XMPP Council - 2013-07-24


  1. MattJ

    Meeting today? Do we have an agenda?

  2. Kev

    Yes, meeting today. I even have something for the agenda.

  3. Kev

    Which is that Fippo thinks the deferred 'you don't want to do this' text is too strong, so we should discuss.

  4. MattJ

    +1

  5. MattJ

    What's next? :D

  6. Kev

    I disagree, incidentally :)

  7. Kev

    But I do have a counter-proposal.

  8. Kev

    That's all I have for the Council meeting. I don't think there are any protoXEPs waiting.

  9. MattJ

    None that I know of

  10. Tobias

    hopefully i can produce a bit more work for the council over the next weeks :)

  11. Kev

    "thanks"

  12. Tobias

    m&m, possible use of posh would require DNA support in the server, right?

  13. m&m

    it would … although DNA is a framework, not an actual protocol

  14. m&m

    at least for the connecting/verifying server

  15. m&m

    (or client)

  16. fippo

    tobias: well, that depends on the verifying server

  17. fippo

    it might be just another way to do dial-back

  18. Tobias

    fippo, just wondering what, most likely, will be the standard way of doing it so i can implement it as a prosody module :)

  19. fippo

    tobias: just try everything before actually doing <db:verify/> :-)

  20. Tobias

    that smells like brute force :)

  21. m&m

    there's not really a better way to go right now, though

  22. m&m

    although, there are only 3 options

  23. m&m

    1) just check the PKIX (-:

  24. m&m

    2) DANE/DNSSEC

  25. m&m

    3) POSH

  26. fippo

    4) cridlands samecert

  27. fippo

    9999) <db:verify/>

  28. m&m

    db:verify is not a prooftype

  29. ralphm

    Unfortunately, I have to pick up my car from repair right about now. I cannot attend the Council meeting in time.

  30. m&m

    arrgh!

  31. ralphm

    I sent a message to the list regarding our notices above Deferred XEPs.

  32. ralphm

    I should be back in time for the Jingle SIG meeting.

  33. Kev

    ralphm: Thanks.

  34. Kev

    m&m: Why is db:verify not a prooftype?

  35. Kev

    I should rather say. I think db:verify is reasonable as a prooftype, as we'd be better off getting adoption by rolling what's actually used in, rather than trying to bootstrap based on things we've yet to manage to achieve.

  36. m&m

    hold on

  37. m&m

    the problem with treating db:verify as a prooftype (as defined in draft-ietf-xmpp-dna) is that you are bootstrapping from untrustworthy sources (DNS)

  38. m&m

    DNSSEC makes those sources more trustworthy, but at that point you've already confirmed the identity

  39. fippo

    m&m: sure. but this offers an upgrade path.

  40. fippo

    and fwiw, I think it can be totally eliminated thanks to samecert

  41. fippo

    (which isn't a prooftype either)

  42. stpeter

    howdy

  43. stpeter

    brb

  44. MattJ

    There was something I didn't like about samecert

  45. MattJ

    but I can't remember what

  46. fippo

    mattj: it's about the same level of proof as <db:verify/>

  47. fippo

    just less roundtrips

  48. fippo

    it still has the problem that it works and therefore hides problems instead of breaking

  49. stpeter

    what is samecert?

  50. MattJ

    stpeter, instead of looking up DNS, check that the certs on the originating and the authoritative are the same

  51. Kev

    I assume, from the name, was the incoming stream negotiated with the same cert I get if I connect to that server.

  52. fippo

    http://jabber.soup.io/post/88601075/Dave-Cridland-Dialback-Now-without-dialback

  53. fippo

    the paragraph that starts with "My first suggestion"

  54. stpeter checks in for his flight to Berlin tomorrow

  55. fippo

    re berlin i might have gotten some of the free software foundation europe people interested in joining the hackfest

  56. stpeter

    fippo: cool

  57. Kev

    Right. It is time, it is time.

  58. Kev

    1) Roll call.

  59. Tobias

    hereo

  60. Kev

    I am here. Ralph is not, as he is seeing to a sick car, and sent apologies just up there ^

  61. m&m

    presente

  62. MattJ

    Here

  63. Kev

    Righty.

  64. Kev

    2) Fippo thought that the deferred text was confusing to people who don't know better, and maybe it shouldn't say that people shouldn't implement.

  65. Kev

    Ralph posted the following suggested text to the list:

  66. Kev

    The above is just one example of confusion about our standards process I've encountered recently, specifically the 'Deferred' state. The current notice reads: WARNING: Consideration of this document has been *Deferred* by the XMPP Standards Foundation. Implementation of the protocol described herein is not recommended. I propose the following instead: WARNING: This document has been automatically *Deferred* after 12 months of inactivity in its previous *Experimental* state. Implementation of the protocol described herein is not recommended for production systems. However, exploratory implementations are encouraged to resume the standards process.

  67. fippo likes the text ralph sent.

  68. m&m

    it's a definite improvement

  69. fippo

    i think we had the reason for deferring it closer to that line before we moved all the changelog stuff to the end

  70. Kev

    I'd like a reference to XEP-0001 for each of the states, not just deferred. I thought Ralph's text was a reasonable base. I'd like to tweak further, but a definite improvement.

  71. stpeter nods

  72. Kev

    Actually, I'm not sure how much I'd like to tweak it further, that text is pretty good.

  73. Kev

    I think we could reasonably wordsmith Experimental too, in similar ways.

  74. stpeter

    quite possibly

  75. stpeter

    I haven't actually read any of that text in quite a while

  76. MattJ

    Yes, Ralph's text is fine

  77. Kev

    Anyway. Would everyone be happy with making the Ralph change + xep1 reference? (Including Peter)

  78. MattJ

    I was wondering about the last sentence, but I think it's ok

  79. Tobias

    i'm okay with ralph's text too...it sure is an improvement

  80. Kev

    I think the last sentence is important in purpose.

  81. m&m

    yes

  82. stpeter

    all of those informational status paragraphs ought to include links to XEP-0001

  83. Kev

    Because 'deferred' is scary when people think it means 'retracted'.

  84. MattJ

    Me too, but I was thinking perhaps s/encourag/welcom/

  85. Kev

    or 'rejected'.

  86. MattJ

    But as I said, I think I've changed my mind, I like it as-is

  87. Kev

    I think we could make this chance and further wordsmith as time goes on.

  88. Kev

    s/chance/change/

  89. m&m

    exactly

  90. Kev

    fippo: And this resolves your complaint, right?

  91. fippo

    kev: absolutely

  92. Kev

    Marvellous.

  93. Kev

    I don't think we had anything else on the agenda, did we?

  94. Kev

    3) Date of next meeting.

  95. Kev

    Next week's presumably going to clash clash with some other thing that's going on.

  96. m&m

    quite possibly

  97. Kev

    Week after, then?

  98. Tobias

    wfm

  99. stpeter

    heh

  100. m&m

    that works for me

  101. stpeter

    I will not be working the following week, but do feel free to have a fun meeting on the 7th

  102. Kev

    If Peter's not working, that probably means we're not going to have new things to talk about.

  103. Kev

    Assuming them popping up next week is unlikely

  104. Kev

    Should we skip a fortnight?

  105. m&m

    I'd rather not skip that long, even if there's nothing much to do

  106. Kev

    OK.

  107. stpeter

    yeah

  108. Kev

    Fortnight today, then.

  109. stpeter

    I'm just a figurehead anyway :P

  110. Kev

    stpeter: If you don't push protoXEPs, we have no work. Having no Council work to do appeals to me :)

  111. Kev

    4) Any other business.

  112. stpeter

    heehee

  113. MattJ

    No other business, but a praise of the people (re)working on Jingle :)

  114. fippo

    kev: i think you might have explained your -1 on sox with your last post to jingle@

  115. MattJ

    Good to see the interest

  116. ralphm

    back

  117. stpeter

    I need to catch up on jinglish things before the meeting 9 minutes from now

  118. Kev

    fippo: As in "My post explained it" or "When I posted, I should have also explained..."?

  119. Kev

    ralphm: We decided to make you chief blurb-writer. That's about it.

  120. fippo

    kev: somewhere in between. i'd note that the sox authors actually have no intention to replace jingle though :-)

  121. Kev

    And yes, I'm very happy that people are working on Jingle in earnest. If only I had time to meaningfully contribute, other than snide remarks here and there.

  122. ralphm

    yay!

  123. Kev

    I think we're done?

  124. MattJ

    +1

  125. Kev

    Thanks all.

  126. Kev bangs the gavel.

  127. Tobias

    thank you

  128. ralphm

    Thanks!

  129. ralphm

    that was pretty efficient :-D

  130. Kev

    fippo: That may be obvious to you, but it's not obvious to the rest of (me) where SoX fits into a world where it improves interop rather than harming it.

  131. Kev

    fippo: And my -1 was provisional until the Jingle SIG produce a coherent argument about where all these things fit, rather than a blanket -1.

  132. ralphm

    with that, any other comments welcome in 5 minutes next door