XMPP Service Operators - 2025-11-05


  1. icebound.dev

    itsjustmyluck@xmpp.is --> PMing NSFW without consent

  2. icebound.dev

    not sure if the admin is here or not :)

  3. icebound.dev retracted a previous message, but it's unsupported by your client.

  4. icebound.dev retracted a previous message, but it's unsupported by your client.

  5. icebound.dev

    Sorry spoke too soon, someone thought they were funny pretending to be a fucking NSFW spammer, disregard my report.

  6. icebound.dev

    shit was meant to retract that not delete it locally 🤦‍♂️

  7. icebound.dev

    I am starting to realise that the servers I often get spam/abuse from are not in this channel, and are the ones which ignore reports :/

  8. broke

    you mean xmpp.is admin?

  9. broke

    is it a single person?

  10. broke

    there's something I wanna ask that seems to fit here the most, I asked this on the prosody chat and they pointed me to a very important thing. And that is server compliance with what restriction I am creating. I personally wanted less ports to be open and wanted to only support legacy ssl, meaning only tls (no starttls). But most servers seems to probe 5222 if there is an SRV with that, and if there is literally no 5269 or any port mentioned for xmpp-server, they reach me late because of doing starttls first. On clients I find the same thing as well. Clients always tries 5222 first. And servers always try 5269 first, Is there a reason to still do this?

  11. broke

    there's something I wanna ask that seems to fit here the most, I asked this on the prosody chat and they pointed me to a very important thing. And that is server compliance with what restriction I am creating. I personally wanted less ports to be open and wanted to only support legacy ssl, meaning only tls (no starttls). But most servers seems to probe 5269 if there is an SRV with that, and if there is literally no 5269 or any port mentioned for xmpp-server, they reach me late because of doing starttls first. On clients I find the same thing as well. Clients always tries 5222 first. And servers always try 5269 first, Is there a reason to still do this?

  12. broke

    And here's the weird part for me...

  13. broke

    If I DO block 5269, sometimes s2s completely fails and servers doesn't even try 5270.

  14. broke

    Some servers do, others don't.

  15. Samson Smith

    icebound.dev what happened?

  16. moparisthebest

    that shouldn't be the case, clients and servers should only try those if all the srv records failed

  17. broke

    moparisthebest: If there is no SRV record for xmpp-server, It completely fails.

  18. icebound.dev

    > icebound.dev what happened? just sexual harassment lol

  19. icebound.dev

    I thought it was a wider attack but it wasnt so I was just gonna drop it

  20. broke

    and If there is, but the port is blocked it _then_ tries 5270.

  21. broke

    and if there is, but the port is blocked it _then_ tries 5270.

  22. broke

    I find this very very weird.

  23. moparisthebest

    > moparisthebest: If there is no SRV record for xmpp-server, It completely fails. that shouldn't be the case either... maybe time to report bugs to servers Conversations is a client that for sure won't do that

  24. broke

    well, c2s works tbh, just not s2s.

  25. broke

    nonetheless you can even specify legacy ssl directly in conversations.

  26. broke

    which I do, that helps in faster connection.

  27. broke

    (which is the main reason to even do this)

  28. icebound.dev

    > there's something I wanna ask that seems to fit here the most, I asked this on the prosody chat and they pointed me to a very important thing. And that is server compliance with what restriction I am creating. I personally wanted less ports to be open and wanted to only support legacy ssl, meaning only tls (no starttls). But most servers seems to probe 5269 if there is an SRV with that, and if there is literally no 5269 or any port mentioned for xmpp-server, they reach me late because of doing starttls first. On clients I find the same thing as well. Clients always tries 5222 first. And servers always try 5269 first, Is there a reason to still do this? I do not see the benefit of reducing ports

  29. icebound.dev

    any decent firewall lets you chain ports

  30. broke

    well, no starttls.

  31. broke

    big plus.

  32. broke

    legacy ssl is always great.

  33. icebound.dev

    funny enough starttls was picked over direct a while back, and up until recently openfire called direct "legacy"

  34. broke

    well, some people told me not to call it "direct tls"...

  35. broke

    ;-;

  36. icebound.dev shrugs

  37. broke sighs

  38. icebound.dev

    anyways startls is too common, disabiling it is like disabling IPv4

  39. broke

    disabling IPv4 is a good thing though.

  40. broke

    I'd expect IPv6s to be the norm starting next decade.

  41. broke

    And deprecate IPv4.

  42. broke

    Please.

  43. broke

    s/next/next next//

  44. icebound.dev

    alright thats a second JID on my server itsjustmyluck@xmpp.is has spammed with nsfw, so I will now put it here in case others have issue with this JID.

  45. Samson Smith

    >> icebound.dev what happened? > just sexual harassment lol WTF

  46. broke

    icebound.dev: err wait, how do you know this?

  47. broke

    that nsfws

  48. icebound.dev

    > that nsfws because they are spamming asking for help c**ing

  49. icebound.dev

    I believe I know which spammer it is too

  50. broke

    spamming where?

  51. broke

    on dms?

  52. icebound.dev

    I thought they gave up months ago but seems like they are back, and im being targeted as I am the one which kept reporting them last time.

  53. icebound.dev

    > on dms? yes, but now has joined channels too

  54. broke

    wait how did they get spammed in dms?

  55. icebound.dev

    because the MUC they joined doesnt protect JIDs :)

  56. broke

    ...

  57. broke

    that's a stupid thing to do..

  58. icebound.dev

    doesnt matter, watch out for that JID

  59. broke

    alr.

  60. broke

    and iirc, gajim has anti-spam things if I heard this correctly.

  61. Samson Smith

    > I believe I know which spammer it is too Who

  62. broke

    they probably should use one of those things now ig.

  63. broke

    I thought pubmucs with open JIDs were only used for private members only encrypted mucs. Don't you think it should be disallowed from pubmucs?

  64. broke

    I thought rooms with open JIDs were only used for private members only encrypted mucs. Don't you think it should be disallowed from pubmucs?

  65. broke

    (Infact, that's required for omemo)

  66. broke

    (Infact, that's required for omemo, I mean encryption)

  67. Menel

    > I thought rooms with open JIDs were only used for private members only encrypted mucs. Don't you think it should be disallowed from pubmucs? Moderators can always see all jids, and public rooms cam be configured that everyone sees that

  68. broke

    The former is acceptable, the latter isn't.

  69. moparisthebest

    your client should warn you before you join if it doesn't report a bug

  70. Menel

    That's for each admin to deside. And your client to warn your when joining broke

  71. broke

    "your client to warn you" oh it does?

  72. moparisthebest

    conversations and forks for sure do

  73. icebound.dev

    > your client should warn you before you join if it doesn't report a bug I don't think gajim does but I know conversations does

  74. icebound.dev

    not hiding JIDs can be useful sometimes, it is not so useful in public channels, but in private channels it can be useful to see other peoples JIDs

  75. TheCoffeMaker

    gajim does tell if MUC is anon, semi-anon or if your jid is publicly listable before u enter to that room

  76. icebound.dev

    > gajim does tell if MUC is anon, semi-anon or if your jid is publicly listable before u enter to that room oh right yeah, it tells you but doesn't warn you, its up for you to review the settings :P