XSF Discussion - 2018-02-22


  1. moparisthebest

    What's that sasl mode that proves the server knows your password?

  2. moparisthebest

    I recall Zash talking about it, like it'd be safe to use with a self signed cert type thing

  3. SamWhited

    moparisthebest: scram is one

  4. Ge0rG

    https://signal.org/blog/signal-foundation/ wow! It's impressive what they've achieved with 2.3 developers, and now they are drowning in money.

  5. Zash

    Ge0rG: Weren't they already drowning in money?

  6. Ge0rG

    Zash: they just got 50M USD on top.

  7. marmistrz

    > by Brian Acton, the co-founder of WhatsApp

  8. marmistrz

    am I the only one to whom it's suspicous? :)

  9. Ge0rG

    marmistrz: he sold it to facebook and made gazillions on it?

  10. Ge0rG

    > sell software for distribution Jaber and base 450000 jabber - ready business for you is$500 a real bargain!

  11. jonasw

    I don’t even get it

  12. marc

    Ge0rG: ?

  13. Zash

    English is hard

  14. Ge0rG

    Just got another incoming XMPP spam flood

  15. jonasw

    I think he’s getting spimmed

  16. Ge0rG

    somebody is selling a jabber spam bot plus a JID database for 500 usd.

  17. marc

    Oh 😮

  18. Seve

    A real bargain!

  19. marmistrz

    Ge0rG: I know they sold it to fb, what I mean that "if it's shaped in the same way whatsapp was shaped"

  20. marmistrz

    I mean, I'm not using signal (phone number as an ID is a bad idea), but if it's backed by the whatsapp founder, I'd have at least a little more suspicion before ever thinking to use it, judging from the way whatsapp has evolved

  21. jonasw

    marmistrz, maybe the whatsapp founder feels bad for what happened :)

  22. marmistrz

    jonasw, by supporting another walled garden?

  23. jonasw

    obviously he’s not opposed to walled gardens

  24. marmistrz

    well, if someone donated 5% of what he did for XMPP client development...

  25. Guus

    There are times that I wonder if our some on our support forums are high on something: "Smack Bubble in Rainbow"

  26. Guus

    (A service provided by openrainbow.com is apparently XMPP based?)

  27. Ge0rG

    marmistrz: there was a time when I respected moxie for what he does. Back then I'd have said that moxie will keep an eye on that money.

  28. intosi

    General service announcement: systemd sucks.</rant>

  29. MattJ

    Old news

  30. intosi

    I know.

  31. Ge0rG

    XMPP sucks, too.

  32. Zash

    Everything sucks.

  33. intosi

    No, no, you're wrong there.

  34. intosi

    My vacuum cleaner doesn't anymore.

  35. jonasw

    well played

  36. Kev

    Mine either, actually, its battery doesn't last a full round of the floor it's on :(

  37. Ge0rG

    Kev: should've bought a Tesla then :P

  38. Kev

    I call it Gir.

  39. Zash

    That's it! What we really need is Internet-connected battery-powered IoT vacuum cleaners that collects dust into a blockchain!

  40. Ge0rG

    Zash: into a quantum blockchain!

  41. Zash

    And battery charging as a service!

  42. intosi

    3) profit!

  43. Ge0rG

    Zash: and huge batteries, because PoW performance

  44. Guus

    Kev: your floor is to large.

  45. Kev

    Very much not!

  46. Guus

    your vacuum cleaner disagrees.

  47. Kev

    My vacuum cleaner is senile, it used to be fine with it.

  48. Guus

    Hmm... Sure your floor didn't grow?

  49. Kev

    Moderately.

  50. Guus

    I concede.

  51. Ge0rG

    Never underestimate continental drift!

  52. Guus

    I unconcede. What he said!

  53. pep.

    Can somebody add me into memberbot's roster? I sent a subscription presence yesterday

  54. jonasw

    pep., ping Alex

  55. pep.

    What he said ^

  56. marc

    Ge0rG, are you going to change your langing page such that it accepts xmpp URIs?

  57. Ge0rG

    marc: are you going to look up how to do a string replace in Erlang?

  58. marc

    Ge0rG, yeah, if you tell me the reason for _not_ using the xmpp URI ;)

  59. Ge0rG

    marc: every character counts. more characters = longer QR codes = less readabiliy = higher chance for typos when entering manually

  60. Ge0rG

    marc: the "xmpp:" is really redundant there.

  61. Ge0rG

    marc: besides, you need to url-encode the JID anyway, so you won't get around custom processing.

  62. Ge0rG

    marc: and you could just make the default prefix "xmpp:" for when constructing the link

  63. marc

    Ge0rG, that's not a good reason for _not_ accepting it ;)

  64. Ge0rG

    marc: yes it is. I'm forcing you to do something that's good for your users.

  65. marc

    ...

  66. marc

    sounds like I need to fork :D

  67. Ge0rG

    marc: or you need to bury your ego and just listen to what my ego says is the right thing™ 😛

  68. marc

    Ge0rG, not sure if this is gonna happen for this thing ;)

  69. jonasw

    marc, FWIW, I think Ge0rG has a point here.

  70. marc

    jonasw, 5 chars?

  71. marc

    are you going to ban too long local parts of a JID too? :D

  72. marc

    or too long domain names?

  73. Ge0rG

    marc: now guess why I own yax.im

  74. marc

    Ge0rG, because it's a funny name?

  75. marc

    probably I'll append the OMEMO keys to the QR code

  76. marc

    We generate QR codes for a XMPP URI and 5 OMEMO keys at the moment

  77. marc

    Works fine

  78. Ge0rG

    marc: to the xmpp URI or to the location URL?

  79. marc

    Ge0rG, to the xmpp URI

  80. marc

    doesn't make sense to transfer OMEMO keys via an "insecure" channel

  81. Ge0rG

    marc: not much probably.

  82. marc

    it's for the QR code only

  83. Ge0rG

    marc: I mean obviously it does make sense to transmit them but not to trust them.

  84. Ge0rG

    just check for MitM

  85. marc

    Ge0rG, nitpicker ;)

  86. Ge0rG

    marc: professional pessimist^W security analyst :P

  87. marc

    Ge0rG, if we put the URI into the fragment part we don't need to URL encode it AFAIK

  88. Ge0rG

    marc: except when we do.

  89. marc

    Ge0rG, hm?

  90. Ge0rG

    marc: as I mentioned some days ago, have a look at typical irc Transport JIDs

  91. Ge0rG

    marc: https://yax.im/i/#%23channel%25irc.man-da.de@irc.transport.server?join

  92. marc

    Ge0rG, wtf should I care about IRC transport URIs for invitation?

  93. Ge0rG

    marc: "#" is a valid local part character

  94. marc

    Ge0rG, "'#" can be used as username?

  95. Zash

    -nodeprep #hello

  96. Bunneh

    Zash: #hello

  97. jonasw

    yeah

  98. Ge0rG really needs to change name into Cassandra.

  99. marc

    Ge0rG, but URL encoding is the wrong thing here, we don't need to escape all characters as not URL encoders do

  100. marc

    ...as normal..

  101. Ge0rG

    marc: good luck figuring out the right subset. Also don't forget to tell me!

  102. marc

    Ge0rG, if you would _really_ care about the chars you would spend your time to figure it out ;)

  103. Ge0rG

    marc: I'm here to point out the problems, not the solutions!

  104. Guus

    wait, I thought you were here because you were our mascot?

  105. Ge0rG

    Guus: what? I'm way too sexy for that! 😁

  106. Seve

    A sexy mascot is what we need so..

  107. Guus

    w00t! just had my first in-client video conference :)

  108. Guus

    it's still very rough around the edges, but still, I'm excited :)

  109. Seve

    Great Guus :D

  110. Ge0rG

    Guus: we are only a decade behind?

  111. Guus

    stop raining on my parade!

  112. Guus

    I've created a beautiful freak!

  113. Guus

    outdated, yes. incomplete, yes.

  114. Ge0rG

    Guus: oh, your own client. Good work, then! :)

  115. Ge0rG

    Time to catch up!

  116. Guus does the happydance

  117. marc

    Guus: Video conference via XMPP/Jingle?

  118. marc

    Ge0rG, After a walk in the sunshine I'm convinced now ;)

  119. Ge0rG

    marc: :)

  120. Guus

    marc, yes (although we're cheating by embedding jitsi-meet in a browser component)

  121. marc

    Guus, is this some closed-source component?

  122. Tobias

    Guus, but doing jingle xmpp protocol?

  123. Guus

    Jitsi Meet is open source, a webrtc-based solution that uses XMPP (COIN/Colibri/Jingle) for negotiation.

  124. Guus

    we've server-sided wrapped that solution in a very-easy-to-install plugin

  125. Guus

    clients will detect its presence, and use it isntead of the public instance

  126. Guus

    from there, it's a matter of opening the correct URL, and let Jitsi do the rest.

  127. Holger

    What's COIN?

  128. Guus

    Added bonus is that users of our client can invite anyone that has a browser.

  129. marc

    Guus, so you had a video conference between browsers running jitsi meet?

  130. Guus

    https://xmpp.org/extensions/xep-0298.html

  131. Guus

    marc, yes.

  132. Holger

    Guus: Ah, thanks.

  133. marc

    Guus, with n > 2?

  134. Guus

    yes. not unlike google talk

  135. Guus

    err

  136. Guus

    hangouts

  137. Guus

    I can try posting an invite here, if you want

  138. Guus

    (beware, alpha-quality software release)

  139. Guus

    Please join me in this web-based video-conference room: https://meet.jit.si/xsf

  140. marc

    Guus, masked camera here ;)

  141. marc

    But I can see you :D

  142. Guus

    I am nice to look at :D

  143. marc

    Nice

  144. marc

    Works fine

  145. Guus

    yeah, one-on-one is peer-to-peer

  146. marc

    Maybe somebody else can join to test with n > 2? :)

  147. Guus

    beyond that, a bridge kicks in

  148. marc

    where is that bridge hosted?

  149. Guus

    I've added a uuid to make the room name somewhat more unique, but neglected doing that in a muc context

  150. Guus

    it's part of the Jitsi setup. We're now using their public service, but that is also deployable as an Openfire plugin

  151. marc

    Nice, now we need this in some desktop / moible client :)

  152. Guus

    (or a standalone installation if you go with their setup)

  153. Guus

    they do have a mobile client

  154. marc

    okay, now we need this in other mobile clients :D

  155. Guus

    they have an API that potentially is usable for that

  156. Guus

    not my area of expertise though

  157. SamWhited

    Jitsi doesn't like my home network for some reason, that's new.

  158. Guus

    I popped out for a second

  159. Guus

    still, you should've been able to join that Meet.

  160. Guus

    ... you know that, having worked at their stuff :)

  161. SamWhited

    one would think, but I've already forgotten most of how it works.

  162. SamWhited

    You're the expert, fix it Guus!

  163. Guus

    if you're trying to use their public service, it should pretty much work out of the box, if you're not blocking outbound traffic and have a relatively sane NAT setup.

  164. Guus

    other than that, Damian is a hero :)

  165. SamWhited

    Indeed

  166. SamWhited

    I'll dig in later, catching the bus now. My network hasn't changed to my knowledge, so maybe it just needs a kick

  167. Ge0rG is just "participating" in a Lync conference call. The audio works for the first second, and then there is silence.

  168. Maranda

    Nice

  169. Ge0rG

    I wonder when my coworkers will figure it out.

  170. Ge0rG

    I'm only 19mins and two reboots in.

  171. Maranda

    Wasn't Lync one of those dismissed things by Microsoft btw?

  172. Maranda

    (it should be "Skype for Bussness" now)

  173. Ge0rG

    Maranda: it's not dismissed, just renamed.

  174. Maranda

    🤣

  175. Maranda

    Touche

  176. Ge0rG

    Third reboot, same result.

  177. Maranda

    Keep going, try with a kick or two since you're at it 🤣

  178. Guus

    Board meeting time

  179. ralphm bangs gavel

  180. ralphm set the topic to

    XSF Board Meeting | Logs: http://logs.xmpp.org/xsf/ | Agenda https://trello.com/b/Dn6IQOu0/board-meetings

  181. ralphm

    0. Welcome and Agenda

  182. ralphm

    Who do we have?

  183. Guus

    Nyco apologised earlier

  184. Guus

    Martin is showing as offline to me

  185. Guus

    Mattj?

  186. MattJ

    Here

  187. ralphm

    ok good

  188. Guus

    shall we?

  189. ralphm

    So I don't see many changes from last time around. Did we progress on any of the items in our Trello, and discuss it?

  190. MattJ

    I think we skipped last week because of lack of people, and I don't remember the week before that

  191. Guus

    we skipped the week before too

  192. Guus

    no progress that I am aware of, although one new item was added

  193. Guus

    the comm-wt card

  194. ralphm

    Ok

  195. Ge0rG

    And I forgot to write to an ML regarding a SPAM SIG :(

  196. ralphm

    Ok, let's start with these then

  197. ralphm

    1. Communications Team

  198. Guus

    if desirable, we can do a post-mortem on the summit/fosdem too

  199. ralphm

    (ok)

  200. ralphm

    I'm with Kev on this, that we don't really need this to be formalized

  201. ralphm

    (at the moment)

  202. ralphm

    I'd rather just first see some stuff happening, without the yak shaving

  203. Guus

    I think it helps to get stuff under way

  204. ralphm

    How?

  205. Guus

    by defining that team members get access to things they want to use to produce stuff.

  206. ralphm

    I strongly believe we should get away from needing some kind of formal structure to be able to do things.

  207. MattJ

    I'm hearing that, combined with "we need to restrict access to XSF resources"

  208. Ge0rG

    you might want to approve the persons named for comm-wt access to github, before they go away in disdain.

  209. Guus

    well, getting access to resources are a pain. I've done that plenty of times recently, and it was always painful and took way to long for comfort.

  210. ralphm

    Ge0rG: wait what?

  211. Guus

    i'm not seeing harm in us formaly defining the team again, and defining that its members get read/write access to our website and social media accounts.

  212. MattJ

    Same here

  213. Ge0rG

    ralphm: referring to JC's last email on that thread.

  214. MattJ

    I agree this shouldn't be so painful, but it appears to me that (re)formation of the team seems the least painful way for everyone involved

  215. Guus

    (also, the overhead of us doing this is minor)

  216. ralphm

    But since this is GitHub, why can't we start with pull requests (that don't require any additional permissions) and move from that? I think that's the point Kev tried to make. I'm not opposed to restart a formal team, but I don't see it being on the critical path.

  217. Ge0rG

    ralphm: the deployment cycle via PRs and editor bugging is long and frustrating

  218. Guus

    ralphm: because it's keeping the team-members dependent of others, and thus less effective. Also, I really don't see the harm in not giving them access.

  219. Guus

    why not facilitate them if it benefits us all?

  220. Guus

    as I said, most of the work for board on this is this debate. After that, it's asking iteam to add some permissions to existing resources.

  221. Guus

    which I assume will take them about five minutes to implement

  222. intosi

    iteam can do that, words need only be said.

  223. ralphm

    Ok, so let me get this straight. The deployment cycle of the site itself is cumbersome and would require an Editor to approve?

  224. intosi

    It does not.

  225. Guus

    no, it now requires <unnamed entity> (usually me) to approve.

  226. intosi

    It requires someone with commit permissions on the website repo to merge the PR.

  227. MattJ

    Which is currently (iteam + editors)?

  228. Guus

    no, it's undefined, as far as I know

  229. Ge0rG

    wouldn't that be another reason to name it "comm-team"?

  230. Guus

    (defining that comm-team gets access would be another nice side-eff... what he said)

  231. Guus

    What would be arguments for _not_ doing this?

  232. ralphm

    Well, it is just a bit of work to get going.

  233. ralphm

    You need a charter, a chair and Members (only).

  234. Guus

    I believe that's neglible.

  235. Guus

    the charter has been proposed, as has the chair and member list

  236. Guus

    (on list, and in @commsteam)

  237. Guus

    @commteam, not @commsteam

  238. MattJ

    I'm also inclined towards "let's just do it"

  239. ralphm

    Ok. I didn't see the chair mentioned. I have no strong objections to having a formal team with only those Members having those accesses for publishing at xmpp.org

  240. Guus

    the chair was mentioned in the MUC only (being JC)

  241. ralphm

    Aha

  242. Guus

    it'd also be good if that team can use our social media accounts (which currently is limited to Twitter, I think?)

  243. Ge0rG

    It would be great to mention all the many XMPP-deploying projects on @XMPP

  244. ralphm

    So the proposal is as the e-mail with JC as chair.

  245. ralphm

    The only objection I have is Simon not (currently) being a Member.

  246. ralphm

    AFIAK

  247. Guus

    ah, good catch.

  248. Guus

    bylaws do say: "Participation in Teams shall be limited to elected Members of the Corporation."

  249. ralphm

    Indeed that was my point

  250. ralphm

    So assuming Guus is motioning this with the exception of Simon, I'm +1

  251. Guus

    he's not listed as a member, nor do find an application for him

  252. Guus

    Ok, I'm motioning.

  253. ralphm

    For the record: Daniel Wisnewski (Daniel_W) Jan-Carel Brand (jcbrand) as Chair Nicolas Vérité (Nyco) Severino Ferrer de la Peñita (Seve or SouL) Charter: The team's mission is to inform the XMPP community and interested parties on news and recent developments within XMPP ecosystem.

  254. MattJ

    So he can still submit PRs/etc. but a member will need to approve

  255. MattJ

    Still an acceptable situation if you ask me

  256. ralphm

    Right

  257. Guus

    agreed

  258. ralphm

    I left off the second part of the charter as it seems superfluous and not alligned with the bylaws w.r.t. openness of the team itself. Of course others are still free to contribute through the Team.

  259. ralphm

    aligned even

  260. Guus

    agreed (as was agreed on in the mail htread)

  261. ralphm

    Yeah, just repeating here for the record

  262. Guus

    I'm +1 to the formation of the WT, the chair and member list, and the charter.

  263. ralphm

    +1 (again)

  264. MattJ

    +1

  265. ralphm

    Yay

  266. Guus

    also, I motion that we grant read/write permission to the source code repository that holds the XMPP.org website code to all members of the Communications work team - which should reflect future team changes.

  267. ralphm

    The motion carries.

  268. ralphm

    Guus: I don't think we need to formalize that per se. But I'm sure Kev and intosi can make sure of the necessary technical requirements for the Team to carry out their task.

  269. Guus

    I'd simply like to avoid a situation where everyone looks at eachother and wonders who is allowed access, by making this explicit.

  270. intosi

    Is there a list of github user ids for the members of the comms team?

  271. Guus

    intosi: I'll ask them to provide that

  272. intosi

    Ta.

  273. ralphm

    I'm sure there will be other things that access is required for and I'm not sure if asking Board every time is needed. If iteam would give out too many creds, that'd be the day.

  274. intosi

    Sure. Comms team can just ask iteam.

  275. Guus

    I was going to follow up with one more similar motion, for social media accounts :)

  276. ralphm

    I trust iteam to handle this all properly.

  277. Guus

    ok, works for me

  278. ralphm

    I think that's enough for today, right?

  279. MattJ

    wfm

  280. Guus

    yup

  281. ralphm

    2. Date of Next

  282. ralphm

    +1W

  283. ralphm

    3. Close

  284. MattJ

    wfm

  285. ralphm

    Thanks all!

  286. ralphm bangs gavel

  287. Guus

    works for me

  288. Guus

    thank you

  289. ralphm set the topic to

    XSF Discussion | Logs: http://logs.xmpp.org/xsf/ | Agenda https://trello.com/b/Dn6IQOu0/board-meetings

  290. Ge0rG

    No spam team this week.

  291. Ge0rG

    *sniff*

  292. Guus

    sorry Ge0rG

  293. MattJ

    Ge0rG, haven't we discussed the spam team 50 times already?

  294. ralphm

    We actually rejected it because there wasn't a feeling of needing a team of SIG for doing things around that topic.

  295. Ge0rG

    MattJ: with the Spam Fighting Manifesto, there is some traction in the open to improve default server configs, write modules to block domains etc.

  296. ralphm

    Great.

  297. Ge0rG

    but that said Manifesto is made by a single volunteer who lacks time to incorporate feedback from server admins already.

  298. MattJ

    Github repo and get people to submit PRs - if unable to merge PRs, find the people that you imagine in the SIG and grant them access to merge the PRs

  299. intosi

    An impromptu SIGish, if you like.

  300. Ge0rG

    I had hoped for something more official than my private github account, but okay.

  301. ralphm

    why?

  302. ralphm

    Please, and I have asked this before, why can't you do any of this without 'something more official'?

  303. ralphm

    I don't understand this at all.

  304. Ge0rG

    I'm doing it already, as far as my time permits. Sorry for bothering you.

  305. intosi

    comm team effective immediately, with the list JC mailed?

  306. intosi

    I note that Winfried replied that he wanted to be on it as well.

  307. Ge0rG

    intosi: without Simon AFAIU

  308. ralphm

    Ge0rG: it is not about bothering. It is about not feeling for doing a lot of bureaucracy just because.

  309. Ge0rG

    ralphm: ah, I see

  310. Guus

    intosi, in my understanding, Winfried was in, Simon out.

  311. jonasw

    couldn’t we put the anti spam manifesto under the xsf account?

  312. ralphm

    intosi: yes. I think Winfried already has permissions, by the way, and I wasn't sure if was actually applying for the team or merely contributing.

  313. Guus

    intosi, scratch that

  314. Guus

    we didn't include Winfried when voting (I overlooked that)

  315. ralphm

    jonasw: why?

  316. jonasw

    ralphm, dunno, if georg is unhappy with having that under his account

  317. ralphm

    The TLS manifesto was also not an XSF activity and worked out fine

  318. jonasw

    I’m just throwing ideas in the room

  319. MattJ

    jonasw, as I understand it some people are opposed to the manifesto. It's just going to cause way more debate and just be counter-productive when it's totally unnecessary in the first place

  320. jonasw

    yeah, true

  321. ralphm

    and what MattJ said applied to that manifesto, too

  322. MattJ

    Right

  323. intosi

    I have added JC as maintainer of the comm team group

  324. MattJ

    I'm not opposed to the manifesto, and Ge0rG, if you need help, I and others would be more than willing I'm sure

  325. intosi

    Invited Nÿco, kept Winfried in it for now, and require two more github user names, for SouL and for Daniel_W.

  326. MattJ

    Thanks intosi

  327. Guus

    SeveFP <-- Soul's github

  328. intosi

    Guus: ta

  329. Ge0rG

    MattJ: thanks for your offer!

  330. winfried

    In is ok for me (I am in a meeting, can't communicate all the time)

  331. intosi

    I suggest that the comm teams adds its own page to the website, and optionally fixes its own wiki page :)

  332. Guus

    it's nice how we keep having the same conversation in two different MUCs :)

  333. Guus

    I just proposed that (sans the wiki, but that's a good idea)

  334. intosi

    Don't know in which muc that happened, but I'm not in it :)

  335. Guus

    commteam@muc.xmpp.org

  336. intosi

    Ah, not something I'd want to be in on a permanent basis currently :)

  337. Guus

    you not being in it made it more impressive

  338. intosi

    Great minds, fools, there was something with a comparison there.

  339. marc

    Ge0rG, fixed my ejabberd implementation but your landing page doesn't work on "firefox klar"

  340. marc

    Ge0rG, no button is clickable

  341. marc

    Ge0rG, can you verify and at least add an issue please

  342. marc

    works fine on firefox (desktop version)

  343. waqas

    marc: Is that different than Firefox Focus?

  344. marc

    waqas, it seems, I don't know if they use a different engine, probably not

  345. marc

    maybe some javascript issue but at least some javascript code works

  346. marc

    the xmpp URI is correct etc.

  347. marc

    waqas, oh, thought you're asking about "normal" firefox

  348. waqas

    https://play.google.com/store/apps/details?id=org.mozilla.focus&hl=en and https://play.google.com/store/apps/details?id=org.mozilla.klar&hl=en look practically the same, both were updated on February 5, and only Focus is listed on the Mozilla page: https://play.google.com/store/apps/developer?id=Mozilla&hl=en

  349. marc

    waqas, does it work for "firefox focus"?

  350. waqas

    I was asking what Klar was, as I've only heard of Focus

  351. marc

    waqas, AFAIK it's just the german version of firefox focus

  352. waqas

    Ah, right you are: https://support.mozilla.org/en-US/kb/difference-between-firefox-focus-and-firefox-klar