[07:02] <hrw> does someone manage unity bugs? it is hard to search in 1700 bugs before reporting own one
[08:00] <AlanBell> hrw: what is the bug?
[08:00] <AlanBell> normally as you file one it will search for things that look like duplicates
[08:02] <hrw> bug 903072
[08:02] <hrw> launchpad does not suggested any sensible duplicate this time
[08:02] <hrw> I am going back to KDE4 soon
[08:02] <hrw> amount of unity bugs affects my work too much
[08:04] <hrw> too often when screen is locked I have to wait up to 1.5 minute just to get 'enter password' dialog
[08:04] <hrw> often before that I get preview of desktop which is privacy issue (reported already by people)
[08:04] <hrw> windows are moved when I unlock screen
[08:05] <hrw> ghosted windows after unminimize: http://marcin.juszkiewicz.com.pl/~hrw/shots/ghostedwindow.jpg
[08:06] <hrw> launcher (or expose) showing incorrect amount of windows: http://marcin.juszkiewicz.com.pl/~hrw/shots/howmanychromiums.jpg
[08:09] <hrw> bug   903085 reported
[08:10] <hrw> apportcollected
[08:14] <hrw> is screen lock handled by lightdm?
[14:51] <AlanBell> JohnLea: did you have any thoughs on bug 855532 about the non-visual design of the shortcut overlay?
[14:54] <JohnLea> AlanBell; hyia
[14:54] <JohnLea> re. the overlay
[14:55] <AlanBell> yeah, looks pretty, but doesn't sound pretty
[14:55] <JohnLea> it is non-focusable, but is there another way to make it screen reader friendly?  perhaps a special 'screen reader' key?
[14:56] <JohnLea> would you want all the content to be read, or only a portion?
[14:56] <AlanBell> should all be readable really
[14:57] <AlanBell> it is really useful information to screen reader users
[14:58] <JohnLea> Could it start being read when the Alt key is held down, and the reading stops when the Alt key is released?
[14:58] <AlanBell> orca doesn't generally work like that
[14:59] <AlanBell> it could be piped through speech dispatcher separate from orca, but I am not sure that is a great solution
[14:59] <AlanBell> do you guys use orca to test your designs?
[15:01] <JohnLea> no, most of the accessibility testing is done in the context of automating tests as that uses the accessibility plumbing, we should though (but we are under massive time/resource pressure)
[15:02] <JohnLea> what would be your preferred solution?
[15:02] <AlanBell> yeah, I know about the tests that used the accessibility plumbing for ubiquity, it was broken through most of the oniric dev cycle because of that!
[15:03] <AlanBell> my preferred solution would be that the non-visual design gets considered as early as possible because it is a lot easier to design it in than tack it on afterwards :)
[15:05] <AlanBell> if the overlay had been implemented as a lens it might kind of work
[15:05] <AlanBell> lenses are now navigable and readable
[15:05] <JohnLea> ;-)  but in this specific case?  as the element is non-focusable, my thought is to pipe it through the speech dispatcher when the Alt key is held down
[15:05] <AlanBell> I would be interested in TheMuso's opinion on that
[15:06] <JohnLea> ok, I'll update the description for now, and we will wait to see what TheMuso says.  Ping me when you see him online
[15:07] <AlanBell> ok, will do, thanks
[15:09] <JohnLea> thx as well!
[16:01] <hrw> http://marcin.juszkiewicz.com.pl/2011/12/12/used-unity-for-over-a-month/ - opinions?
[16:21] <mhr3> hrw, what was the alt+f2 issue?
[16:21] <mhr3> the lens was just dead?
[16:25] <hrw> mhr3: no, xkill does not have .desktop file
[16:25] <hrw> app lens works but for *.desktop apps
[16:26] <mhr3> ah, yea, i misread, thought you used alt+f2 twice and the first time it wasn't there..
[16:29] <leo__> hi
[16:29] <leo__> anybody here?
[16:34] <AlanBell> leo__: no, just you
[16:45] <om26er> that "Only window decoration is shown after un-minimize" could be/is specific to ATI
[16:46] <om26er> any unity-2d guys here?
[16:46]  * om26er thinks Kaleo rotated
[16:47] <om26er> anyways would it be fine for you guys to give me bug status rights for unity-2d? i could sometimes help sync status accross unity(s)
[16:52] <om26er> greyback, ^^
[18:15] <om26er> Trevinho, you getting somone to sponsor https://code.launchpad.net/~3v1n0/ubuntu/oneiric/bamf/libreoffice-fixes/+merge/85199 ?
[20:47] <Trevinho> om26er: how should I?
[21:20] <TheMuso> AlanBell, JohnLea, I think that trying to let users know about the shortcut keys when the overlay is shown is not the solution. However, I currently can't think of anything better to suggest, and piping it through speech-dispatcher directly is not the solution.
[21:21] <TheMuso> I think we need something that allows users to find out the shortcut keys for icons on the launcher, but not when the overlay is shown. I'm trying to think of a way it could possibly be done when the launcher has focus and the user is navigating through the launcher.
[21:35] <AlanBell> it is a puzzle indeed
[21:36] <AlanBell> I don't like the idea of a separate user interface for it very much, it won't get maintained right if it is two bits of code
[21:36] <TheMuso> Right.
[21:37] <AlanBell> void UnityScreen::InitHints() has to be shared at least
[21:37] <TheMuso> I've ogt an idea. Perhaps we extend the unity a11y code to put the shortcut for the launcher icons in the description of the ATK object for that icon. We then write an Orca script that allows a user to press a shortcut key, and Orca can read the shortcut key for tha tlauncher icon to the user.
[21:38] <AlanBell> the overlay does quite a bit more than the super+1-9 shortcuts
[21:39] <TheMuso> So, consider this. A user knows there are shortcut keys to launch some stuff on the launcher. So they focus the launcher, and find the app they are intersted in. They then press an Orca key combination to get the keyboard shortcut read to them.
[21:39] <TheMuso> Oh, what else does it show?
[21:39] <AlanBell> that sounds like a good thing, but not the same as the overlay
[21:40] <TheMuso> RIght, so what does the overlay show?
[21:41] <AlanBell> the overlay has categories of shortcuts for the launcher, dash, top bar, switcher, workspaces and windows
[21:41]  * TheMuso sighs.
[21:41] <TheMuso> The only way we can make good use of that is if we can explore it one item at a time.
[21:41] <AlanBell> it is a reference card for 33 different keyboard shortcuts for unity and the desktop
[21:44] <TheMuso> Well the shortcuts will need to e documented.
[21:44] <TheMuso> Unless the DX guys are willing to implement the ability to navigate around the shortcut hints window when accessibility is enabled...
[21:44] <TheMuso> I can't think of any other way of solving this.
[21:45] <TheMuso> Sure we could add something to orca etc, but if new shortcuts are added to unity in the future, we need to then keep another copy in sync.
[21:45] <AlanBell> https://docs.google.com/a/canonical.com/document/d/1jqeKtIJwqLtl58Wk_fqjr9Rrgxn9zsouCYOo-cZsLSE/edit?authkey=CLGG9NkJ&hl=en_GB
[21:45] <AlanBell> that google doc is the list of shortcuts in the launcher
[21:45] <TheMuso> Right.
[21:48] <TheMuso> Hang on, thats got an error..
[21:48] <TheMuso> alt + F10 doesn't open indicators, F10 gets to the top menu bar.
[21:48] <TheMuso> Alt + F10 currently maximises a window.
[21:49] <AlanBell> it is right in the overlay, wrong in the google doc
[21:50] <TheMuso> Ok.
[21:57] <AlanBell> andyrock is building it and JohnLea designed it I think
[21:58] <TheMuso> Right.
[21:59] <AlanBell> just wish it was designed to sound as pretty as it looks :(
[21:59] <TheMuso> I can understand that. As I said above, the only way I think this could be done is f it were rendered as a window, with the shortcuts presented as a list, and one could use arrow keys to move between the shortcuts in the list, and Orca would read them out.
[22:00] <AlanBell> yeah, just wondering if I can hack the code to stop it going away when you let go of super
[22:01] <TheMuso> But you still need code to deal with the focus changes etc.
[22:01] <TheMuso> And keyboard navigation.
[22:02] <andyrock> AlanBell, TheMuso hi
[22:02] <AlanBell> hi andyrock
[22:02] <andyrock> we cannot use super + cursor key
[22:02] <andyrock> really...
[22:03] <TheMuso> andyrock: Thats not what I was thinking.
[22:03] <AlanBell> andyrock: line 2445 of unitshell.cpp ->  // Setup Shortcunt Hint
[22:03] <AlanBell> might want to fix that little typo ;)
[22:03] <andyrock> i'm a kind of busy (with dash keynav) right now, so i've not read above
[22:04] <andyrock> AlanBell, i'm going to work on the shortcut overlay tomorrow ;) thk btw
[22:06] <andyrock> TheMuso, what were you thinking?
[22:07] <andyrock> AlanBell, ah it's a comment :) i'll fix it tomorrow
[22:08] <AlanBell> andyrock: yeah, I don't *normally* point out typos in comments, but on this occasion . . .
[22:09] <andyrock> yeah, there're some typos also in the text...
[22:09] <andyrock> i'll fix it as soon as possible
[22:09] <andyrock> dash key nav is tricky :/
[22:12] <TheMuso> andyrock: For screen reader users, we only have 2 options for helping users find out about keyboard shortcuts. Document them, and point users to the documentation, or make the overlay behave differently if accessibility is enabled. This differing behavior could be a window like the dash, that can be navigated with arrow keys. The shortcuts would be listed in a list view of some kind.
[22:13] <TheMuso> We could also maintain sed window as a part of Orca, but then we would need to keep checking every release to see if there were new/changed shortcuts.
[22:13] <TheMuso> Actually... THis still may be useful in Orca anyway, because looking up keyboard shortcuts for other desktops may be something people want to do as well.
[22:14] <andyrock> TheMuso, what about sharing code betwee orca and unity?
[22:14] <andyrock> we can do it using libunity
[22:14] <TheMuso> andyrock: Orca is in python./
[22:14] <andyrock> ah...
[22:14] <TheMuso> Yeah libunity cold work, but I think we need something thats more useful accross all desktops.
[22:14] <TheMuso> Yeah libunity cold work, but I think we need something thats more useful accross all desktops./c
[22:14] <TheMuso> gah
[22:15] <TheMuso> I think for this cycle, its something we will have to put into Orca, and work a better solution for the future.
[22:15] <TheMuso> Having said that, I won't have time to do it. :)
[22:16] <andyrock> and i've no idea how to do it ;)
[22:16] <andyrock> TheMuso, you will be in Budapest?
[22:18] <TheMuso> andyrock: Yes.
[22:19] <AlanBell> andyrock: so would UnityScreen::InitHints() be callable from python via libunity and return something that can be itterated over to get the shortcut contents?
[22:19] <andyrock> TheMuso, we can talk about it in Budapest
[22:19] <TheMuso> AlanBell: libunity has GIR bindings.
[22:19] <andyrock> AlanBell, not
[22:19] <TheMuso> ah sorry misread.
[22:19] <andyrock> we cannot call UnityScreen::InitHints here
[22:19] <andyrock> from python
[22:19] <andyrock> but i can change something
[22:19] <AlanBell> well, can that data structure be exposed somehow
[22:20] <andyrock> yeah...
[22:20] <andyrock> or we can...
[22:20] <andyrock> create the model in a file
[22:20] <andyrock> i mean:
[22:20] <AlanBell> yeah, that works
[22:21] <andyrock> right now we have something like this in unityshell
[22:21] <andyrock> .cppp
[22:21] <AlanBell> as long as it is one place to maintain the shortcuts, and I can get the active shortcuts and translated descriptions
[22:22] <andyrock> we add a shortcut to the shortcut model
[22:22] <andyrock> most of the time it's of the type COMPIZ_OPTION
[22:22] <andyrock> but there is the support for an HARDCODED_OPTION
[22:22] <andyrock> and if future we can add the support for a GSETTINGS_OPTION
[22:22] <andyrock> etc.
[22:23] <andyrock> the list of the shortcuts (but non the values) is hardcoded
[22:23] <andyrock> but i can make it possible to read if from a file
[22:23] <andyrock> a xml file should be good
[22:24] <AlanBell> JSON even better
[22:25] <andyrock> yeah we already have in unity a json parser
[22:25] <andyrock> btw i've to finish the shortcut overlay first of all
[22:26] <andyrock> then we can add the support for a json file later
[22:26] <andyrock> i prefer to do everything step by step
[22:27] <TheMuso> Makes sense.
[22:29] <andyrock> and we can share the code using LibUnity
[22:29] <AlanBell> {"Launcher":{"shortcut":"Super (Press)","description":"Open Launcher,display shortcuts", . . . }, "Dash":{. . . } . . .}
[22:30] <AlanBell> if libunity can pass the decoded text with all the shortcuts looked up and inserted that would be ideal
[22:31] <andyrock> AlanBell, mmmm
[22:32] <andyrock> i think that it's better sharing the shortcuts list in a file
[22:32] <andyrock> sharing the code to read the option using libunity
[22:33] <andyrock> there is only a problem... there is not a libcompiz right now
[22:33] <andyrock> and i'm using CompOption in ShortcutHint.cpp
[22:34] <andyrock> so for the moment we cannot add it to another lib
[22:35] <andyrock> btw we can read the compiz options in orca using gconf
[22:37] <AlanBell> not dconf?
[22:38] <andyrock> dconf = gsettings
[22:38] <andyrock> in this case
[22:38] <AlanBell> ok
[22:38] <TheMuso> Yes, gsettings.
[22:38] <TheMuso> One should not use dconf directly.,
[22:39] <TheMuso> andyrock: Is unity switching to gsettings this cycle? Or should I say, compiz?
[22:39] <andyrock> TheMuso, you should say compiz :)
[22:39] <andyrock> i've no idea about this, but there are some branches about it
[22:40] <TheMuso> ok
[22:40] <andyrock> i can ask sam as soon as it log in
[22:40] <andyrock> *he
[22:40] <andyrock> sorry :)
[22:42] <TheMuso> Its fine, was just curious.
[23:56] <andyrock> TheMuso, likely Compiz will switch to gsettings according to Sam :)