16:12:13 <API> #startmeeting
16:12:13 <tota11y> Meeting started Thu Sep 20 16:12:13 2012 CET.  The chair is API. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:12:13 <tota11y> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:12:29 <API> #topic Marketing and Fundraising
16:12:31 <API> jjmarin, ?
16:14:10 <mgorse> Okay. Sorry for being so late.
16:14:19 <API> mgorse, np
16:15:11 <jjmarin> #info because there isn't any change in the features for GNOME 3.6, the Release Notes are finished with the correction added in the last meeting
16:15:19 <jjmarin> done 1
16:15:23 <jjmarin> done !
16:15:37 <clown> congrat jjmarin, and thanks for writing those notes.
16:16:15 <jjmarin> usual
16:16:25 <jjmarin> ;)
16:16:44 <API> so ...
16:16:53 <API> anything else? moving to next (or previous) topic?
16:17:26 <jjmarin> nothing more from my side
16:18:01 <API> ok
16:18:16 <API> #topic GNOME 3.6: Code Freeze is here. Accessibility is AlwaysOn(tm). Anything critical?
16:18:31 <API> #info my action of review pef/memory consumption is still pending
16:18:59 <API> #info William Jon and others are doing a leak review of all the stack, several bugs and patches committed
16:19:09 <API> #info also some bugs caused, still pending to be solved
16:19:10 <API> done
16:19:14 <API> mgorse, ?
16:19:32 <API> I thinkm that you have more detailed information about all this
16:20:00 <mgorse> #info mgorse applied several patches that fixed leaks. Some of those patches caused other problems that should be fixed, but, regardless, hoping that there aren't any new problems.
16:20:32 <mgorse> #info also planning to install the interim caret focus tracker into the bin directory; unsure if this constitutes a freeze break.
16:20:34 <mgorse> done
16:20:57 <joanie> mgorse: did you see but 684434
16:21:02 <joanie> mgorse: did you see bug 684434
16:21:03 <tota11y> 04Bug https://bugzilla.gnome.org/show_bug.cgi?id=684434 normal, Normal, ---, liyuan, UNCONFIRMED, Segmentation fault when exiting GTK applications in GTK 3.5.18 with an AT client registered.
16:21:30 <mgorse> oh, I hadn't seen that
16:21:39 <joanie> matthias triaged it
16:21:43 <joanie> it's a typo
16:21:51 <joanie> but a big deal of a typo
16:21:55 <mgorse> #info also, bug 684434 needs to be fixed
16:22:06 <clown> matthias has eagle eyes.
16:22:49 * mgorse didn't know that atk_bridge_adaptor_cleanup was being called
16:23:10 <API> joanie, typo?
16:23:18 <API> mgorse, it was added recently
16:23:21 <joanie> did you see the bug comment API?
16:23:34 <API> it seems that I didn't read the last ones
16:23:55 <joanie> for (ls = clients; ls; ls = ls->next)
16:23:56 <joanie> g_free (l->data);
16:24:07 <API> mgorse, on the bugm Luke mention that happens after a specific commit
16:24:13 <clown> why does g_free even work?  (What is l?)
16:24:15 <mgorse> yeah, that's an easy fix; needs freeze break but I don't see that being a problem
16:24:19 <API> that commit was done by William Jon as part of his cleaning effort
16:26:10 <mgorse> clown: l is a GList pointer. ls is a GSList pointer (though not sure if there's a good reason for me to be using both in my code). Anyway, I suspec t that it never worked, but atk_bridge_adaptor_cleanup wasn't being called, so the code wasn't being tested
16:26:33 <clown> ah — (we need unit tests).
16:26:45 <clown> thanks mgorse.
16:27:37 <API> well, yes, nobody never uninstalled the module, so yes
16:27:40 <API> probably not tested
16:27:51 <API> so, any other want to add something to this point?
16:28:02 <joanie> i've got just one
16:28:19 <joanie> #info Joanie needs to look into bug 684474.
16:28:19 <tota11y> 04Bug https://bugzilla.gnome.org/show_bug.cgi?id=684474 normal, Normal, ---, orca-maint, UNCONFIRMED, screen reader doesn't read on the login screen
16:28:39 <joanie> #info It may not be an Orca bug, but she is going to triage it before passing it along.
16:28:41 <joanie> (done)
16:29:26 <clown> if it matters, gs-mag works from the login screen.
16:29:40 <joanie> heh
16:29:46 <clown> oh, but, that's because it's part of gnome-shell
16:29:56 <joanie> does tracking, etc.?
16:29:57 <clown> doesn't really depend on a11y libraries.
16:30:01 <joanie> i.e. are you getting events?
16:30:39 <clown> good question.  How would I start the interim focus tracker from the login screen?  Is that even possible?
16:31:19 <clown> I suppose I could rebuild gs with the built-in focus tracker (the real one), and see what happens.
16:31:38 <joanie> don't spend time on it if you don't need to
16:31:45 <joanie> I'll look into it
16:32:03 <clown> sure.  If I think of anything that can help, I'll let you know, joanie.
16:32:07 <joanie> :)
16:32:08 <joanie> thanks!
16:32:23 <clown> or, if you want me to try something, ask away.  (I might say "no").
16:32:26 <clown> ;-)
16:32:27 <joanie> :)
16:33:23 <API> api the evil talking after the silence: anything else?
16:33:55 <jjmarin> Not critical, what is going on with the "Enable by Keyboard" option in the a11y menu ? It seems an useless option with the Accessibility Always On, anyway, it doesn't hurt anybody :)
16:34:14 <clown> I want to say something about the interim focus tracking, as the last sub-topic, API
16:34:49 <API> jjmarin, I still don't know what that feature does
16:34:57 <API> activate a11y with a keyboard combination?
16:35:17 <clown> is that the onscreen keyboard?  Or something else?
16:35:56 <API> clown, no, we are talking about a new/moved option at the universal access settings
16:36:09 <clown> something else then..
16:36:14 <API> at typing, there are a "enable by keyboard" checkbox
16:36:17 <API> anyway,
16:36:35 <API> #action API will investigate that "Enable by Keyboard" thing
16:36:36 <API> so
16:36:39 <API> clown, your turn
16:37:18 <clown> okay, regarding mgorse's last info about putting the interim focus tracker into bin, here is some background.
16:37:57 <clown> #info Bjorn Lie had a problem building a package for openSUSE, and traced it to the interim focus tracker.
16:38:02 <clown> #info https://bugzilla.gnome.org/show_bug.cgi?id=682636#c11
16:38:03 <tota11y> 04Bug 682636: normal, Normal, ---, liyuan, UNCONFIRMED, Interim focus tracking for gnome-shell magnifier via D-Bus
16:38:28 <clown> #info I found a way to fix the specific problem.
16:39:13 <clown> #info but that fix may not be legitimate — it's a change ot the makefile such that the focus tracker code's executable bits are on when it's copied to site-packages.
16:39:36 <clown> #info  mgorse suggested that we really should be using EXTRA_DIST in the makefile for this code.
16:40:09 <clown> #info the problem there is the tracker isn't installed anywhere even if you have pyatspi2 installed.
16:40:34 <clown> #info which makes it very complicated for an end user to find the tracker and launch it.
16:40:52 <clown> #info mgorse's latest suggestion is to install it in /usr/bin.
16:41:29 <clown> #info but I'm not sure that's not too heavy handed for a hack interim solution.
16:41:41 <clown> done — any questions?
16:42:39 <API> not from my side
16:43:19 <jjmarin> so you think /usr/bin is not good udea because it is an interim solution ?
16:43:42 <mgorse> I'm just wondering if I need to ask permission from the release team to do this--"simple build fixes" are allowed without asking, but I'm not sure if installing something in /usr/bin goes beyond a simple build fix
16:44:02 <clown> jjmarin:  I'm not totally against it, but isn't is odd to put something like that in a global bin folder?  Maybe I"m too cautious.
16:44:14 <mgorse> Also, the original code installed it in /usr/lib/python-(version)/site-packages/examples, which isn't really the right place to install it
16:45:16 <API> mgorse, I agree with you
16:45:23 <jjmarin> I think we are delivering betas in gnome like boxes or clocks, so I don't thing this is a problem unless it has security problems or something
16:45:26 <API> I mean that we didn't ask any kind of permission
16:45:33 <clown> anyother way to look at it:  If the relase team says not to put it in /usr/bin, where can it go so that it's relatively easy for a user to find it and run it?  What's plan B?
16:45:33 <API> because it was just a test
16:45:55 <jjmarin> clown: ok :-)
16:46:21 <API> clown, I guess that plan B is a kind of manual
16:47:19 <jjmarin> and we want people try it easily
16:47:34 <API> jjmarin, well yes
16:47:35 <API> but
16:47:35 <clown> API, you mean instructions like "clone the pyatspi2 from git, then navigate to the examples folder and run mayFocustracker.py"?  Something like that?
16:47:45 <API> this is already a plan B
16:47:53 <API> we were not in time for a feature
16:48:05 <API> and this was just a way to ship this
16:48:27 <API> if the concluded that the feature is not ready, we can't just install that on /usr/bin
16:48:27 <jjmarin> API: exactly
16:48:50 <API> clown, well, yes, something like that (probably more elaborated)
16:49:00 <API> I know that it is not really user-friendly
16:50:05 <clown> yes, it is hardly user-friendly.
16:51:27 <clown> I was hoping that the instructions would be simply, "run '../some/common/directory/magFocusTracker.py', and to put that line in your .bash_profile, if you want it on everytime you log in".
16:51:29 <API> I will use again the card "this is already plan B"
16:52:30 <API> mgorse, in which directory is installed using EXTRA_DIST?
16:52:47 <clown> API, that's the point — it's not installed if its' in EXTRA_DIST.
16:52:54 <API> ah ok
16:53:10 <clown> you must git the source if you want to see the examples folde.
16:53:11 <API> but there isn't any kind of /.../share
16:53:14 <API> directory to install
16:53:15 <clown> "folder".
16:53:21 <API> like documentation or so on?
16:53:25 <mgorse> I just labeled it EXTRA_DIST because that is what some other modules do to handle their "examples"
16:53:36 <mgorse> but this case may or may not be different
16:54:36 <API> well, Im sorry
16:54:43 <API> but as I said at the beginning of the meeting
16:55:01 <API> I need to go before usual meeting time
16:55:03 <API> so about this
16:55:17 <API> clown, mgorse could you investigate any other option to ship that stuff?
16:55:27 <clown> API, yes.
16:55:28 <API> I still don't buy the /usr/bin thing
16:55:33 <API> and having said so
16:55:36 <API> I must go
16:55:42 <clown> understood.
16:55:51 <jjmarin> API: bye !
16:55:57 <API> if anyother one want to led the meeting ...
16:55:59 <API> so bye
16:56:17 <joanie> did he not end the meeting
16:56:25 <joanie> #endmeeting
16:56:29 <joanie> nice
16:56:44 <clown> we can't end the meeting?  Only the chair?
16:56:52 <clown> …has that authority?
16:56:59 <joanie> you can add chairs
16:57:13 <joanie> I need to look at the bot docs :-/
16:57:21 <clown> http://wiki.debian.org/MeetBot
16:57:25 <joanie> thanks
16:57:49 * clown doesn't see anything about adding chairs.
16:57:58 * clown or tables for that matter.
16:58:09 <joanie> there's a command
17:00:36 * jjmarin wonders is someone enter in the room names API can do the trick :-)
17:00:43 <clown> addchair <channel> <network> <nick>
17:01:05 <clown> addchair clown
17:01:10 <clown> #addchair clown
17:01:18 <joanie> you have to be chair to do that
17:01:23 <joanie> so, lemme deal with this
17:01:24 <clown> !
17:01:40 <joanie> as the bot owner if everyone leaves me alone I should be able to solve it
17:01:50 <joanie> (api is pinging me via jabber asking where I am)
17:07:15 <joanie> #endmeeting