XSF Editor Team - 2014-03-11


  1. stpeter

    hmm

  2. stpeter

    m&m let's discuss in the meeting tomorrow

  3. m&m

    I need to put together an agenda for today

  4. winfried

    some suggestions:

  5. winfried

    - post mortum

  6. m&m

    remember we only have 30 minutes!

  7. winfried

    - access

  8. m&m

    - BOSH

  9. winfried

    - dividing work

  10. winfried

    - how to communicate with the council

  11. winfried

    - structuring changelogs and tagging changes as major/minor

  12. winfried

    - XEPS we have to act on

  13. winfried

    - the next humorous xep

  14. Steffen Larsen

    +1

  15. m&m

    that is more like 120 minutes of stuff

  16. Steffen Larsen

    :-)

  17. m&m

    we're not talking about it all

  18. Steffen Larsen

    its actually hard to divide work when we use email as our XEP queue

  19. Steffen Larsen

    unless we are talking together and splitting work in our muc here

  20. Lloyd

    Agree with Steffen Larsen

  21. Steffen Larsen

    well I am off.. talk to you in a couple of hours guys

  22. Kev

    Isn't it just a case of replying to an item on the list before you deal with it? It hardly affects me what tools you folks use, but still...mailing lists seem like they should work fine.

  23. m&m

    Lloyd: would you track down Ash for us, please?

  24. winfried

    I Have to excuse Stefan Strigler, he probably can't make it today

  25. Lloyd

    He;s around

  26. Steffen Larsen

    excused! :-)

  27. Lloyd

    just poked ash to join early

  28. Ash

    Hello!

  29. winfried

    hi!

  30. m&m

    my clock sais we still about about two minutes

  31. m&m

    says even

  32. m&m

    clearly that is demonstrating the elite typing skills of an editor

  33. Lloyd

    :)

  34. Steffen Larsen

    :-)

  35. Ash

    :)

  36. Steffen Larsen

    hello Ash and Winfried!

  37. winfried

    counting down for banging the gavel

  38. winfried

    hi steffen ;-)

  39. Steffen Larsen

    2

  40. Steffen Larsen

    1

  41. Steffen Larsen

    0

  42. winfried

    *BANG*

  43. m&m bangs gavel

  44. m&m

    alright, so the agenda for today

  45. m&m

    1) Hand-off Portmortem ------------------------------

  46. m&m

    I kept meaning to write this up while in London last week, but failed miserably

  47. stpeter

    hi :-)

  48. winfried

    how much is there to say?

  49. m&m

    XEP-readme was updated to reflect the latest

  50. Steffen Larsen

    yes it looks more modern now :-)

  51. m&m

    it still requires a good amount of manual steps

  52. m&m

    and special access

  53. m&m

    list administration for standards@

  54. stpeter

    right

  55. m&m

    WordPress (until Simon and Laura replace the WP-based site with something static)

  56. m&m

    git

  57. m&m

    shell

  58. Ash

    Not sure but I guess the new site will still be wordpress

  59. stpeter

    yeah, too many disparate tools

  60. m&m

    I think there's still a couple of bugs in the XEP-readme, and I'll submit a patch to address that

  61. Steffen Larsen

    yes a lot could be done automatically

  62. m&m

    the shell and python scripts have some documentation, which probably should go into the readme (or someone to update the scripts)

  63. Steffen Larsen

    but lets do that when we understand the process better

  64. m&m

    right

  65. winfried

    +1

  66. m&m

    I started looking into what it would take to use git hooks

  67. winfried

    would love to be able to test / run all scripts and so locally

  68. m&m

    I think it's possible, but I have some concerns if there are non-Editor team people with push rights

  69. m&m

    yes

  70. Steffen Larsen

    how about also using merge request from gitorious?

  71. m&m

    I'm not a Git expert, but wouldn't that require us to have the canonical repo in gitorius?

  72. winfried

    Steffen: maybe better understand the current process before doing so

  73. Steffen Larsen

    sure

  74. m&m

    I think the current readme covers the steps fairly well now

  75. Steffen Larsen

    yes

  76. m&m

    the only bug I've found are references to a local git working copy that is not in the right place

  77. Steffen Larsen

    my biggest concern is not the process.. but more the queue and the many mails that confuses me..

  78. winfried

    the readme still needs to be build into html and updated on the site ;-)

  79. Ash

    Steffen: +1

  80. m&m

    Steffen Larsen: one thing at a time!

  81. Steffen Larsen

    thats done by a script right

  82. Steffen Larsen

    m&m: right

  83. m&m

    yes, XEP-readme is a XEP, so gen.py can do the right thing with it

  84. m&m

    I'm not sure there's much more to say about the postmortem

  85. m&m

    stpeter?

  86. stpeter

    sounds about right to me

  87. winfried

    ok, that brings us to the access rights....

  88. stpeter

    I'd like to get more people working on things (see next topic) because different people will find different process bugs

  89. m&m

    2) Access Control ------------------------------ + Website + Git + Mailing lists

  90. m&m

    I lumped website together

  91. m&m

    that's really shell (for the static-ish side) and WP

  92. m&m

    to get access, you need to send Kev your SSH key

  93. m&m

    to get git and shell

  94. winfried

    ok

  95. Ash

    Kev had concerns with lots of people having shell access. Has that been resolved?

  96. m&m

    he's nervous about signing everyone up for that, though

  97. m&m

    no it has not

  98. Steffen Larsen

    ok

  99. Steffen Larsen

    so we will ship him our key?

  100. m&m

    Steffen Larsen: correct

  101. m&m

    I understand Kev's concerns about shell access

  102. winfried

    so there is no editor group with proper limits?

  103. m&m

    but, we need to be able to do the job

  104. m&m

    so unless we're going to do a potentially massive tooling effort right now ...

  105. m&m

    … the i-team is going to have to grant limited access

  106. m&m

    there are limits, though

  107. Steffen Larsen

    ok

  108. Lloyd

    if we implemented git hooks then shell access would become a moot point

  109. m&m

    we really just need to have /bin/bash as user "xsf"

  110. m&m

    Lloyd: most likely, yes

  111. m&m

    I'm not sure how much effort is required there

  112. Steffen Larsen

    for what? updating the local repo and generating the html ?

  113. m&m

    Steffen Larsen: yes

  114. Kev

    If you all want shell, can you let me know which commands you need to run as xsf, please?

  115. Kev

    Just granting lots of extra people access to bash doesn't sound ideal.

  116. Steffen Larsen

    ok. we might just be able to do that by cron? and let it check out the latest master from the repo?

  117. m&m

    Steffen Larsen: slow down a little (-:

  118. Steffen Larsen

    sorry.. just trying to make it easier by not having ssh access

  119. m&m

    if we're willing to limp along with a couple of people having shell...

  120. Lloyd

    I'd prefer not to have shell if possible, there should be tools/procedures in place to make things as simple as possible. (happy to have shell I'm talking the global "I")

  121. m&m

    … and others (possibly those same people) working on more automated tooling ...

  122. Lloyd

    m&m +1

  123. m&m

    … we could get by with a lot fewer shell rights

  124. Steffen Larsen

    yes!

  125. winfried

    ok so we should divide efforts: tooling and current queue

  126. Lloyd

    yup

  127. m&m

    winfried: +1

  128. Steffen Larsen

    +1

  129. winfried

    +1

  130. winfried

    ;-)

  131. m&m

    the tooling group should be able to work with a local clone of git

  132. Ash

    In terms of tooling, we could make use of Atlassian's tools which are free for open source projects

  133. m&m

    Ash: you need to clear that with the i-team

  134. m&m

    they would be the ones to maintain the base software

  135. m&m

    installs, updates, runtime, etc

  136. Ash

    Their cloud service is free

  137. m&m

    if the Board is not comfortable moving git to a cloud service, I don't see this being acceptable either

  138. winfried

    guess some git-hooks being build on the current scripts should do?

  139. Ash

    Jira for ticket management, Bamboo for CI (to do the builds)

  140. m&m

    s/Board/XSF membership/

  141. Ash

    Fair enough. Just an idea :)

  142. m&m

    I do appreciate it

  143. stpeter

    Ash: we have previously used Atlassian tools so I think we could use them again if desired / needed

  144. Lloyd

    Not a fan of atlassian, but at least it would give us an easy integrated solution. However if we were going that far I'd prefer github + travis - with the added benefit of visibility :)

  145. m&m

    but I appreciate the arguments against better (-:

  146. stpeter

    Lloyd: +1 to transparency

  147. m&m

    as long as we can run it on athena, and the i-team is willing to deal with the software installs/updates

  148. m&m

    to me, scripts are one thing, but software packages are another

  149. stpeter

    nod

  150. winfried

    can you elaborate that a bit?

  151. m&m

    and I don't think it's appropriate for the editor team to maintain software packages unless we can isolate it to just the editor team function (write-wise)

  152. Lloyd

    I think maybe we're getting a little sidetracked with solutions rather than tasks.

  153. winfried

    yup

  154. m&m

    Lloyd: I think you're right (-:

  155. m&m

    anyway, let's figure out who wants to do what

  156. stpeter

    engineers love to talk about their tools :-)

  157. m&m

    who wants to look into improved automation?

  158. Ash

    :)

  159. Lloyd

    stpeter :)

  160. winfried

    *silence*

  161. Lloyd

    So how about those with current access work on current queue and others in tooling team (discuss via mailing list) as a first start and then anyone who wants to switch (or do both) please say

  162. Lloyd

    m&m I'll happily do that.

  163. m&m

    thanks Lloyd

  164. Lloyd

    m&m: you might not like my suggestions ;) I have strong opinions sometimes

  165. m&m

    Lloyd: I am bigger than you, though (-:

  166. Lloyd

    ..and across an ocean :P

  167. stpeter

    heh

  168. Kev

    When we've used Atlassian previously, it has been a services we've hosted, and the maintenance was not fun.

  169. m&m

    frankly, as long as it gets the job done and the maintenance is reasonably isolated, I'm open

  170. stpeter

    Kev: agreed

  171. m&m

    s/reasonably isolated/reasonable/

  172. m&m

    as for queue management, I've found email to be sufficient for small teams like this

  173. Kev

    I'm not saying 'iteam is not willing to do this', just noting that it's a significant time drain.

  174. stpeter

    email + this chatroom are fine by me

  175. stpeter

    but I'm old-fashioned

  176. winfried

    stpeter +1 (on both)

  177. m&m

    stpeter: apparently I am too

  178. Lloyd

    mostly works for me, if it doesn't I'm sure we'll shout.

  179. m&m

    just be good about checking your email d-:

  180. m&m looks at the agenda again

  181. m&m

    is there more about access we want to talk about?

  182. m&m

    mailing lists

  183. stpeter

    right, I sent mail about that

  184. m&m

    I think we should all have admin rights on at least editor@, and probably standards@ too

  185. stpeter

    I think it would spread the load

  186. m&m

    stpeter: that's why I have it in the agenda (-:

  187. stpeter

    (we need to get some greylisting and other spam-fighting things on atlas while we're at it)

  188. winfried

    wondering what will be coming my way, but lets hit it ;-)

  189. m&m

    winfried: you can help with tooling (-:

  190. m&m

    and/or list moderation

  191. m&m

    stpeter: does the editor team need to be involved with that anti-spam effort?

  192. winfried

    thats ok, was just wondering what list moderation would end me up with ;-)

  193. m&m

    winfried: so far, it's periodically checking the editor@ list and junking a bunch of emails

  194. m&m

    really not more than once a day

  195. stpeter

    m&m I don't think so

  196. winfried

    ok

  197. m&m

    standards@ has maybe one thing a week that needs attention

  198. m&m

    from what I've noticed

  199. m&m

    ok, we're behind schedule

  200. winfried

    yup

  201. m&m

    any more on access?

  202. winfried

    nope

  203. Lloyd

    (shake)

  204. Ash

    Don't think so

  205. m&m

    there is a wiki that Steffan got started, so we can capture some of the decisions made here on that page

  206. m&m

    moving on

  207. m&m

    3) Outstanding Work ------------------------------ + BOSH + Deferred XEPs + Website updates

  208. Ash

    + XEP-0060

  209. winfried

    BOSH: I send a patch to the editor@ fixing a minor type

  210. winfried

    s/type/typo/

  211. m&m

    winfried: as soon as possible, please (-:

  212. m&m

    There is a list of XEP to defer

  213. winfried

    send it on feb 17th

  214. winfried

    will resend it...

  215. m&m

    oh, send not sent

  216. m&m

    ah

  217. m&m

    yes, please … it probably got lost in the shuffle from stpeter to !stpeter (-:

  218. stpeter

    nod

  219. winfried

    (and I don;t have rights) After that BOSH can be voted upon for final

  220. m&m

    stpeter: you have the list of XEPs to defer handy?

  221. stpeter

    m&m: yep, please hold

  222. m&m

    winfried: /nod

  223. m&m

    I suggest that we deal with deferment once a month

  224. Ash

    Sounds sensible

  225. m&m

    it means some Experimental XEPs will have extended times, but that seems fine to me

  226. winfried

    that is already current practice not?

  227. m&m

    winfried: the current practice is that stpeter gets to the deferred XEPs when he can (-:

  228. stpeter

    the only XEPs to be deferred that I see are 305, 316, and 317

  229. m&m

    we'll try to be a little more regular with it

  230. m&m

    stpeter: thanks

  231. m&m

    for website, I think the only thing left right now is automation, which we already discussed

  232. m&m

    all the other WP updates I had in my local .todo have been taken care of

  233. winfried

    for my picture: who has access and can work on the current queue right now?

  234. Lloyd

    not I

  235. m&m

    definitely myself and stpter

  236. Ash

    Nor me

  237. m&m

    I thought winfried had git access, but I guess that's not the case

  238. winfried

    nope

  239. m&m

    I think all of us should have git rights

  240. m&m

    then you can ping myself or stpeter for the stuff that requires shell access

  241. m&m

    at least until we get better automation going

  242. stpeter

    that seems fine for now

  243. stpeter

    nod

  244. m&m

    *hint hint Lloyd (-:

  245. Lloyd

    I will start looking into tooling this week and report back via email/next meeting

  246. m&m

    thanks

  247. m&m

    Ash: you mentioned XEP-0060 … are the patches pending?

  248. Ash

    Not yet. I have started, but don't have much time.

  249. m&m

    ok, so you're working on patches, but don't have them yet

  250. Ash

    Yeah

  251. m&m

    so there's no action for the rest of the editor team (-:

  252. Ash

    So not to worry

  253. Ash

    :)

  254. m&m

    I think that covers #4

  255. m&m

    er #3

  256. m&m

    4) Future Work ------------------------------ + Automation options + Humorous XEP

  257. m&m

    we've already figured out automation

  258. Lloyd

    Ash, remind me to talk about patches dwd and I discussed last week

  259. m&m

    regarding a humorous XEP …

  260. Ash

    Lloyd: Will do

  261. m&m

    do we want to try and do one this year? or does someone already have a writeup ready for council sign-off?

  262. winfried

    well... I suggest we vote on a humorous xep before sxep-0001 is changed :-P

  263. m&m

    heh

  264. m&m

    we need an idea first

  265. stpeter

    I haven't done one of those in years

  266. stpeter

    they were fun, though

  267. winfried

    was thinking about ciricular approving bodies for approving humorous xeps

  268. stpeter

    in London we got to talking about the morse code transport binding - who needs WebSocket? ;-)

  269. m&m

    winfried: (-:

  270. m&m

    either of those are fine with me

  271. m&m

    I think I might have been the one to suggest the morse-code transport, too (-:

  272. m&m

    if something is to come from this body, I think we would need it ready before next Friday (03/21)

  273. m&m

    it's just under two weeks, but I know I don't have the time myself

  274. winfried

    was brainstorming once with steffen on xmpp over humanpowerd two wheeled pedal transport

  275. winfried

    but I want to have that one implemented before submitting ;-)

  276. m&m

    haha

  277. m&m

    remember — "XEP or it doesn't exist". Someone needs to write it before we can publish

  278. m&m

    any volunteers?

  279. stpeter

    when was the last humorous XEP published?

  280. m&m takes a look

  281. winfried

    to be honest, xmpp over morse code gives me a bit of a feeling: "been there before"

  282. stpeter

    2009

  283. stpeter

    2011

  284. m&m

    yes, XEP-0295

  285. winfried

    commentory on some current topic (like the json xep) is great

  286. m&m

    that or another security one

  287. winfried

    yeah!

  288. m&m

    the anti-privacy XEP

  289. stpeter

    the problem with humorous XEPs is that you need to be inspired to write one, and I guess I haven't felt inspired in that way since 2009 :(

  290. m&m

    stpeter: exactly

  291. m&m

    and I take the lack of a hand or "I'll do it" to mean we have no volunteers to write it

  292. winfried

    if someone volunteers with me, I want to write an anti-privacy xep

  293. winfried

    got some great ideas bubbling up

  294. m&m

    I don't have the time to do that

  295. stpeter

    the best humorous specs also have a germ of truth in them

  296. m&m

    I signed up for too much extra work last week (-:

  297. winfried

    neither have I ;-)

  298. stpeter

    m&m: you did!

  299. m&m

    ok, moving on

  300. m&m

    5) Next Meeting ------------------------------

  301. winfried

    yeah, ping me if anybody wants to be involved

  302. Ash

    Automated carbons to your local government agency, to save taxpayers money?

  303. Lloyd

    :)

  304. winfried

    yep, that kind of stuff@!

  305. m&m

    I think this meeting is useful

  306. m&m

    so we should keep doing it

  307. Ash

    Agree

  308. m&m

    but do we need/want to do it weekly or bi-weekly?

  309. winfried

    +1

  310. stpeter

    shall we add it to the calendar?

  311. m&m

    monthly is too long

  312. winfried

    agree

  313. m&m

    stpeter: I think that makes sense

  314. Lloyd

    fortnightly seemed to make sense

  315. Ash

    +1

  316. Lloyd

    At least until processes/automation settles?

  317. Kev

    FWIW, I think very short meetings frequently make a lot of sense. But that's just me.

  318. m&m

    Lloyd: I was about to suggest weekly until we have automation (-:

  319. winfried

    I have to leave right now...

  320. m&m

    we're 17 minutes over

  321. winfried

    I will read it back in the logs, this time usally fits, otherwise best on monday or friday

  322. m&m

    is everyone ok with resuming next week?

  323. Lloyd

    m&m fine with that, as long as they aren't all 1hr long :)

  324. m&m

    no, please no!

  325. m&m

    30 minutes

  326. Ash

    OK

  327. m&m

    and I'll have a better agenda for next week, really (-:

  328. m&m

    ok, next meeting on 2014-03-18 @ 16:00 UTC

  329. stpeter

    yep, cut it off at 30 and we can delay until the next one

  330. m&m

    exactly

  331. Lloyd

    oh, a pep node for "selfies"

  332. m&m

    ok, I say we're done

  333. m&m bangs gavel

  334. m&m

    thanks everyone

  335. Lloyd

    Cool see you all next week - and every working day until then :)

  336. stpeter

    Lloyd: :-)

  337. m&m puts Lloyd on next weeks agenda (-:

  338. Ash

    Cheers!

  339. Lloyd

    You can discuss me all you want

  340. m&m

    (-:

  341. m&m

    holy crap … I'm already 31 emails behind since I last checked

  342. stpeter

    BTW I just edited the postfix config a bit on atlas, but no promises that it will help

  343. m&m

    which was just before 16:00 UTC

  344. m&m

    stpeter: we'll see (-:

  345. m&m

    everyone, be sure to get Kev your info for git access

  346. Lloyd

    Kev how best to do this?

  347. Kev

    Send me it gpg signed using the key you had me sign at the summit? ;)

  348. m&m wanders off for coffee