13:35 < Zdra> smcv, The spec says that MC must keep the value of nickname
property on Account and set it after each connection. Is it also
true if I set alias on self handle?
13:36 < Zdra> smcv, I mean, setting alias on self handle should probably be
defined to have the exact same behaviour than setting nickname on
account? or not?
13:36 < smcv> Zdra: I don't know
13:36 < smcv> Zdra: quality-of-implementation issue
13:37 < smcv> Zdra: I believe MC picks up self-alias changes and stores them in
13:37 < Zdra> smcv, you don't think that's something we should force at the
13:37 < smcv> Zdra: it's not something I've thought about in enough detail to
have a rationale either way
13:38 < Zdra> smcv, atm empathy set alias on self handle. but if MC does not
store that, I should change to set nickname on account
13:38 < smcv> Zdra: if you do, please file a spec bug asking that it be
clarified, and explain which interpretation you think is better
13:38 * sjoerd would like the demands for a minimal spec compliant
AccountManager to be not too high
13:38 < sjoerd> In practise we know which implementation we'll use on the gnome
desktop, so that's fine for us
13:38 < smcv> Zdra: please consider nickname changes from other resources when
13:39 < smcv> I think that might be a good rationale for MC's current behaviour
13:39 < smcv> suppose I'm logged in using Telepathy on N900 and Pidgin on laptop
13:40 < smcv> and I change my nickname in Pidgin
13:40 < smcv> MC on the N900 should pick up the change for next time, ideally?
13:40 < sjoerd> does the alias-before-connect override the server alias? should
13:40 < smcv> but then there's also an interaction with changing your nickname
on the laptop while the N900 is offline, and then logging in with
13:40 < smcv> sjoerd: yes exactly
13:41 < smcv> I think the right thing for Empathy to do is to set Nickname and
let the AM propagate the change, tbh
(In reply to comment #0)
> 13:40 < sjoerd> does the alias-before-connect override the server alias? should
> it ?
I think it should override what's on the server if you've changed it while offline, or the server should override the local setting if the local setting has not changed since the last time we connected (this is the same logic we use for avatars). Unfortunately, Aliasing makes this impossible, because we can't tell for sure whether the alias is server-stored; we should fix this in Names.
That means that (to continue the example from IRC discussion) if you alternate between Pidgin on laptop and Telepathy on N900, with one at a time connected, occasionally changing your name on the current device, the right things will happen.
> 13:39 < smcv> I think that might be a good rationale for MC's current behaviour
> 13:39 < smcv> suppose I'm logged in using Telepathy on N900 and Pidgin on laptop
> 13:40 < smcv> and I change my nickname in Pidgin
> 13:40 < smcv> MC on the N900 should pick up the change for next time, ideally?
Implementation detail: it does, mostly in mcd-connection.c.
> 13:40 < smcv> but then there's also an interaction with changing your nickname
> on the laptop while the N900 is offline, and then logging in with
> the N900
However, this is currently mis-handled in MC due to deficiencies of Aliasing. The N900's MC will override what the laptop set:
/* FIXME: ideally, on protocols with server-stored nicknames, this should
* only be done if the local Nickname has been changed since last time we
* were online; Aliasing doesn't currently offer a way to tell whether
* this is such a protocol, though. */
-- GitLab Migration Automatic Message --
This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.
You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/telepathy/telepathy-spec/issues/51.