes … +1 to the SIG

  244. stpeter


  245. ralphm

    +1, obviously

  246. stpeter

    Dave Cridland: that was before I drank the IETF kool-aid

  247. m&m


  248. Dave Cridland

    stpeter, Remember, you're part of the IETF establishment, now.

  249. m&m

    just for the minutes ...

  250. stpeter

    Dave Cridland: yeah, I'm part of the problem for sure!

  251. ralphm

    Guess that's it

  252. ralphm


  253. stpeter

    sheesh, Kev is gone for a week and we have a palace revolt!

  254. m&m

    SIG leaders are Dave Cridland (chair), Philipp Hancke, Lance Stout, and Ralph Meijer?

  255. ralphm

    I'm sure Kev will be delighted what happened in his absence

  256. ralphm

    m&m: yeah

  257. stpeter


  258. m&m


  259. Tobias


  260. m&m

    then I think we're done

  261. m&m bang gavel … 15 minutes late

  262. Tobias

    cya & thanks m&m

  263. m&m

    thanks to all

  264. ralphm


  265. m&m

    now to work on my other chair (-:

  266. m&m

    minutes will be out presently

  267. ralphm

    Talking about chairs, guess who gets to write an e-mail now

  268. stpeter

    m&m: thanks for chairing!

  269. MattJ

    +1, thanks :)

  270. ralphm cheers for Dave

  271. stpeter


  272. Dave Cridland

    ralphm, I think I should probably wait until Kev states his opinion.

  273. m&m

    if I run for the next council, maybe I'll ask to chair that session

  274. m&m


  275. ralphm

    Dave Cridland: that'd be civil, indeed

  276. stpeter

    now what I want to know is, what does this have to do with that atramentous software Dave works on these days?

  277. Dave Cridland


  278. ralphm

    pretty much nothing

  279. stpeter


  280. stpeter

    love that word

  281. stpeter

    it means "inky black"

  282. ralphm is in e-mail these days, too

  283. Dave Cridland

    That's excellent.

  284. Dave Cridland

    I normally refer to it as Cthulu-Mail, but that's possibly better.

  285. ralphm

    I'm not sure if I want my e-mail client to be light-sucking

  286. stpeter

    I'd prefer that a lot of my email disappear on the other side of the event horizon ;-)

  287. fippo

    stpeter: inky will sort your emails by relevane!

  288. fippo


  289. stpeter

    Inky does it all, eh?

  290. stpeter

    might need to give it a try :-)

  291. Dave Cridland

    stpeter, I'm hurt that you haven't already.

  292. stpeter

    Dave Cridland: I seem to be stuck in an email rut with Thunderbird :-)

  293. MattJ

    It's not open-source, is it? :/

  294. Dave Cridland

    MattJ, Nope. It's Cthulu-mail, it'll eat your soul.

  295. MattJ

    I'm afraid it's going to take quite something to get me move to another hosted provider

  296. Dave Cridland

    MattJ, It doesn't host your mail; it's just an IMAP client.

  297. MattJ

    Hmm, ok

  298. MattJ

    But IMAP from The Cloud?

  299. MattJ

    or IMAP from my local machine?

  300. Dave Cridland

    MattJ, IMAP from your machine. Arcode/Inky never gets to see your mail, all the magic processing happens on your laptop.

  301. Dave Cridland

    MattJ, Architecturally, it's all very '90's.

  302. MattJ

    Then what is the "Cloud-Enabled" part about? :)

  303. Tobias

    Dave Cridland, any e2ee (PGP/S/MINE) support coming? :)

  304. Tobias


  305. Dave Cridland

    MattJ, Configuration and things.

  306. MattJ


  307. m&m

    "and things"

  308. stpeter flags http://hancke.name/jabber/jingle-hill for reading after his next conference call

  309. Dave Cridland

    Tobias, We'd like to.

  310. Tobias

    Dave Cridland, so ACAP didn't succeed? :)

  311. m&m hands Dave a compress for that burn

  312. Dave Cridland

    m&m, Yeah, we actually keep your credentials in the cloud, encrypted using a key derived from SRP on your Inky password, so we don't even have the ability to read that one.

  313. Dave Cridland

    And as for ACAP, I've not yet converted Inky. But soon, soon... Muahahahaha!

  314. fippo

    stpeter: i suppose cullen (is he called dr fluffy now?) will really appreciate the request to produce a definition of a webrtc sdp :-)

  315. stpeter

    fluffy has always been his nick, AFAIK

  316. fippo

    stpeter: right, but his bangok slides have a "dr" in front of his realname which he either recently acquired or omitted from other slides.

  317. stpeter

    yes he has a PhD

  318. Dave Cridland

    stpeter, We're considering calling our next release "Atramentous Cephelopod", now.

  319. stpeter

    most excellent!

  320. fippo

    stpeter: you could have told us that stox basically also has an jingle<->sdp mapping issue ;-)

  321. stpeter

    fippo: yeah, I think that item might get dropped from the charter

  322. fippo

    well, to me it does make little difference whether it's done at the ietf or the xsf

  323. stpeter

    fippo: well, as I said on the STOX list just now, there's a lot of flux here and it will be interesting to see how it all ends up -- but our little SIG at the XSF might be part of the story :-)

  324. fippo


  325. stpeter

    I'd like to hear what developers really care about -- e.g., is SDP really going away for end-to-end communication?

  326. fippo

    i'm not sure what the guys over at rtcweb have for use-cases

  327. fippo

    but for anything 1-1 telephony, file transfer o/a is the way to go imo

  328. fippo

    groupchat/conference is harder, but plan b or no plan (my favorite now) provide everything necessary

  329. fippo

    i doubt that anyone can come up with something that is less painful than sdp

  330. stpeter

    yeah, agreed!

  331. fippo

    and heck

  332. fippo

    it only took me three days to do trickle ice with pranswer

  333. MattJ

    I think it took me less time to implement dialback

  334. fippo

    dialback is sane compared to trickle + pranswer ;-)

  335. fippo

    stpeter: is STOX meeting in berlin?

  336. ralphm

    fippo: IETF 87 is in Berlin, yes