16:07:43 #startmeeting 16:07:43 Meeting started Thu Nov 15 16:07:43 2012 CET. The chair is joanie. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:07:43 Useful Commands: #action #agreed #help #info #idea #link #topic. 16:08:04 #topic Java ATK Wrapper: Abandonware? 16:08:20 #info There has been some recent talk on the Orca list about Java accessibility. 16:08:35 #info One user indicated that he could not build the latest. 16:08:54 #info Joanie just tried and confirmed that it doesn't build (autotools issue probably). 16:09:13 what is it that wrapper for? I thought java had its own a11y interface that was auto-published? Or is it this wrapper that exposes java a11y info to atspi? 16:09:22 #info But this reminds us (yet again) that the former Sun guys are gone. :( 16:09:37 clown: we don't have the bridge in *nix 16:09:50 ah... 16:09:59 different from windows 16:10:07 we used to have a java atk bridge I think 16:10:11 I do see the odd question about the bridge on windows on the java access list. 16:10:20 but that's windows, not *nix 16:10:20 and then when the bonobo deprecation happened 16:10:25 and it's once in a blue moon. 16:10:37 we wound up with this java atk wrapper 16:10:41 that I know nothing about 16:10:50 other than it seems to be abandonware 16:10:53 you know more than me :-) 16:11:16 #info Joanie wonders what we want to do about this, adding that perhaps we should wait until Piñeiro is back to chime in. 16:11:28 so, the question is: do we maintain it? Or do we ask Peter what Oracle's intention is here? Or do nothing? 16:11:44 clown: exactly 16:11:49 I could info that as possible questions to ask. 16:11:55 sure 16:11:58 not knowing the answers... 16:12:03 :) 16:12:06 I don't think we'd lose anything by asking Peter 16:12:18 Indeed. 16:12:32 #info questions to ask are: should we (gnome a11y developers) attempt to maintain the java atk wrapper? Or, 16:12:49 though every time we ask we get people saying nice things like they'll take care of it. 16:12:53 #info should we touch base with Peter and ask him what Oracle's intentions are here? Or, 16:12:54 But in the end, they don't 16:13:04 #info should we simplay abandon it? 16:13:20 s/simplay/simply/ 16:13:21 So we can ask Peter and he can say nice things, and he can ping the guys behind it, and they can say they will do something 16:13:29 but in the end, kiddos, it's us 16:13:44 #info Joanie is going to see if she can sort out the build issues. 16:13:53 #info re: asking Oracle, we could publish that request on the java-access list to make the request more public. 16:14:38 (done) 16:14:41 clown: sure (and I have no objections to that). But what I said above still stands. 16:15:03 joanie: we are on the same page. 16:15:04 I doubt they will say "screw you, we're abandoning it" 16:15:28 they will say nothing, or say something untrue ("we care about accessibility, and we'll take care of this") 16:15:31 I kinda wish they would if that's what they actually intend (perhaps with a bit more tact). 16:15:44 :) 16:16:01 So if I get the fricken thing built I'll likely push that change 16:16:25 it might be that we all just have to pitch in and keep the thing sufficiently alive 16:16:31 * joanie shrugs 16:16:33 * clown wonders the status of java on linux/gnome these days, never mind a11y. 16:16:42 who knows 16:17:10 As I understand it, Eclipse has their own implementation 16:17:25 that is (as I understand it) better than official java 16:17:38 and doesn't require the wrapper 16:17:51 mgorse: any knowledge on any of this? 16:18:11 as I recall, Eclipse uses SWT for the GUI, which probably means native GTK widgets with a java wrapper. 16:18:20 yeah 16:19:14 so it's only pure snorkel java that requires this jaw? 16:19:42 it's the Java Swing GUI. 16:19:48 that needs a bridge. 16:20:00 clown: info that? 16:20:03 pls 16:20:07 okay... 16:20:41 #info Eclipse doesn't have this problem probably because it uses the SWT GUI for java, which means native GTK widgets with a java wrapper. 16:21:06 #info the problem occurs for Java Swing GUI since that GUI is "pure java" widgets. 16:21:26 #info pure java widgets are not realized by any native widgets, so they need an a11y bridge of some sort. 16:21:27 I guess swt is similar to wx (ie, uses the platform's native widgets) 16:21:37 mgorse: right. 16:21:51 although I've never actually looked at how to code with SWT. 16:22:32 hence, my "probably" hedge. 16:22:43 anyhow, done with my infos. 16:22:47 so in terms of conclusions, aside from me trying to get the blessed thing built.... 16:22:56 you think it's worth asking on a public list now? 16:23:09 or shall we wait until fearless leader returns? 16:23:28 i'm inclined to wait for API's opinion. 16:23:43 ok 16:23:46 just in case he knows something we don't. 16:23:57 he knows many things, being all wise and crap 16:23:58 :P 16:24:13 and well read (?) 16:24:19 ;-) 16:24:34 about jaw or the log? 16:25:02 dunno about the former; the latter, yes, I suspect he will see that. 16:25:15 about many things. he has a tendancy to surprise me, and I have a "he knows about that too?" moment. 16:25:46 aha. Yes, he knows about many things. It doesn't surprise me. 16:25:58 that's why he's lead and I am but a humble minion. 16:25:59 ;) 16:26:04 anyhoo, moving on 16:26:11 #topic Python 3 NOW 16:26:23 which is largely for javi who is not here, but.... 16:27:00 #info This weekend, Matthias stated in #a11y that it would be ideal for us to have the entire accessibility stack Python 3 ready for 3.7.2's release. 16:27:25 #info The 3.7.2 release tarballs are due Monday the 19th. 16:27:44 #info The only non-ready module from our team is Accerciser. 16:27:59 #action Joanie will ping Javi yet again about this. 16:28:13 other things to say on this point? 16:28:23 what's the worst case scenario? That accerciser ceases to function? 16:28:44 yeah, I suppose 16:28:49 wow... 16:28:53 mind you, I did all the patches ages ago 16:29:02 some may need updating now 16:29:10 there may be python 3.3 things 16:29:16 but I did the fricken work 16:29:21 I do a lot of ARIA testing using accerciser. Mind you, I'm sticking with F17 for the moment. 16:29:39 and I clearly identified what was python 2 compatiable 16:29:49 so Javi could have already reviewed and committed it 16:29:54 joanie: Have you tested Orca with Python 3.3? I'm realizing that I haven't done any testing with it so far 16:30:07 so I suppose the worst, worst case scenario is hostile module takeover 16:30:18 in which someone pushes the changes and does the release 16:30:22 mgorse: I have been 16:30:23 yes 16:30:30 for a while now 16:30:39 okay; that implicitly tests pyatspi somewhat then 16:30:50 will the interim python-based focus tracker be affected? 16:30:53 you need to build speech-dispatcher, and liblouis, and brlapi, etc. 16:31:04 clown: Pyatspi has been set for a while now 16:31:12 * clown whew! 16:31:19 so you should triple-check 16:31:37 but Mike and I did our stuff before the 3.6 release 16:31:43 I kinda suspected that, but I want to be doubly sure. Okay, I'll try to give it a test. 16:31:47 clown: I haven't tested the focus tracker itself, though 16:31:48 thanks! 16:31:59 where "try" == "try to find some time". 16:32:40 just to speed things up: what do I need to install to make this a valid test? 16:32:42 okay, moving on to a related topic (which may or may not have content) 16:32:46 * joanie waits 16:32:59 clown: just build it all with python 3 16:33:10 joanie: what is "all"? 16:33:17 i.e. PYTHON=/usr/bin/python3 ./autogen.sh bla bla bla 16:33:30 pyatspi for sure 16:33:39 okay, anything else? 16:33:40 and probably your focus tracker 16:33:50 that's part of pyatspi 16:34:00 module wize yes 16:34:03 so if I build pyatspi, I build the focustracker. 16:34:08 really? 16:34:13 okay 16:34:21 yup. It's in the "examples" folder of pyatspi. 16:34:23 I'll help you look after the meeting if you'd like 16:34:25 It's in an examples directory 16:34:39 thanks mgorse! 16:34:53 but mgorse did his conversion prior to it being added 16:35:04 joanie: I have another meeting after this one. and am extremely busy this afternoon. Mayber tomorrow we can talk. 16:35:08 i.e. it should be tested 16:35:11 ok 16:35:22 I may look anyway and propose patches if needed 16:35:31 yes, I will test (sometime in the next week [i hope]). 16:35:44 so moving on to a related topic.... 16:35:50 #topic GNOME 3.8 updates 16:35:58 other than the Python 3 stuff.... Anything else? 16:37:06 #info mgorse is working on having keystroke listeners preserved/reinstated if the registry daemon dies 16:37:23 yay! 16:37:44 (which means that libatspi would have to keep track of what it has registered, which it currently doesn't) 16:37:58 btw, is there some way to kick modules again into being accessible? 16:38:19 i.e. if the registry dies, running apps are not accessible until those apps are restarted 16:38:56 Do you mean keystroke listeners? Or programs getting kicked off the bus if they send bad data? Or something else? 16:39:06 something else I *think* 16:39:24 for instance, let's say I'm running Orca in gnome-shell and using gnome-terminal 16:39:39 if the registry dies and I launch firefox 16:39:44 firefox will be accessible 16:39:47 The keystroke listeners are what I know about that goes away if the registry dies, since they are stored in the registry process 16:39:55 but gnome-shell and gnome-terminal won't be 16:40:33 if I then restart gnome-shell and/or gnome-terminal, those restarted apps will be accessible again 16:40:57 Have you reproduced this? I'd need to try to figure out exactly what is going on 16:41:02 and this is why users on the Orca list have been saying they need to log out and log in when the registry crashes 16:41:17 reproduce the crash or the apps not being seen? 16:41:35 either, I guess 16:41:58 well, the latter happens reliably (i.e. *when* the registry crashes, the running apps are no longer accessible) 16:42:07 I rarely get registry crashes however 16:42:27 When I've seen the registry crash or go away, it usually seemed to coincide with gnome-shell crashing 16:42:44 it happens to me independent of that 16:43:12 #action Joanie will try to find a way to reproduce apps going away when the registry crashes and file a bug. 16:43:22 I dunno if you'll be able to fix that or not, but .... 16:43:45 I'll need to try more to reproduce. A related issue, I guess, is that apps need to be re-registered, since the registry maintains a list of children 16:44:04 * joanie nods 16:44:12 it's the re-registering I'm talking about 16:44:14 i guess 16:44:15 :) 16:44:40 i.e. the registry crashing is not ideal, but it's not the end of the world if it auto-restarts itself and all the apps continue to be accessible 16:44:54 it's the Orca users having to log out and log in again that concerns me 16:45:13 Maybe. On a related note, it would help if someone had a core when the registry crashes 16:45:47 I think I'm going to start working with the registry running under gdb 16:45:55 that sounds snarky but i don't mean it that way 16:46:01 i usually have a terminal running 16:46:15 and get a crash here and there, so I might as well be prepared for it 16:46:25 and we cannot ask users to do this 16:46:50 * joanie writes a note to do this 16:47:34 okay other stuff? 16:48:07 though some users might be able to save a core and bring up a trace with gdb, if needed. I need to figure out how to modify my settings, actually--right now I just get a file called "core", meaning that it will be overwritten if I have more than one of them, so I've never ended up with a core for the registry daemon, even if I should have one 16:49:26 i wonder if I have any cores from what's it's name? (fedora's crash reporter) 16:49:30 I'll look for those later 16:49:40 * joanie notes that down too 16:49:57 okay, moving on I guess 16:50:07 #topic W3C updates 16:50:10 two parts 16:50:28 #info Joanie emailed the Board list about GNOME joining the W3C. 16:50:39 #info Hopefully it will wind up on the Tuesday agenda. 16:50:42 (done) 16:50:49 clown: you have group updates? 16:51:01 some info. 16:51:28 #info One of the groups that Joseph participates in is the Independent User Interface" group. 16:51:38 #iinfo Also known as "IndieUI". 16:52:08 #info it is mostly public: the mailiing list is public, the meeting minutes are public, and the two documents are public. 16:52:28 s/#iinfo/#info/ 16:52:46 #info the mailing list is here: http://lists.w3.org/Archives/Public/public-indie-ui/2012Nov/ 16:53:01 whoops, that just november. 16:53:11 #info the mailing list is here: http://lists.w3.org/Archives/Public/public-indie-ui/ 16:53:33 #info: latest meeting minutes (from yesterday) are here: http://lists.w3.org/Archives/Public/public-indie-ui/2012Nov/0035.html 16:54:28 #info: the lastest draft describing the proposed UI independent DOM events is here: https://dvcs.w3.org/hg/IndieUI/raw-file/tip/src/indie-ui-events.html 16:54:59 #info: the latest draft describing how user preferences affect web apps is here: https://dvcs.w3.org/hg/IndieUI/raw-file/tip/src/indie-ui-context.html 16:55:38 #info Joseph also particpates in the PFWG — the group defining the ARIA standard 16:56:13 what's the pf? 16:56:13 #info no news this week. We continue to test implementations of ARIA in different browsers/operating systems/desktops. 16:56:32 #info "PFWG" is "Protocols and Formats Working Group" 16:56:37 aha thanks 16:57:06 #info they are responsible for making sure that other W3C standards (e.g. SVG) do not inadvertantly create a11y problems. 16:57:32 #info but, the PFWG major thrust currently is getting ARIA 1.0 out. 16:57:33 (cool! and sounds hard) 16:57:50 questions? 16:58:04 not yet. I should read the documents/drafts 16:58:10 thought of another info. 16:58:11 * joanie notes that down as well 16:59:00 #info the IndiUI discussion list is public in terms of reading. If you want to post to the list, you do have to sign up. It's not unlike signing up of gnome discussion lists. 16:59:26 s/signing up of gnome discussion/signing up for gnome discussion/ 16:59:31 (done) 16:59:38 thanks!! 16:59:45 you're welcome. 17:00:12 having these summaries and docs we should read is quite helpful 17:00:45 anyone have questions for clown? 17:01:32 no from me 17:01:36 okay, given the time..... 17:01:41 #topic Marketing 17:01:46 jjmarin: ? 17:02:17 ok 17:03:22 #info I've realised that we have updated yelp gnome info that I think is good for using for the web and remove the current content of the GNOME a11y user guide 17:05:11 jjmarin: you mean this: http://library.gnome.org/users/gnome-help/stable/a11y.html ? 17:05:25 #info Juanjo will contact the documentation for put this documentation online and I will help to update to gnome 3.6 and future releases 17:06:16 yes 17:06:35 so what do you mean by put it online? 17:06:46 ok 17:07:52 I didn't know it was online 17:07:58 :) 17:08:00 this is the yelp documentation 17:08:11 the idea is using this instead of http://library.gnome.org/users/gnome-access-guide/stable/ 17:08:37 aha 17:08:57 yeah, the g-a-g is old and needs to be removed 17:09:06 exactly 17:09:19 so that involves filing a bug 17:09:56 I'm pretty sure that library.gnome.org (unlike the projects.gnome.org) is something not under our control 17:10:09 and that bugs are preferred to us just doing it 17:10:19 assuming we even *can* just do it 17:10:35 I think so 17:10:47 you think which? 17:11:40 I think we must ask 17:11:48 * joanie nods 17:11:54 file a bug really 17:11:57 I've had to do that 17:12:14 https://bugzilla.gnome.org/enter_bug.cgi?product=website 17:13:00 okay, so other topics on this front? 17:13:02 I don-t know if it is good remove Do you think is good just to remove http://library.gnome.org/users/gnome-access-guide/stable/ or link to the new page 17:13:16 the g-a-g has gok on it 17:13:25 and states that orca does magnification 17:13:26 yes 17:13:28 and is about gnome 2 17:13:29 very old 17:13:39 so the g-a-g should be gagged 17:13:41 ;) 17:13:56 and a redirect could be done 17:14:01 so as not to break any links 17:14:37 brb 17:15:10 so given the time.... 17:15:21 #action Juanjo will open a bug for removing http://library.gnome.org/users/gnome-help/stable/a11y.html or link to the gnome-access-guide 17:15:40 (really, it should be both) 17:16:08 ok 17:16:15 and wait 17:16:20 you have it backwards 17:16:33 the gnome-access-guide is uber outdated 17:16:40 sorry 17:16:41 it needs to be removed 17:16:44 I re-action 17:16:46 NOT the gnome-help stuff 17:17:15 * clown is back 17:17:28 * joanie waits for the re-action 17:18:13 #action Juanjo will open a bug for removing http://library.gnome.org/users/gnome-access-guide/stable/ and link to http://library.gnome.org/users/gnome-help/stable/a11y.htm 17:18:26 thanks :) 17:18:37 * jjmarin is slow today :-) 17:18:41 #topic Reminder Q3's 17:18:47 https://live.gnome.org/Accessibility/QuarterlyReports/2012/Q3 17:18:51 https://live.gnome.org/GnomeMarketing/QuarterlyReports/2012/Q3 17:19:05 #info There still is no due date from the Marketing Team 17:19:17 * joanie is grateful for small miracles 17:19:24 #topic Miscellaneous Time 17:19:35 anything to add that wasn't on the agenda? 17:20:36 not from me. 17:21:00 * jjmarin doesn-t remember anything 17:21:10 :-) 17:21:18 Miscellaneous Time is more fun when serious Piñeiro is here and we can be silly. 17:21:30 exactly 17:21:45 who is silly, exaclty? eh? 17:21:55 #info Joanie notes that Miscellaneous Time is far less fun when Fearless Leader is away. 17:22:02 it's like a class without a teacher :-) 17:22:12 hah 17:22:34 okay, on that note 17:22:39 #endmeeting