XSF Discussion - 2022-01-11


  1. bung

    test

  2. Guus

    A+

  3. emus

    Hi Kev, I am wondering if you have notes left somewhere from previous GSoC applications. Additionally, I wanted to ask whether you have input regarding experiences in organsiation / mentoring?

  4. Guus

    emus: flow has also been involved in GSoC organizing in the past. If you haven't already, you might want to poke him too.

  5. Daniel

    emus: if you have a specific question feel free to ask me too. I've ran my own org for two years. But if you are just after some general information I think there are some ebooks out there on mentor first timer and org first timer

  6. Daniel

    I don't have the link anymore but I read those when I Frist started

  7. flow

    emus already reached out to me and I suggested that Kev may have some additional notes

  8. Guus

    Why is the certificate of this server trusted by xmpp.net: https://xmpp.net/result.php?domain=psynet.su&type=client While the certificate of this is not? https://xmpp.net/result.php?domain=igniterealtime.org&type=server Same cross-signed chain, as far as I can tell?

  9. Guus

    (note: the latter domain consists of two servers, which confuses things a little)

  10. emus

    Daniel: Thank you Daniel. I just wanted to ask for experience or anything (lessons learned) that one wanted explicit to tell or recommend. I need to prepare the application now

  11. jonas’

    https://github.com/xsf/xeps/pull/1144 do people really need to sign the CLA for this trivial stuff? cc @ board

  12. dwd

    My ex-board view: They don't need to "sign the CLA", but they do need to agree to the IPR policy somehow.

  13. jonas’

    that is sign the CLA in our case

  14. jonas’

    I think in *this* jurisdiction the change wouldn't have the necessary schöpfungshöhe in order to even qualify for copyright

  15. dwd

    Well, that's what I mean - the CLA automates the process, but if you can get a statement in any other form (including over IM) that should be sufficient.

  16. dwd

    "You OK with our IPR policy?" "Yeah".

  17. debacle

    About XEP-0060: When a JID, subscribed to a node with "Whitelist Access Model", is removed from the list, but not explicitely unsubscribed, should it get more items or not?

  18. MattJ

    Without looking at spec or code, I'm pretty sure that if it no longer has permission to be subscribed, it should be unsubscribed

  19. debacle

    MattJ IMHO, the spec is not entirely clear. At least not to me: > A node access model under which an entity may subscribe and retrieve items only if explicitly allowed to do so by the node owner (subscription requests from unauthorized entities are rejected). The first part backs your view, but the last sentence blurs the picture. (I'm not a native speaker of English, blame me!)

  20. MattJ

    Native speaker or not, I see no sense in allowing someone to subscribe, removing their permission and having them still be subscribed

  21. debacle

    MattJ Me neither! But it wouldn't be the first time, XMPP holds surprises :-)

  22. MattJ

    :)

  23. Holger

    This error is explicitly specified to be returned to non-whitelisted JIDs requesting items: https://xmpp.org/extensions/xep-0060.html#subscriber-retrieve-error-whitelist

  24. Holger

    This error is explicitly specified to be returned to non-whitelisted JIDs requesting items: https://xmpp.org/extensions/xep-0060.html#subscriber-retrieve-error-whitelist

  25. Holger

    > XMPP holds surprises :-) At least ejabberd does! Non-boring implementation.

  26. jonas’

    Holger, but that doesn't imply that they don't still *receive* items :)

  27. jonas’

    broadcasting items in reply to a publish is not, technically, retrieving

  28. MattJ

    Holger, what does ejabberd do in the case where a JID's access is removed?

  29. MattJ

    I'm fairly sure (without checking code) that Prosody will unsubscribe it

  30. MattJ

    But of course I may be wrong

  31. jonas’

    … let's better look now than tomorrow

  32. jonas’

    it does seem to unsubscribe

  33. Holger

    MattJ: Better ask debacle about the details 🙂 They stumbled over the behavior, and I won't be able to check before tomorrow.

  34. Holger

    > Native speaker or not, I see no sense in allowing someone to subscribe, removing their permission and having them still be subscribed Me neither, the desired/intended behavior seems obvious.

  35. debacle

    looks like a bug in ejabberd, if my colleagues did observe carefully