XSF Discussion - 2020-05-24


  1. xnamed^

    https://xmpp.org/extensions/xep-0045.html#example-9 What about adding to the room features: <feature var='http://jabber.org/protocol/muc#admin'/> When user affiliation is "owner" or "admin", and <feature var='http://jabber.org/protocol/muc#owner'/> When user affiliation is "owner"?

  2. MattJ

    xnamed^: what for exactly? Because this can change, so you would have to keep checking it

  3. MattJ

    Or track your affiliation, in which case it's not needed anyway

  4. flow

    so I have xep313 latest (mam:2) and 0.5 (mam:1) side by side open, what exactly caused the namespace bump. IIRC it was only a tiny thing, but the examples all look the same to me (besides s/mam:1/mam:2/)

  5. flow

    so I have xep313 latest (mam:2) and 0.5 (mam:1) side by side open, what exactly caused the namespace bump? IIRC it was only a tiny thing, but the examples all look the same to me (besides s/mam:1/mam:2/)

  6. Zash

    That stanza-id was required

  7. flow

    ahh, § 3.5, not sure why we did a namespace bump for that, that appears to have been a mistake in hindsight

  8. Zash

    So that clients knew they could count on stanza ids being archive ids

  9. flow

    right, but that would have been simply annouced via an extra feature, no?

  10. flow

    right, but that could have been simply annouced via an extra feature, no?

  11. Zash

    Dunno. Maybe.

  12. jonas’

    flow, yes, could’ve been

  13. MattJ

    +1, mistake in hindsight once that became the only change

  14. pep.

    Is there a quick way to emulate / write a module for a server that automatically returns an error for sasl? I'm trying to see which clients display the error text (and which don't)

  15. MattJ

    Probably 3-4 lines in prosody

  16. pep. reading through the doc

  17. MattJ

    module:hook("...event...", function (event) return event.origin.send(st.stanza("failure", {xmlns=...}) end)

  18. MattJ

    event can be figured out from https://prosody.im/doc/developers/events or mod_saslauth code

  19. lovetox

    you really shoud think about that, lets just add a feature strategy

  20. lovetox

    especially on experimental XEPs

  21. lovetox

    because then we could argue, there is always only a :1 we never need :2 because we just do a <feature name="mam:2">

  22. lovetox

    and this does work for all changes, not only for additional btw

  23. lovetox

    you really should write down a guidline when to use features and when to use a namespace bump

  24. lovetox

    and then people should honor it, if they like it or not

  25. lovetox

    so its consistent

  26. Kev

    It works whenever the behaviour as allowed but not required by the previous spec.

  27. Kev

    It works whenever the behaviour was allowed but not required by the previous spec.

  28. flow

    lovetox> and this does work for all changes, not only for additional btw I doubt that is true, there are certainly valid reasons for a version bump

  29. lovetox

    a feature works for something thats a feature, hence OPTIONAL

  30. lovetox

    but thats not the reason they are used lately

  31. lovetox

    its, oh there is some sensible change that XEP needs, everybody needs that change, but there is some impl out there that does not want to upgrade, so lets make the required change OPTIONAL to not break some client

  32. lovetox

    and i get this notion for a XEP that is in Draft or whatever final status there is

  33. lovetox

    but not for Experimental

  34. lovetox

    but anyway i can also parse 27 features before i know the server does what i want, it doesnt matter much

  35. emus

    Hello guys, just a reminder to put your news into the upcoming XMPP Newsletter for this May if you have anything to say or already published! 📬 We draft the Newsletter on Github and will archive to the Wiki once released: https://github.com/xsf/xmpp.org/pull/706 Merge requests are welcome! 👋

  36. Neustradamus_

    emus: How we can edit a draft?

  37. emus

    You should basically be able to create a merge request. If you don't want you can just comment a section. But merge request would be easier

  38. wurstsalat

    Neustradamus_: pull requests on github are the quickest way emus: have you had a look at mine?

  39. emus

    Ehm, I though I merged everything, or?

  40. emus

    Okay, thats really odd and happened before - I don't get ANY email or notficiation about those pull requests (anyway thanks for you work!

  41. emus

    Okay, thats really odd and happened before - I don't get ANY email or notficiation about those pull requests (anyway thanks for you work!)

  42. Link Mauve

    Hi, how can one add @xml:lang support to XML Schema?

  43. Link Mauve

    Do we have any such example?

  44. Link Mauve

    Apparently I have to define xmlns:xml='http://www.w3.org/XML/1998/namespace', which goes against the XML spec AIUI.

  45. emus

    wurstsalat - thanks, is merged

  46. Link Mauve

    Indeed it does, xmllint rejects it.

  47. wurstsalat

    emus, maybe you have to adjust notifications for your own fork?

  48. emus

    will check, but thats weird acutally 🤨️

  49. Zash

    Re the Bookmarks thread, I remembered this exists: https://xmpp.org/extensions/xep-0194.html

  50. mathieui

    wasn’t this intended for contacts?

  51. Zash

    Seems to be

  52. lovetox

    Zash, not sure how you think this has anything to do with the thread

  53. lovetox

    this tells your contacts that you joined a MUC

  54. Zash

    Why is that so different from telling your own clients?

  55. pep.

    Might be a subset of it? :/

  56. lovetox

    Zash we already have a mechanism to tell our own clients

  57. lovetox

    its called Bookmarks

  58. Zash

    :(

  59. xnamed^

    MattJ, as you said "track your affiliation" it's needed to let client know that to show actions for those features which can be accessed without having to join the MUC

  60. flow

    Link Mauve> Hi, how can one add @xml:lang support to XML Schema? Isn't xml:lang implictly everywhere?

  61. Zash

    flow: I'd expect to be implicitly allowed everywhere since it's a namespaced attribute

  62. flow

    Zash, yep, me too

  63. pep.

    What if there's a limited number of occurs, don't we need semantics to say "it's fine if there's a different @xml:lang"?

  64. Zash

    That seems more likely

  65. pep.

    What do you mean

  66. Link Mauve

    flow, yes, but how do you make the schema aware that an element can be duplicated if the @xml:lang is different?

  67. Zash

    Can schema even express that?

  68. Link Mauve

    I can’t find any hint that they do.

  69. pep.

    Zash, that was my comment as well, is it possible? :p