14:36:18 <API> #startmeeting
14:36:18 <tota11y> Meeting started Thu Aug 25 14:36:18 2011 UTC.  The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:36:18 <tota11y> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:36:19 <joanie> I'll fix the bot
14:36:21 <API> guys remember:
14:36:25 <API> http://live.gnome.org/Accessibility/Meetings
14:36:33 <API> so
14:36:46 <API> #topic gobject-introspection and related issues
14:37:02 <API> I guess that this is more meaningful that what we have on the agenda
14:37:03 * joanie observes API doesn't like her topic text
14:37:14 <API> so, joanie jhernandez|eee ...
14:37:19 <API> and mgorse
14:37:25 <joanie> I'll start I guess
14:38:14 <joanie> #info When 3.1.90 goes out, there will be two versions of pygobject: 2.28.6 will *not* use introspection. 2.90.x *will* use introspection. And the two versions can co-exist.
14:38:45 <joanie> #info Related with this change, any application which attempts to import pyatspi cannot import gtk via static bindings
14:39:07 <joanie> #info This doesn't impact most applications
14:39:24 <joanie> #info But it definitely impacts Orca and Accerciser because they need to import both pyatspi and gtk.
14:39:46 <joanie> #info Joanie made it an agenda item just in case there are others impacted by this change whom she forgot about.
14:40:04 <joanie> #info Joanie and Javi are actively working on their introspection ports now and plan to be done for 3.2.
14:40:07 <joanie> (done)
14:40:10 <API> and why two versions?
14:40:13 <clown> if there is time:  how does one import gtk non-statically?
14:40:20 <API> I mean, why it is required 2.28.6 and 2.90.x?
14:40:23 <joanie> API because of Cairo I think
14:40:33 <joanie> mgorse might remember more
14:40:37 <API> clown, with gobject-introspection ;)
14:40:49 <joanie> clown: in python: 'from gi.repository import Gtk'
14:40:49 <clown> d'oh!
14:40:49 <API> statically gtk was the old bindings
14:40:58 <mgorse> 2.28.6 is used to support static bindings, and they don't want to have one code base supporting both I guess. Not sure other than that
14:41:00 <clown> thanks joanie
14:41:12 <API> clown, in summary until know you could have a mixed environment
14:41:26 <API> with static (gtk) and gobject-introspection based (at-spi2) bindings
14:41:44 <joanie> I also recall something being said in #python a week or so back that there were two applications (which were not Accerciser or Orca) which still required static bindings.
14:42:11 <API> joanie, but I guess that they can use just static bindings
14:42:18 <joanie> yeah
14:42:20 <API> and not require gobject-introspection ones
14:42:25 <joanie> IF pygobject supports static bindings
14:42:30 <joanie> which 2.28.6 didn't used to do
14:42:39 <joanie> but now it does
14:42:48 <joanie> when the .90 release comes out anyway
14:42:56 <API> ok, thanks so in summary:
14:43:03 <joanie> we're all doomed
14:43:23 <API> #info 2.28.6 support static bindings, but no one of them support a mixed environment
14:43:32 <clown> joanie, you're being to positive, again...
14:43:47 <API> well, I miss other info, from what we talked the other day
14:44:06 <API> about what is preventing the port to gobject-introspection bindings
14:44:09 <joanie> sorry clown. I'm just a regular pollyanna
14:44:12 <API> those crashing bugs
14:44:22 <joanie> API good point
14:44:29 <API> joanie, could you add that here?
14:45:08 <joanie> #info The port to gobject-introspection is blocked by a few segfaults in pygobject from master.
14:45:26 <joanie> #info John Palmieri and Mike are working on providing patches for these bugs
14:45:44 <joanie> mgorse: I didn't have time to test your patch, but I will after the meeting
14:46:20 <API> ok, so in summary, and trying to elaborate that doomed part
14:46:27 <API> you are working on that port
14:46:35 <API> but unless those bugs get solved
14:46:40 <joanie> 14:40:04 < joanie> #info Joanie and Javi are actively working on their introspection ports now and plan to be  done for 3.2.
14:46:44 <API> gnome 3.2 will have a crash happy
14:46:47 <joanie> We will succeed
14:46:52 <joanie> if I have to kill people
14:46:53 <API> orca and accerciser
14:46:53 <joanie> ;-)
14:47:04 <joanie> we cannot ship GNOME 3.2 without a functioning screen reader
14:47:20 <API> ok, so lets summarize that with
14:47:34 <API> getting a functioning screen reader for 3.2 will be more complex than expected
14:47:43 <API> anything else in this poing?
14:47:45 <API> point?
14:47:51 <joanie> nope
14:47:56 <API> people?
14:48:00 <API> questions, doubts?
14:48:01 <jhernandez|eee> I'm still working on it
14:48:37 <jhernandez|eee> and I'm still ensuring that accerciser works properly with latest pygobject
14:49:08 <API> jhernandez|eee, what about those pygobject crashes?
14:49:29 <API> is add a workaround a possible plan B?
14:50:32 <jhernandez|eee> yes, in fact I have a B plan, but to remove some features isn't aa good idea
14:50:36 <jhernandez|eee> so ...
14:50:57 * jhernandez|eee shrugs
14:50:59 <API> ok, thanks
14:51:08 <API> well, should we move to next point?
14:51:17 * joanie nods
14:51:23 * jhernandez|eee nods too
14:51:27 <API> #topic Marketing and Fundraising
14:51:52 <API> #info Karen Sandler asked for end-oriented a11y goals for the a FoG campaing
14:52:14 <API> in summary, something that can be specifically funded
14:52:18 <API> joanie, is this correct?
14:52:25 <joanie> API ask jjmarin
14:52:45 <jjmarin> ji
14:53:11 <jjmarin> Karen says as a first step we'll need
14:53:20 <jjmarin> to clearly set out the goals of the campaign
14:53:31 <jjmarin> and how the money would be used
14:53:47 <jjmarin> we list sponsoring attendees at hackfests and at CSUN as a past success of FoG
14:54:07 <jjmarin> but she asking about a more end-oriented goals we can use instead
14:54:17 <joanie> jjmarin: I think we need to focus on development
14:54:25 <API> a past success of FoG?
14:54:30 <API> I don't understand
14:54:43 <API> this is the first time that a FoG for a11y was created
14:54:44 <joanie> i.e. development-related tasks required to make GNOME compellingly and fully accessible
14:54:45 <API> AFAIK
14:55:27 <jjmarin> She want a goald for a a11y campaign
14:55:48 <jjmarin> something well defined
14:56:03 <jjmarin> we need xxx $ to do this
14:56:11 <jjmarin> that's my impression
14:56:14 <API> ah, you mean use that as a example of how this FoG money will be used
14:56:24 <jjmarin> exactly
14:57:04 <jjmarin> and we the this goal is done, we can inform. We did it !
14:57:17 <jjmarin> s/and we/and when/
14:57:45 <API> jjmarin, ok, but as Joanmarie said, we should also include specific projects
14:57:54 <API> so, using our beloved #info's
14:58:08 <API> #info Karen want specific goals
14:58:20 <API> #info jjmarin mention hackfests, CSUN and other conferences
14:58:26 <jjmarin> We need somethin like this http://stormyscorner.com/2010/03/one-step-closer-to-a-sys-admin.html
14:58:43 <API> #info joanie mention that it is also required to use this opportunity to develop
14:59:27 <API> well, yes, although this include more a high level description
14:59:33 <API> but it is a good point to start
14:59:49 <API> so, going on with that
14:59:56 <API> #topic 2.1 Bochure for fundraising
15:00:05 <API> hmm, we already talked about that, right?
15:00:11 <API> so
15:00:25 <API> #topic 2.3 André Klapper asked for a11y achievements to be included in the 3.2 release notes
15:00:38 <jjmarin> What about the content of this brochure ?
15:00:45 <jjmarin> sorry
15:00:49 <jjmarin> I'm slow
15:01:01 <API> jjmarin, ah, I though that the content of the brochure was define those specific goals
15:01:10 <API> am I wrong?
15:01:38 <jjmarin> well, it's brochure for attracting companies
15:01:50 <jjmarin> to help to develop a11y
15:02:06 <API> it was also mentioned on the marketing mailing list?
15:02:12 <jjmarin> yes
15:02:48 <jjmarin> you also need brochures for people in a11y conferences
15:02:56 <jjmarin> s/you/we(
15:03:13 <clown> actual physical paper, jjmarin?
15:03:14 <jjmarin> IS there anything similar written?
15:03:32 <jjmarin> Well, we need to start with the content :_)
15:03:38 <clown> right...
15:03:52 <clown> but is the end goal a glossy handout?
15:04:11 <jjmarin> yes and no
15:04:24 <jjmarin> we can put this content in web I think
15:04:34 <jjmarin> opinions ?
15:04:46 <jjmarin> is it worth ?
15:04:50 <joanie> I think we should start with content
15:04:50 <Pendulum> I think it might be good to have something that's both on web, but can easily translate to a print handout
15:04:57 <API> well, it would be good to ask on the marketing.-list about the nature of those brochures
15:05:01 <Pendulum> but, yeah, content is the best first step
15:05:14 <jjmarin> ok
15:05:14 <joanie> and then worry about different formats (e.g. braille, large print, various languages)
15:06:06 <joanie> so given that we are starting with figuring out the content....
15:06:13 <joanie> jjmarin: what do you need from us?
15:06:13 <API> Pendulum, yeah you are right
15:06:18 <jjmarin> #Info Juanjo we'll ask for help to write these brochures
15:06:27 <API> we could try to see what we have
15:06:32 <API> what we need to add
15:06:40 <API> and meanwhile check for the format
15:06:49 <Pendulum> I'd be happy to help with the gathering content part if someone points me in the right direction :)
15:07:23 <clown> would our 3.2 "must have" page be a starting point?
15:07:33 <jjmarin> Pendulum: Ok, I'll count with you
15:07:37 <jjmarin> :)
15:08:12 <clown> http://live.gnome.org/Accessibility/ThreePointTwo/Issues
15:08:28 <joanie> clown: I think that might be more of a "what needs funding by FoG"
15:09:08 <API> yes I agree, now that I understood that this brochure thing
15:09:39 <API> ok, so anything else about the brochure?
15:10:15 <jjmarin> No, we'll come with this again
15:10:16 <jjmarin> :)
15:10:28 <jjmarin> when get more background
15:10:35 <API> #topic 2.3 André Klapper asked for a11y achievements to be included in the 3.2 release notes
15:10:42 <jjmarin> ok
15:10:48 <API> not sure if that fits here or not
15:11:02 <joanie> it's marketing
15:11:03 <API> I mean that those release notes are required, no matters
15:11:12 <API> if there is a marketing thing or not
15:11:18 <joanie> and it was raised by Andre in the context of Juanjo's marketing message
15:11:20 <jjmarin> It was mentioned by Andre Klapper in the marketing list
15:11:49 <jjmarin> he says that there didn't
15:11:51 <andre_> it can be marketing to developers and users, yes
15:12:03 <jjmarin> mention a11y
15:12:11 <jjmarin> in 3.0
15:12:22 <andre_> we dropped a11y from 3.0 release notes as it wasn't something to be proud of at that time, IIRC
15:12:30 <jjmarin> because the support wasn't
15:12:34 <jjmarin> quite good
15:12:40 <jjmarin> He want to be sure
15:12:42 <jjmarin> we
15:12:53 * joanie observes that this statement should in and of itself be indicative to Karen as to why we need a FoG for a11y
15:13:05 <Pendulum> joanie: +1
15:13:17 <jjmarin> are working on having good release notes
15:13:43 <jjmarin> For making easy
15:13:59 <joanie> andre_: We're working hard on introspection ports :-P See earlier agenda item. ;-)
15:14:06 <jjmarin> I can take charge of writing the summary
15:14:09 <jjmarin> so
15:14:18 <jjmarin> we can write drafts in https://live.gnome.org/Accessibility/ReleaseNotes/ThreePointTwo
15:14:26 <jjmarin> and I can write a summary in https://live.gnome.org/Accessibility/ReleaseNotes/
15:14:26 <API> jjmarin, you mean release notes summary?
15:14:30 <andre_> it really does not have to be much. maybe two or three things if something comes to anybody's mind that's worth to mention
15:14:54 <API> andre_ joanie jjmarin : thanks for the explanation
15:14:55 <jjmarin> yes, only the a11y part, of course
15:15:03 <API> yes, it would be a good idea to include the release notes
15:15:05 <jjmarin> to be revised by Joanie and API
15:15:06 <joanie> andre_: Right. I think clown has some things worth mentioning w.r.t. gnome-shell mag
15:15:13 <API> it is true that we would like a better status
15:15:26 <API> but it is worth to mention which things have improved
15:15:29 <joanie> and API has been working on the gnome-shell general accessibility
15:15:33 <joanie> API exactly
15:16:03 <API> not as much as I wanted, but yes I did something there
15:16:09 <jjmarin> Do we agree ?
15:16:24 <joanie> #action Team members should add items they'd like included on https://live.gnome.org/Accessibility/ReleaseNotes/ThreePointTwo
15:16:27 <API> jjmarin, we agree that it is worth to mention a11y on 3.2 release notes
15:16:43 <joanie> #action Juanjo will then draft a summary for the GNOME release notes
15:16:51 <clown> deadline for adding to release notes?
15:17:02 <joanie> #action Joanie will help with proofreading, etc. of the release notes summary.
15:17:19 <jjmarin> they are going to publish GNOME 3.2 release notes in about five weeks
15:17:36 <clown> thanks jjmarin
15:17:36 <API> lets try to finish this item, as we have 15 minutes, lets try to finish marketing point
15:17:38 <API> sorry guys
15:17:54 <API> #topic 2.4 List of a11y devices compatibles with GNOME ?
15:18:06 <API> jjmarin, could you elaborate that?
15:18:29 <jjmarin> It was mention by Sri that we don't have a list of devices who works with GNOME
15:18:58 <jjmarin> It sounds good to me to have such list
15:19:14 <jjmarin> think like braille readers and so on
15:19:25 <jjmarin> s/think/things/
15:19:35 <API> hmm yeah, but it can be a really hard work, due all the amount of hw vendors possible
15:20:16 <joanie> jjmarin: I agree to a certain extent, but much of what we do is software so I think that is what we want to focus on -- both for what we have, and (by virtue of what will not be on the list) what we lack
15:20:44 <joanie> in other words, we want to get potential funders to see what cool things we have and wonder things like, "so... where's the speech recognition support?"
15:20:47 <joanie> ;-)
15:21:48 <API> ok, so in summary,
15:21:52 <joanie> but we could write something generic (not vendor specific) about things like refreshable braille displays
15:21:54 <API> that it would be good to add that
15:22:03 <API> but doesn't require to be extensive on the list
15:22:06 <Pendulum> I think having something generic is good
15:22:16 <API> joanie, yes exactly
15:22:46 <joanie> #action Team members who know about particular hardware types should get this information to Juanjo
15:22:46 <API> so, anything else about this point?
15:22:51 <jjmarin> We can start with something, and find ways to getting better.
15:22:59 <joanie> i.e. I can write up braille displays
15:23:20 <joanie> #action Joanie will generate content about braille displays for Juanjo
15:23:21 <joanie> (done)
15:23:50 <API> ok, less that 10 minutes till the end
15:23:55 <API> #topic libatspi documentation
15:24:00 <API> alibezz, mgorse ?
15:24:12 <alibezz> hi everybody
15:24:28 <alibezz> so... I did some adjustments on libatspi documents
15:24:37 <alibezz> mgorse and joanie are aware of them
15:24:56 <alibezz> last week someone suggested
15:25:12 <alibezz> - sorry, I can't remember now who did it, but I think it was tbsaunde -
15:25:29 <alibezz> that the libatspi constants that arre also present in other documents
15:25:34 <alibezz> such as atk
15:26:02 <alibezz> should be documented in one single place and simply referred in the documents that use them
15:26:11 <alibezz> it would be good for consistency, for example.
15:26:21 <alibezz> I think that's a good ide ain general:
15:26:34 <alibezz> thinking about shared points in core a11y documents
15:26:40 <alibezz> and putting them in one single place
15:26:58 <alibezz> I don't know yet which would be a good way of doing it
15:27:08 <alibezz> but it really sounds like a good idea to me.
15:27:24 <API> hmm
15:27:42 <API> alibezz, yes, but at this moment we lack that shared place
15:27:58 <API> does mean that you removed that documentation from your libatspi2 documentation?
15:28:11 <alibezz> API: yeah, that's what I was thinking... I don't know what would be a proper place to put all shared content
15:28:19 <alibezz> no, I didn't :)
15:28:47 <mgorse> It reminds me of company suggesting/thinking a while ago that atk should be used from the AT side as well
15:28:48 <alibezz> it's just an idea for the future, if feasible.
15:28:54 <API> well, the issue is that although we try to get ATK and atspi as synced as possible, thats
15:28:59 <API> not true at this moment
15:29:08 <API> mgorse, yes thats a good point
15:29:21 <mgorse> and I *think* this would basically work, since atk is mostly just a set of interfaces, although it might require a few API changes (ie, what if an application like Orca is both an implementor and an AT?)
15:29:29 <API> if in any specific future at-spi and atk got merged the shared place is obciuos
15:29:30 <API> obvious
15:29:36 <mgorse> so I wonder if that's a direction to take in the future, but might require a longer discussion / more thought
15:30:06 <API> mgorse, yes, I agree
15:30:14 <clown> what about other a11y apis that use atspi?  like javax.accessibilty?
15:30:27 <alibezz> I agree, it's just to check if this is a crazy thought or not :)
15:30:33 <API> in summary, although that proposal makes sense
15:30:49 <API> probably in the short term if would be required to maintain the current status
15:31:10 <mgorse> clown: if you mean java-atk-wrapper, it doesn't use AT-SPI directly
15:32:07 <clown> mgorse, I was thinking a little more high level but trying a concrete example.
15:32:12 <API> clown, in the old times, its true that java uses at-spi
15:32:22 <API> but after at-spi2 they moved to use a wrapper
15:32:22 <clown> so, atspi is a service for all a11y apis to publish through.
15:32:33 <clown> and ATK is just one a11y api that's published.
15:32:52 <API> hmm, well
15:33:02 <API> at this moment libatpi2 is used mainly by ATs
15:33:16 <API> ATK is just a abstract bunch of interfaces
15:33:24 <API> so the at-spi2-atk bridge
15:33:34 <API> uses it to communicate the apps with at-spi2
15:33:40 <clown> truth be told, I'm leaning towards everything heading to something like aria, but that's way in the future.
15:34:04 <mgorse> clown: Okay; I think I know what you mean. There would still need to be IPC; what I was suggesting wouldn't even necessarily mean changing the IPC, although it could
15:34:36 <mgorse> ie, libatspi could define its objects based on atk interfaces, and they would have virtual functions that would make IPC calls
15:34:53 <clown> mgorse, I like that.
15:35:06 <API> and taking into account that we are 5 minutes over time, and we are entering on offtopic places ...
15:35:06 <API> ;)
15:35:15 <API> anything else about libatspi documentation?
15:35:17 <joanie> clown: I dig the aria idea. :-)
15:35:30 * joanie would appreciate an #info for the minutes
15:35:32 <clown> there should be some #info's for what alibezz wrote about...
15:35:32 <mgorse> FYI, I committed what alibezz has sent me so far, so it'll be in the next release on Monday.
15:35:43 <alibezz> ok
15:36:45 <alibezz> #info The idea of documenting shared parts of core a11y libs in one single place, for consistency, ws discussed.
15:37:33 <alibezz> #info The consent is that it's hard to think about it in a short term, because most a11y libs are not sync'd as they should ideally be
15:38:03 <alibezz> #info also, the libatspi documentation was updated with some alibezz' patches
15:39:50 <API> alibezz, anything else?
15:39:58 <alibezz> nope :)
15:40:05 <API> alibezz, ok thank you
15:40:10 <API> so, lets finish the meeting
15:40:19 <API> as it is one of our favourite sections
15:40:24 <API> #topic miscellaneous time
15:40:31 <API> any comment not
15:40:34 <API> included on the agenda?
15:40:37 <API> (short) comment
15:42:08 <mgorse> FYI, I implemented part of fregl's proposal for determining/setting accessibility
15:42:25 <API> mgorse, do you plan to include it for 3.2?
15:42:30 <mgorse> ie, the bus launcher now has a org.a11y.Status.IsEnabled. Currently it can only be read, not written to
15:42:38 <API> I saw your message on the release-team ml without too much feedback
15:42:46 <joanie> #info the CSUN call for proposals officially came out a few hours ago. Anyone wishing to present a paper at CSUN should start drafting. ;-)
15:42:55 <fregl> mgorse: cool. I finally pushed the change to take the dbus a11y-address from dbus
15:43:10 <mgorse> API: Matthias told me that the freezes and next release were pushed back a week and so I didn't need to ask for permission, so I just committed it
15:43:49 <API> mgorse, ah ok, I understood that you were working on that, but it was not finished yet
15:43:55 <mgorse> fregl: Are you still checking the X property first?
15:44:13 <fregl> mgorse: no, I check the dbus and then as fallback the x property
15:44:32 <fregl> mgorse: I need to get some stuff ready for wayland. using dbus there is very much welcome.
15:44:41 <mgorse> API: Well, reading works, except that enabling/disabling at runtime still doesn't really work. Writing to it is sitll TODO
15:44:51 <API> and it seems that you have some people testing it ;)
15:44:53 <API> mgorse, ok thanks
15:45:03 <API> well, and taking into account that we are 15 minutes over time
15:45:08 <API> Can I close the meeting?
15:45:12 <fregl> did I mention? mgorse is my hero :)
15:45:27 <mgorse> I should #info that
15:45:34 <fregl> definitively ;)
15:45:48 <mgorse> #info mgorse implemented org.a11y.Status.IsEnabled into the bus launcher. Currently it can be read but not written. Writing is still TODO
15:46:01 <fregl> oh, I thought the hero bit.
15:46:10 <clown> fregl, you should #info *that*
15:46:30 <mgorse> wanted to get that in so that Firefox, etc can have an API to start coding to
15:46:45 <mgorse> other than looking at GSettings, which I don't know will work outside of GNOME
15:46:52 <fregl> #info mgorse is fregl's hero
15:47:19 * joanie grins
15:47:27 <joanie> I am leaving that in the official minutes
15:47:28 * clown go mgorse go!
15:47:45 <joanie> And we shall always remember the heroism of mgorse for years to come
15:47:56 <API> a good conclusion for the meeting ;)
15:48:02 <joanie> +1 API
15:48:03 <API> #endmeeting