(15:01:11) API: well, meeting time
(15:01:32) API: agenda here
(15:01:35) API: http://live.gnome.org/Accessibility/Meetings
(15:01:47) API: probably we should change the channel topic with this link
(15:01:59) joanie: API that's done already
(15:02:06) joanie: unless you want it just that topic
(15:02:26) msanchez [~msanchez@155.99.117.91.static.mundo-r.com] entered the room.
(15:02:34) API: ah ok
(15:02:42) joanie: I didn't want an uber-long topic
(15:02:44) API: well, lets start with
(15:02:52) API: please raise your hands if you are here
(15:03:03) ***fer does
(15:03:04) ***joanie raises her hand
(15:03:20) ***mgorse raises his hand (not that anyone can see him doing it)
(15:03:24) ***msanchez did it, but nobody was looking at him
(15:03:39) ***API I looked to msanchez ... he is a liar
(15:05:26) API: ok, 5 introduction minutes are enough
(15:05:31) API: lets start the project
(15:05:34) API: ups
(15:05:37) API: start the meeting
(15:05:44) ***eeejay raises hand
(15:05:49) API: 1. Documentation
(15:05:54) ***joanie hopes shaunm will raise his
(15:05:58) API: shaunm, are you here? you are the star of this point
(15:08:19) API: well, I guess that shaunm is not right now here
(15:08:25) API: so lets move the point for later
(15:08:29) shaunm: oh oh!
(15:08:30) shaunm: sorry
(15:08:31) joanie: lol
(15:08:35) ***joanie kicks shaunm
(15:08:37) ***fer waves
(15:08:38) API: ah
(15:08:43) API: cool
(15:08:55) API: please joanie and shaunm lets go with the documentation item
(15:09:28) shaunm: so
(15:09:42) shaunm: I've been working on stubbing out pages in the desktop help for basic a11y features
(15:09:47) shaunm: http://people.gnome.org/~shaunm/a11y-help.png
(15:10:07) shaunm: (ignore the double posts of links under color-blindness. that's a yelp bug I've already fixed)
(15:10:32) ***clown raises his hand (sorry for being a little late).
(15:10:40) shaunm: the idea is that we will at least present every accessibility feature in the main desktop help
(15:11:12) shaunm: things that need lots more documentation (orca) may have their own application help, but we'll still present orca in the desktop help, then link to the app help for more information
(15:12:26) shaunm: my organization right now is more or less feature-based, though I've tried to categorize them by need
(15:12:55) shaunm: devoid of real information on what types of people need what, the feature breakdown is the only way I can present things
(15:13:09) shaunm: (and maybe that's ok)
(15:13:49) joanie: I don't think it's a bad thing
(15:14:01) clown: I'm wondering if the organization should match the new preferences gui?  Seeing, for example: http://live.gnome.org/Accessibility/NewPreferencesGUI#Seeing_Tab
(15:14:23) clown: actually, I think it sort of does match, already.
(15:14:41) shaunm: clown: yeah, I think the new prefs gui goes a long way towards organizing things the way I'd organize the help anyway
(15:15:05) clown: cool
(15:15:29) shaunm: though there are things I don't see in there
(15:15:53) fer: yeah, like colorblind stuff :)
(15:16:15) fer: who is implementing universal preferences dialog now?
(15:16:29) joanie: william jon mccann
(15:16:31) shaunm: for example, the Locate Pointer feature, and increasing the pointer size. in gnome 2, we tie those to the mouse prefs, but they're really for people with vision problems
(15:16:38) clown: http://live.gnome.org/Accessibility/GNOME3#New_Universal_Access_Preferences_UI
(15:16:41) clown: fer ^
(15:16:57) shaunm: fer: do we have anything for color-blind people, short of telling us that we can change colors in some apps?
(15:16:58) clown: although that section needs updating.
(15:17:23) fer: shaunm: in GNOME 2 we had colorbling applet
(15:17:46) shaunm: fer: didn't that just change screen colors to simulate color-blindness?
(15:17:57) joanie: the applet did
(15:18:01) joanie: but Orca actually uses it
(15:18:12) joanie: and provides color filtering through libcolorblind
(15:18:15) shaunm: that's more of a feature for developers, I would think. so they can test their apps.
(15:18:43) joanie: I mean Orca actually uses the library, not the associated applet
(15:19:00) joanie: but it's optional as libcolorblind is (to my knowledge) not a blessed dependency
(15:19:12) joanie: and I gather from Fer (I think it was Fer) that it's in bad shape. :-(
(15:19:13) shaunm: all right, well, if there's something to put under there, we'll put it there. the heading won't show up if there's no links
(15:20:05) shaunm: so, things I would like to know:
(15:20:55) shaunm: what's up with caribou? gok had all sorts of stuff for using a switch. it was way more than a clicky-clicky gui keyboard
(15:21:23) joanie: yeah, eeejay, what's up with caribou?
(15:21:25) joanie: ;-)
(15:21:31) fer: gok had switch support, but it won't work on gnome 3
(15:21:40) shaunm: also, I'd like to know which of our a11y features are likely to be interesting to people using a switch
(15:21:44) eeejay: well, i have been putting some time in cleaning the current code base up
(15:22:00) eeejay: scanning support is in
(15:22:17) eeejay: right now i am migrating to pygi
(15:22:26) eeejay: that's about it..
(15:22:52) eeejay: so shaunm, caribou has initial switch support, it's not completely fantastic yet
(15:23:30) shaunm: ok
(15:24:26) shaunm: more generally, I'd love to have a better sense of what kinds of mobility problems people might have, what sorts of hardware they use to help them, and which features are interesting to which people
(15:24:45) shaunm: which, joanie, I think goes along with the a11y personas we talked about the other day
(15:24:51) joanie: indeed
(15:25:17) joanie: the blindness and low vision ones are on my to-do list after docs and merging a new feature into Orca
(15:25:23) joanie: but I cannot get you other personas
(15:25:50) joanie: I can help think about them; just I'm not the expert for the non-vision-related stuff
(15:26:03) eeejay: you guys know of the aegis personas, right?
(15:26:10) ***joanie doesn't
(15:26:18) shaunm: me neither
(15:26:22) ***clown ashamed to say I don't
(15:26:24) ***joanie has been largely unaffected by aegis :-/
(15:26:36) eeejay: http://www.aegis-project.eu/index.php?option=com_content&view=article&id=63&Itemid=53
(15:27:05) joanie: sweet
(15:27:09) eeejay: tada!
(15:27:09) joanie: eeejay ftw
(15:27:23) shaunm: oh nice
(15:27:31) eeejay: and they have some liberal license
(15:28:41) shaunm: this is helpful
(15:28:46) API: I don't this point, but we are reaching the half hour deadline
(15:28:55) API: I think that it is worth to start to think
(15:28:56) API: in 
(15:28:58) API: conclusions
(15:28:59) API: actions
(15:29:23) API: and then try to coordinate this point out of the meeting, if more level of detail is required
(15:29:47) ***API this "I don't this point" should be "I don't want to spoil this point"
(15:29:52) shaunm: right
(15:30:08) joanie: so tentatively shaunm: Are you set with personas?
(15:30:12) joanie: for now?
(15:30:14) shaunm: for now
(15:30:19) ***joanie nods
(15:30:32) shaunm: I would like to determine if any other areas need their own help documents
(15:30:46) shaunm: possible candidates are the magnifier and caribou
(15:30:57) joanie: what about accerciser?
(15:30:57) shaunm: plus other alternative inputs like dasher
(15:31:07) shaunm: more of a developer feature, right?
(15:31:10) joanie: yup
(15:31:27) joanie: still needs a bunch of documentation though
(15:31:45) shaunm: so it should be talked about in the accessibility developer's guide (on my long list)
(15:31:53) joanie: gotcha. my bad.
(15:32:17) shaunm: um
(15:32:32) eeejay: accerciser has it's own docs as well
(15:32:39) shaunm: and I guess just one more question (which can be answered later), then I'm done
(15:32:55) shaunm: eeejay: right, and it should. but it should be presented in the devel guide
(15:33:08) eeejay: right
(15:33:22) shaunm: to what extent should I present or talk about gnome-friendly a11y tech that isn't in our modulesets?
(15:33:35) fer: dots?
(15:33:41) shaunm: mousetrap, opengazer, any of the crazy cool things those chaps from cambridge showed me
(15:33:43) shaunm: dots
(15:33:45) joanie: ugh, guys, you know....
(15:33:53) joanie: fer is here to talk about API Changes
(15:34:01) joanie: and eeejay CSUN
(15:34:04) joanie: so....
(15:34:07) joanie: in summary,
(15:34:08) fer: and for the rest of the meeting :)
(15:34:08) shaunm: move on
(15:34:12) joanie: Shaun doesn't need personals
(15:34:14) shaunm: I'll send a list of questions to the list
(15:34:15) joanie: ;-)
(15:34:17) joanie: super
(15:34:21) joanie: (sorry shaunm!)
(15:34:21) fer: great
(15:34:24) fer: so I take it then
(15:34:29) eeejay: shaunm: unrelated to docs, i would push those for module inclusion if they are ready
(15:34:36) eeejay: shaunm: and then happily document
(15:35:24) ***API guessing if point 1 is over or not
(15:35:31) joanie: it is
(15:35:34) ***API 25 till the end
(15:35:39) API: ok, thanks
(15:35:44) API: lets move
(15:35:46) API: point 2
(15:35:47) fer: ok, so my take
(15:35:52) fer: API changes
(15:35:52) API: 2 AT-SPI/Atk API changes proposed (fer) 
(15:36:10) API: fer this is related with this¿
(15:36:13) API: http://live.gnome.org/Accessibility/Roadmap#Augmented_and_Consistently-Implemented_Atk
(15:36:27) fer: currently I have two things I would like to chagne, but also I would like to discuss the better way to do it and what other things you may want to change
(15:36:58) fer: well, some issues are covered there
(15:37:15) API: fer, hmm but this sound as a long discussion about a specific technical aspect
(15:37:24) API: and weekly meetings purpose is not that ;)
(15:37:30) fer: yeah, I can keep technical details for the mailing list
(15:37:40) fer: but I would like to discuss about API breaks vs. API aditions
(15:38:17) API: fer ah ok
(15:38:21) fer: that is, at this point we can change atk/at-spi API because GNOME 3 (gtk) is going to do it anyway
(15:38:26) API: about atk2 or just improve atk
(15:38:44) fer: but my question is... is it worth for so small changes?
(15:38:47) ***joanie wouldn't mind both places
(15:38:50) joanie: fer it is
(15:39:02) fer: ok
(15:39:11) API: you mean if it is worth to make API breaks for small changes?
(15:39:11) joanie: text changed is broken for java, and sounds like gecko
(15:39:28) joanie: text selection sucks right now, and doesn't work in Orca+Gecko
(15:39:30) API: joanie, but this is a problem with ATK?
(15:39:32) fer: and then, probably, as the break window is open, we may want to change some other stuff at the same time, so I'll make a call
(15:39:41) API: or it is because the current implementation?
(15:39:44) joanie: API both Atk and AT-SPI
(15:39:52) API: fer at this point
(15:40:01) API: we will not arrive at time for a ATK API break
(15:40:12) fer: really?
(15:40:13) API: gtk3 people are starting to closing the doors
(15:40:23) API: in fact this is the reason they are already planning gtk4
(15:40:27) joanie: lol
(15:40:29) API: because they detected some 
(15:40:32) API: things
(15:40:38) API: that will not arrive in time for gtk3
(15:40:44) joanie: API then I still think we aim for backwards compatibility
(15:40:46) API: gtk3 needs to be finished before gnome3
(15:40:53) joanie: and add new stuff
(15:41:06) fer: yeah, and mark the other stuff and deprecated
(15:41:09) joanie: because we have existing problems in other toolkits than Gtk
(15:41:13) fer: and remove it for atk4 :)
(15:41:17) joanie: lol
(15:41:25) mgorse: Is there a date for API freezes? I think it usually concides with .5
(15:41:30) API: well but fer mentioned this breakage because gtk3 ;)
(15:41:49) API: mgorse, I don't know, I have never be too good for that
(15:41:55) API: but the issue is 
(15:42:04) API: if we are planning to improve ATK and at-spi API
(15:42:08) joanie: mgorse: I think you're right. Which is in like 3 or 4 weeks
(15:42:12) API: that could lead to some breakage
(15:42:16) API: this will require debate
(15:42:20) API: this will mean some weeks
(15:42:22) API: etc
(15:42:26) ***joanie frowns
(15:42:27) API: so
(15:42:31) API: we can start that work
(15:42:32) fer: Jan 31  is API freeze for gnome 3
(15:42:45) API: but plan this new atk2, or atk3 or atk4
(15:42:47) mgorse: TwoPointNinetyone lists the API/ABI freeze as on 2.91.6 / Jan 31
(15:42:54) API: for gnome3 would not be really sensible
(15:43:01) API: IMHO of course
(15:43:21) mgorse: but, yeah, there are potential issues with Gecko, OpenOffice, etc., which have release cycles independent of GNOME
(15:43:32) joanie: ditto for Java
(15:43:37) fer: if we don't fix it now, we are going to have broken apps out there for many years
(15:43:46) API: joanie, about my question
(15:43:50) API: the text selection stuff
(15:43:55) API: is a problem of ATK?
(15:44:01) API: on Java, Gecko etc
(15:44:05) joanie: more complex than that
(15:44:11) joanie: so it's two-fold
(15:44:15) mgorse: I don't see a reason not to add things, but I think we need to preserve backwards compatibility for now
(15:44:19) joanie: in part the current implementation just bites
(15:44:40) API: mgorse, I agree
(15:45:05) joanie: in part, Gecko's broken caret navigation means that if we had better support from Atk/AT-SPI, Orca would automatically support reporting Gecko selection
(15:45:16) joanie: I don't want to deep dive in this
(15:45:17) fer: ok, so the plan could be, add new stuff, and mark the old api as deprecated in order to remove it in the future?
(15:45:20) joanie: given the hour
(15:45:28) joanie: +1 fer
(15:45:29) API: well, and joanie take into account that we put the "ATK revamp" for april 2011 ...
(15:45:39) joanie: true
(15:45:49) fer: ok, so that is a good solution
(15:45:51) joanie: I think we should have an Atk hackfest
(15:45:56) joanie: seriously
(15:46:09) API: joanie, so no a accessibility hackfest with a ATK "party"?
(15:46:16) API: a specific atk hackfest?
(15:46:16) joanie: no
(15:46:17) fer: I'll mail to the list proposing addition of new signals/prototypes for text-inserted, text-updated, text-removed and selection-changed
(15:46:18) joanie: correct
(15:46:29) clown: fer, which list?
(15:46:32) mgorse: Do we want to do that at CSUN (possibly lead into next topic)? Or in Boston for instance?
(15:46:33) clown: gnome-a11y?
(15:46:42) fer: wow, having a atk hackfest with the java guys + mozilla + any other parties would be SO awesome
(15:46:46) fer: clown: yup
(15:46:51) joanie: or spain
(15:46:51) clown: fer, thanks
(15:46:54) API: fer, when you mean new? you mean new signals with new names, or change the meaning of the current ones?
(15:47:14) fer: new signals with new names is the only way to not break API/ABI
(15:47:15) joanie: because many of the people who need to be in spain are there. 
(15:47:17) API: well, indirectly this would be also at-spi related hackfest
(15:47:28) joanie: of course
(15:47:30) API: ie, I think that it would be good 
(15:47:34) joanie: Atk+AT-SPI
(15:47:44) API: making at-spi API more similar to the ATK one
(15:47:49) API: I know that this is hard
(15:48:00) API: as at-spi is more general than ATK
(15:48:06) API: but just an idea
(15:48:12) API: well, 10 minutes to go
(15:48:16) API: so as I said before
(15:48:17) API: we need
(15:48:20) API: conclusions
(15:48:21) API: actions
(15:48:33) API: conclusions: we can't break ATK API breakage now
(15:48:51) API: 2. It shouldn't a problem to add things and mark methods/signals as deprecated
(15:49:12) API: 3. This can be the initial work for the following ATK2 or atk4 work
(15:49:22) API: I say atk4, because with this gtk4 thing
(15:49:30) API: gtk people are thinking in a glib4
(15:49:36) Pendulum: eep sorry I forgot about the meeting!
(15:49:41) API: Pendulum, no worry
(15:49:58) API: 4. Debate about a specific atk hackfest
(15:50:02) API: I forget something?
(15:50:14) fer: nop, very good summary
(15:51:20) API: ok, so lets move
(15:51:24) API: btw: eeejay 
(15:51:32) eeejay: ?
(15:51:32) API: we usually don't add items on the agenda 
(15:51:36) API: during the meeting :P
(15:51:45) eeejay: API, i didn't
(15:51:47) API: we have miscellaneous time for that kind of things ;)
(15:51:51) API: ah ok
(15:51:58) API: sorry
(15:51:59) eeejay: API, added it yesterday
(15:52:08) API: I misunderstood joanie
(15:52:12) API: 10 minutes
(15:52:22) API: so please eeejay wait a moment
(15:52:28) API: I have a fast item (I hope)
(15:52:36) API: 3. gtktreeview
(15:52:48) API: as you already know gtktreeview is one of the stars of gtk
(15:52:54) API: and a really huge widget
(15:52:55) fer: and of gail! :)
(15:53:09) API: yes, and a huge piece of code on gail
(15:53:14) API: right now they are revamping it
(15:53:20) API: gtktreeview and all the cell stuff
(15:53:25) API: and they are finishing it
(15:53:44) API: in theory things should still work
(15:53:44) ***joanie hopes it won't be any less performant than the beast already is
(15:53:57) API: but, of course they didn't test it
(15:54:02) fer: is that for gtk3, right?
(15:54:04) joanie: (continue to work and suck like they did before you mean?)
(15:54:06) mgorse: Yeah. The gail treeview code has serious performance issues; it would be nice if it could be improved somehow, but I've never looked at it that carefully
(15:54:21) API: Tristan words summarizes it:
(15:54:27) API: I expect libgail to not need any changes, we paid special attention
(15:54:27) API: that focus navigation was going to work the same way using the new
(15:54:27) API: code... however there could be some fallout I'm not aware of...
(15:54:33) API: well, yes performance is a issue
(15:54:45) API: but the thing is that now we should check that it works
(15:54:50) fer: but it is not a regression :)
(15:54:51) fer: agreed
(15:54:53) joanie: super
(15:55:07) API: so
(15:55:14) fer: API: are those changed on gtk3 trunk? or in a separate branch?
(15:55:21) API: conclusion: gailtreeview is not tested with this revamped gtktreeview
(15:55:42) API: fer: a separate branchh for the moment, although this guy didn't confirm my question about the specific one
(15:55:53) API: action: anyone with time should test it
(15:56:08) API: and I will send a mail to Li Yuan about it, as he is not here
(15:56:12) API: 5 minutes to go!!
(15:56:20) API: eeejay, your move
(15:56:22) mgorse: We need to know which branch and/or when it'll be merged into master
(15:56:27) eeejay: ok!
(15:56:35) API: mgorse, when==soon
(15:56:47) eeejay: we have CSUN coming up, as I understand there won't be a hackfest there which is great
(15:57:26) eeejay: i am sooo swamped that bryen took on a lot of the organizing, so i am not sure who besides him and me are volunteering
(15:57:38) eeejay: in any case we will be about 4-5 people
(15:57:50) eeejay: i guess we should check with the travel committee?
(15:58:13) eeejay: bryn. me and korn will be presenting
(15:58:22) eeejay: um
(15:58:25) fer: would we have a booth there?
(15:58:36) eeejay: fer, yeah, that is the main thing
(15:58:49) eeejay: fer last year we had booth + presentations + hackfest
(15:59:21) eeejay: otoh, if anyone is interested in meeting and collaborating there on project let me know
(15:59:32) API: eeejay, when joanie and me contacted the foundation about a buget
(15:59:40) eeejay: when is the next a11y meeting?
(15:59:41) fer: I may go to CSUN, but I don't know yet
(15:59:41) API: we include dCSUN
(15:59:56) API: eeejay, we have one each Thursday ;)
(16:00:00) joanie: but not 4 or 5 people
(16:00:08) API: http://live.gnome.org/Accessibility/Meetings
(16:00:23) eeejay: great! Next thursday i will have a clearer agenda re:csun, this is more of a heads up
(16:00:40) eeejay: joanie: how many people?
(16:00:40) mgorse: eeejay: It's actually 11pm Wednesday your time
(16:00:48) joanie: 3
(16:00:57) API: anyway, this budget this is somewhat stucked, I guess that they are waiting for a final roadmap
(16:01:06) eeejay: joanie: bryen is taken care of, no need for funding
(16:01:08) joanie: They have our roadmap
(16:01:10) API: and these days gpoo is mssing
(16:01:11) API: missing
(16:01:23) eeejay: i think we have someone else on board, someone from india?
(16:01:46) joanie: eeejay: I was planning on going on the gnu dime, but that looks like it's going to not happen
(16:01:51) joanie: (was going to be on a panel)
(16:02:06) joanie: someone from india?
(16:02:43) eeejay: i forgot his name
(16:02:53) joanie: does he expect funding from us?
(16:03:09) eeejay: joanie: i'll get back to you on this
(16:03:16) joanie: or to put it another way....
(16:03:34) joanie: staffing a booth is not something that requires international travel
(16:03:40) joanie: which tends to be costly
(16:03:50) eeejay: joanie: right
(16:03:55) ***eeejay agrees
(16:04:01) joanie: and given a limited (extremely limited) budget
(16:04:08) joanie: I think we fund active team members 
(16:04:19) joanie: over people whose names we can't even remember ;-)
(16:04:27) eeejay: oh, maybe he is stateside, rohan?
(16:04:33) joanie: stateside is better
(16:05:06) joanie: anyhoo, API is going to crack the whip and demand conclusions
(16:05:08) joanie: ;-)
(16:05:11) API: eeejay, in which gnome project this india guy is related?
(16:05:18) eeejay: yeah, he is in india, ok. we will have to disappoint him..
(16:05:27) ***API API using a totally broken grammar sentence
(16:05:41) eeejay: API, he is Bryen's protege :)
(16:05:42) joanie: the meaning behind it is still valid
(16:06:38) API: eeejay, ok, but he is working on a specific gnome project? 
(16:08:32) eeejay: API, can't answer that. looks like he is done a bunch of research things. in any case the travel expenses and limited involvement probably does not justify it
(16:08:51) API: eeejay, ok, as you said, you need some details
(16:08:56) API: so, for the minutes sake
(16:09:04) API: can you summarize your point?
(16:09:06) API: conclusions
(16:09:08) API: actions
(16:10:03) eeejay: csun is in march, we have space for 2 more people, let me know if you want to volunteer, more details next week
(16:10:25) mgorse: eeejay: I'm fairly certain I'll be there. Should not need funding, though
(16:10:25) API: eeejay, ok, thsnks
(16:10:28) API: thanks I mean
(16:10:35) eeejay: mgorse: sweet!
(16:10:42) API: well 10 minutes over the time
(16:10:49) eeejay: not bad
(16:10:51) API: I think that we can declare the meeting over
(16:11:07) ***clown listens for the sound of a gavel
(16:11:08) API: sorry if someone was not able to say anything
(16:11:17) API: but you have the ml 
(16:11:23) API: meeting over