XMPP Summit - 2026-01-29


  1. emus

    Good morning, Im at the venue already 😌 When you get in you can just head right and take the stairwair to the 4th floor. Further right should be an elevator. If nothing works, just aks at the reception, they are very kind. See you later!

  2. emus

    Room is 4.02 (just ahead from the end of the stairways)

  3. Edwin

    I assume the summit will be held in Dutch this year, given the location.

  4. Link Mauve

    Just like it was held in German last year? :p

  5. emus

    https://jabbers.one:5281/file_share/019c08b1-f177-784c-afa3-57c28d39e19a/zb2rhdW9cgvoGYLWDD9sKRL6WtP975sWpNN7AhZrtmqmhWBTp.jpg

  6. emus

    > I assume the summit will be held in Dutch this year, given the location. I practiced last night watching Spongebog in Dutch ✔️

  7. Arne

    Some delay 😬

  8. zeank

    Am I the only one in the Jitsi room?

  9. emus

    > Am I the only one in the Jitsi room? Likely

  10. Rémi

    I’m not there yet, want me to join?

  11. emus

    > I’m not there yet, want me to join? You can, but we havent started. So time to relax

  12. zeank

    Now that I typed in the name by hand, I'm waiting to be let in ... (seems at least like there's an existing room)

  13. Rémi

    Yeah yeah I was just proposing if you need to check something. Tell me when to join

  14. Daniel

    I'm in the lobby. You didn't leave already did you?

  15. singpolyma

    good morning, all

  16. lissine

    Good evening

  17. Edwin

    Wifi password: HVNB-P@rtn3rs#23

    👍 2
  18. mathieui

    all topic links should be up-to-date now

    👍 1
  19. singpolyma

    The fewer RFCs we have the less modular it is, and the less things can be clearly optional (I know we can just mark every section of the RFC as MAY...) and we end up with the opposite problem we have now, instead of "too many extensions" it becomes like OpenPGP "too big RFC"

  20. Rémi

    Shoot you’ve started already! I thought you’d mention it here 🥲

  21. singpolyma

    just barely started

  22. singpolyma

    right. I think eg replacing sasl1 with sasl2 might be a reason to do RFC

  23. goffi

    emus, Is that you who rite the points of each speaker in real time? It's super useful, thank you.

  24. goffi

    emus, Is that you who write the points of each speaker in real time? It's super useful, thank you.

  25. emus

    Nope, and I am happy someone does it ❤️

  26. goffi

    OK, so thank to whover does that.

  27. emus

    Please continue with that. I won't be able to do all at once

  28. goffi

    OK, so thanks to whoever does that.

  29. Rémi

    I’ll try to help taking notes but the sound detection online is a bit too low I guess so it’s hard to follow far participants

  30. emus

    > I’ll try to help taking notes but the sound detection online is a bit too low I guess so it’s hard to follow far participants Thats already a great help!

  31. debacle

    Slightly off-topic, but still interesting to most, I guess: I just got a affirmative email by Wikimedia, that we most likely can get their wonderful room again for an XMPP sprint. Date must still be arranged with them. It will be one weekend of 8..10, 15..17, 29..31 of May or 5..7, 12..14, 19..21 of June. Further discussion here: xmpp:sprints@muc.xmpp.org?join Preliminary wiki page here: https://wiki.xmpp.org/web/Sprints/2026-05_Berlin

    👍 1
  32. singpolyma

    I'm not sure we're talking about XMPP 2.0 as though a new RFC would make us "more standard"... we already have an IETF RFC standard. we can revise it, but it won't be any "more" standard than the existing one

  33. Rémi

    For those online, disable Also wik’s microphone it’s a lot better! (Otherwise there’s echo)

    ❤ 1
  34. goffi

    emus, I won't be able to do the Forums lighning talk today, tomorrow should be fine.

    👍 1
  35. goffi

    and email gateway is actually the same talk.

  36. goffi

    as well as standard@ mailing list ;)

  37. emus

    ah, can you update?

  38. goffi

    sure

    ❤️ 1
  39. singpolyma

    one RFC per XEP!

  40. emus

    > one RFC per XEP! The pad is yours 😉

  41. mathieui

    (be careful to not wipe the pad :p)

    🫣 1
  42. goffi

    I suggest that we take the occasion to rename XMPP to eXtensible Message and Pubsub Protocol.

  43. emus

    Regarding sponsoring: What about other funds, like https://www.opentech.fund/funds/internet-freedom-fund/ I believe there used to be also a German internet security fund

  44. jabber

    https://datatracker.ietf.org/doc/draft-melnikov-sasl2/

  45. zeank

    Or this maybe (for funding): https://www.sovereign.tech

  46. Thilo Molitor

    https://www.ietf.org/archive/id/draft-melnikov-sasl2-02.txt

  47. Thilo Molitor

    https://datatracker.ietf.org/doc/draft-melnikov-sasl2/

  48. Daniel

    is there a call for adoption already?

  49. Edwin

    https://calls.disroot.org/XMPPSummit28 for those who haven't found the video stream yet,

    🙏 2
  50. egnun

    I can have an eye on the computers 😉

  51. Rémi

    How long will likely be your break? 2 hours? Please keep us posted when you start again 🙂

    👍 1
  52. emus

    1hr

  53. emus

    Please come down to the cafertia

  54. emus

    Wait

  55. emus

    Food is down stairs in the cafeteria

  56. emus

    Starting in ~10

  57. Edwin

    A/V should be back.

  58. singpolyma

    not so much controversial as everyone wants it to mean something different ;)

  59. singpolyma

    note that hierarchical discovery of services (including MUCs) has nothing to do with MUC server. This is disco XEP. MUC server just happens to implement this but it is not tied to that

  60. singpolyma

    and also obviously we need to have implementations that allow non-techical users to create their own MUC server, but that's a totally separate consideration from spaces

    ✔️ 1
  61. emus

    singpolyma, shall I raise this to the discussion?

  62. singpolyma

    I think at least my first point is relevant, that we have this protocol and it's not tied to MUC server as was said

  63. Rémi

    Are we discussing the original XEP or the PR?

  64. singpolyma

    the XEP has been rewritten completely a few times so I'm not sure

  65. MattJ

    From the room: "there is no PR" :)

    👍 2
  66. singpolyma

    does that require a "root MUC" / lobby MUC or really the app can discover that this MUC is in a space from any given MUC. so you get invited to any MUC in the space and poof the app finds the others

  67. singpolyma

    discord actually doesn't have a lobby. though I think Matrix does

  68. gnemmi

    And so does Telegram

  69. gnemmi

    No matter how you look at it, you need an entry point to discover the spaces inside it

  70. emus

    singpolyma gnemmi - topics to raise? Please expose clear to me then I will raise

  71. Syndace

    Audio disappeared

  72. Syndace

    back

  73. debacle

    Lip reading ftw.

    😅 2
  74. Rémi

    Can people come back to help on the Pad? It’s pretty empty atm 😬

  75. MattJ

    On it

    ❤️ 2
  76. singpolyma

    wildcard certs from LE are easy, that's all I use from them

  77. emus

    singpolyma let me know when to expose to the round

  78. Rémi

    > From the room: "there is no PR" :) > I _just_ realized that when joining I had opened my note with questions about XEP-0372: References + PR 🤦🏻‍♂️ Sorry for the confusion ahah

  79. emus

    All good 🙂

  80. singpolyma

    MattJ: for a gateway / component that wants multiple domains for muc services is it finally time for https://xmpp.org/extensions/xep-0225.html ? ;)

  81. MattJ

    I'm actively working on it, the past few weeks

    🎉 1
  82. singpolyma

    Discord does not make you join all the room on a server. very far away from that there is a list of "default channels" in a discord server that new users see and users can choose which ones they see or don't

    💯 1
  83. Rémi

    Maybe someone could display a Discord screenshot / recording / live demonstration for people who don’t use it?

  84. Rémi

    Maybe someone could display a Discord screenshot / recording / live demonstration in the room for people who don’t use it?

  85. Rémi

    I feel some misunderstanding on the minimal goal for an “iterative approach” (btw +1 for that)

  86. singpolyma

    https://upload.mboa.dev/c39dffc4-fc98-4152-8e38-59bb20fbbf3f/Screenshot_2026-01-29_07-59-05.png

  87. singpolyma

    discord screenshot for you

  88. gnemmi

    Spaces are a key feature for any project (regardless of its nature), working group, teams, and even events like congresses, conferences, summits and even meetings (like Fosdem itself).

  89. gnemmi

    Any software project would greatly benefit from having different spaces for “Core”, “UI”, “UX”, “Proposals”, “New Features”, “Bug Reports”, “Support”, “Off Topic”.

  90. gnemmi

    Any software project would greatly benefit from having different spaces for “Core”, “UI”, “UX”, “Proposals”, “New Features”, “Bug Reports”, “Support”, “Off Topic”. Fosdem itself could have made use of such feature, creating a “Fosdem” MUC containing a space for each of the “Tracks”, that in turn would contain a space for each of the events that would take place in each track. And the same can be said for the upcoming DebConf26.

    💯 1
  91. gnemmi

    That's my user case

  92. Rémi

    > I feel some misunderstanding on the minimal goal for an “iterative approach” (btw +1 for that) > What I meant is if we go the iterative route (let’s do something simple now and improve later), I don’t think everyone has the same end-user features in mind.

  93. gnemmi

    A door to access other doors that lead me to the topic I want to be involved

  94. larma

    singpolyma, if you deselect a room in the browse channel list, will you: - still receive mentions in that room? - still be displayed as a member of the room when opening the member list of the room?

  95. larma

    singpolyma, if you deselect a room in the browse channel list, will you: - still receive mentions in that room? - still be displayed as a member of the room when someone else opens the member list of the room?

  96. Rémi

    > I feel some misunderstanding on the minimal goal for an “iterative approach” (btw +1 for that) > What I meant is if we go the iterative route (let’s do something simple now and improve later), I don’t think everyone has the same end-user features in mind. Edit: for the first step*

  97. gnemmi

    Matrix got it so well that projects like Fedora and Kde are using it for that

  98. larma

    I think this checkbox is not about "being a member in the room", but rather "show the channel in the list"

  99. larma

    https://upload.xmpp.ipora.org/file_share/0697b5b4-5929-7b44-877d-e901104d714b/6ff02663-2a37-46f5-8c2b-0461f42921db.png

  100. singpolyma

    I can't test the mentions question right now, but I expect yes. but you don't show in the participant list I just checked that

  101. larma

    the wording is even about "hiding" and they say you still get mentioned

  102. larma

    > I can't test the mentions question right now, but I expect yes. but you don't show in the participant list I just checked that hmm, I didn't check with another user, but I just opened a server which has language-specific channels where all of them are hidden by default. I made a few visible just to check and they all have the exact same member list as the #general channel on the same server. Maybe it's different per server?

  103. Rémi

    I just checked on a Discord server and when checking the list of members per channel I could see different numbers… but I’m not an expert and maybe it just changed in the fex seconds it took me to switch (though unlikely)

  104. Kev

    (Work meeting, haven't been listening) There's differences in Discord between public channels that people can include or not include in their sidebars, and private channels that are managed by (often self-assigned) roles (hats), and membership is limited to the relevant roles/individuals.

    💯 1
  105. singpolyma

    there are also "community servers" and "personal servers" which work differently

  106. larma

    yes, we already established that a direct affiliation sync across all rooms of a space would not allow to replicate all of Discord's features.

  107. Kev

    I had a protospec about this, written up in my Obsidian vault, which Obsidian sync wiped out and lost all history for, in one of my rare annoyances with Obsidian, based on top of https://github.com/swift/protoxeps/blob/master/acl.md and https://github.com/swift/protoxeps/blob/master/related-entities.md

  108. Kev

    (Not helpful particularly, just sharing my annoyance)

  109. singpolyma

    MUC PEP time

  110. emus

    Sorry, is it about this XEP? https://xmpp.org/extensions/xep-0469.html

  111. MattJ

    No

    👍 1
  112. MattJ

    I'm not aware of any XEP

  113. emus

    thanks

  114. singpolyma

    MAM doesn't even have a way to fetch a single item by ID, does it?

  115. MattJ

    singpolyma, https://xmpp.org/extensions/xep-0313.html#example-11

  116. singpolyma

    oh ok. not sure why I thought that one was missing

  117. MattJ

    It was a later addition, only 6 years old

  118. MattJ

    https://blog.prosody.im/multi-user-chat-gets-rich/

  119. Rémi

    That synchronized “yes” was amazing

  120. singpolyma

    Daniel: for avatars in MUC the participants also you can only get the hash of avatar updates from presence using the vcard thing right? But do you take that hash and fetch if via pep now?

  121. lissine

    From a UX perspective, how are you going to deal with perfect forward secrecy in encrypted chats? For example, if I join an OMEMO-encryped group chat, is the expectation that I'm able to read existing pinned messages?

  122. Daniel

    singpolyma: I was primarily talking abitf about the room itself

  123. singpolyma

    Right, I understand that. but you said you were getting rid of vcard everywhere and participants sort of need it to so I was curious

  124. Daniel

    Ah yes. I think I misspoke on that part

  125. Edwin

    tel: 32168

  126. edhelas

    o/

  127. dwd

    In other news XEP-0316 already exists.

  128. MattJ

    Marvellous

  129. egnun

    Bless you.

  130. Edwin

    Dzięki

  131. Rémi

    Do we need password recovery if we go passwordless? (Legitimate question)

  132. MattJ

    Depends who "we" is. Some deployments may go passwordless, but some will probably always use passwords.

  133. singpolyma

    Need account recovery somehow

  134. singpolyma

    Passwordless arguably needs it even more

  135. MattJ

    Yeah

  136. Rémi

    Right

  137. larma

    https://xmpp.org/extensions/xep-0389.html also has account recovery

  138. egnun

    > Dzięki Nie ma problemu

  139. mathieui

    maybe we can send a vcard to the server to setup recovery info

  140. singpolyma

    IBR already asks for this kind of info. Email address etc

  141. singpolyma

    What we lack is a command to trigger recovery flow from in an app or what

  142. singpolyma

    What we lack is a command to trigger recovery flow from in an app or whatever

  143. mathieui

    "<topic> is quite simple" famous last words

    🙃 1
  144. Rémi

    I had a question regarding XEP-0372: References, which could be addressed in the revision (it’s just a detail, no need to talk about it here IMO): Would ‘begin’ and ‘end’ work with messages with alternate encodings (XEP-0481: Content Types in Messages, e.g. Markdown + plain)? Is it possible to have multiple ranges like hreflang works?

  145. singpolyma

    XHTML-IM has a similar question. Would make sense to be able to specify eg an html class name or id that goes with a particular reference maybe

  146. emus

    Kindly help me and start you comments to raise at the summit with: SUMMIT: <your thought>

    👍 1
  147. Rémi

    +1 on the “replace with mentionned name with name from address book”. I don’t know about the implementation details but I believe a lot of apps send an id and the mention looks different for everyone (e.g. WhatsApp)

  148. emus

    Kindly help me and start your comments to raise at the summit with: SUMMIT: <your thought>

  149. singpolyma

    Yeah this is similar to how I reformat xmpp iri links already

  150. mathieui

    attention works for muc, it’s just something like "@all" currently with no options :p

  151. singpolyma

    Yup

  152. Edwin

    https://ralphm.net/blog/2019/09 for those playing along at home.

  153. ralphm

    Actually https://ralphm.net/blog/2019/09/09/fastening, but I expect no more blog posts in September 2019

  154. egnun

    Break? Time for a dance.

  155. emus

    > Break? Time for a dance. Please!

    😅 1
  156. emus

    I heard people talking about Disco already 🥲

  157. jp

    > Break? Time for a dance. XMPP Tiktok Dance

  158. egnun

    Like this? (needs sound on) https://inv.nadeko.net/watch?v=iu6SYk9diWg

  159. mathieui

    > I heard people talking about Disco already 🥲 You sure it wasn’t Cisco?

  160. zeank

    I like what I hear, but I don't like it means "someone" has to implement that for ejabberd if I want to have oAuth too

  161. zeank

    I like what I hear, but I don't like it means "someone" has to implement that for ejabberd too, if I want to have oAuth too

  162. zeank

    I like what I hear, but I don't like it means "someone" has to implement that for ejabberd, if I want to have oAuth too

  163. emus

    zeank, if you want to speak up please indicate with "SUMMIT"

  164. zeank

    Thanks! No intent at this point.

    👍 1
  165. mathieui

    > WebTransport provides support for multiple streams, unidirectional streams, and out-of-order delivery. It enables reliable transport via streams and unreliable transport via UDP-like datagrams.

  166. singpolyma

    Yes. WT is just another name for QUIC. But it has a handshake to trick webservers into thinking it might be http3

  167. moparisthebest

    WebTransport been implemented and working in xmpp-proxy for years, your server can s2s to my mucs with it

  168. moparisthebest

    converse-tauri and fluux with my patch can c2s to xmpp.beer with it ;)

  169. zeank

    Have fun at the dinner! Bon appetit! Thanks for this great day!

  170. Intosi

    Dinner: meet at the NH hotel at 18:50, or at Bruxelles Central @19:05. We will take metro 5 to Hermann-Debroux at 19:13, then bus R75 to Overijse, after which we'll have a 7 minute walk to the restaurant Auberge Bretonne.

  171. mathieui

    and if we meet angry farmers we can have a 45 minutes walk to Auberge Bretonne

  172. debacle

    For those *not* going to the official dinner: Some of us go to eat at Soleil Afrique, 19:30 hrs, which one or the other person might remember from last year.

  173. debacle

    https://www.openstreetmap.org/node/4666814252

  174. emus

    We'll take you from the NH Collection Brussels Grand Sablon hotel to the restaurant, L'Auberge Brétonne (Brusselsesteenweg 670 – 3090 OVERIJSE), for our traditional annual XMPP Standards Foundation Dinner. We leave from the Hotel at 06:45 pm. Please be on time if you did signed-up and intend to take the taxi. https://www.openstreetmap.org/node/12076635837 You can also come to Centraal line 5 at 07:00 pm and join the group. If you don't come by taxi, please let us know so we don't wait for you. Dinner starts at 8 pm. Please be especially on time here.

  175. emus

    https://jabbers.one:5281/file_share/019c0a83-4834-7d74-8db1-9c7013be9c32/3d441dea-eb7e-4a55-8b4d-5d5f0a5bb99b.png

  176. mremond

    I had sign up for a departure from NH hotel, but will go directly from centraal line 5

    👍 1
  177. mremond

    [Corrected] I had signed up for a departure from NH hotel, but will go directly from centraal line 5

  178. egnun

    Enjoy your meals, everyone!

  179. egnun

    Don't drink too much. It's only the first day! See you all tommorrow. 😜

    😵‍💫 1
  180. emus

    *Servings for Day 2* Today I bought bananas in addition to our booked servings. Anything you missed and I can get for tomorrow? I heard already: - mandarines

  181. jsautret

    I'll meet at centraal too

  182. goffi

    ralphm: as much as I wanted to, I'm in no condition to come. Sorry for the inconvenience.

    ❤️‍🩹 1
  183. ralphm

    Be well Goffi!

  184. goffi

    Thanks

  185. goffi

    I don't know if I can make it tomorrow, we'll see.

  186. mathieui

    emus: you're talking about taxi, but we're taking the bus right?

  187. emus

    Yes, sorry. That point was outdated when zi wrote the text.

    👍 1
  188. emus

    Yes, sorry. That point was outdated when I wrote the text.

  189. Holger

    I'll meet you at Centraal.

  190. emus

    We are sitting in the foyer 👋

  191. emus

    of NH

  192. Arne

    on the way to Nh now

  193. emus

    We are leaving to the Centraal soon ❗

  194. Arne

    Eh

  195. Arne

    I'm at NH now

    ✔️ 1
  196. MattJ

    Leaving NH in a couple of minutes

  197. Arne

    Where are you?

  198. jsautret

    I'm at centraal line 5

  199. emus

    Coming, had some detour

  200. Holger

    Seems I ended up at the wrong side of the station.

  201. emus

    Holger where are you now?

  202. Holger

    Starbucks.

  203. emus

    https://jabbers.one:5281/file_share/019c0af1-81a1-7c37-8029-6a516c1d99ac/zb2rhhX4yQQPU9uX1NWqb98ui5CUNcrjcVKtNAoxKXvZ2BTXm.jpg

  204. emus

    Really till the end

  205. Holger

    Ok.

  206. Holger

    In the M5 now.

  207. ralphm

    We're waiting upstairs

    👍 1
  208. emus

    (Sublevel still though)

  209. emus

    > *Servings for Day 2* > > Today I bought bananas in addition to our booked servings. > > Anything you missed and I can get for tomorrow? > > I heard already: > - mandarines Reminder for your wishes. Otherwise I can print more handouts if SCAM & Board allow

  210. mremond

    For me it was all great, no special wish

    🤩 1