jdev - 2024-02-29

  1. lovetox


  2. lovetox

    i feel like received/displayed in the marker XEP is what most people need, just the info what was the last message received/displayed

  3. lovetox

    with the acknowledged functionality its different, there we need to store the flag for every message

  4. lovetox

    compared this to the Receipts XEP, which sends a receipt for *every* message. I feel this is not really useful.

  5. singpolyma

    Yes, I think storing just received/read up to is sufficient for anything I can presently imagine

  6. singpolyma

    larma: I just saw https://wiki.xmpp.org/web/Gsoc2024/Dino/Inline_link_preview and just wanted to say we have some features around this in cheogram android already and if this goes forward I'd love to collaborate so we end up using the same or similar formats

  7. kapad

    i want to report a typo error, at the xep documetation, - reading https://xmpp.org/extensions/xep-0092.html, at section `4. Determining Support` -> `Example 3` , the iq query miss the `type` attribute, so that leads to an error like, ``` <error type="modify"> <bad-request xmlns="urn:ietf:params:xml:ns:xmpp-stanzas"/> <text xmlns="urn:ietf:params:xml:ns:xmpp-stanzas" xml:lang="en">Missing attribute 'type' in tag &lt;iq/&gt; qualified by namespace 'jabber:client'</text> </error> ```

  8. lovetox

    about https://xmpp.org/extensions/xep-0461.html can i expect that fallback body start will always be 0

  9. lovetox

    and why again do we need to set a to=jid when we reply?

  10. lovetox

    what would i do with this information as receiving client?

  11. singpolyma

    No falback start is definitely not always zero

  12. MattJ

    kapad: thanks! I will try and prepare a patch this week if nobody else does.

  13. kapad

    ok. !

  14. lovetox

    singpolyma, how so? and what do you expect the receiving client to do?

  15. singpolyma

    To remove the part of the body between start and end from display if they support the feature in for

  16. lovetox

    but how does this make sense from a sending perspective

  17. lovetox

    seems your GUI misleads the user into writing a message that can never be displayed on the receiving side in the same way

  18. singpolyma

    Nothing in the fallback sections is composed by the user

  19. lovetox

    ok lets take a step back, please describe the situation where your client sends a fallback start that is not 0

  20. lovetox

    and how this can happen

  21. singpolyma

    I'll admit usually when something is before a fallback it's usually a different fallbrck (with r different for)

  22. singpolyma

    I'll admit usually when something is before a fallback it's usually a different fallbrck (with a different for)

  23. lovetox

    "different for" ?

  24. lovetox

    whats that

  25. lovetox

    why is there another fallback before another fallback

  26. singpolyma

    Like the fallback has a for= attribute to specify what it is a fallback for. Reply, reaction, oob, etc

  27. lovetox

    ok didnt see that, and what would be an example where another fallback is before the reply fallback?

  28. singpolyma

    First one I can think of is cheogram.com sends fallback text for extended addressing, which if present is always first in our case and reply fallback if presint would be second in that case

  29. singpolyma

    Fallback has its own xep I think replies just references it as something you could use if you want

  30. lovetox

    hm ok, i guess there is no way around for me to store the start then

  31. lovetox

    singpolyma, do you happen to know what the story with the "to" attribute on the reply is?

  32. lovetox

    if i have the stanza id, i know anyway for whom the message was

  33. lovetox

    if i dont have it .. not sure the jid matters

  34. singpolyma

    Yeah no idea. Maybe a sanity check?

  35. singpolyma

    Or maybe for notifications in MUC? I dunno

  36. singpolyma

    It makes it cheaper to tell if a reply is to you I guess, without implementing the whole spec or if you don't have logs

  37. lovetox

    i suspect its trying to help the case when you dont have the stanza id

  38. lovetox

    but full jid in muc ..

  39. lovetox

    not sure if occupant id would help

  40. lovetox

    you could also not know the occupant id

  41. lovetox


  42. lovetox

    is this maybe for crossposting replies to other chats?

  43. singpolyma

    Oh maybe

  44. lovetox

    then the receiver does definitly not have the stanza id, but he could draw the same UI, and put a jid or nickname in the top or something like that

  45. singpolyma

    Start a new muc with first message a replybto message in other MUC, like discord "threads"

  46. singpolyma

    Might even have the stanza id, if they are in both mucs, and now knows what context to look in

  47. lovetox

    good point

  48. kapad

    MattJ: is that how things work: https://ua3.anondns.net/im-storage/p/892560/9c04/fix-xep-0092-xml-typo.diff ?