XSF Discussion - 2014-02-22


  1. ralphm

    https://news.ycombinator.com/item?id=7281237

  2. Neustradamus

    note: https://xmpp.net/ has a problem: Code d'erreur : ssl_error_no_cypher_overlap I am not alone to have this problem.

  3. Simon

    How is everyone's test day going?

  4. intosi

    Too early to tell for j.org, I only just flipped the switch.

  5. Simon

    intosi: excellent!

  6. Neustradamus

    now all are good for me :)

  7. Neustradamus

    for S2S

  8. Simon

    I've basically left all the buddycloud server set to force since the last test day.

  9. Neustradamus

    I have only problems with: xmpp:dev@muc.jappix.org (Metronome) xmpp:support@muc.jappix.org (Metronome) xmpp:grimoire@muc.metronome.im (Metronome) xmpp:beem@conference.elyzion.net (?)

  10. Simon

    Is Metronome the Prosody fork?

  11. Neustradamus

    yes!

  12. intosi

    So far, the usual suspects are failing on my home interop testing servers.

  13. Simon

    gmail.com and?

  14. intosi

    gmail, cisco.

  15. Neustradamus

    "about rooms"

  16. intosi

    Neustradamus: from your own server? No need reporting unrelated issues that have been reported before and are looked into. That only muddles the feedback on the TLS testing day.

  17. Neustradamus

    reply from Metronome dev: [12:23:37] <Marco Cirillo> you have a problem only joining mucs on Metronome servers which run bidi [12:24:02] <Marco Cirillo> try with your account on jappix.com [12:24:13] <Marco Cirillo> you'll see it works

  18. Neustradamus

    it is a bug before the switch like I have sent

  19. Neustradamus

    *** 2014-02-21 [08:43:05] <Neustradamus> *** 2014-02-21 [01:54:23] <Maranda> stpeter btw, j.o's seems to have some blocking issues regarding bidi, some users (e.g. Neustradamus) can't join mucs on my server because, even if it requested bidi & authenticated, j.o still bounces the incoming stanzas it receives on the bidi enabled stream.

  20. intosi

    Neustradamus: stop doing that. The jabber.org admins know about the issue, and you know that. It is not related to the TLS testing day.

  21. Neustradamus

    no problem, strange that it is a little hours before...

  22. intosi

    As expected, the number of active s2s connections on jabber.org is dropping, where on normal Saturdays, the number should still be increasing. Still too early to tell the difference between this day and the previous one.

  23. Simon

    intosi - are you or Peter going to fix the cipher list on jabber.org?

  24. intosi

    Eventually, yes.

  25. intosi

    Though not today.

  26. intosi

    With eventually, I really mean Real Soon Now™ ;)

  27. Simon

    great. I always notice it in the test results.

  28. intosi

    There is some discussion on what impact this will have on older clients and clients that cannot do HGE.

  29. Simon

    HGE?

  30. intosi

    High Grade Encryption

  31. intosi

    And whether we should care about that.

  32. Simon

    There's a high chance that older clients are set to autoconnect and have sat untouched in the background of a machine for years.

  33. Zash

    xnyhps, is xmpp.net broken? submit.php and result.php gives empty replies

  34. Kev

    Yes.

  35. Kev

    You're welcome.

  36. Zash

    .

  37. intosi

    But it's more secure that way ;)

  38. ralphm

    fippo, did you see this update: https://code.google.com/p/webrtc/issues/detail?id=2923

  39. fippo

    ralphm: yeah... we should probably work on giving them a... say swiften-based... example :-)

  40. Kev

    Sounds like something for the XSF to fund ;)

  41. fippo

    nah, for google!

  42. fippo

    damn, that would have been a nice gsoc project actually

  43. Kev

    It was on the list of ones I was going to put up for Swift, until I didn't have time to mentor.

  44. fippo

    ralphm: i'd actually say that the signalling components of libjingle are not useful :-)

  45. ralphm

    :-D

  46. ralphm

    That's kinda funny

  47. Simon

    xnyhps if you are there, - running a test against any domain on xmpp.net is failing. some Secruity test day folks are confused.

  48. Kev

    Yeah.

  49. Kev

    Peter asked if the OS could be updated. Turns out that the observatory doesn't like the new packages much.

  50. Kev

    I've been looking at it, but I've got stuff to do this afternoon.

  51. Kev

    Thijs knows about it (but I think is AFK, hasn't been responding for a few hours).

  52. Simon

    Thanks Kev.

  53. Simon

    If there is anything I can do, I'm happy to jump on.

  54. Kev

    I don't think so, thanks.

  55. Kev

    At least we're A on ssllabs now.

  56. Kev

    Weren't looking so good with openssl0.9.8

  57. Simon

    I understand. And if I remember correctly, upgrading openSSL basically implies a new distro version.

  58. Kev

    Nothing more relaxing after a long week at work than a remote dist-upgrade on an old old machine.

  59. Kev

    Other than, you know, not that.

  60. Simon nods. On my own server where I know all the gotchas… fine. But you never know what undocumented workarounds someone else has done in the heat of the moment and then forgotten to go in and fix properly.

  61. Kev

    OS did OK in the end...other than that the only thing running on the machine doesn't run on the new packages.

  62. Kev

    That's got things a /little/ better.

  63. Simon

    Thanks Kev!

  64. Kev

    It's hard not to hate PHP sometimes.

  65. Zash

    only sometimes?

  66. Kev

    I don't spend much of my time thinking about it.

  67. Kev

    Still a load of errors in the logs, but I don't know if they're normal or not.

  68. Kev

    Yeah, a number of them seem to be general bugs - errors in the logs going back days.

  69. Kev

    Not others, though.

  70. Kev

    Oh, no, that error's been there for days too.

  71. Kev

    Right. I can't find any errors in the logs that haven't shown up before.