fluid-work IRC Logs-2012-09-04

[08:39:29 CDT(-0500)] <cindyli> michelled: did someone take over the video player IE8 support jira? wonder if I shall continue with it or work on something else

[08:40:45 CDT(-0500)] <michelled> cindyli: continue with it for now - no one picked it up

[08:40:50 CDT(-0500)] <cindyli> ok

[09:10:22 CDT(-0500)] <alexn> michelled: could you check my pull request. I have this change https://github.com/fluid-project/videoPlayer/pull/43 but feel free to give me a feedback since I'm not absolutely sure about this one

[09:11:11 CDT(-0500)] <michelled> sure alexn

[09:11:41 CDT(-0500)] <alexn> michelled: thank you ! and I also added comments here http://issues.fluidproject.org/browse/FLUID-4743

[09:19:16 CDT(-0500)] <alexn> michelled, cindyli: I think we have a regression in a videoPlayer. I just checked out latest demo branch and videoPlayer does not look "healthy" in IE9

[09:19:47 CDT(-0500)] <alexn> controls look slightly broken and I remember that it was not like that last week or so

[09:21:52 CDT(-0500)] <alexn> michelled, cindyli: oops my bad. please ignore me

[09:22:15 CDT(-0500)] <alexn> (smile) I just had a wrong branch checked out. I guess my coffee has not kicked in yet.

[09:34:52 CDT(-0500)] <jhung> justin_o: I had emailed Hasan about Ocropus 0.6. He said he'd prefer to migrate to that version than keep going with Ocropus 0.5.x.

[09:36:28 CDT(-0500)] <Justin_o> jhung: okay.. that's great.. makes sense

[09:42:54 CDT(-0500)] <michelled> alexn: there are a couple things you can do to improve that pull request

[09:43:18 CDT(-0500)] <michelled> firstly, let's not hard code the string 'Volume'

[09:45:57 CDT(-0500)] <michelled> in fact, it's already in the volumeControls defaults, so we should just reuse that

[09:47:41 CDT(-0500)] <michelled> secondly, I would put the binding of the mouse enter and mouse leave functions into the bindDOMEvents function

[09:49:19 CDT(-0500)] <alexn> michelled: thx for feedback. Will do this, it sounds very reasonable

[09:49:29 CDT(-0500)] <michelled> alexn: lastly, let's try and find another name for the contentFunction option. In the context of the 'toggle button' component, content function implies content of the button instead of what it really is - content of the toggle button

[09:49:37 CDT(-0500)] <michelled> oops, I mean tool tip (smile)

[09:51:00 CDT(-0500)] <alexn> ok will do! Also I started checking my screen reader issues and with latest demo updates some of the problems do not look so bad. I updated this one http://issues.fluidproject.org/browse/FLUID-4751 and will send a pull request soon

[09:51:16 CDT(-0500)] <michelled> thx

[10:31:45 CDT(-0500)] <anastasiac> fluid-everyone: stand-up, anyone?

[11:47:57 CDT(-0500)] <anastasiac> michelled, I've issued a pull request for the videoPlayer transcript area keyboard a11y: https://github.com/fluid-project/videoPlayer/pull/53

[11:49:08 CDT(-0500)] <michelled> thx anastasiac

[11:55:39 CDT(-0500)] <michelled> anastasiac: did you mean to include DebugFocus.js in your pull request?

[12:29:07 CDT(-0500)] <anastasiac> michelled, thanks for catching that. I did NOT mean to add that file. I've removed it

[14:08:52 CDT(-0500)] <alexn> michelled: could you please take another look at my pull request at let me know if anything looks confusing or questionable https://github.com/fluid-project/videoPlayer/pull/43

[14:08:56 CDT(-0500)] <Justin_o> jhung: should i just close this one as won't fix http://issues.fluidproject.org/browse/DECA-291

[14:41:06 CDT(-0500)] <alexn> michelled: I retested and updated http://issues.fluidproject.org/QuickSearch.jspa?searchString=FLUID-4750 & http://issues.fluidproject.org/QuickSearch.jspa?searchString=FLUID-4751 , as well as submitted pull requests for those 2 jiras

[14:44:19 CDT(-0500)] <michelled> thx alexn

[14:58:17 CDT(-0500)] <alexn> michelled, cindyli: a good hint for testing browsers with JAWS - Do not touch or click anywhere in the JAWS windows when your 40 minutes session is expired. This way JAWS will continue to work and you can continue screen reader testing as long as you need. If you click anything in those JAWS popup windows it will stop working and you will have to restart

[15:05:17 CDT(-0500)] <anastasiac> alexn, I just saw your pull request that adds some ARIA to the language menu

[15:05:24 CDT(-0500)] <anastasiac> I was just considering that issue myself

[15:05:54 CDT(-0500)] <anastasiac> I'm wondering of you think that the language menu should have a role of menu, and that the individual menu items should have roles of menuitmecheckbox?

[15:06:30 CDT(-0500)] <alexn> well it is a little bit of both

[15:06:41 CDT(-0500)] <alexn> it is a menu with some items you can check and uncheck within

[15:07:06 CDT(-0500)] <alexn> since in the end we want to tell user the list of the items in the menu but also communicate on what options is being selected

[15:07:19 CDT(-0500)] <alexn> at least it is my perception of the purpose for those menus

[15:08:07 CDT(-0500)] <anastasiac> right - I was suggesting that the <ul> itself be given a role of "menu" and each <li> be given a role of "menuitemcheckbox." Does that jibe with your understanding of what those roles are meant for?

[15:08:55 CDT(-0500)] <anastasiac> alexn, for the record, I've been using http://www.w3.org/WAI/PF/aria/roles#menu (and related areas on that page) for my reference

[15:09:41 CDT(-0500)] <alexn> anastasiac: that is a very good find

[15:09:53 CDT(-0500)] <alexn> I will be honest with you I have not tried this one

[15:10:12 CDT(-0500)] <anastasiac> also, alexn, the menu should probably have it's 'aria-hidden' attar set properly

[15:11:03 CDT(-0500)] <alexn> anastasiac: could you be a bit more precise

[15:11:26 CDT(-0500)] <anastasiac> http://www.w3.org/WAI/PF/aria/states_and_properties#aria-hidden

[15:12:01 CDT(-0500)] <anastasiac> alexn, the menus are initially hidden, so the aria-hidden property should be set to "true"; when it's shown, the property should be updated to "false", etc

[15:12:11 CDT(-0500)] <alexn> anastasiac: I was mostly working with a screen reader and what it plays to the user. For setting aria attributes properly I though we have a separate JIRA http://issues.fluidproject.org/browse/FLUID-4786

[15:12:19 CDT(-0500)] <alexn> *thought

[15:12:26 CDT(-0500)] <anastasiac> ok, alexn, that was going to be my next question (smile)

[15:12:36 CDT(-0500)] <anastasiac> I'm happy to keep working under FLUID-4786

[15:12:56 CDT(-0500)] <anastasiac> and we can merge our stuff, depending on whose branch goes into demo first (wink)

[15:13:06 CDT(-0500)] <alexn> we definitely should have proper aria attributes set but currently the one you mentioned does not impact what is being played by a screen reader

[15:13:20 CDT(-0500)] <alexn> I mean it works as intended

[15:13:42 CDT(-0500)] <anastasiac> ok, alexn, I'll use your branch as an indication of what's required for the screen readers to behave reasonably, and add in other stuff

[15:13:44 CDT(-0500)] <anastasiac> thanks!

[15:13:54 CDT(-0500)] <alexn> np (smile)

[15:14:31 CDT(-0500)] <alexn> yes that JIRA I mentioned above goes through more components within the videoPlayer no matter if they are being played or not… I guess

[15:15:09 CDT(-0500)] <alexn> I will try menuchecked and see if it can resolve current problems with screen readers. Thanks a lot for a suggestion !

[15:15:18 CDT(-0500)] <jhung> Justin_o: I closed DECA-291 you mentioned to me earlier.

[15:17:41 CDT(-0500)] <anastasiac> alexn, I'm curious about your use of aria-selected. You always have it the same as aria-checked. I was assuming that aria-selected would indicated something like which element is currently selected via hover or keyboard access, which might not be the same as the one that's currently checked.

[15:17:48 CDT(-0500)] <anastasiac> I looked at the OAA example at http://www.oaa-accessibility.org/example/25/, and they don't use aria-selected

[15:17:51 CDT(-0500)] <anastasiac> what do you think?

[15:18:13 CDT(-0500)] <anastasiac> maybe I'm interpreting things wrong...

[15:18:20 CDT(-0500)] <alexn> anastasiac: well actually this one is a good question I want to double check with michelled about this case

[15:18:49 CDT(-0500)] <alexn> basically aria-checked is a proper attribute to use for the most of the cases and screen reader will work with it "somehow" but as I mentioned in a JIRA there are few corner cases

[15:19:07 CDT(-0500)] <alexn> first is Safari

[15:19:25 CDT(-0500)] <alexn> Safari + VoiceOVer ignores aria-checked at all and only reads out aria-selected

[15:19:36 CDT(-0500)] <anastasiac> ah, interesting

[15:19:40 CDT(-0500)] <anastasiac> that sucks

[15:20:01 CDT(-0500)] <alexn> I know basically I'm stuck between 2 fires here

[15:20:29 CDT(-0500)] <anastasiac> yeah, I guess it makes sense to work around the browser inconsistencies, even if it's not quite the exact perfect use of the attribute

[15:20:29 CDT(-0500)] <alexn> on one hand I want screen reader to read something to user to let the user know what is going on within the menus

[15:20:37 CDT(-0500)] <anastasiac> thanks for the clarification

[15:20:45 CDT(-0500)] <alexn> and on the other hand I want to use only one attribute aria-checked or aria-selected

[15:20:58 CDT(-0500)] <alexn> so not sure….

[15:21:03 CDT(-0500)] <alexn> michelled: what do you think about this one

[15:21:32 CDT(-0500)] <anastasiac> hm. This is odd… my Safari is not using our custom controls… why would that be??

[15:22:04 CDT(-0500)] <anastasiac> well, it's not hiding the other controls; it's got both. weird...

[15:22:13 CDT(-0500)] <alexn> I saw it before. they are used by rendered on top of the default Safari controls

[15:22:30 CDT(-0500)] <alexn> there should be a JIRA for this somewhere…. I think

[15:22:42 CDT(-0500)] <anastasiac> ah, ok, so this happens to you too, alexn?

[15:24:13 CDT(-0500)] <alexn> yes

[15:24:21 CDT(-0500)] <anastasiac> ok, so it's not just me (smile)

[15:24:59 CDT(-0500)] <michelled> anastasiac: that's because of the track support that has been added to webkit

[15:25:00 CDT(-0500)] <alexn> worry not, anastasiac. i think everyone has the same behaviour

[15:25:13 CDT(-0500)] <michelled> it's a problem in chrome too if you turn the flag on

[15:25:52 CDT(-0500)] <michelled> I thought we had a JIRA for it

[15:26:03 CDT(-0500)] <alexn> michelled: I also thought so

[15:26:16 CDT(-0500)] <michelled> I can't find it now - can you alexn?

[15:26:25 CDT(-0500)] <alexn> it feels like I saw it somewhere

[15:26:28 CDT(-0500)] <alexn> let me search for it

[15:27:35 CDT(-0500)] <michelled> alexn: as to your question above - I would prefer to see more of the screen-reader + browser combinations working

[15:28:56 CDT(-0500)] <alexn> michelled: cannot find… maybe the JIRA about implementing our own controls for a fullscreen mode stuck in my head and I assumed that we have the one for overwriting Safari controls

[15:29:06 CDT(-0500)] <alexn> weird :S

[15:30:54 CDT(-0500)] <michelled> alexn: could you file it please?

[15:31:12 CDT(-0500)] <alexn> yeah will do. Should I add it to the FIP as well ?

[15:31:36 CDT(-0500)] <anastasiac> alexn, I notice that you're setting aria-checked and aria-selected to false for everything, but not updating the currently selected language to true on initial start-up

[15:32:07 CDT(-0500)] <alexn> anastasiac: let me double check since it should because screen readers work

[15:32:08 CDT(-0500)] <alexn> (wink)

[15:32:14 CDT(-0500)] <alexn> *should work I mean

[15:32:34 CDT(-0500)] <michelled> alexn: yes, please add it to the page

[15:35:11 CDT(-0500)] <anastasiac> ah, alexn, probably my mistake, I haven't fully merged yet

[15:35:12 CDT(-0500)] <michelled> anastasiac: I wasn't able to tab into the transcript area at all in FF

[15:35:35 CDT(-0500)] <michelled> anastasiac: when I was testing your pull request for 4782

[15:35:42 CDT(-0500)] <anastasiac> michelled, odd… it worked for me...

[15:35:48 CDT(-0500)] <anastasiac> let me re-test

[15:35:56 CDT(-0500)] <alexn> michelled, anastasiac: created http://issues.fluidproject.org/browse/FLUID-4787

[15:36:36 CDT(-0500)] <alexn> anastasiac: I will test aria things a bit later since while yura1 has some time I want to check some of the cspace code with him.

[15:37:21 CDT(-0500)] <anastasiac> michelled, I just checked again. I'm able to tab to the transcript area in FF no problem. I can get to it, arrow around with it, hit enter to scrub to a particular spot...

[15:38:17 CDT(-0500)] <michelled> thx alexn - I modified it slightly

[15:39:04 CDT(-0500)] <michelled> anastasiac: weird. I'll try again

[15:39:24 CDT(-0500)] <anastasiac> np, alexn. FYI, the checked and selected are updating just fine. it was my mistake

[15:40:52 CDT(-0500)] <michelled> alexn: I put some comments on the 4744 pull request - let me know if they aren't clear

[15:55:51 CDT(-0500)] <michelled> anastasiac: still not working for me - maybe you can show me tomorrow

[15:56:18 CDT(-0500)] <anastasiac> yeah, hat's weird michelled. Does it work for you in other browsers? Are you getting any console error messages?

[15:56:41 CDT(-0500)] <anastasiac> can you tab to the transcript area language menu and close button, michelled?

[15:57:50 CDT(-0500)] <alexn> michelled: thx for editing my JIRA. anastasiac: glad to hear that aria attributes work for you

[15:58:01 CDT(-0500)] <alexn> michelled: I will take a look into 4744

[15:59:33 CDT(-0500)] <michelled> anastasiac: no console errors. I can tab to the transcript language selection, close and main area, but I can't arrow around at all

[15:59:51 CDT(-0500)] <anastasiac> odd. ok, let's look at it tomorrow

[16:00:02 CDT(-0500)] <anastasiac> oh, what about other browsers, michelled?

[16:00:37 CDT(-0500)] <michelled> same behaviour in chrome and safari

[16:00:41 CDT(-0500)] <michelled> I haven't tried IE

[16:00:51 CDT(-0500)] <anastasiac> ok

[16:03:01 CDT(-0500)] <anastasiac> ah, michelled: in your FF General preferences, do you have "Always use the cursor keys to navigate within pages" turned on or off?

[16:03:18 CDT(-0500)] <anastasiac> if it's on, try turning off and testing again

[16:05:23 CDT(-0500)] <alexn> michelled: I was only confused by the last comment in my pull request but other ones are pretty clear

[16:07:23 CDT(-0500)] <anastasiac> alexn, regarding the aria-checked and aria-selected: they don't seem to work for the 'show/hide captions/transcripts' menu item, only the languages

[16:07:59 CDT(-0500)] <michelled> alexn: I replied to your comment on the pull request

[16:08:41 CDT(-0500)] <alexn> anastasiac: they do not work since I did not include those menu options to have aria attributes on purpose

[16:08:57 CDT(-0500)] <alexn> you cannot select or unselect. Check or uncheck those menu options

[16:09:12 CDT(-0500)] <alexn> these menu options change their label and this label is read to the user

[16:09:41 CDT(-0500)] <anastasiac> alexn, you did give them the attributes, you just don't update them

[16:09:46 CDT(-0500)] <alexn> only languages are menu options which could exclude each other

[16:09:54 CDT(-0500)] <anastasiac> it's probably correct for them to have the attributes

[16:09:59 CDT(-0500)] <anastasiac> but they should be updated

[16:10:32 CDT(-0500)] <alexn> well if you start clicking Shoe or Hide Captions - with English caption selected. English caption will still be selected

[16:10:51 CDT(-0500)] <anastasiac> hm

[16:10:54 CDT(-0500)] <anastasiac> interesting

[16:10:58 CDT(-0500)] <alexn> also how would you envision adding those attributes to those buttons ?

[16:12:12 CDT(-0500)] <alexn> the only one reason why I even added those aria attributes to the options is the following

[16:12:14 CDT(-0500)] <anastasiac> alexn, I see your point (I just also see mine, as well (smile) I guess I was thinking of show/hide as a checkbox: checked when the captions are showing, unchecked when they're hidden. But I do see your point

[16:12:28 CDT(-0500)] <anastasiac> hm

[16:12:33 CDT(-0500)] <alexn> screen readers will say that there are 3 options in the appeared menu instead of saying that there are 2

[16:13:22 CDT(-0500)] <alexn> wait a second I'm confusing myself (smile)

[16:13:34 CDT(-0500)] <alexn> role=option is the one which will tell how many menu items is there

[16:14:15 CDT(-0500)] <alexn> I guess I added those attributes to those menu items just to follow the common structure between the menu items with the Captions/Transcripts seciotns

[16:14:18 CDT(-0500)] <alexn> *sections

[16:14:51 CDT(-0500)] <anastasiac> alexn, ok, I'll do some screenreader testing with the menuitemradio role instead of option, see what it says. This also might be a discussion for jameswy: is the show/hide conceptually one of the "options" or not? hm… all these metaphysical questions of existence...

[16:16:07 CDT(-0500)] <alexn> anastasiac: sure we can bring more people to discuss this, since I do not have an exact answer for this one. But it seems reasonably for me not to set change those aria attributes even because it just sounds confusing when you have 2 selected items in a menu and one of them is something like Hide All

[16:16:28 CDT(-0500)] <alexn> and another one is English

[16:16:57 CDT(-0500)] <anastasiac> yes, alexn, leave it as is for now. I'll give it some more testing with the menu role on the menu.

[16:18:59 CDT(-0500)] <alexn> anastasiac: agreed! let's bug james tomorrow about this one (smile)

[16:19:08 CDT(-0500)] <alexn> to be on a safe side