16:05:55 <API> #startmeeting
16:05:55 <tota11y> Meeting started Thu Aug 22 16:05:55 2013 CET.  The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:05:55 <tota11y> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:06:09 <API> #topic GSoC updates
16:06:12 <API> magpie_desktop, ?
16:06:19 <magpie_desktop> #info this week was very unproductive and the focusCaretTracker and magnification edits did not make 3.9.90
16:06:38 <magpie_desktop> #action Magdalen will make sure the changes are in time for 3.9.91
16:06:48 <magpie_desktop> #info the delay was due to a number of factors:
16:06:57 <magpie_desktop> #info 1) The information Jasper gave me about how to run gnome-shell in jhbuild was not
16:07:00 <magpie_desktop> correct and this messed up the symlinks to libaries that are needed. Magdalen is not sure why that happened but she suspects it was because Jasper having too many conversations at once.
16:07:11 <magpie_desktop> #info 2) The discussion about whether to make the tracker a singleton or not went on for too long.
16:07:22 <magpie_desktop> #info Magdalen feels she should be involved in decisions about work that she has to implement especially when the decisions make things more difficult to implement
16:07:44 <magpie_desktop> #info  3) Magdalen broke her computer before guadec and when she finally got it working that was a terrible connection
16:08:28 <magpie_desktop> #info Magdalen apologises for the delay and has nothing further to add
16:08:35 <magpie_desktop> questions?
16:08:42 <API> magpie_desktop, about 2)
16:08:49 <API> when the thread at gnome-accessibility-devel
16:08:50 <API> started
16:09:03 * clown waves
16:09:04 <API> I assumed that you were subscribed to that list
16:09:24 <API> taking into account some of the comments at bugzilla and in private emails
16:09:27 <API> it seems that not
16:09:39 <API> my idea was including you on the discussion since the beginning
16:09:40 <magpie_desktop> yeah but clown specifically told me not to involve myself
16:09:51 <API> I was surprised for not seeing any mail from you
16:09:55 <API> ah ok
16:10:00 <API> then never mind
16:10:03 <API> in any case
16:10:08 <API> gnome-shell reviewer (Jasper)
16:10:15 <API> already mentioned to be pragmatic
16:10:18 <API> something we all agreed
16:10:25 <API> so for the moment lets not focus on that
16:10:36 <API> so, my other comment
16:10:38 <API> about this:
16:10:42 <API> <magpie_desktop> #action Magdalen will make sure the changes are in time for 3.9.91
16:10:55 <API> fwiw, the freeze for 3.9.90
16:10:57 <magpie_desktop> that's the next cycle
16:11:00 <API> no
16:11:03 <API> is not the next cycle
16:11:06 <magpie_desktop> i mean bit
16:11:08 <API> is still 3.10 cycle
16:11:15 <magpie_desktop> whatever it's called
16:11:16 <API> so is next step on current cycle
16:11:25 <magpie_desktop> right API thanks
16:11:31 <magpie_desktop> so that is
16:11:35 * magpie_desktop getting link
16:11:39 <API> https://wiki.gnome.org/ThreePointNine
16:11:41 <API> so
16:11:52 <API> 3.9.90 included several freezes
16:12:01 <API> that included feature freeze
16:12:07 <API> so as here is explained:
16:12:32 <API> https://wiki.gnome.org/ReleasePlanning
16:12:33 <magpie_desktop> the 2nd sept but It;ll be in before then
16:12:38 <API> in order to get that into 3.9.91
16:12:45 <API> we need to ask for a freeze break
16:12:49 <magpie_desktop> no we don't
16:12:58 <magpie_desktop> it's not subject to the freeze
16:13:00 <API> magpie_desktop why not?
16:13:07 <magpie_desktop> they've told me
16:13:15 <API> who told you that?
16:13:29 <magpie_desktop> jasper. mclaren quite a few
16:13:47 <API> while, I will clarify with them
16:13:54 <API> but being strict with the "law"
16:14:00 <API> tracking is listed here:
16:14:03 <API> https://wiki.gnome.org/ThreePointNine/Features
16:14:05 <API> as a feature
16:14:20 <API> and current freeze is also feature freeze:
16:14:24 <API> " Feature Freeze: new functionality is implemented now"
16:14:36 <API> in any case
16:14:45 <magpie_desktop> well why was it not in the tiimeline then?
16:14:46 <API> what I wanted to say, with my release team hat on
16:14:56 <API> is that I could handle the request for feature freeze
16:15:27 <API> so, I can change that to "If needed, I will handle the request fore feature freeze break"
16:15:30 <magpie_desktop> everyone here seemed to suggest it wouldn't be subject to this freeze so this is the first i've heard that
16:15:48 <API> magpie_desktop, what do you mean that it was not in the timeline?
16:15:53 <API> is in the timeline here:
16:15:58 <API> https://wiki.gnome.org/ThreePointNine
16:16:06 <API> see 3.9.90
16:16:22 <API> and it was already implicitly suggested that was included
16:16:32 <API> because we always told you to try to get this for 3.9.90
16:16:34 <API> in any case
16:16:38 <API> as I said
16:16:38 <magpie_desktop> well you all helped me write it and nothing about this is in there
16:16:58 <API> *if needed*, I can request the feature breeze break
16:17:16 <API> so summarizing this with pretty bot-labels
16:17:30 <magpie_desktop> API that's not true
16:17:41 * API holding
16:17:44 <API> magpie_desktop, what is not true?
16:17:53 <magpie_desktop> clown asked me to wait till he got back on 19th to even commit
16:18:09 <API> ah, ok
16:18:11 <API> then sorry
16:18:11 <magpie_desktop> so no, that's not what you always said
16:18:17 <API> you got mixed messages
16:18:35 <API> for example mclasen suggested to have this for 3.9.90 on some of the bugs
16:18:40 <API> our bad
16:18:50 <clown> yes, I think I misread the timeline.  I was fixated on the code freeze date of Sep 16.
16:19:10 <clown> I did not think that Aug 19 was a deadline for this work.
16:19:21 <API> clown, ok,
16:19:22 <API> in any case
16:19:27 <clown> however,...
16:19:29 <API> it seems that jasper and mclasen
16:19:37 <API> seems ok to have this for sep 16
16:19:40 <API> so
16:19:45 * clown shuts up.
16:19:46 <API> as I was saying about pretty info
16:20:00 <magpie_desktop> 12 August - 23 August: Complete non-UI Tasks
16:20:04 <magpie_desktop> Note: GNOME's Freeze period begins 19 August. This will not give me enough time to complete the UI for the settings for the 3.10 release. But my mentors feel that the focus and caret tracking itself can be finished in time for 3.10 and that it is important to do so. So I will focus on that.
16:20:13 <magpie_desktop> that;s what the timeline says
16:20:32 <API> #action API will confirm if in order to have magpie work for 3.9.91, it is needed a feature freeze request
16:20:48 <API> #action if needed, as soon as magpie work is ready, he will do that request
16:21:00 <joanie> so the timeline mentions the freeze
16:21:36 <API> magpie_desktop, joanie are my actions a proper summary to what we were saying before? or do I need to add something else?
16:21:51 <magpie_desktop> yes and having the tracker in for 3.10
16:22:02 <joanie> I will leave it to you and magpie_desktop to decide that API
16:22:18 <API> magpie_desktop, well, but that was already summarized by you
16:22:27 <magpie_desktop> ok thanks API
16:22:39 <API> so, using magpie_desktop words
16:22:45 <API> more questions/doubts?
16:23:08 <magpie_desktop> API
16:23:36 <magpie_desktop> what do I do when it's ready ? just put a patch on and then ask or ask and then put a patch?
16:23:44 <magpie_desktop> in what order I mean
16:24:05 <API> magpie_desktop, the order is
16:24:12 <API> 1. magpie finish the work
16:24:19 <API> 2. upload a patch for review
16:24:30 <API> 3. patch is reviewed
16:24:40 <API> 3.1 if needs more changes goto1
16:24:46 <API> 4. patch is accepted
16:24:59 <magpie_desktop> 3. rinse and repeat
16:25:04 <API> (during 1-4 API will confirm if the feature break request is needed)
16:25:05 <magpie_desktop> ok thank you API
16:25:30 <API> 5. if needed, API will ask for the feature break request
16:25:42 <API> magpie_desktop, fwiw, every time you request a break
16:25:59 <API> feature freeze break, ui freeze break
16:26:00 <API> etc
16:26:07 <API> you need something to justify the break
16:26:16 <API> and the best thing to justify it
16:26:21 <API> is a working and accepted patch
16:26:22 <magpie_desktop> ok well they accepted this project and the timeline
16:26:44 <clown> emphasis on "working" :-)
16:26:46 <magpie_desktop> and nothing has deviated from that so i don't see why they would have an issue
16:27:06 <API> the only issue that they could have is the lack of a patch ;)
16:27:10 <magpie_desktop> clown: ?
16:27:20 <API> sorry, but without patch they are not going to accept that
16:27:23 <API> said captain obvious
16:27:35 <magpie_desktop> :-) on that note
16:28:03 <magpie_desktop> i am happy if everyone is to change topic??
16:28:35 <API> I have no questions
16:28:42 <API> magpie_desktop, you have a new topic to change to?
16:28:47 <magpie_desktop> well done your slider got committed API
16:29:11 <API> wow, a totally unrelated topic change
16:29:14 <API> in any case, thanks
16:29:27 <API> but there are three bugs related with sliders, that was only one
16:29:34 <API> having said so
16:29:38 <magpie_desktop> nope i am finished with update as long as nobody has questions
16:29:46 <magpie_desktop> that's what i meant :-)
16:30:03 <API> I will change to next (included on agenda) topic
16:30:06 <magpie_desktop> two went in
16:30:25 <API> #topic Testing Evolution Redux
16:30:29 * API one go out
16:30:36 * API sorry for the jokes
16:30:38 <API> so, joanie ?
16:30:47 <joanie> jjmarin: you want to cover this one?
16:30:53 <joanie> since you mentioned it to me in the first place
16:31:04 <joanie> i'll take the action item to test
16:31:50 <joanie> ok, if not....
16:32:11 <jjmarin> yes#info aruiz ask in th
16:32:15 <joanie> #info Juanjo mentioned to Joanie last week that Alberto Ruiz wanted to know if the Evolution accessibility bugs are still present.
16:32:34 <joanie> #info Joanie is sure that they are, but is happy to update the (5+ year old) bugs with new test cases.
16:32:51 <joanie> #action Joanie will do some testing and update the Evolution a11y bugs.
16:32:52 <joanie> done
16:33:05 <jjmarin> thanks joanie !
16:33:10 <joanie> jjmarin: you bet :)
16:33:49 <API> questions time?
16:34:00 <jjmarin> joanie: I think is good idea to ping aruiz when you finished your tests
16:34:12 <joanie> I assume he is cced on the bugs
16:34:16 <joanie> if not, he should be
16:34:30 <joanie> as I update bugs and file any new ones, he'll get notified
16:34:32 <jjmarin> yes, sounds sensible :-)
16:34:36 * API waiting for questions time
16:34:37 <mgorse> Is he an evolution maintainer now?
16:34:54 <jjmarin> he is a manager of applications in red hat
16:36:08 <jjmarin> red hat contributions to evolution, libreoffice and mozilla firefox are under his management AFAIK
16:36:37 <mgorse> ok
16:37:09 <API> as nobody is talking I assume that I can talk now :P
16:37:12 <API> so my question
16:37:27 <API> as far as I remember there are a bazillion of evolution accessibility bugs
16:37:38 <API> is the idea of this task triagge all those bugs?
16:37:44 <joanie> I'll start there
16:37:51 <API> where is there?
16:38:00 <joanie> triaging all the existing bugs
16:38:08 <joanie> then I'll look for others
16:38:13 <jjmarin> AFAIR, he asked about the email composer
16:38:25 <joanie> we have bugs about that
16:38:27 <joanie> :)
16:38:41 <API> well, just worried because that seems a big task
16:38:52 <joanie> API didn't say I would do it *tomorrow* :)
16:38:52 <API> and probably evolution guys could do some of the triagging work
16:39:03 <joanie> if they know about a11y, sure
16:39:06 <API> as far as I see
16:39:14 <API> they see a bazillion bugs on their own component
16:39:19 <joanie> updated (new) listeners might be in order
16:39:21 <jjmarin> we can ask him to start with the areas he is most interested of
16:39:27 <API> and ask others to confirm them
16:39:33 <API> in any case, this is somewhat ranty
16:39:34 <joanie> but if you all want me to not do anything
16:39:46 <joanie> I can delete my action item
16:39:57 <API> joanie, well, if you are willing to do that
16:40:03 <API> I can't object
16:40:09 <joanie> you could
16:40:10 <joanie> :)
16:40:12 <API> just ranting a little
16:40:18 <API> joanie, well, if you are willing to do that
16:40:21 <API> I don't want to object
16:40:25 <joanie> ;)
16:40:28 <joanie> fair enough
16:40:29 <API> so just ranting a little
16:40:32 <API> so no more questions from me
16:40:36 <API> from others?
16:42:40 <API> nobody talking
16:42:43 <API> #topic W3C updates
16:42:46 <API> clown, joanie ?
16:43:05 <joanie> I never have updates on this front
16:43:09 <clown> I am still in catch-up mode this week, so I only have one thing to report.
16:43:32 <clown> #info there is an effort to define a new aria-describedat attribute.
16:43:50 <clown> #info originally, the value would be an url that pointed to a off-page description.
16:44:10 <clown> #info it is similar to longdesc, but could be used for any element.  longdesc is for <img> element only.
16:44:54 <clown> #info it is also contrasted with the existing aria-describedby which takes an idref to an element on the current page.
16:45:19 <clown> #info however, this is still very preliminary.  Some of the ideas are documented here:  https://dvcs.w3.org/hg/aria-unofficial/raw-file/tip/describedat.html
16:45:22 <clown> done.
16:46:16 <API> seems a lot of different ways to describe an element
16:46:52 <API> sometimes giving a lot of options means confuse implementors about which to use
16:47:14 <clown> the history goes something like this.
16:47:14 <API> anyway, if being defined I guess that is because is needed
16:47:51 <clown> 1.  first there was longdesc, which very few authors used correctly, and user agents implemented didn't implement the same way.
16:47:56 <joanie> so if it's described at another page
16:48:13 <joanie> that means a11y cannot do anything until the browser implements support for it
16:48:13 <clown> 2.  aria-describedby was suggested next, and was used consistently by authors and user agents.
16:48:27 <joanie> i.e. going off to a different page and grabbing the info
16:48:41 <clown> 3.  but, authors wanted a way to put the description on a separate page, and you can't do that with aria-describedby.
16:48:41 <joanie> and then exposing it via the a11y framework of the platform in question
16:49:30 <API> joanie concern is also true
16:49:44 <API> not sure how ATs like screen readers could use that attribute
16:49:58 <clown> oh, yes, joanie.  There are questions as to how browsers should implement describeat.  But that's been true of longdesc for a long time.
16:50:38 <joanie> yeah, but longdesc doesn't require the useragent to go to a non-loaded page to fetch stuff
16:50:43 <joanie> right?
16:50:46 <clown> for example, FF implements longdesc as an accessilbe action.
16:50:56 <clown> that is a "go to this page" action.
16:51:08 <clown> it's very similar to a link.
16:51:13 <joanie> ah
16:51:28 <joanie> I guess something similar could be done here then
16:51:37 <clown> But IE simply stuck the url of the longdesc into the accessible description.
16:51:38 <API> clown, so if a screen reader want to use it, it will call the browser to move to a different page
16:51:39 <joanie> sucks that you have to leave the current page to get the info though
16:51:53 <clown> joanie, absolutely correct
16:51:56 <joanie> anyhoo, what do I know?
16:51:57 <joanie> :)
16:52:08 <clown> that's why (IMO) aria-describedby works
16:52:42 <API> well, it seems to work because the info is at the same page
16:52:45 <API> but you said
16:52:50 <clown> right, API
16:52:52 <API> "but, authors wanted a way to put the description on a separate page"
16:52:58 <API> so we are again on the same problem
16:53:04 <clown> yes.  authors keep asking for that.
16:53:07 <API> unless screen readers start to do some magic stuff
16:53:31 <API> like "wget url_at_aria-describedat, then expose"
16:53:43 <API> something that seems more like a browser responsibility
16:53:46 <API> in any case
16:53:49 <API> as you said
16:53:50 <clown> I think it's okay if the AT tells the user somehow that this is going to result in a context change, are you sure you want to do that?
16:53:52 <API> this is still preliminary
16:54:11 <API> so I should not be surprised by the open questions
16:54:15 <clown> well, what do ATs do with ordinary links now, API?
16:54:24 <API> well, but ordinary links
16:54:33 <API> are there to move to a different page
16:54:39 <API> because after all
16:54:43 <API> web pages is about that
16:54:48 <API> moving between pages
16:54:50 <API> but here
16:55:12 <API> we are talking about moving to a different page, just to get the description of a element of the page you are currently reading
16:55:23 <API> doesn't seems really user-friendly to me
16:55:33 <clown> the current though on describedat is that it is for moving to a different page for further informaiton.  That page coudl be a full HTML page.
16:55:33 <API> disclaimer: I just thought about that 2 minutes ;)
16:55:54 <API> in any case
16:56:01 <clown> sure.  it's good to have this kind of discussion around these preliminary ideas.
16:56:02 <API> using again the card of "preliminary"
16:56:11 <API> I will not complain a lot more
16:56:23 <API> and after all, meeting is almost finished
16:56:29 <clown> wow, time flies.
16:56:40 <API> so
16:56:42 <joanie> u.s. airways planes out of philly don't
16:56:47 <API> #topic Marketing
16:56:48 <joanie> <snark />
16:56:51 <API> (aka engagement)
16:56:54 <API> jjmarin, ?
16:56:59 <joanie> aka evangelism
16:57:22 <jjmarin> #info Juanjo has been working on updating some gnome a11y entries in the wikipedia: http://en.wikipedia.org/wiki/Linux_Screen_Reader http://en.wikipedia.org/wiki/Gnopernicus http://en.wikipedia.org/wiki/Orca_%28assistive_technology%29  http://en.wikipedia.org/wiki/Screen_reader http://en.wikipedia.org/wiki/List_of_screen_readers . Work in progress, more changes soon :-)
16:58:03 <jjmarin> and this all my evangelism for this week :-)
16:58:10 <joanie> :)
16:58:15 <API> gnopernicus!
16:58:22 <joanie> heheh
16:58:30 <API> in any case
16:58:33 <API> jjmarin, ok thanks
16:58:39 <API> so questions, doubts?
16:58:46 <clown> does gnopernicus still work?
16:58:51 <jjmarin> yes, I have to put in the list of dead screen readers :-)
16:59:00 <joanie> define still
16:59:08 <joanie> it never did remotely as much as Orca
16:59:14 <joanie> and is of course corba-based
16:59:58 <jjmarin> clown: you need an old gnome version, then it works :-)
17:00:07 <clown> still='up to the present time'.  How is that?
17:00:20 <clown> corba.  doesn't still work, then.
17:00:25 <API> last commit on gnopernicus:
17:00:26 <joanie> up to the present time in a really old gnome, yes
17:00:26 <API> https://git.gnome.org/browse/archive/gnopernicus/
17:00:31 <API> 5 years ago
17:00:56 <joanie> and mostly translation
17:01:04 <mgorse> I thought development stopped more than five years ago. Probably that commit was a translation update or something
17:01:08 <jjmarin> the real last commit is August 21, 2006
17:01:10 <joanie> yup
17:01:22 <joanie> anyhoo, we have another meeting
17:01:24 <jjmarin> s/is/was
17:01:26 <joanie> so misc time?
17:01:28 <clown> 5 years ago + 1 day.
17:01:38 <jjmarin> yes !
17:01:52 <API> #topic miscellaneous time
17:01:59 <API> gnopernicus is dead
17:02:01 <API> anything else?
17:02:10 <clown> long live gnopernicus~
17:02:10 <joanie> #info gnopernicus is dead
17:02:20 <clown> #action long live gnopernicus!
17:02:24 <joanie> haha
17:02:34 <jjmarin> LSR is dead too, long live to Orca !!!
17:02:44 <joanie> :)
17:03:04 <API> miscellaneous time was dead time today
17:03:13 <API> all stuff really relevant
17:03:16 <API> so relevant that ...
17:03:19 <API> #endmeeting