Gajim - 2017-04-04

  1. tanager lovetox: i wasn't aware debian packaged the plugin, that's good
  2. tanager There do seem to be a few experimental pidgin omemo plugins floating around but none "official" yet afaik?
  3. tanager The more clients implement it, the better, as i see it
  4. tanager « you can also download the plugin from git if that makes it more secure for you It will if the git repo uses gpg-signed commits and tags? 😇
  5. tanager I understand that projects are often low on resources but i do think it's a good idea for all projects to be implementing secure update mechanisms if they are independently obtaining executable code at all. (Making implementation relatively easy is what projects like TUF are about)
  6. tanager I wasn't so much asking for my own use as much as for the general systemic health of the users
  7. tanager « so to invest that much time into an update mechanism is i believe a waste of time The main gajim update mechanism for many users will be APT or DNF etc., and these systems already have good security. The question here is really for the plugins specifically, which constitutes something of a 'backdoor' where untrusted executable code can enter a system that otherwise contains only trusted code
  8. tanager A stop-gap measure to implementing a secure plugininstaller could be to ship more plugins with the base gajim installation, as is done for the pluginmanager plugin already. That way the security of the code is managed by the distros
  9. tanager s/security/integrity
  10. tanager ^ looking at this link it seems the weak spots are windows, fedora, and the "Distributions without a package". But the major linux distros are mostly taken care of
  11. tanager « also, neither the signal lib that gajim omemo uses, or the plugin is audited Ah. That is something of a problem 😟 « also gajim itself has probably multiple security problems Is there anything you're thinking of specifically; or just because it hasn't been audited?
  12. tanager Lack of audits is a problem, but this is an independent problem of its own, distinct from secure update mechanisms. Secure update mechanisms guarantee the integrity of the code rather than the security; a crypto flaw exposing some messages is a different problem to malicious code delivery by someone who e.g. hacks the gajim ftp server