+++ This bug was initially created as a clone of Bug #28866 +++ See that bug for initial discussion.
Next steps here are (quoting cassidy and myself from <https://bugs.freedesktop.org/show_bug.cgi?id=28866#c10>): > I fixed this one (not sure if that's the best way, this code looks like a bit > like a can of worms to me tbh :\ ) This code *is* a can of worms :-( See smcv/wip-req-export, in which I tried to split McdChannel into the bits that are really a channel and the bits that are a channel request, and have the latter be exported as the ChannelRequest; in particular, the top commit, which is called "WiP: doesn't work" and is something like my 5th attempt at it. Your code changes look good, except for the last commit where I need to read the code a bit more to work out wtf is going on. I suspect that what you did is necessary and correct, even though it makes little sense. Would you mind if I pick up this branch and make it use an in-tree extension instead of the patched telepathy-glib you're presumably using, so we can merge it before undrafting the spec, and continue to refactor?
The referenced branch takes all the commits I didn't object to from Guillaume's branch, and replaces the last commit (which we were both suspicious about) with a version that I think makes more sense.
Looks good (assuming all the tests pass now). I'm happy to see that the general desgin was actually as crack as I thought. :) Your comment makes things a bit clearer but it would be good to properly refactor all this mess at some point.
(In reply to comment #3) > Looks good (assuming all the tests pass now). They do. Fixed in git for 5.5.4; I've left this bug open for implementation of the final spec for this. > I'm happy to see that the general desgin was actually as crack as I thought. :) > Your comment makes things a bit clearer but it would be good to properly > refactor all this mess at some point. Yeah, I know; I introduced McdRequest as a way to work towards that, and will continue in that direction.
Bug #30092 continues the refactoring. The branch referenced here, which follows on from the one on Bug #30092, adds request-properties to the Handler_Info as well as the Observer_Info, which is mandated by current spec git (the forthcoming 0.19.12); this will probably be good to have before undrafting this branch.
+ * @paths_out: (transfer container) (element-type utf8): Requests_Satisfied + * @props_out: (transfer container) (element-type utf8 GHashTable): + * request-properties for Observer_Info or Handler_Info Shouldn't you add (out) as well? (We don't really care here but if we use gir annotation, best to get them right). Except that, branch looks good.
Branch merged to master with the obvious change; it'll be in 5.7.0. Leaving this bug open for the final version of the spec, when that lands.
My 'hints' branch updates MC to my spec proposal from Bug #28866 (read that bug first).
++
Thanks, merged.
http://git.collabora.co.uk/?p=user/cassidy/telepathy-mission-control;a=shortlog;h=refs/heads/hints-stable-30000 implements the stable API.
Merged to master for 5.7.1
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.