14:05:32 <API> #startmeeting
14:05:32 <Services> Meeting started Thu Oct 16 14:05:32 2014 UTC.  The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:05:32 <Services> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:05:33 <clown> ok move that discussiong to misc time, mgorse.
14:06:26 <API> #topic Progress towards 3.16
14:06:39 <API> #info this is the first meeting where we discuss 3.16 cycle
14:07:17 <API> #info I will personally just point work of interest, in a brainstorming mode, eventually (soon) we would need to decide on the priorities
14:07:21 <API> #info Wayland
14:07:42 <API> #info wayland in general was discussed on the recent GNOME Boston summit
14:08:04 <API> #info http://blogs.gnome.org/mclasen/2014/10/12/gnome-summit-update/
14:08:36 <API> #info Matthias Clasen revamped the current wayland status:
14:08:39 <API> #info https://wiki.gnome.org/Initiatives/Wayland/Status3.14
14:09:00 <API> #info that is just a list of bugs, some of them relevant (crashed included)
14:09:29 <API> #info IMHO, it would be good if we revisit our wayland-accessibility page, and see what of all them is still true
14:09:42 <API> #info AFAIR, that was done testing 3.9
14:10:22 <API> #info Proposed AI: <someone> test current gnome+wayland, update https://wiki.gnome.org/Accessibility/Wayland if needed
14:10:36 <joanie> I can do some of that
14:10:38 <API> questions comments at this point?
14:11:01 <joanie> not necessarily all of it
14:11:21 <API> ok, in parallel due the lack of updates on Maliit, that was in theory our reference to wayland extension related with keyboard input
14:11:37 <API> #info Tizen, a linux-wayland based
14:11:43 <API> #info ecosystem?
14:11:51 <API> #info also has some virtual keyboards
14:12:25 <API> #action API will take a look to tizen repositories, to check if there is anything more updated to have a reference for GNOME
14:13:03 <API> joanie, as you raised your hand for the previous proposed action
14:13:09 <joanie> Most of it
14:13:12 <joanie> :)
14:13:16 <joanie> want me to AI myself?
14:13:25 <API> yes please
14:13:27 <joanie> k
14:13:59 <joanie> #action Joanie will begin reviewing the existing list of Wayland-related issues we identified that impact our assistive technologies.
14:14:22 <joanie> mgorse: will you have time to resume looking into some of the things we flagged as missing
14:14:35 <joanie> like mouse location and event synthesis?
14:14:47 <mgorse> hoping to later this month--we have "hackweek" coming up
14:15:03 <joanie> hackweek?
14:15:26 <mgorse> yeah, it's a week when people can work on whatever projects they want to work on
14:15:34 <joanie> nice
14:15:36 <mgorse> so hopefully will help me get started anyhow
14:15:49 <mgorse> but I need to learn the codebase
14:16:42 <API> ok thanks, so unless someone complains I will keep moving on the current topic
14:17:04 <API> based on agenda, next subtopic should be "Feature proposals "
14:17:09 <API> but that doesn't apply anymore
14:17:30 <API> so I will create subtopic
14:17:34 <joanie> features were deprecated?
14:17:41 <API> well, first
14:17:55 <API> release team wants to rethink the concept of feature proposal
14:18:05 <API> it didn't have too much success sicne the beginning
14:18:11 <API> and faded really badly
14:18:22 <API> I think that almost nobody proposed new features
14:18:27 <API> well, I will info all this
14:18:46 <API> #info "feature proposals" period is being rethinked by the release team
14:19:02 <API> #info mostly due the lack of success of this way to propose new stuff to the platform
14:19:16 <API> #info partly due the vague definition of "feature" under this
14:19:30 <API> #info Allan Day has some ideas in relation of how to approach new cycles
14:19:41 <API> #info fwiw, Allan Day asked to join release team
14:19:58 <API> additionally, as far as I know, please correct me
14:20:13 <API> we don't plan to ship new features, but just improve the platform as it is
14:20:14 <API> right?
14:20:34 <joanie> I think that's probably a fair statement
14:20:53 <API> then if you don't mind, I will not add a subtopic about new features ;)
14:20:59 <joanie> ok
14:21:14 <API> #info Subtopic: relatively new ATK APIs and their spread
14:21:40 <API> #info during previous releases new interfaces and methods new methods were added, other were deprecated
14:21:52 <API> #info but not all of them are already implemented on the toolkits
14:21:58 * joanie nods
14:22:18 <API> #info examples: new, shiny and easier to understand AtkText methods are only (afaik) implemented on webkitgtk
14:22:39 <API> #info new AtkValue methods are not implemented at all (in fact, I think that are not even supported on at-spi2)
14:23:18 <joanie> if I had to bet, Gtk+ doesn't implement AtkTableCell
14:23:50 <API> ah true, thanks
14:24:02 <joanie> I need to see if Gecko does
14:24:04 <joanie> WebKitGtk does
14:24:16 <API> #info new AtkTable+AtkTableCell are only implemented at webkitgtk
14:24:25 <joanie> wait
14:24:29 <API> well, although in the case of AtkTable, it is mostly deprecations
14:24:29 <joanie> new AtkTable?
14:24:34 <joanie> ok
14:24:43 <API> yes it is not really new worded
14:24:47 <API> s/new/well
14:24:53 <joanie> np
14:25:17 <API> well, not sure if getting all of them implemented for 3.16 is too much
14:25:32 <API> but at least I would like to get some of them
14:25:36 <joanie> +1
14:25:50 <API> taking into account that the support of atkvalue on gnome-shell is somewhat recent, and done by me
14:26:08 <API> and that the old methods on atkvalue, due gvalue, led to implement it using a workaround
14:26:50 <API> #action API: during this cycle will implement the new AtkValue methods on gnome-shell, replacing the old ones. Would also implement the at-spi2 stuff if they are still not implemented
14:27:01 <API> note that Im talking only about gnome-shell
14:27:17 <API> I would like to not volunteer for too much and then not getting it done
14:27:25 <API> additionally
14:27:53 <API> #action <someone>: it would be good to create the bugs on the different toolkit bugzillas in order to track that there are new interfaces, new methods, deprecated methods
14:28:04 <joanie> ok
14:28:26 <clown> it would be nice to have a list of the affected toolkits.
14:28:42 <API> well, any toolkit implementing atk would be affected ;)
14:28:55 <clown> yup...
14:28:56 <joanie> #action Joanie will create a new wiki page and begin filing bugs in the various issue trackers regarding who needs to implement/deprecate what wrt ATK API changes.
14:29:08 <API> so just a quick list: gtk, clutter, webkitgtk, libreoffice, mozilla
14:29:15 <clown> java?
14:29:19 <API> although as I mentioned, in some cases it is already implemented
14:29:23 <joanie> qt's at-spi2 bridge?
14:29:38 <API> clown, hmm, yes java-atk-wrapper should be also affected
14:29:51 <API> but probably that module has greater problems that new ATK interfaces
14:29:52 <clown> does "clutter" include the ST toolkit?
14:30:09 <API> and after all, we are conservative, they are deprecated but would be still working for a while
14:30:23 <API> clown: hmm, well, add "gnome-shell" to that list
14:30:28 <API> as I also added libreoffice
14:30:37 <API> for me st is  not exactly a "real" toolkit
14:30:42 <API> only gnome-shell is using it
14:31:01 <API> joanie, yes good point
14:31:10 <clown> I thought it was based on some mobile toolkit (meego?), and they were trying to keep them in sync.
14:31:21 <API> it would be good if qt also update it
14:31:25 * joanie nods
14:31:34 <joanie> this will take me a little while
14:31:39 <joanie> but I'll try to get on it soon
14:31:49 <API> clown, they said that they planned to sync to mx (that toolkit coming from meego)
14:31:53 <API> they never did it
14:32:16 <clown> okay API, but...
14:32:18 <API> at this point mx and st diverged so much that I don't see a trivial amount of work syncing them
14:32:28 * API waiting a but...
14:32:34 <clown> in a related note, I just went to "meego.com", and was redirected to tizen.com.
14:33:19 <API> yes meego got sucked by tizen
14:33:25 <API> or meego was renamed to tizen
14:33:30 <clown> which you mentioned earlier in the context of wayland.
14:33:43 <API> I don't remember exactly what was the details of tizen announcement
14:33:50 <API> tizen doesn't use mx at all
14:34:00 <API> afaik
14:34:09 <API> tizen doesn't use clutter at all, afaik
14:34:14 <clown> okay, I'm way behind I guess.  ancient history in the intenet age...
14:34:22 <API> it was somewhat replaced by efl
14:34:35 <API> but depending on the tizen profile, efl is there or not
14:34:44 <API> fwiw, tizen has the concept of profiles
14:35:01 <API> tizen ivi, tizen mobile, tizen wearables
14:35:09 <API> and tizen common as the, well, common parts
14:35:37 <API> having said so, moving to next subtopic ?
14:35:44 <clown> fine with me.
14:35:46 <joanie> not yet
14:35:47 <joanie> maybe
14:36:00 <joanie> this may be just a related aside
14:36:05 <API> shot then
14:36:17 <joanie> but also on my to-do list, and I really need to force myself to make it a priority now is:
14:36:38 <joanie> There are a few things for which new API in ATK and AT-SPI2 are (imho) required
14:37:11 <joanie> so if there's not any problem with that, I think it makes sense to do that ASAP and then lump it into what we just discussed
14:37:34 <joanie> I am happy to do the initial proposals, patches, etc.
14:37:43 <joanie> (done)
14:37:54 <API> can you elaborate that "lump into what we just discussed"?
14:37:59 <API> lump as ...?
14:38:27 <joanie> well, for instance, I think we need some new API related to position in set and set size. It came up as a result of W3C/WebApp stuff
14:38:32 <joanie> but it applies across the board
14:38:44 <joanie> so I'm suggesting I propose a solution (with patches) now
14:38:57 <joanie> if blessed, then that's another case of toolkits need to implement it
14:39:01 <joanie> we need to track it
14:39:03 <joanie> etc., etc.
14:39:26 <clown> probably too much detail, but do you mean something like the groupPosition() functin in IA2, joanie?
14:39:33 <clown> *function
14:39:35 <joanie> something like, yes
14:39:40 <API> well, if I just focused on start the spread to already existing but not implemented APIs was just to be conservative
14:39:41 <clown> okay, I understand.
14:39:43 <joanie> whether or not it's exactly like, I'm not sure
14:39:55 <API> joanie, we can add those new APIs to the discussion
14:40:02 <joanie> thanks
14:40:08 <API> but we would need also someone to work on them
14:40:14 <joanie> I would work on them
14:40:18 <joanie> is what I'm saying
14:40:41 <API> ok, so could you info a first list of to-be-added ATK new interfaces/methods
14:40:55 <joanie> not yet
14:40:59 <joanie> because I need to make that list
14:41:01 <joanie> ;)
14:41:27 <API> ok, so as in any case that would be a new subtopic
14:41:35 <API> #info Subtopic: new ATK interfaces
14:42:12 <API> #info joanie pointed to the need of new ATK interfaces, some of then detected as a need as a result of W3C collaboration
14:42:52 <API> #info on next week meeting, we will list the concepts to new APIs that we have around, and prioritize/assign-worker on them
14:42:59 <API> joanie, how that sounds?
14:43:05 <joanie> fine
14:43:10 <API> ok
14:43:28 <API> so moving to next subtopic?
14:43:41 <joanie> sure
14:43:46 <API> sorry if Im pushing too much, but it seems that this topic will eat a lot of the meeting (is already doing)
14:43:58 <API> #info Subtopic: at-spi2-cache
14:44:20 <API> #info As discussed on the past, it would be good to evaluate the real beneficts of at-spi2 cache
14:44:58 <API> #info it is not clear the performance benefits it brings, and there are now and then cache coherence related bugs
14:45:42 <API> #info problem: right now at-spi2 doesn't work properly disabling all the cache, at least after some tests that Piñeiro made while working on the "gtktreeview children-changed flood" bugs
14:46:16 <API> #action API will create a bug about at-spi2 not working if the cache is disabled
14:46:37 <mgorse> Thanks. I thought it should work, even if it's a little slow, but haven't tried to test in a while
14:46:38 <API> #info Piñeiro thinks that that is needed in order to evaluate cache vs non-cache
14:46:59 <API> mgorse, well, when I tested the app got blocked on the get_name method
14:47:03 <API> but I didn't have too much to test
14:47:25 <API> implicitly that "open a bug" would include some little confirmation that that is the case
14:47:52 <API> #info being conservative, I will not propose more work on this issue for now
14:48:08 <API> well, and this was my last subtopic on the big "Progress towards 3.16" topic
14:48:16 <API> so, comments, questions, doubts, more subtopics?
14:50:22 <API> seems that everybody is overwhelmed for so  big topic
14:50:30 * clown is reeling.  not.
14:50:44 <joanie> or read your earlier, we need to push on
14:50:45 <joanie> ;)
14:50:56 <API> so I will move to next topic
14:51:01 <API> #topic W3C updates
14:51:09 <API> sorry clown and joanie, today you don't have too much time
14:51:21 <joanie> clown: as *always* I defer to you
14:51:43 <clown> #info there was no meeting on Mon, and the User Agent meeting was cancelled due to lack of quorum.  No updates
14:51:54 <clown> over to you joanie
14:52:02 <joanie> heh
14:52:09 <joanie> We can move on
14:52:24 <API> well, so this week was a good week to not have too much time to the w3c updates
14:52:29 <API> well, jjmarin is not here
14:52:38 <API> so no marketing
14:52:45 <API> #topic miscellaneous time
14:52:51 <API> anything miscellaneous to mention?
14:52:54 <joanie> yes
14:53:10 <joanie> mgorse: did you see my ping about the AtspiTableCell iface docs not being generated?
14:53:22 <joanie> I started looking into why and then other stuff came up
14:53:47 <mgorse> oh, I'll take a look
14:53:53 <joanie> thanks!
14:54:09 <joanie> and sorry for the "other stuff". I really did plan to just give you a patch
14:54:23 <joanie> ok, that's my only misc
14:54:26 * joanie shuts up
14:54:34 <mgorse> I'm giving a Linux presentation next week at the local accessibility/usability meetup. Does anyone have any suggestions for what to talk about?
14:55:01 <mgorse> also need to figure out if impress is accessible
14:55:18 <joanie> mgorse: it could be more accessible
14:55:24 <joanie> some issues are in LO; some in Orca
14:55:29 <mgorse> ok
14:55:36 <joanie> if you find issues, ping me and I'll either fix them or file them
14:55:41 <mgorse> thanks
14:55:50 <joanie> and I'll try to take a look this weekend
14:56:12 <joanie> once your preso is created, Evince + Orca should work
14:56:26 <joanie> and Orca should present the slides in gnome-documents too
14:56:44 <mgorse> ok, I'd better play around/practice some this weekend
14:56:49 <joanie> in gnome-documents it should auto present as you advance to each slide
14:56:53 <mgorse> would be good practice making presentations, anyhow
14:57:09 <joanie> in evince, Orca doesn't (yet) try to detect you are in slide mode
14:57:15 <joanie> I'll try to fix that this weekend too
14:57:20 <joanie> or maybe today or tomorrow
14:58:07 <joanie> TODO list longa, vita brevis
15:00:48 <API> end meeting time is here
15:00:51 <API> so something else?
15:00:59 <API> closing meeting?
15:01:06 <clown> sure
15:01:33 <API> #closemeeting
15:01:37 <API> ups
15:01:39 <API> #endmeeting