14:06:12 <API> #startmeeting
14:06:12 <Services> Meeting started Thu Jul 10 14:06:12 2014 UTC.  The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:06:12 <Services> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:06:22 <API> #topic Progress towards 3.14
14:06:57 <API> #info API started to review wayland-dev mailing lists and wikis; he didn't read in a while and a lot of traffic; no conclusions yet
14:07:14 <API> and unfourtunately thats it
14:07:15 <API> next one?
14:07:38 <clown> sure
14:08:46 <API> nobody else what to share any update?
14:08:48 <API> joanie, mgorse ?
14:08:59 <joanie> for what it's worth, I'm updating all of orca's regression tests so I can get a valid and thorough baseline do to the cache testing
14:09:03 <joanie> but not finished yet
14:09:20 <API> for minutes sake, do you mind to #info that, I think that is worthy to mention
14:09:25 <joanie> ok
14:09:51 <joanie> #info Joanie is updating all of orca's regression tests so she can get a valid and thorough baseline do to the cache testing.
14:10:12 <joanie> #info Joanie is not finished with this yet, so she has nothing to report about cache findings.
14:10:18 <joanie> (done)
14:10:33 <API> mgorse, something from your side or moving to next topic?
14:10:57 <mgorse> #info mgorse worked on some atk/gtk/at-spi code to implement a click method, as a work-around, but it needs to be tested
14:11:02 <mgorse> done
14:11:24 <mgorse> also monitoring wayland-devel, and some people are asking about virtual keyboards, which seems slightly relevant to us, but no conclusions from what I can tell
14:11:28 <API> mgorse, do you plan to share that as patches on any bug?
14:11:50 <mgorse> yeah, I should create a bug for it, but should try to test it first
14:12:05 <API> mgorse, yeah, in general there are a lot of non-conclusions on wayland-devel, and suddenly stuff appears on the repositories
14:12:58 <API> having said so
14:13:03 <API> moving to next topic then?
14:13:07 <mgorse> ok
14:13:37 <API> #topic W3C updates
14:13:39 <API> clown, ?
14:14:06 <clown> The high light (for me) this week was discussing something we decided we wouldn't do.
14:14:14 <joanie> heh
14:14:24 <clown> But, it's probably a good idea to let the a11y team here know about it.
14:14:54 <clown> #info Steve Faulkner of the Paciello Group proposed a new role:  "disclosure".
14:15:13 <clown> #info his working example is here:  http://thepaciellogroup.github.io/disclosure-button/
14:15:47 <clown> #info  The two buttons in that example have the role="disclosure".
14:16:32 <API> hmm
14:16:36 <clown> #info the role requires aria-expanded (true/false) and aria-controls (points to the content that is disclosed/hidden depending on the state of the button.
14:16:40 * API shutting up
14:17:24 <clown> #info the working group felt that a button role as it currently is in the spec is sufficient, and agreed that aria-expanded and aria-controls should be added to a button used for disclosing.
14:18:05 <clown> #info Michael (W3C person) noted that Apple's a11y API does have a disclosure role, but none of the other a11y APIs do.
14:18:13 <clown> done, questions.
14:18:16 <clown> ?
14:18:24 <API> well, as I was about to say after my "hmm"
14:18:52 <API> that disclosure role is somewhat not-needed as soon as you add aria-expanded (true/false)
14:18:56 <API> because afaiu
14:19:08 <API> on aria, having those properties even on false
14:19:15 <API> is a way to say that the element is expandable
14:19:24 <API> (on atk we need the state expandable)
14:19:28 <API> but in any case
14:19:29 <clown> good point.
14:19:44 <API> another problem with that role
14:19:50 <API> as far as I see from the example
14:19:51 <clown> in aria, if aria-expanded is *not* present, then the object is not expandable.
14:19:58 <API> it is thought to be used on buttons
14:20:11 <API> so a specific kind of buttons
14:20:23 <API> but buttons are not the only thing that can expand stuff
14:20:37 <API> im thinking on some menu items for example
14:20:48 <API> ie: some gnome-shell menu-items
14:21:08 <API> unless Im understanding wrong
14:21:18 <API> and that disclosure is for everything that can "disclose"
14:21:37 <clown> treeitems also support aria-expanded (i'm pretty sure).
14:21:41 <API> but in that case, imho, knowing that an element is a menu item is more important that an element is a disclosure
14:21:50 <API> well, this became a long rant
14:22:10 <clown> essentially, API, you agree with the majority of the working group.
14:22:14 <API> but in summary: I agree with the workig gorup conclusion
14:22:17 <joanie> :)
14:22:26 <API> without typos too
14:22:31 <clown> "great minds ..."
14:22:45 <jjmarin> hehe
14:22:52 <jjmarin> API pass the exam :-)
14:22:54 <API> so... anything else? more questions?
14:23:17 <clown> joanie, any other highlights that you can think of?
14:23:24 <joanie> nope
14:23:36 <clown> oh, one more thing about disclosure.
14:24:21 <clown> #info the motivation was to the replace the HTML 5 new <details>/<summary> elements.  They are supposed to operate as a disclosure as well.
14:24:58 <clown> I'm not well versed in these new elements, but they are used in the aria user agent implementation mappings.
14:25:06 * clown looks for url
14:26:17 <clown> #info The entries in the role mapping table are all done using <details>/<summary> (and a lot of javascript):  http://www.w3.org/TR/core-aam-1.1/#mapping_role_table
14:26:40 <clown> #info the look and act like "disclosure" buttons.
14:26:51 <clown> #info they look and act like "disclosure" buttons.
14:27:24 <clown> done.
14:27:35 <jjmarin> Nice to see ATK/AT-SPI Roles in the table :-)
14:28:11 <API> ok thanks
14:28:15 <API> so moving now?
14:28:23 <clown> fine with me.
14:28:28 <API> (no questions from my side btw)
14:28:28 <API> k
14:28:33 <API> so a surprise topic
14:28:37 <API> #topic GUADEC
14:28:44 <API> not really a lot to say but just in case
14:28:46 * clown is stunned
14:28:51 * clown :-)
14:28:57 * joanie screams in horror
14:29:13 <API> #info API will be present at the evince hackfest, and at the first days of the GUADEC (core days)
14:29:32 <API> #info taking into account the usual dates for the AGM meeting probably he will not be around for that
14:30:10 <API> #info and taking into account that usually bofs are at the end, he will not available then
14:30:31 <API> #info although this year, the a11y bof is somewhat replaced by the evince hackfest, as I would focus on evince a11y
14:30:33 <joanie> agm is on the 28th
14:31:18 <API> joanie, then confirmed I will not be at the agm
14:31:35 <joanie> we can come up with slides before you leave
14:31:59 <API> well yes, as usual, instead of going to presentations we can be at the corridors working on slides and stuff :P
14:32:23 <API> having said so, yes, this year plan is less ambitious that other years, but it has the evince a11y stuff
14:32:28 <API> so I think that could be a good GUADEC
14:32:37 <API> so, that was my personal plan
14:32:47 <API> anyone else want to say something about GUADEC?
14:32:48 <jjmarin> #info Juanjo will also be present at the Evince hackfest, and at the GUADEC core days (I'll leave the 30th)
14:33:14 <joanie> #info Joanie will be at the Evince hackfest, core days, and the first non-core day.
14:33:53 <clown> #info Jospeh will not attend GUADEC
14:34:03 <clown> or "Joseph", actually...
14:34:12 <joanie> hehehe
14:34:24 <joanie> when you cannot spell your own name, it's time for more coffee
14:34:27 <joanie> :)
14:34:31 <clown> indeed!
14:34:52 <joanie> said joasomething
14:35:20 <API> yes, more coffee would give you the ability to write your own name wrong, but faster
14:35:34 <API> so, if nobody has anything else to say about GUADEC, I will move to next topic
14:36:11 <API> #topic Marketing
14:36:13 <API> jjmarin, ?
14:36:26 <jjmarin> #info Juanjo has put the new Orca, AT-SPI and ATK entries in the Spanish wikipedia in a decent state
14:36:48 <jjmarin> #info Juanjo will ask to change the title of the a11y article in the annual report from "Accessibility campaign" to "Accessibility achievements" because only a part of the (final after shorteninh) text is about the Accessibility campaign
14:36:50 <joanie> yay
14:36:54 <clown> cool
14:36:59 <jjmarin> #info This short article for the Annual Report is in the following PDF in page 13 (BTW, not sure if it is an accessible PDF) https://github.com/gnome-design-team/gnome-marketing/raw/master/annual-report-2014/GAR2013.pdf
14:37:25 <clown> jjmarin:  do you changes to the Spanish wikipedia have any impact on the English version?
14:37:35 <clown> *sigh*  "your changes"
14:38:00 <jjmarin> some small changes in Spanish are reverted into English
14:38:28 <jjmarin> and viceversa
14:38:34 <clown> very nice!
14:39:26 * API needs to read that entries
14:39:29 <API> jjmarin, thanks for that
14:39:38 <API> that==everything you reported
14:40:03 <jjmarin> I was also doing some cleaning where Gnopernicus and LSR are still mentioned as default screen reader :-)
14:40:12 <jjmarin> and things like that
14:40:17 <API> gnopernicus ftw!!
14:40:23 <joanie> lsr was never the default
14:40:25 <joanie> ;)
14:40:40 <jjmarin> touché ! :-)
14:41:03 * clown remembers Gnopernicus...
14:41:32 <API> clown probably still uses gnopernicus
14:41:38 <API> is something that nobody would understand
14:41:43 <API> he used it before it was cool
14:41:55 <API> in fact, it was never cool
14:41:56 <clown> he never used it; he just read about it.
14:42:14 <jjmarin> clown doesn't like mainstream stuff
14:42:16 <jjmarin> :-)
14:42:19 <API> well, as this is starting to be really miscellaneous
14:42:23 <joanie> heh
14:42:24 <API> #topic miscellaneous time
14:42:32 <API> and before going to the total misc stuff
14:42:41 <jjmarin> Orca is so mainstream :-)
14:42:53 <API> #topic These days peter vagner, an orca user, was taking a look to a 3rd party library called venom
14:42:54 <clown> actually I have a question about Orca, but API has the floor.
14:42:59 <API> argh
14:43:01 <API> #topic miscellaneous time
14:43:07 <API> #info These days peter vagner, an orca user, was taking a look to a 3rd party library called venom
14:43:20 <API> #info this library is used for some 3rd party applications
14:43:22 <API> hmm
14:43:57 <API> #info like transmission-gtk
14:44:08 <API> #info that defines a custom gtktreeviewcellrenderer
14:44:31 <API> #info the interesting part of this, is that is was written on vala, raising some interesting questions about how to extend
14:44:45 <API> #info default gtk accessibility support on vala
14:45:14 <API> #info also raised some bugs, as some stuff is not packaged on gtk vala (the a11y apis)
14:45:42 <API> #info also pointed the outdated guides to extend gtk accessibility (ie: they still mention the use of factories)
14:46:23 <API> #info hopefully all this would lead to that application more accessible, better knowledge of how to do some stuff on vala, and a list of documents to deprecate/update
14:46:24 <API> done
14:46:59 <mgorse> Thank you for helping him, btw. There's a learning curve, and I'm really glad that more people are being involved and figuring out how to do things
14:47:12 <joanie> +1 to that
14:47:50 <jjmarin> +1
14:48:04 <jjmarin> even in Vala :-)
14:48:06 <API> no problem, but my inner evil teacher inside sometimes have the feeling that he should investigate a little more before asking
14:48:46 <API> anyway, he also mentioned that it is  not needed to answer so quickly if Im busy, so I guess that I can continue to answer as far as I have time to
14:49:31 <jjmarin> ask to your inner angel by now :-)
14:50:15 <API> so, anything else miscellaneous to share?
14:50:53 <jjmarin> nope
14:52:09 <clown> yes
14:52:29 <clown> I discovered an oddity in the relationship between gsettings and orca.
14:52:51 <clown> It's a small thing, but I thought it qualified for misc.
14:53:06 <clown> Step 1:  gsettings set org.gnome.desktop.a11y.applications screen-magnifier-enabled true
14:53:13 <clown> The result is that orca launches.
14:53:23 <joanie> really?
14:53:26 <jjmarin> wow
14:53:31 <clown> Step 2:  kill -s SIGTERM orca-pid
14:53:32 <joanie> it shouldn't
14:53:39 * joanie looks
14:53:48 <clown> The result is that orca quits like it usually does.
14:53:59 <clown> Step 3:  gsettings get org.gnome.desktop.a11y.applications screen-magnifier-enabled
14:54:26 <clown> The result is "true" — the gsettings still believes that orca is running when it isn't.
14:54:39 <joanie> so two different things
14:54:44 <joanie> enabled is not really enabled
14:54:47 <joanie> I learned that the hard way
14:54:48 <clown> re:  "joanie:  it shouldn't"
14:54:55 <clown> I did this on F20.
14:54:57 <joanie> enabled means as a session service
14:55:19 <joanie> so orca doesn't toggle the enabled setting
14:55:26 <joanie> because I was told not to
14:55:38 <joanie> the "it shouldn't" ==
14:55:51 <joanie> I'm pretty sure that Orca is listening to the screen reader setting
14:55:54 <joanie> but if that is not the case
14:55:55 <joanie> I will fix it
14:56:07 <joanie> maybe it's an issue in the settings daemon
14:56:10 <joanie> I am looking
14:56:13 <clown> so, you think this is an Orca bug?
14:56:16 <joanie> but step 3 is by design
14:56:24 <joanie> which?
14:56:27 <joanie> step 1 or step 3?
14:56:47 <clown> Okay, one other fact:  If orca is running and I do gsettings set org.gnome.desktop.a11y.applications screen-magnifier-enabled false
14:56:57 <clown> the result is Orca is shut down.
14:57:42 <clown> the "bug" (if it is a bug) is the one case where the setting (enabled) and the state of orca (not running) are inconsistent.
14:57:54 <joanie> so let's back up
14:57:58 <joanie> do you mean screen-magnifier
14:58:01 <joanie> or screen-reader?
14:58:09 <clown> gah.  no.
14:58:16 * clown reallly needs more coffe
14:58:22 * joanie rolls her eyes
14:58:25 <joanie> so with this in mind
14:58:32 <joanie> 16:54:48 <clown> re:  "joanie:  it shouldn't"
14:58:35 <clown> gsettings set org.gnome.desktop.a11y.applications screen-reader-enabled true
14:58:43 <joanie> was my "magnifier != screen reader"
14:58:44 <clown> sorry
14:58:51 <joanie> and orca should not start up when you start the magnifier
14:59:03 <clown> yeah, that shouldn't happen with the magnifier.
14:59:05 <joanie> the other issue is that I was told that running != enabled
14:59:11 <joanie> lemme search for the bug
14:59:23 <clown> so forget the magnifier setting.  All those steps actually invovled the screen-reader-enabled setting.
14:59:36 <joanie> ok, then the behavior is correct
14:59:49 <joanie> because enabled == as a start up service
14:59:52 <joanie> for the user session
14:59:57 <joanie> and not running
15:00:06 <joanie> says people who are not me
15:01:02 <clown> So, one can set the setting to false, and orca stops (in one scenario), but leave the setting alone and shut down orca itself (in a another scenario), and you get different pairs of settings/states.
15:01:10 <clown> and that make sense?
15:01:11 <joanie> i know
15:01:21 <joanie> if you kill orca
15:01:31 <joanie> does that mean you don't want to start it next time you log in?
15:01:46 <joanie> if you need to install something on the system of a user who is blind
15:01:50 <joanie> and kill orca
15:01:59 <joanie> and it toggles the service off
15:02:06 <clown> I have no idea.  The scenario I'm working within doesn't involve logging out.
15:02:14 <joanie> exactly
15:02:23 <joanie> so lemme try a third time ;)
15:02:36 <joanie> you are talking about a gsetting
15:02:43 <clown> right.
15:02:48 <joanie> that gsetting controls what hapens when the user logs in
15:02:57 <joanie> does orca start ("enabled")?
15:03:06 <joanie> if you quit orca and orca turns "enabled" off
15:03:12 <joanie> next time you log in: No orca
15:03:12 <clown> let's say that it's off when one logs in.
15:03:18 <clown> the setting, I mean.
15:03:23 <joanie> if it is off when the user logs in
15:03:48 <joanie> and you start orca as an executable
15:03:54 <joanie> it should not change the setting
15:04:00 <joanie> if you use gettings to start orca
15:04:00 <clown> I go to the control panel and (essentially) set the setting to "on".  Orca starts.
15:04:11 <joanie> that is the setting for the service
15:04:29 <joanie> it is not "just running orca"
15:04:37 <joanie> and I have no control over that
15:04:42 <joanie> I just do what I am told
15:04:45 <joanie> but I'm looking for the bug
15:04:49 <clown> okay.
15:04:54 <joanie> in which I say "really? that makes sense??"
15:04:57 <joanie> just like you are
15:05:09 <joanie> and was told "yes. you are dumb joanie. of course it makes sense"
15:05:10 <joanie> :)
15:05:45 <API> hmm
15:05:53 <API> this seems too much deep/specific for the meeting
15:05:57 <clown> Well,...
15:05:58 <API> and more for misc time
15:06:10 <API> and it seems that there are several questions on the air
15:06:27 <API> so, could we fisnih the meeting, resume this on IRC/mailing list?
15:06:29 <clown> It is my understanding that the un-technical user is being told:  to start orca, go to the control panel and enable the screen reader.
15:06:40 <clown> API, sure.
15:07:12 <API> clown, ok thanks, and sorry for interrupting you in the middle of the thought
15:07:13 <API> having said so
15:07:16 <API> #endmeeting
15:07:16 <joanie> clown: I hear you
15:07:18 <API> hmm
15:07:21 <API> #endmeeting