16:07:07 <joanie> #startmeeting
16:07:07 <tota11y> Meeting started Thu May 24 16:07:07 2012 CET.  The chair is joanie. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:07:07 <tota11y> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:07:28 <joanie> #topic GNOME 3.6 - pyatspi2 deprecation?
16:07:55 <joanie> #info We can now use at-spi2 via gobject introspection.
16:08:24 <joanie> #info It seems to Joanie that there is not, therefore, a need to also maintain pyatspi2.
16:08:44 <joanie> #info Joanie is going to try migrating Orca to at-spi2 via introspection.
16:09:24 <joanie> #info What do people think about deprecating pyatspi2, giving people a year (tops) to convert, and then stop maintaining it?
16:09:32 * joanie yields the floor to comments
16:10:27 <fer> differences are similar to pygtk <--> gtk with g.o. instrospection?
16:10:42 <fer> (hi, btw!)
16:10:47 <joanie> hey btw!
16:11:00 <joanie> as far as I know, it would be similar but easier
16:11:08 <fer> nice then
16:11:09 <mgorse> fer: Yeah, it's similar. Pyatspi right now is a wrapper to implement the old API
16:11:28 <joanie> mgorse: and it does it via introspection, right?
16:11:45 <aleiva> it is
16:11:50 <joanie> i.e. we already know at-spi2 via introspection works
16:12:03 <joanie> whereas in other areas, we've thought -- or hoped -- it worked
16:12:08 <joanie> and then found major bugs
16:12:17 <fer> so not big risks here
16:12:22 <aleiva> who is using pyatspi beyond orca and accerciser?
16:12:23 <mgorse> and I'd need to look at interfaces again.Ypatspi has classes right now that wrap some interfaces because it was causing problems to try to call a function from an interface with the same base name as a function elsewhere (ie, get_attributes functions)
16:12:23 <mgorse> so I need to check if that works any better now and, if not, if something can be done about it
16:12:24 <joanie> at-spi2 is already tested introspection-wize
16:12:42 <joanie> mgorse, mind infoing that?
16:12:49 <joanie> for the sake of the meeting minutes
16:12:51 <fer> aleiva: probably all the testing projects
16:13:03 <joanie> but w.r.t. testing projects.....
16:13:08 <joanie> dogtail is dead I think
16:13:18 <joanie> so there's just ldtp
16:13:35 <joanie> and we can give people a year (if we must) before dropping it outright
16:13:50 <mgorse> #info mgorse will need to test whether interfaces still cause problems in terms of function names conflicting (ie, *_get_attributes functions) and, if so, if something can be done about it
16:14:17 <joanie> So if I'm in favor of it, Javi is in favor of it, we can convince Naggapan, and mgorse thinks it's doable pending his investigation....
16:14:26 <joanie> that would be all of the "stakeholders"
16:15:02 <joanie> and imho, if there are others.... Deprecations, much like sh*t, happens ;)
16:15:26 <fer> :)
16:15:52 <joanie> btw I already gave API_afk a heads-up about this yesterday (he's in a meeting atm).
16:16:06 <joanie> and fearless leader did not say "zomg! don't do that!!"
16:16:23 <joanie> so....
16:16:31 <joanie> is there anyone opposed to this move?
16:17:02 * joanie waves to heidi
16:17:09 * heidi waves to joanie
16:17:39 <joanie> #info No one seems to be opposed to the tentative deprecation of pyatspi2.
16:18:05 <joanie> #info Next steps: Wait for Mike to look into the potential issue he raised.
16:18:15 <joanie> mgorse: when do you think you can look into that?
16:18:56 <mgorse> I'll check before the next meeting, anyway / figure out what needs to be done, if anything
16:19:20 <joanie> suhweet
16:19:21 <joanie> thanks!
16:19:57 <joanie> #info Mike believes he will have an preliminary answer by next meeting, so we will table this discussion for one week.
16:20:13 <joanie> i.e. let's not stress Nagapan before we have to. ;)
16:20:25 <joanie> aleiva: could you please double-check with Javi on this?
16:20:32 <joanie> I suspect he will be in favor of it
16:20:35 <aleiva> yup
16:20:38 <joanie> but I don't want to guess wrong
16:20:40 <joanie> awesome thanks!!
16:20:41 <aleiva> I think so
16:20:59 * joanie prepares to move on
16:21:09 <joanie> to a somewhat related topic
16:21:51 <joanie> #info GNOME seems to want us to migrate to Python 3 (though it's not an official GnomeGoal yet).
16:22:24 <joanie> #info Canonical is targetting their next stable release as only Python 3 -- at least in terms of the live media -- *as far as I understand it*
16:22:34 <joanie> #info Therefore, we kinda sorta need to do this.
16:22:56 <joanie> #info Mike has begun (and might be finished??) with the work in pyatspi.
16:23:11 <joanie> #info Joanie has been working on this for Orca
16:23:47 <joanie> #info Liblouis has a prelimary version committed
16:24:06 <joanie> #info Speechd has a patch attached to a bug.
16:24:24 <joanie> #info brlapi's project lead thinks he can do this soon.
16:24:42 <joanie> so.... I am *thinking* and hoping we can be Python 3ized for gnome 3.6
16:24:52 <joanie> if Javi can do this. /me looks at aleiva
16:24:53 <joanie> :)
16:25:02 * joanie yields the floor for comments
16:25:28 <mgorse> Luke will be really happy if we are
16:25:39 <joanie> mgorse: indeed he would
16:26:39 <joanie> aleiva: can you add this to the "ask Javi" list please?
16:27:54 <aleiva> yup
16:28:06 <aleiva> in fact we talked about that
16:28:32 <joanie> cool
16:28:56 <joanie> so... anyone else with comments, questions, whatevers about targetting 3.6 for our team being python 2-free?
16:30:04 <joanie> silence == consent ;)
16:30:24 <joanie> #info Our team will target GNOME 3.6 for being migrated to Python 3.
16:31:01 <joanie> #info Juanjo is unable to be here today and has no updates.
16:31:28 <joanie> #info Please please please add your input to the Open Help Conference page
16:31:36 <joanie> https://live.gnome.org/Accessibility/AccessibleHelp
16:32:14 <joanie> anything else we need to talk about?
16:32:19 * joanie yields the floor
16:33:31 <joanie> I'm taking this as a "no"
16:33:50 * joanie raises gavel and looks for fingers
16:34:05 <joanie> #endmeeting