XSF Discussion - 2017-11-13

  1. Link Mauve

    “01:44:52 Zash> Link Mauve: Can't you just chmod a+r the xep84 node, like people want with omemo?”, and willingly give away your JID in your MUC presence so that other participants can find it? ^^

  2. Link Mauve

    I just added a page for an event we’ll be part of, https://wiki.xmpp.org/web/Capitole_du_Libre_2017

  3. Link Mauve

    It’s next weekend, Guus, could we get a stock of stickers in time for this event? :x

  4. Link Mauve

    I’ll try to get other things as well, for example a French translation of Arc’s Prosody booklet, and maybe stickers of ours.

  5. Guus

    Link Mauve: sure. Send me a shipping address.

  6. edhelas

    Guus o/

  7. edhelas

    we need to prepare a bit for T-DOSE as well :)

  8. Guus

    edhelas: yes. Suggestions?

  9. Arc

    holy crap, im on live internet video

  10. Arc


  11. jonasw

    congrats or so? :-)

  12. Guus

    If I click this link, everyone better had their clothes on, Arc...

  13. jonasw


  14. jonasw

    Arc, which one of that is you? :-)

  15. Guus

    He's the one sitting down.

  16. Guus

    did he bring his own french coffee press?

  17. Guus

    oh, shh, he's looking!

  18. Guus

    oh, false alarm.

  19. Guus

    I could do this all day! ;)

  20. Arc

    i brought it here to stay here, but yes. its for tea

  21. Guus

    woah - quite a bit of lag.

  22. Arc

    yea, a few seconds

  23. jonasw


  24. jonasw

    building a tea-machine with CLI interface so that I can write down tea-instructions as makefiles is on my long-term todo :)

  25. Guus

    jonasw, you need a (different) hobby.

  26. Guus

    I mean, CLIs are so passé.

  27. edhelas

    jonasw why not write a XEP for that ?

  28. jonasw

    Guus, trivial to make anything on top of CLI :)

  29. Zash

    Guus: NO U

  30. edhelas

    Makefile Over XMPP

  31. jonasw

    edhelas, HTCPCP over XMPP?

  32. Arc


  33. Zash

    Arc: Moldy tea? Sounds awful

  34. Guus

    He just put away his laptop - probably won't respond :)

  35. Guus

    I like my new Big Brother role :)

  36. jonasw

    Big Guus

  37. jonasw

    Guus, edhelas, also, a CLI tool could just be a frontend to an XMPP client sending commands ;-

  38. jonasw

    Guus, edhelas, also, a CLI tool could just be a frontend to an XMPP client sending commands ;-)

  39. Guus


  40. arc


  41. arc


  42. Link Mauve

    Guus, do you prefer one in France or in the UK? (CC: mathieui.)

  43. mathieui

    one what?

  44. Link Mauve

    Shipping address for SCAM material.

  45. Link Mauve

    Or maybe it would be more prudent to ship directly to the friend hosting us?

  46. Link Mauve

    Guus, how long would it take to ship them, generally?

  47. Guus

    Link Mauve: I don't care. You should account for a couple of days delivery time

  48. mathieui

    well, unless it arrives on saturday, it doesn’t change much

  49. Link Mauve

    For me it has to arrive on Friday morning at the latest.

  50. Guus

    Link Mauve: I have little experience - I think Daniel received his after two days. That was NL->DE

  51. Ge0rG

    wouldn't it be cheaper and faster to order material from a local printing shop?

  52. Guus

    Link Mauve: no promises other than that I will send it today.

  53. edhelas

    Guus you'll bring goodies for T-DOSE BTW ?

  54. Guus

    Georg: I have them pre-printed, shipping doesn't cost much.

  55. Link Mauve

    Ge0rG, it may be, and we’re going to print some ourselves as well (French conference).

  56. Ge0rG

    Link Mauve: what's the difference, accents all over "XMPP"? :D

  57. Link Mauve


  58. edhelas


  59. edhelas


  60. daniel

    printing something that is comparable in quality is pretty expensive. while shipping is not

  61. Ge0rG


  62. Link Mauve

    Ge0rG, JabberFR is pretty well-known around here, so we’ll probably stickers with our logo, and also a translated flyer of Arc’s Prosody setup.

  63. Link Mauve

    Ge0rG, JabberFR is pretty well-known around here, so we’ll probably print stickers with our logo, and also a translated flyer of Arc’s Prosody setup.

  64. Ge0rG

    X̲͍̬̺M̰͔͉̼̹̲̯Ṕ̭ͭ̿P̟͉̤̮̥̺̍ :D

  65. Link Mauve

    Guus, ok, so: Emmanuel Gil Peyrot 19 Victoria Road CB4 3BW Cambridge United Kingdom

  66. Link Mauve

    Guus, ok, so: Emmanuel Gil Peyrot 19 Victoria Road CB4 3BW Cambridge, Cambridgeshire United Kingdom

  67. Guus

    Link Mauve: just the stickers, or also Arc' leaflet (I've got English ones only)

  68. Link Mauve

    Hmm, maybe a few English ones, I’m sure there will be some non-French people around.

  69. Link Mauve

    mathieui, any idea for the amount? I’ve never been to that particular event.

  70. Guus

    I've sent Daniel 100, I think.

  71. Guus

    If you don't put them on the table all at once, they'll last longer :)

  72. Guus

    (stickers, that is)

  73. mathieui

    100 sound good

  74. zinid

    what stickers/

  75. zinid

    is it a meme or something here?

  76. Link Mauve

    Guus, sure. :)

  77. Link Mauve

    zinid, the things you can put on the back of your laptop.

  78. Ge0rG

    I didn't know XMPP had stickers support

  79. zinid

    but I don't have a laptop...

  80. zinid

    fridge is ok?

  81. Ge0rG


  82. Guus

    100 stickers, 20 english folders (the folders are bulkier to send - I'd be happy to send as many as you want, but if you don't need them, I'd rather prevent have to find suitable envelopes :)

  83. Guus

    will that do, Link Mauve?

  84. Link Mauve

    Guus, yup, should be fine, thanks!

  85. Guus

    ok. Will send them this afternoon.

  86. mathieui


  87. Guus


  88. Guus

    Jonasw the library list on the website isn't validated against the list of known platforms - is that by design?

  89. Guus

    (I'm not bothered by it - just checking)

  90. Ge0rG

    I think we had a discussion about "platforms" for libraries being harder to pinpoint than for applications

  91. jonasw

    Guus, isn’t it? I thought it was

  92. jonasw

    ah right

  93. Ge0rG

    MattJ: what about making MAM response messages type=headline?

  94. Guus

    Link Mauve: two envelopes on their way.

  95. Link Mauve

    Yay, thanks!

  96. moparisthebest

    jonasw: is there ever more to making tea than boiling water and dropping a tea bag in there for a bit?

  97. Zash

    Dangerous statement to make near any brits

  98. mathieui

    teabags? blasphemy

  99. Link Mauve

    My monocle just fell.

  100. SamWhited

    > My monocle just fell. I'm really glad that stereotype translates across the channel as well as it doors across the Atlantic, that made me laugh.

  101. Ge0rG

    I just wondered if that stereotype will be still understandable in the next 10-20 years, but then I realized that most of us know monocles from TV anyway, not from real-life.

  102. Zash

    I should get monocle

  103. mathieui

    My monocycle just fell

  104. Zash


  105. jonasw

    moparisthebest, :-O

  106. jonasw

    (a) timing, and (b) teabags?! heresy!

  107. moparisthebest

    Zash: I refrained from mentioning I then ice it and put in copious amounts of sugar for fear of British retribution

  108. jonasw

    moparisthebest, that has its uses and would also be amazing if automated

  109. dwd

    moparisthebest, Making tea is just boiling water, a teabag, a mug, and some milk once ready.

  110. Ge0rG

    ewww, milk!

  111. dwd

    moparisthebest, I know some people like to faff about a bit with loose tea and so on, but it basically tastes the same whether decent tea leaves are in a bag or not.

  112. dwd

    Ge0rG, Yes, of course. Otherwise Wrong.

  113. moparisthebest

    some people put milk in coffee here, never seen that with tea, it's funny how little things like that are totally different in different places

  114. Ge0rG

    I've tried tea with milk once or twice, and it was a shocking experience.

  115. jonasw

    it really depends on the tea

  116. dwd

    moparisthebest, Totally wrong in some places you mean.

  117. jonasw

    some Assam with milk -- great. earl grey with milk -- eww

  118. moparisthebest

    is there more than one kind of tea? /sarcasm :)

  119. dwd

    jonasw, Ah, I was worried you were about to suggest that fruit infusions were tea.

  120. jonasw

    dwd, eww fruit infusions

  121. dwd

    jonasw, In which case, obviously, you'd have been Wrong many times over.

  122. moparisthebest

    I only ever drink warm/hot tea when I'm sick once or twice a year

  123. moparisthebest

    sweet tea though, that's a different story

  124. dwd

    moparisthebest, That is a bizarre American thing. Probably not helped by being inflicted by Lipton's Yellow Label and not any of the decent stuff.

  125. Ge0rG

    do two Wrongs make one Right?

  126. dwd

    Yellow Label is both the world's most popular tea, and simultaneously not sold in the UK because we like halfway decent stuff.

  127. moparisthebest

    is uh, 'sun tea' a thing elsewhere?

  128. dwd

    Ge0rG, Well, tea without milk is Wrong, and fruit infusions are Wrong, obviously, but fruit infusions without milk is curiously less Wrong than with milk.

  129. moparisthebest

    on a hot summer day we put tea bags and water in a clear glass jar and sit it in the sun all day

  130. moparisthebest

    then of course sugar and refrigerate it

  131. dwd

    moparisthebest, Not seen that one before. Can't really see why you'd want to do that over just boiling water through teabags. (Through would have the same effect of reducing the tannins, which you probably want for ice tea variants).

  132. mathieui

    I ordered iced tea in the US, was expecting the lipton peach soda, came out with half a gallon of bad tea with ice

  133. SamWhited

    > Probably not helped by being inflicted by Lipton's Yellow Label and not any of the decent stuff. That's the point, sweet tea requires a little bit of really crappy tea with your sugar! It's just how things work.

  134. SamWhited

    If you're using good tea you're doing it wrong.

  135. Ge0rG

    it looks like the controversy about tea is even larger than the controversy about MIX

  136. mathieui

    maybe this could go in the council agenda

  137. SamWhited

    Although, while Lipton is acceptable, you need Tetly if you want really good sweet tea. Trust a man who grew up in Georgia.

  138. jonasw

    what’s georgia

  139. jonasw


  140. dwd

    You have Tetley's?

  141. dwd

    jonasw, Country just south of Russia.

  142. SamWhited

    We are the home of sweet tea, comrade!

  143. Ge0rG

    In Soviet Russia, there was only bad Chinese tea.

  144. arc

    OK that's it, xmpp tea room must happen now

  145. SamWhited

    dwd: I'm not sure if it's the same, it's the generic bad black tea blend though, not any directrix type

  146. SamWhited

    *sigh* phones.

  147. SamWhited

    Can't type.

  148. dwd

    I think they only do the one thing. Presumably not advertised by Yorkshiremen in flat caps in Georgia, though?

  149. SamWhited


  150. arc

    That's actually kind of a brilliant thing for a con, instead of a plain old booth, host a tea room.

  151. dwd

    SamWhited, Same company, all right. Here it's https://www.tetley.co.uk/our-teas/our-full-range

  152. SamWhited

    Ah yup, that's the one. They're also very popular here (but only for iced and sweet tea)

  153. SamWhited

    Well, "here", I'm in Texas now and that's not really the same, sweet tea confuses places here.

  154. mathieui

    arc, offering hot tea at fosdem sounds like a plan

  155. Link Mauve

    arc, the Quadrature du Net does that at CCC.

  156. dwd

    arc, We could use Lloyd's XMPP code for controlling the kettle, too.

  157. mathieui

    (tbh the concept has been quite pioneered by the "quadrateahouse" at each international CCC event)

  158. Link Mauve

    It’s a great way to get random people to discuss together.

  159. Zash

    XEP it!

  160. SamWhited

    I thought teapots all implemented HTTP via RFC 2324?

  161. Zash

    Extensible Tea and Coffee Protocol

  162. Link Mauve

    Zash, sounds way too syn and ack ridden.

  163. Zash

    Pfft, HTTP, who uses that anymore?

  164. Ge0rG

    can't we just have somebody publish the beverages?

  165. Zash

    <stream:tea to="ge0rg">...

  166. arc

    Bulk tea is cheap. Dirt cheap

  167. Ge0rG

    Is that RFC 2324 compatible?

  168. SamWhited

    I do like this idea … xmpp tea room. I hope I can actually make it to the summit, we'll see. Trying to convince my company, but probably only can if one of my talks is accepted.

  169. dwd

    Erm. Did disco#item elements once have an "action" attribute?

  170. dwd

    Probably one for ralphm, Kev.

  171. dwd

    Perhaps zinid too. Basically it's before my time anyway.

  172. Zash


  173. Zash

    Looks like a yes

  174. Zash

    No older versions than 2.1 available? :(

  175. dwd

    Ah, part of disco#publish. Also supported by Openfire. :-)

  176. Zash

    Wait, what

  177. dwd

    Zash, Our Pubsub code has various bits in saying the spec isn't stable and might change, too. It's like archeology sometimes.

  178. Zash

    Version 2.4 (2008-06-06) Per Council consensus, removed disco-publish feature (implementations should use publish-subscribe instead).

  179. dwd

    Zash, Yup. Openfire did make that jump. But still has disco#publish support knocking about - mosly, I suspect, where it hasn't been removed.

  180. Zash

    I wasn't even aware this had existed

  181. jonasw


  182. dwd

    Zash, I was dimly aware of disco#publish. Enough to vote to strip it while I was on Council, apparently.

  183. Zash

    Before my time

  184. dwd

    Zash, I stood for the first time in 2007. Fiercly contested election, back then: https://wiki.xmpp.org/web/Board_and_Council_Elections_2007

  185. ralphm

    dwd: but even disco push didn't require an action attribute

  186. ralphm

    Don't think we had that attribute

  187. ralphm

    Oh, just saw the old version. Never mind

  188. dwd

    ralphm, Yeah. I don't think Openfire has this quite right even so, but it's probably been lying in wait for years.

  189. Ge0rG

    Can we have that "MUC" vs "Group Chat" vs "Conference" discussion on standards@ instead of on twitter, please?

  190. moparisthebest

    I suppose it depends on whether your intended audience is twitter users or not

  191. jonasw

    which discussion?

  192. daniel

    can we have the official twitter account RT please?

  193. daniel

    not sure whats wrong about twitter

  194. daniel

    you should be glad i didn't post that on instagram

  195. zinid

    daniel: 140 characters

  196. daniel

    zinid, 280!

  197. zinid


  198. daniel

    and it's a poll

  199. daniel

    you should just say a or b

  200. jonasw

    daniel, what twitter?

  201. jonasw

    I have no idea what you’re all talking about (that’s wrong with twitter)

  202. zinid

    what is twitter?

  203. daniel


  204. Kev

    daniel: "Chat" if you want the least confusing to users, in my experience. Although we call them Rooms in Swift.

  205. Ge0rG

    My twitter client didn't show it's a poll - nor the options.

  206. daniel

    is your twitter client an rss reader?

  207. Ge0rG

    no, it's an actual twitter client, from a third party.

  208. zinid

    damn, I don't even have an account in twitter

  209. Ge0rG

    Why isn't there an XMPP-based microblogging plattform?

  210. mathieui

    zinid, get on with the times, grandad

  211. Ge0rG

    Kev: isn't "Chat" ambiguous, in that it might mean a "private chat" as well?

  212. Kev

    It is.

  213. zinid

    mathieui: so you call this bloated crap with retarded limitiations "the times", okay...

  214. edhelas

    Ge0rG you mean a Web client that implement the XEP 277 ? :-°

  215. daniel

    ironically i think twitter is mostly millennials and older?

  216. daniel

    i think kids these days are on other platforms

  217. Ge0rG

    edhelas: no, I mean something that has a name and that will be recognized by most people in this MUC

  218. edhelas feels disapointed :(

  219. jonasw

    Ge0rG, does that ambiguity matter?

  220. Ge0rG

    jonasw: I'm not sure.

  221. edhelas

    Ge0rG you don't think that Movim fullfill your requirements ?

  222. Ge0rG

    I've been in that position ("how to name MUCs") some months ago, and I went with "Group chat"

  223. Ge0rG

    edhelas: I tried it once, but it didn't allow me to see NSFW content because of prosody :(

  224. Ge0rG

    And then my phone died or somesuch.

  225. edhelas


  226. Holger

    Movim kills phones?

  227. Ge0rG

    No, unrelated event.

  228. zinid

    so sad story

  229. Ge0rG

    Yeah. Sad.

  230. daniel

    there is nsfw content on movim?

  231. Ge0rG

    daniel: I never was able to figure that out.

  232. edhelas

    who merged this feature in Movim, that's you Holger ?

  233. zinid

    daniel: mostly gayish

  234. daniel


  235. edhelas

    there's no NSFW content on Movim, there's NSFW content on the XMPP network :o

  236. zinid

    daniel: do you really want a link? :)

  237. Ge0rG

    I think that the more important thing than naming a MUC is naming a JID.

  238. daniel

    i'd prefer not to ender my xmpp credentials into some website...

  239. daniel


  240. zinid

    and rename XEP back to JEP, plz

  241. mathieui

    Ge0rG, "address"

  242. edhelas

    there's 2 difficults things in XMPP, naming things and send bold characters in chat

  243. Ge0rG

    mathieui: "user address"? "account address"? "jabber address"?

  244. Ge0rG

    mathieui: my address is a postal one.

  245. Ge0rG

    edhelas: LOL

  246. jonasw

    oh dear, I’m afraid of this becoming a meme :)

  247. Ge0rG

    jonasw: fortunately, there is no way to send links to XMPP messages.

  248. jonasw

    isn’t there?

  249. jonasw


  250. mathieui

    jonasw, btw, didn’t you have to make some tweaks to CCG after implementing it in poezio?

  251. jonasw

    mathieui, just pushed them, waiting for the docker hub build

  252. Ge0rG

    jonasw: we don't even have solved message IDs.

  253. jonasw

    Ge0rG, before solving message IDs, let’s solve message routing

  254. Ge0rG

    jonasw: right.

  255. daniel

    is this scheduled before or after world peace?

  256. jonasw

    difficult to compare things to NaN

  257. ralphm

    Ge0rG: XMPP is all about distributed systems and decentralization. Why would we need to discuss particular topics in one place only?

  258. jonasw

    ralphm, especially in this situation having a central discussion venue is vital.

  259. jonasw

    it helps the various distributed entities to have a common view on things

  260. jonasw

    which is important for interop

  261. Ge0rG

    ralphm: nothing wrong with that, except that most of our distributed venues are not based on XMPP anyway.

  262. daniel

    well twitter is mostly about user engagement and marketing. it's not meant to have a meaningful discussion

  263. ralphm

    I don't see a problem with that. Each communication medium has different properties. I like e-mail for long-form discussions, but don't like it for IM-style usage (which I see a lot in corporations). Twitter is nice for public snippets. XMPP is good as a basis for building IM systems.

  264. ralphm

    There are a bunch of things I like about Slack

  265. ralphm

    (which could totally be build on top of XMPP)

  266. Ge0rG

    if only we had a proper slack clone on top of XMPP

  267. ralphm

    Go fund a team to make it

  268. ralphm

    It is not a technology issue at all

  269. Ge0rG

    I know.

  270. Ge0rG

    People are reinventing IM all the time, and burning venture capital on the way.

  271. ralphm

    I disagree this is a waste of money. Also I don't care how VCs spend their money.

  272. ralphm

    Slack has a useful proposition that makes it more viable for teams like mine at work to collaborate than if we had to use any current offering based on XMPP.

  273. ralphm

    Of course there are a bunch of things I don't like about it.

  274. ralphm

    E.g. I think their channels should be more like MIX, with orthogonal streams of data

  275. ralphm

    Now I have to ~force~ educate people to not put notifications of integrations into the same channel as where discussions happen.

  276. Ge0rG

    ralphm: I'm pretty sure there is a market niche for a Slack clone that is deployed on premise and can be configured to follow archival / access policies.

  277. ralphm


  278. Ge0rG

    And XMPP has all the building blocks for that.

  279. moparisthebest

    what does it give you other than the bots for integration? (over a XMPP MUC)

  280. Ge0rG

    But people rather do it on top of HTTP REST JSON.

  281. ralphm

    moparisthebest: what Slack offers to us?

  282. moparisthebest


  283. Ge0rG

    moparisthebest: Slack is easy to "deploy"

  284. zinid

    Because xmpp is complex

  285. moparisthebest

    in the same way as it's easy to just use a public xmpp server and muc I guess, that can't be the reason?

  286. Ge0rG

    moparisthebest: have you ever used Slack? And any of the XMPP client offerings?

  287. Ge0rG

    moparisthebest: there is a world of a difference, UX wise

  288. ralphm

    moparisthebest: consistent UI across platforms, including desktop, easy to configure integrations (even by non-technical people), search, rich(er) formatting, snippets, mentions, team-mentions, (automatic) snoozing, reminders, (custom) emoji, emoji responses, unlimited editing/deletion of previously sent messages.

  289. moparisthebest

    no I haven't used slack that's why I'm asking :)

  290. Ge0rG

    moparisthebest: try it out for a week or two

  291. moparisthebest

    unlimited editing/deletion of previous messages isn't a security problem?

  292. moparisthebest

    especially for companies that must keep records or whatever?

  293. ralphm

    moparisthebest: at this point in time, anyone developing XMPP clients or servers, or participates in the standards discussions thereof, should use Slack to understand what we're up against.

  294. ralphm

    I am tempted to disqualify people's opinion based on that.

  295. Ge0rG

    ralphm: I agree

  296. jonasw

    good, we need a team which does that kind of research

  297. jonasw

    because I cannot take any client serious which only runs in the browser.

  298. Ge0rG

    Slack is the mirror that's held up to XMPP client developers to show how they failed.

  299. moparisthebest

    what if it's against your religion to use proprietary software? :)

  300. ralphm

    moparisthebest: enlightenment?

  301. Ge0rG

    moparisthebest: then pay somebody to use it and look at them doing it.

  302. ralphm

    Ge0rG: I think limiting that to just XMPP client devs is not fair.

  303. Ge0rG

    ralphm: maybe, but in the context of the XSF it is a very valid limitation.

  304. moparisthebest

    other's opinions are helpful anyway, I can open slack, but if I'm not using it daily or in a team or whatever I can't really get a good feel over what it does better or worse

  305. ralphm

    jonasw: totally, you at least need browser, iOS and Android. Desktop apps seem to be a thing still. I heard that my Mac using colleagues use the desktop version of Slack. I just run it in the browser.

  306. moparisthebest

    plenty of people seem to like riot.im, I found it confusing after a few minutes and quit

  307. Ge0rG

    I'd even go as far as to say that renaming from JSF to XSF and thus shifting focus away from software to protocol was a bad idea.

  308. ralphm

    moparisthebest: I agree it helps if you need it for work or another shared project where there is regular usage.

  309. SamWhited

    You should start a new JSF, the XMPP Software Foundation (also called XSF).

  310. ralphm

    Ge0rG: that's disingenious. The XSF/JSF has always been about standards.

  311. Ge0rG

    ralphm: I think that it's rather easy to port a modern web app onto the Electron operating system, so having [web, ios, android] might suffice.

  312. SamWhited stops being unhelpful and goes back to actual work.

  313. Ge0rG

    SamWhited: I'd call it the Jabber Software Foundation and focus on IM client interop.

  314. ralphm

    Ge0rG: we didn't do much about software, but named it like Apache. The furthest we got was hosting JabberStudio.

  315. moparisthebest

    ah yes that's what I want, a chat client that uses more resources than my java IDE

  316. moparisthebest

    oh wait I have that, it's called running a windows 7 VM just for lync :'(

  317. Ge0rG

    ralphm: my point is: we are sorely missing an entity that has a focus on software and UX of XMPP

  318. Kev

    Ge0rG: Me :p

  319. moparisthebest

    I thought I'd seen a lot of focus on UX lately Ge0rG

  320. Ge0rG

    moparisthebest: you mean the Styles XEP?

  321. moparisthebest

    not everyone has to focus on the same thing

  322. Ge0rG

    Kev: then you know the answer to daniel's initial question on how to name MUCs?

  323. moparisthebest

    well yea that's a good example, but also UX considerations in other XEPs

  324. Ge0rG

    moparisthebest: I'm trying to push UX in the XSF for around two years now, and more often than not I hear the "we are about protocols, not UIs" response.

  325. Kev

    Ge0rG: I use 'rooms'. Users tend to use different terms based on their backgrounds - we come up against 'chats' quite a lot.

  326. moparisthebest

    ah I don't think I've seen that

  327. ralphm

    'room', 'chat room', 'group', 'group chat', 'or 'channel' are just fine

  328. ralphm


  329. Ge0rG

    ralphm: just fine for whom?

  330. moparisthebest

    I don't know why you couldn't make a few UX standards recommendations in XEPs, some do right?

  331. ralphm

    Ge0rG: people. WhatsApp uses group. Slack uses channel.

  332. Ge0rG

    ralphm: if you are helping a friend configure an XMPP client over the phone, do you tell them to look for rooms, chatrooms, group chats, channels or all of the above?

  333. ralphm

    Facebook had a thing called Rooms

  334. Ge0rG

    I think it wouldn't hurt to standardize on a name for those, and even to provide some i18n lookups for that name.

  335. moparisthebest

    Ge0rG, you ask what client first and go from there I guess

  336. ralphm

    So this is the problem right, unlike Slack or WhatsApp, there's not a single entity creating XMPP-based IM clients.

  337. Ge0rG

    moparisthebest: so I need to remember the client->name mapping for all commonly used XMPP clients?

  338. jonasw

    moparisthebest, haha

  339. moparisthebest

    or, you do what you suggested Ge0rG , and write one-client-to-rule-them-all using electron or whatever the hot framework of the day is, and have everyone use that?

  340. Ge0rG

    moparisthebest: that's not what I suggest.

  341. jonasw

    moparisthebest, I find it amazing how ignorant people can be about the software they’re using and how hard it can be to figure out what they use.

  342. Ge0rG

    moparisthebest: I suggest to have a common glossary for client developers, see https://wiki.xmpp.org/web/Usability/Glossary

  343. jonasw

    they just use jabber and it takes time to figure out they run xabber/yaxim/conversations.

  344. intosi

    Ge0rG: if I'm helping friends, I tend to use the wrong terms anyway, as different operating systems use different names for everything, and different translations vary wildly. My friends and family can usually cope just fine :)

  345. Ge0rG

    intosi: I'm sure you have smart friends & family :)

  346. moparisthebest

    if I'm helping a friend set up xmpp over the phone I'll be telling them to install conversations and then go from there most likely

  347. moparisthebest

    if you are walking them through at that level you have to know more details meh

  348. ralphm

    I would probably not ever suggest 'conference' as a name, unless it was about AV ones

  349. moparisthebest

    Ge0rG, I think a glossary is a great idea

  350. Ge0rG

    It's called Multi-User Chat.

  351. ralphm

    I have a hard time helping my family and friends anyway, because they use apps in Dutch, which I hate.

  352. Ge0rG

    or short "MUC" - a term that really nobody outside of the core XMPP community can relate to

  353. intosi

    ^ with ralphm on that

  354. Ge0rG

    If we had a database of jabber-related terms and their translations, you could use those when on support duty :P

  355. moparisthebest

    that's a different problem actually, translating technical terms sounds rough

  356. ralphm

    That's why you shouldn't use them

  357. ralphm

    Group or Room translate really well

  358. Zash

    ralphm: And then comes "MIX"

  359. moparisthebest

    maybe you provide 1 UI since the user doesn't care, it's all a 'Group' to them

  360. moparisthebest

    I am basically the last person to try to make UI decisions, I can tell if I like something or not, but that's it

  361. moparisthebest

    I'm the only one I know personally that's been using KDE for the last 11 years, for example

  362. ralphm

    Zash: for MIX the concept is still Room

  363. ralphm

    You can even make MIX rooms be accessible via MUC if you put in the engineering effort

  364. dwd

    ralphm, I think MIX is Channel, but yeah.

  365. ralphm


  366. Ge0rG

    dwd: is that different from MUC?

  367. ralphm

    So yeah, we have MUC Rooms and MIX Channels

  368. dwd

    Ge0rG, MUC talks about Rooms. MIX talks about Channels, as I recall.

  369. Ge0rG

    dwd: why should we use different terms? To a user, it's all the same

  370. ralphm

    It doesn't matter much, and if you really care, we could change it before going to Draft

  371. ralphm

    You'd have to rename some elements, though

  372. Ge0rG

    I'm not sure you need to map XEP language into the UI 1:1

  373. moparisthebest

    what does what the protocol calls protocol-level things have to do with UI ?

  374. Ge0rG got to go. read you later.

  375. dwd

    Nice if they align, though. Makes it much simpler for newcomers.

  376. Flow

    lovetox: regarding your question yesterday: Does that https://github.com/xsf/xeps/pull/541 suit your needs?

  377. jonasw

    Flow, any reason you went for a patch-level version number increase?

  378. jonasw

    I would like to get semver-like versioning in XEPs, if possible.

  379. jonasw

    this feels more like a 0.3 than a 0.2.1 to me

  380. lovetox

    flow yes this solves the problem, but then it seems even harder to solve the use case about invalidating keys

  381. lovetox

    because now you have no notifications at all anymore after sub

  382. lovetox

    i think if we design this, we should at least make distributing and retracting keys have a good workflow

  383. lovetox

    i think there is no way around having something like a metanode

  384. lovetox

    that tells us new keys are there and old keys should be rewoked

  385. Flow

    lovetox, the service still would send notifications if a new item is added, no?

  386. lovetox

    em no?

  387. lovetox

    last published item is set to on_sub or never

  388. Flow

    jonasw, I don't care that mutch, but then again, it rather had the deferred status change a patch level

  389. Flow

    lovetox, isn't send_last_published_item a per node setting?

  390. jonasw

    Flow, did you mean "I’d rather have had the deferred status change a patch level"?

  391. Flow

    and, I think a PubSub server would still send notifications on publish, regardless of the value of send_last_published_item

  392. Flow

    jonasw, that yes :)

  393. lovetox

    but we dont receive that publish if we are not online

  394. jonasw

    Flow, I’m not set on the version change level for deferrals, really. I’m not sure if minor or patch makes more sense.

  395. Flow

    lovetox, why not?

  396. jonasw

    I’m happy with changing it to patch, nobody else voiced concern about that yet.

  397. Flow

    …if your service stores offline messages

  398. jonasw

    Changing normative language should, imo, most of the time be minor level or major level, depending.

  399. lovetox

    flow this is seems like totally new territory

  400. lovetox

    now suddenly the xep depends on offline message, and mam

  401. Flow

    lovetox, also you could still query the node, that is how it was meant to be used primarly

  402. lovetox

    i thought we polling xeps are bad

  403. Flow

    You shouldn't really try to rely on having received the latest status of the node

  404. Flow

    lovetox, depends on the case i'd say

  405. jonasw

    polling is annoying

  406. Flow

    all generalization are bad

  407. jonasw

    that’s not true! ;-)

  408. lovetox

    Flow, it seems you try really hard to make your approach at this work

  409. Flow

    lovetox, I haven't heard an alternative

  410. Flow

    you said something like the "do the OMEMO approach", but I think OMEMO also uses PEP?

  411. lovetox

    i did that at least 3 times now

  412. Flow

    lovetox, must have missed it

  413. lovetox

    introduce a metanode

  414. lovetox

    clients publish valid keyids to it

  415. Flow


  416. lovetox

    clients receive these on every start

  417. Zash


  418. lovetox


  419. lovetox

    then clients do a poll at node:keyid if the dont know the key id

  420. lovetox

    and get the key

  421. lovetox

    that way you have minimal traffic, 2 or 3 keyids is small

  422. lovetox

    and only have to pull keys if they are new to you

  423. lovetox

    at the same moment you could say, every key that is not inside the metanode, has to be retracted

  424. Flow

    I had to think about it a bit more, and I'd like to see a list discussion regarding this

  425. Zash

    What's the context here?

  426. Flow

    Zash, OX sending the full OpenPGP pubkey on every available presncen if you don't configure the node to not do so

  427. Flow

    lovetox, but it doesn't sound like a bad idea

  428. Flow

    please post on list

  429. lovetox

    i already have haha, but i dig it out again and write something :)

  430. Zash

    So you'd like something like how the avatars and stuff works?

  431. lovetox

    yes shit, why didnt i thought of it, thats way easier way to describe it

  432. lovetox

    its exactly like avatars work

  433. jonasw


  434. jonasw

    I wonder whether there should be an informational XEP for bulk storage in PEP

  435. jonasw

    so that people writing XEPs have something to base their work on

  436. lovetox

    Flow, we really should get this going, people are annoyed by OMEMOs perfect forward secrecy

  437. Zash

    jonasw: Like 222/223, or what do you mean by bulk?

  438. moparisthebest

    I also think OX is perfect for building Onion XMPP on top of, fyi

  439. jonasw

    Zash, I don’t think that 222/223 address the costs of sending the whole contents on each connection.

  440. jonasw

    (so maybe an informational one which addresses the issues of on_presence + large amounts of data)

  441. jonasw

    (e.g. base64’d crypto keys, avatars, …)

  442. Zash

    I believe you can have notifications that don't include the payload.

  443. jonasw

    which doesn’t help because PEP/PubSub often only allows for one item

  444. jonasw

    (okay, the @id is probably enough if it’s randomly generated by the server to know that you need to poll the data…)

  445. Zash

    Did I post to the list about me thinking that PEP should have even fewer required features, and a separate extended PEP for more advanced usecases?

  446. Zash

    Like 222/223, if those weren't Informational

  447. jonasw

    Zash, yes, I thnik so

  448. zinid

    Zash: why? users will anyway annoy you to death and you will end up implementing advanced version

  449. Zash

    zinid: They already do and I'm already on it.

  450. moparisthebest

    could just discover PGP keys via DNS instead, at least it's already spec'd out and implemented

  451. zinid

    Zash: so why do you need this separation then? :)

  452. Zash

    Because I like separation

  453. Zash

    A bunch of the simpler things like Mood and Activity don't require all the features mandated by PEP, so why not make the minimal baseline be smaller so new implementations won't have to figure out what MUSTs they can ignore.

  454. Zash

    And a more clear target for stage two, or whatyoucallit

  455. Ge0rG

    And then people end up storing their bookmarks in private XML.

  456. Zash

    It'll inevitably end up being all of xep-60 in the end anyways

  457. Zash

    Ge0rG: make bookmarks depend on 22[23] (whichever was the private one)

  458. Zash

    -xep 222

  459. Bunneh

    Zash: Persistent Storage of Public Data via PubSub (Informational, Active, 2008-09-08) See: https://xmpp.org/extensions/xep-0222.html

  460. Zash

    -xep 223

  461. Bunneh

    Zash: Persistent Storage of Private Data via PubSub (Informational, Active, 2008-09-08) See: https://xmpp.org/extensions/xep-0223.html

  462. SamWhited

    After our discussion this morning I came over all British and had my afternoon cup of of Earl Grey with milk in it… dwd is a corrupting influence.

  463. moparisthebest

    I always assumed Earl grey was French because of jean-luc Picard liking it so much

  464. SamWhited

    I always assumed it was from space.

  465. waqas

    Something something Plan 9

  466. SamWhited

    Did Picard work at Bell labs? I don't remember that being a part of the series.