| Summary: | Incorrectly reports offline contacts as unknown if presence arrives before the roster | ||
|---|---|---|---|
| Product: | Telepathy | Reporter: | Will Thompson <will> |
| Component: | gabble | Assignee: | Telepathy bugs list <telepathy-bugs> |
| Status: | RESOLVED FIXED | QA Contact: | Telepathy bugs list <telepathy-bugs> |
| Severity: | normal | ||
| Priority: | medium | Keywords: | patch |
| Version: | git master | ||
| Hardware: | Other | ||
| OS: | All | ||
| URL: | http://cgit.collabora.com/git/user/wjt/telepathy-gabble-wjt.git/log/?h=fd.o-38603-initial-contact-presence | ||
| Whiteboard: | fun with pronouns | ||
| i915 platform: | i915 features: | ||
|
Description
Will Thompson
2011-06-23 05:30:45 UTC
Here's a fix. I'm not really all that happy with it.
As I say in one of the commit messages, I think it would be better if GabblePresence represented peers' presences using the constructs available on XMPP (available/unavailable/error, show={...} or null) rather than GabblePresenceID, which adds a bunch of other states to muddy the waters (like invisible, which makes no sense for peers, and unknown vs. offline). I looked briefly into fixing this, but it's really invasive, not least because of the self presence (on which invisible makes sense, and on which the presence ID may be outside the range of GabblePresenceID to allow for plugin-supplied presences…).
It occurs to me that there's an edge-case which this won't cover: if we get <presence from='zie@foo.com' type='unavailable'><status>baiii</status></presence> before we receive the roster, Zie will be in the cache as unknown, and _maybe_remove() will not discard that entry because Zee had a non-NULL message. So then Zie will continue to be Unknown rather than Offline if the roster arrives and turns out to have Zie on it. I checked RFC3921, and such presences are totally legal <http://xmpp.org/rfcs/rfc3921.html#stanzas-presence-children> but I can't bring myself to care that much.
I've amended a couple of patches in the branch: • “PresenceCache: discard UNKNOWN presences in maybe_remove” <http://cgit.collabora.com/git/user/wjt/telepathy-gabble-wjt.git/commit/?h=fd.o-38603-initial-contact-presence&id=d1fd28d> has had its commit message reworded—I was wrong, this is the commit that actually fixes this bug. • “GabblePresence: start in state Unknown” <http://cgit.collabora.com/git/user/wjt/telepathy-gabble-wjt.git/commit/?h=fd.o-38603-initial-contact-presence&id=ba15096> has been amended to contain a fix to test-presence.c, and to remove misleading remarks about “slightly surprising side-effects” (since actually the previous commit fixed the bug that it would be surprising for this commit to fix). I think the fix is *still* surprising, but less so. ;) R+ Merged; will be in 0.12.3 and 0.13.2. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.