XMPP Council - 2013-08-14


  1. Kev

    Same as last week's unmeeting, the only thing I see in the agenda for today is http://xmpp.org/extensions/inbox/dynamic-forms.html

  2. Tobias

    ok

  3. Zash

    Will m&m be around for that IETF summary?

  4. m&m

    I will try

  5. Zash

    :D

  6. m&m

    I started the day quadruple-booked for meetings … now I'm down to double-booked (-:

  7. Tobias

    heh

  8. m&m

    if I'm not here … I am −1 on dynamic forms, unless the author can explain why ad-hoc commands is not suitable

  9. Kev

    Meeting in 9 minutes.

  10. Kev

    m&m: Do you believe that dynamic-forms can be done using -50 then, without all the extra business logic that's the core bit of the proposal?

  11. m&m

    yes

  12. Kev

    I don't see it.

  13. Kev

    I mean, I can see how you could build the same stuff on top of 50 that this builds on top of 4, but it still seems to be building stuff on top of XX.

  14. ralphm

    Present

  15. Kev

    ralphm: You've got 6mins yet :)

  16. Kev

    m&m: I don't see how -50 gives you ways of the client telling the server "I've got a value in this field now, but the form's not complete yet" or of the server replying with "Oh, OK, add these fields to the form, then".

  17. m&m

    -50 supports multiple stages, which could fulfill the same role

  18. Kev

    Yes and no.

  19. Kev

    Because if you want the -50 stages to support that, you then lose the ability for -50 stages to support multiple stages.

  20. Peter Waher

    hello

  21. m&m

    that makes no sense

  22. Kev

    Peter Waher: Howdy.

  23. Kev

    m&m: It makes lots of sense :)

  24. m&m

    why can't the next stage send the form up to the point it knows of?

  25. m&m

    I've done this in the past

  26. m&m

    and now I'm in my next meeting

  27. Kev

    m&m: If you use multi-stage to be used for incomplete data submission, like the proposal needs, you can't also use multi-stage for multi-stage forms in their traditional sense.

  28. m&m

    don't know if I'll be paying enough attention here … sorry

  29. stpeter is indeed on a conference call and won't have much attention for the Council meeting

  30. ralphm

    stpeter: booh

  31. Kev

    Because you lose the ability to say "This is the next stage and I really mean it this time, please give me the next stage not the updated version of this stage".

  32. Kev

    And ding, it's time.

  33. Kev

    So 1) Roll call.

  34. Tobias

    here

  35. m&m

    not really heere

  36. Kev

    m&m is absent with apologies in a strange manner.

  37. m&m

    double-booked meetings

  38. m&m

    I can't help it today

  39. m&m

    nor next week

  40. ralphm

    here

  41. Tobias

    maybe we should find a different time then, at the end of this meeting?

  42. Kev

    MattJ: Poke.

  43. Kev

    I'm amenable to other times.

  44. Kev

    Or other days, at least. The time slot is probably as good as we'll get.

  45. Kev

    MattJ only just joined, he can't be not here can he?

  46. stpeter

    ralphm: it seems that the UPnP folks might be interested in using XMPP to cloud-enable their technology, so I'm helping them through some of the issues they're facing

  47. Kev

    I guess he's not.

  48. Kev

    2) http://xmpp.org/extensions/inbox/dynamic-forms.html Accept as Experimental.

  49. ralphm

    stpeter: nice

  50. Kev

    m&m is -1, to post to list.

  51. Kev

    So I guess that makes everyone else irrelevant :)

  52. Peter Waher

    not necessarily

  53. Peter Waher

    since it might be relatively easy to respond to m&m's objections

  54. Kev

    Peter Waher: A single -1 is blocking, so it doesn't matter what the rest of us think.

  55. Peter Waher

    it would be good to hear what all think

  56. MattJ

    Present

  57. Tobias

    i'm +0, post-back is some strange term (don't know a better one off hand) and if xep-0050 is not suitable it could at least mentioned in the introduction why not, since data-forms and ad-hoc are deployed and used

  58. MattJ

    Sorry, clock I was watching seems to be a few mins out

  59. ralphm

    for what it is worth, I am not really convinced that the dynamic stuff needs additional protocol wrappers

  60. Kev

    I don't see how this stuff can be done with what we have without a spec of some sort or another (be it significant business logic on top of existing protocol, or new protocol). So I'm not objecting.

  61. MattJ

    I'm +0, since I haven't read the document yet, I'll read it and post my concerns to the thread if I have any

  62. Kev

    ralphm: So you're blocking it too, or not?

  63. ralphm

    That said, I am not objecting to publishing.

  64. Peter Waher

    the ad-hoc command was considered in the beginning, but was too different for the purpose. If m&m objects on list I can respond to why

  65. Peter Waher

    considered writing something about it in the xep

  66. Peter Waher

    but didn't

  67. Peter Waher

    can easilly add why it was not used

  68. Kev

    OK.

  69. Kev

    So, 3) Date of next meeting.

  70. Kev

    Given m&m's the person with the conflicts, and he's not here, I don't see much point changing the date. Let's go for next week at 15:10UTC as usual, and we can change during the week if we want.

  71. ralphm

    Peter Waher: design choices are always great to have inside specs like this

  72. ralphm

    +1

  73. Tobias

    Kev, fine with me

  74. Peter Waher

    yes, I see that now. Thought about it, but for some reason I didn't add it. I'll correct that in my response to m&m on list

  75. Kev

    OK. 4) AOB?

  76. Tobias

    none here

  77. MattJ

    None here

  78. Kev

    Good enough.

  79. Kev

    Thanks all.

  80. Kev bangs the gavel.

  81. MattJ

    Thanks Kev

  82. Tobias

    thanks

  83. ralphm

    Thanks!

  84. Tobias

    ralphm, you planned some changes to XEP's HTML rendering?

  85. ralphm

    oh, right

  86. Tobias

    in what direction?

  87. ralphm

    basically some additional CSS

  88. Peter Waher

    there are problems with rendering of images

  89. Tobias

    Peter Waher, got an example?

  90. Peter Waher

    img tags are rendered with width="" height=""

  91. Peter Waher

    since width and height attributes are forbidden according to the schema

  92. Peter Waher

    this works in FireFox

  93. Peter Waher

    but makes images invisible in IE

  94. Peter Waher

    (it's treated as width="0" height="0")

  95. Tobias

    Peter Waher, ah..right

  96. Tobias

    wonder if we can do something about that :)

  97. Peter Waher

    if you want, I can mail you a mail I sent to Peter Saint-André earlier

  98. Tobias also needs to check how that looks in the pdf rendering :)

  99. ralphm

    you probably need to look at extensions/xep.xsl, where the attribute is rendered

  100. Peter Waher

    yes, I've proposed a correction

  101. Tobias

    great

  102. Peter Waher

    to whom do I mail it?

  103. Tobias

    can mail it to me

  104. bear

    if you cc me then I can look into fixing it

  105. Tobias

    tm@ayena.de

  106. Peter Waher

    bear: What's your mail address?

  107. bear

    bear@bear.im

  108. Peter Waher

    mail forwarded

  109. Tobias

    thx

  110. bear

    got it

  111. Peter Waher

    (y)

  112. Zash

    delivery receipts over xmpp! :D

  113. bear

    tobias - I pushed the static site to the XSF github repo

  114. bear

    I am using the github repo just as a working area, the final repo will be our normal one

  115. bear

    i'll be adding more to the README to document what each area is

  116. bear realizes he is spamming the wrong folks and will send email

  117. stpeter promises to read the meeting log once he's finished with this conferene call...

  118. Tobias

    bear, we want to mix standards and our website repo?

  119. bear

    mix?

  120. Tobias

    well..do both stuff in a single repo

  121. bear

    that probably needs to be talked about - a lot of it has just been a few of us just doing things

  122. bear

    my reasoning for using github right now is just pure convienence, i'm in it every day

  123. Tobias

    heh

  124. bear

    and this test site needs some serious css love

  125. Tobias

    MattJ, got a link to that repo?

  126. ralphm

    Tobias: I see the gen.py depends on mercurial for versioning information?

  127. MattJ

    Which?

  128. Tobias

    ralphm, yes

  129. ralphm

    so you can't even build the XEPs from the git checkout

  130. Tobias

    you can using XSLT

  131. ralphm

    right

  132. Tobias

    gen.py renders the last version of a XEP that doesn't have an <interim/> tag in it

  133. ralphm

    or all

  134. bear

    part of why I spoke up to help - this needs to be "solved" for the static site generation

  135. bear notes email sent

  136. ralphm

    for some reason the XEP generated from the repo have a smaller font size than on the web site

  137. Tobias

    ralphm, basically i went with hg in python because of the easy py api, i think back then there wasn't a good one for git

  138. Tobias

    i can try changing it to dulwich, but that's at the end of my todo :)

  139. ralphm

    I have a personal preference to hg, too

  140. ralphm

    but our generation process should depend on it

  141. Tobias

    should?

  142. ralphm

    not. I pull and push from github using mercurial with ease

  143. Kev

    Let's not go changing our repo format :)

  144. Tobias

    Kev, hahaha

  145. Tobias

    Kev, ever used dulwich?

  146. Kev

    I don't have a problem with the 'official' generation having dependencies, either. It's easy to generate the XEPs yourself when you're editing them.

  147. Kev

    Tobias: No.

  148. m&m finally has time to catch up …

  149. m&m

    2) I will post my objections to the list by tomorrow (MDT)

  150. m&m

    3) apologies for next week; conflicts with JSON WG interim meeting

  151. m&m

    4) I can send a summary of the POSH BoF to some list, if people care

  152. bear

    m&m - please do send POSH summary

  153. Zash

    +1

  154. m&m

    ok, I'll send that at least to council@, fwd standards@ … sometime this week

  155. stpeter finally reads http://logs.xmpp.org/council/130814/ and sees that he has no action items so moves on ;-)

  156. Kev

    I need to do minutes. Probably several sets.

  157. stpeter

    I think at this point you probably need to sleep :P

  158. m&m

    sleep is for the weak

  159. Kev

    It's only 21:30.

  160. m&m

    (-:

  161. Kev

    I'll try and get minutes done tomorrow.

  162. stpeter

    Kev: thanks

  163. Kev

    The reason I've missed them is almost comical.

  164. Kev

    It's been hot in the UK (although not any more).

  165. Kev

    So I had my VM server shut off to save heating my study more than I needed. My TODO lives on one of the VMs. So I had nothing telling me to do minutes :)

  166. stpeter

    hehehe

  167. m&m

    sounds perfectly logical to me (-:

  168. stpeter

    and yes, m&m and I experienced the tail end of the heat when we were in Berlin

  169. m&m

    ugh

  170. stpeter

    we Americans are spoiled by the ubiquitous and excessive Air Conditioning™

  171. m&m

    in our defense, it approached 37C outside …

  172. Tobias

    just one day :)

  173. m&m

    it wasn't that much colder the rest of the week! *

  174. m&m

    * not including the deluges

  175. Tobias

    not much but a bit, one day was quite rainy though

  176. m&m

    it was interesting how walking through the InterContinental, there were definite climate zones

  177. Tobias

    heh

  178. Tobias

    saw a tornado or so?

  179. m&m

    a few steps one way, and it went from fairly warm to chilly

  180. stpeter

    fascinating and totally off-topic: http://twistedsifter.com/2013/08/maps-that-will-help-you-make-sense-of-the-world/

  181. Tobias

    i find 9 particular surprising

  182. stpeter

    I liked the literal translations of Chinese names for European countries :-)

  183. Tobias

    moral-land, hahaha

  184. m&m

    Braveland!

  185. MattJ

    That's us!

  186. stpeter

    cf. map #3 :P

  187. stpeter

    heh, http://www.slate.com/blogs/the_vault/2013/08/12/the_1931_histomap_the_entire_history_of_the_world_distilled_into_a_single.html is cool too

  188. stpeter

    although it seems to end in 1931

  189. stpeter wonders who ever dreamed up the idea that PowerPoint was a good way to define technical architectures and protocols and such -- yikes!

  190. Kev

    I'm not convinced that claiming the UK uses the metric system is entirely right.

  191. stpeter

    the UK is in transition, I suppose

  192. MattJ

    "transition" :)

  193. Kev

    It's not, though, in any meaningful sense, transitioning.

  194. Kev

    It's halfway and stuck.

  195. stpeter

    ah

  196. MattJ

    As a result, I'm hopeless at kilometres, but useless at yards

  197. stpeter

    many years ago one used to seek some road signs in kilometers in America, but not often any longer -- and that won't change until Satan buys a snow shovel

  198. Kev

    Driving distances/speeds are always imperial. Weight of a person is usual imperial. Cooking is a mix. Most things you buy are metric, but often in imperial round numbers. You buy a pint of beer (if your tastebuds are broken). Short distances could be either. Etc.

  199. stpeter

    s/seek/see/

  200. stpeter

    I'm looking forward to visiting England again for the IETF meeting next spring (although at the same time I'm working hard to finish up all my IETF work!)

  201. MattJ

    I'm looking forward to that too, I might finally make it

  202. stpeter

    I'll definitely plan an XMPP hackfest of some kind

  203. MattJ

    Excellent :)

  204. stpeter

    maybe I'll stay busy with IETF stuff through the Hawaii meeting ;-)

  205. Tobias

    stpeter, hehe

  206. Tobias

    stpeter, who around the IETF do we have to nag about updating jabber.ietf.org?

  207. stpeter

    updating the website or the xmpp server?

  208. Tobias

    xmpp

  209. Tobias

    they are running some ancient ejabberd

  210. stpeter

    I'll ping my friends at AMS about that

  211. Tobias

    thanks :)

  212. stpeter

    I offered to help them

  213. Tobias

    they don't have to switch to prosody...but they could at least update to the current ejabberd release

  214. ralphm

    stpeter: you have people living at the airport?

  215. stpeter

    switching to Prosody is the plan

  216. ralphm

    Kev: for what it is worth, even though the Netherlands is mostly metric, there are various things measured in non-round values

  217. MattJ

    stpeter, maybe poke when 0.9 is out

  218. stpeter

    MattJ: good plan -- I know they have some hardware migrations to finish first

  219. stpeter

    note to self: don't set passwords longer than 40 characters, they might be easy to remember but they're hard to type

  220. Neustradamus

    stpeter: About XSF servers, have you news about updates?

  221. stpeter

    hmm

  222. stpeter

    what kind of updates?

  223. Neustradamus

    Debian 7, M-Link...

  224. stpeter

    M-Link isn't an XSF matter

  225. stpeter

    but that's being worked on

  226. stpeter

    Jonathan Siegle has been working on Debian upgrades

  227. Neustradamus

    it is on it like ejabberd before... ;)

  228. Neustradamus

    ok !

  229. ralphm

    stpeter: I recently learned that three simple dictionary words, separated by spaces, should have enough bits of entropy to be reasonably secure

  230. ralphm

    Wonder if that's true

  231. Kev

    I think the usual pattern is four.

  232. Kev

    But I'll believe that three is sufficient.

  233. Kev

    Four certainly has more entropy than an 8character alphanumberspecial using the usual rules.

  234. stpeter

    ralphm: well, this password is for my PGP key, so I was being paranoid :-)

  235. ralphm

    right

  236. Kev

    But I think the maths is fairly easy.

  237. Kev

    Assume 100 valid chars in the usual passwords, so that's 100^8 combinations.

  238. Kev

    sigsegv:devel/swift/swift/> wc -l /usr/share/dict/words 8:59pm 235886 /usr/share/dict/words

  239. Kev

    Assume four words, so that's 235886^4 combinations.

  240. Kev

    >>> 100**8 10000000000000000 >>> 235886**4 3096054976816887360016L

  241. ralphm

    http://xkcd.com/936/

  242. Kev

    Yes.

  243. ralphm

    for those visual learners

  244. stpeter

    heh, someone pinged just now about an XMPP implementation that doesn't really handle presence correctly for multiple resources -- but it seems that we're not clear about such things in RFC 6121, perhaps because we just assumed that clients would do the right thing about displaying presence correctly (or perhaps we need a XEP on this topic)

  245. stpeter

    things like "don't show the user as offline just because one of the user's resource sent unavailable"

  246. ralphm

    hah

  247. ralphm

    I've heard many people refer to our priority stuff to bat shit crazy, too

  248. Tobias

    stpeter, hehe

  249. Tobias

    i think i stumbled upon that strange sentence while reading up probes stuff