fluid-work IRC Logs-2012-10-05

[08:37:48 CDT(-0500)] <cindyli> Justin_o: the push down of the zip file for the request to fetch captures is done. together with the merge of your dewarp server, has been pushed to my bitbucket branch

[08:42:57 CDT(-0500)] <Justin_o> cindyli: thanks..

[08:43:02 CDT(-0500)] <Justin_o> did you want to come for the coffee walk

[09:57:45 CDT(-0500)] <michelled> alexn1, colinclark: I was looking at the pull request for 4748 which is about improving the screen reader experience when a user tabs to the video player

[09:57:49 CDT(-0500)] <michelled> https://github.com/fluid-project/videoPlayer/pull/55/files

[09:58:15 CDT(-0500)] <michelled> I'm thinking that this doesn't actually improve the experience much

[09:58:22 CDT(-0500)] <colinclark> oh yes, I remember this one

[09:58:30 CDT(-0500)] <colinclark> Tell us more, michelled

[09:58:34 CDT(-0500)] <michelled> with the pull request in, tabbing to the video player reads out "Infusion video player"

[09:58:48 CDT(-0500)] <michelled> but as a user, I don't really care about that - I'm more interested in what the video is

[09:58:57 CDT(-0500)] <colinclark> Yeah, the default description isn't terribly descriptive, is it?

[09:59:00 CDT(-0500)] <michelled> so on the mammals page - I'd like it to read something different

[09:59:03 CDT(-0500)] <michelled> for each video

[10:00:14 CDT(-0500)] <colinclark> I guess my first question is, from a functional perspective, why does a user tab to the VideoPlayer container itself?

[10:00:15 CDT(-0500)] <michelled> I guess I'm wondering whether what we really want is an option to set the title of the video and set that as the aria description under the hood

[10:00:17 CDT(-0500)] <colinclark> Or do they?

[10:01:44 CDT(-0500)] <michelled> they do tab to the container itself - which Is what we want I think - tab to the thing that wraps all our functionality, our controls, transcripts etc

[10:03:33 CDT(-0500)] <colinclark> How do they then navigate around it from there?

[10:04:56 CDT(-0500)] <michelled> tab takes them through the controls and they can interact with specific controls with arrows, space and enter depending on the control

[10:06:28 CDT(-0500)] <colinclark> So why focus the container at all?

[10:06:45 CDT(-0500)] <colinclark> What does the container do?

[10:07:21 CDT(-0500)] <michelled> when you focus the container, it's not supposed to show our controls immediately - it's a bug that it does

[10:07:42 CDT(-0500)] <michelled> upon focus on the container, the user can start and stop the video with enter

[10:08:03 CDT(-0500)] <michelled> the next tab press shows the controls and takes the user to the first control

[10:08:39 CDT(-0500)] <michelled> http://wiki.fluidproject.org/display/fluid/Floe+video+player+interactions

[10:08:51 CDT(-0500)] <colinclark> Right, I remember there being a lot of subtleties here

[10:09:09 CDT(-0500)] <colinclark> what is the "container" this page is referring to, specifically?

[10:09:10 CDT(-0500)] <jhung> cindyli and justin_o, I am going to push up my first pass at a Capture UI to my repo. It's really rough but intentionally done this way in case someone picks up the Javascript development task.

[10:09:25 CDT(-0500)] <Justin_o> jhung: cool, thanks

[10:09:45 CDT(-0500)] <Justin_o> jhung: can you comment on the jira about it being up on your repo with a link to it

[10:10:16 CDT(-0500)] <colinclark> michelled: The actually container of the whole component, or the <video> element?

[10:10:29 CDT(-0500)] <colinclark> "actual," that is

[10:10:50 CDT(-0500)] <michelled> the actual container - because keyboard focus highlight should look like transcripts is part of the thing with focus

[10:12:11 CDT(-0500)] <colinclark> ok

[10:12:13 CDT(-0500)] <jhung> justin_o: I will do that.

[10:12:17 CDT(-0500)] <colinclark> so focus lands on the container

[10:12:32 CDT(-0500)] <colinclark> the user can play and pause with spacebar and enter, is that correct?

[10:12:38 CDT(-0500)] <michelled> yep

[10:12:51 CDT(-0500)] <colinclark> How would they know they know they could do that?

[10:13:37 CDT(-0500)] <michelled> colinclark: anastasiac has an outstanding pull request which I believe puts in some aria to tell them so

[10:13:45 CDT(-0500)] <colinclark> how?

[10:13:50 CDT(-0500)] <Justin_o> jhung: thanks

[10:14:13 CDT(-0500)] <michelled> as for non screen reader users - we are supposed to have an overlay with a play button on it implying that it can be played

[10:15:17 CDT(-0500)] <colinclark> right

[10:15:21 CDT(-0500)] <michelled> colinclark: good question - I hadn't actually looked at her pull request

[10:15:31 CDT(-0500)] <michelled> it seems I was wrong - it only adds ARIA to the controls

[10:15:32 CDT(-0500)] <colinclark> I'm curious, just to know if these issues interact at all

[10:16:00 CDT(-0500)] <colinclark> Is there anything in the HTML5 media spec related to titles or descriptions for <audio> and <video> tags?

[10:17:01 CDT(-0500)] <michelled> I didn't see it - I would think people would use the title attribute that's available on all elements

[10:20:46 CDT(-0500)] <michelled> I had thought that we should grab the title from the video markup - but of course we are generating the video tag

[10:24:39 CDT(-0500)] <colinclark> ok

[10:24:52 CDT(-0500)] <colinclark> So, I think I agree with your idea

[10:25:05 CDT(-0500)] <colinclark> That we actually ask the user for a "video title" option

[10:25:10 CDT(-0500)] <colinclark> which they can specify if they like

[10:25:14 CDT(-0500)] <colinclark> If it's present, we insert it

[10:25:44 CDT(-0500)] <colinclark> The only question I have is, should it be in the title attribute or in an aria-label attribute?

[10:26:05 CDT(-0500)] <michelled> colinclark: I was thinking it should be in the title for the video

[10:26:15 CDT(-0500)] <colinclark> I guess we'd have to test it to see how it works

[10:26:16 CDT(-0500)] <michelled> and then we could see if aria labelled by would read that

[10:26:21 CDT(-0500)] <michelled> when the container gets focus

[10:26:34 CDT(-0500)] <colinclark> what about the ARIA application role we're adding?

[10:26:38 CDT(-0500)] <colinclark> Is it appropriate?

[10:26:48 CDT(-0500)] <michelled> hmmm… interesting

[10:27:11 CDT(-0500)] <michelled> I'm thinking it might not be since video players are plain old HTML now.

[10:27:22 CDT(-0500)] <michelled> but I guess we need to see how the screen readers react

[10:27:31 CDT(-0500)] <michelled> I'll try some stuff out

[10:28:00 CDT(-0500)] <michelled> I think I'll close the pull request and comment on the JIRA

[10:29:48 CDT(-0500)] <colinclark> I guess there are distinctly application-like interactions inside the video player

[10:29:56 CDT(-0500)] <colinclark> Sliders, menu buttons, and the like

[10:30:33 CDT(-0500)] <colinclark> Pretty interesting comment from Rich about Opera's support for ARIA, as well as the role of the Indie UI work at W3C: http://www.paciellogroup.com/blog/2012/10/notes-on-accessibility-of-microsoft-coms-new-home-page/comment-page-1/#comment-21610

[10:30:36 CDT(-0500)] <colinclark> He doesn't mince words

[10:31:59 CDT(-0500)] <michelled>

[11:58:40 CDT(-0500)] <colinclark> Justin_o, jhung, cindyli: I chatted with Iris yesterday about equipment for Decapod

[11:58:46 CDT(-0500)] <colinclark> The short answer is: go for it

[11:58:59 CDT(-0500)] <Justin_o> colinclark: that's great..

[11:59:01 CDT(-0500)] <Justin_o> thanks

[11:59:04 CDT(-0500)] <jhung> cool

[11:59:06 CDT(-0500)] <jhung> thanks

[11:59:06 CDT(-0500)] <colinclark> Choose what you need in terms of two DSLRs and two point and shoots

[11:59:15 CDT(-0500)] <colinclark> and Pat, I think, can help get two competitive quotes

[11:59:18 CDT(-0500)] <colinclark> that should be it

[11:59:23 CDT(-0500)] <jhung> ok

[13:26:14 CDT(-0500)] <jhung> cindyli and justin_o, I was thinking of some error cases for dewarp….

[13:26:40 CDT(-0500)] <Justin_o> jhung: great.. we were going to be getting to that soon

[13:26:46 CDT(-0500)] <Justin_o> jhung: what are you thinking

[13:27:03 CDT(-0500)] <jhung> Mostly related to the integrity of the archive...

[13:27:04 CDT(-0500)] <jhung> so.

[13:27:07 CDT(-0500)] <jhung> 1. Some images have errors (corrupted / unreadable, unmatched pair)

[13:27:14 CDT(-0500)] <jhung> 2. Calibration data is missing

[13:27:23 CDT(-0500)] <jhung> 3. Invalid archive file.

[13:27:50 CDT(-0500)] <jhung> Can't think of others at the moment.

[13:30:26 CDT(-0500)] <jhung> 4. Unsupported files (resolution: ignore them)

[13:46:57 CDT(-0500)] <Justin_o> jhung: we're thinking of just ignoring the unmatched ones

[14:20:04 CDT(-0500)] <jhung> justin_o: thanks. I'm heading offline shortly

[14:20:25 CDT(-0500)] <Justin_o> jhung: sure.. no problem.. did you see my message earlier?

[14:20:34 CDT(-0500)] <jhung> yep

[14:20:38 CDT(-0500)] <Justin_o> so that's fine?

[14:20:52 CDT(-0500)] <Justin_o> jhung: also, not sure how fine a detail we can get on the errors

[14:21:28 CDT(-0500)] <Justin_o> since we will be running this through the command line

[14:26:21 CDT(-0500)] <jhung> justin_o: I don't even know if knowing some of these errors will help the user.

[14:26:56 CDT(-0500)] <jhung> I think unmatched pairs / invalid images is good to convey to user.

[14:27:22 CDT(-0500)] <jhung> This way they can correct the problem?

[14:27:57 CDT(-0500)] <Justin_o> jhung: so we shouldn't just ignore them?

[14:28:08 CDT(-0500)] <Justin_o> the unmatched pairs

[14:28:17 CDT(-0500)] <Justin_o> jhung: also what do you mean by invalid images?

[14:28:44 CDT(-0500)] <jhung> invalid images: corrupt image files, images not jpeg, png, or tiff.

[14:29:47 CDT(-0500)] <Justin_o> jhung: okay

[14:57:24 CDT(-0500)] <alexn1> colinclark: ayt ?

[14:57:36 CDT(-0500)] <colinclark> yep, for a few more minutes, anyway

[14:57:43 CDT(-0500)] <alexn1> I have a tiny question

[14:58:20 CDT(-0500)] <alexn1> whom should I ask for reviewing any of my pull requests which may happen next week while you are away?

[14:59:34 CDT(-0500)] <colinclark> I guess it depends on the nature of the pull request

[14:59:44 CDT(-0500)] <colinclark> Do you figure you'll be finished with the ATutor work?

[15:00:37 CDT(-0500)] <alexn1> I have ATutor pull request which was reviewed by Cindy today and I have a patch already but I still need to do few more tests

[15:01:12 CDT(-0500)] <alexn1> plus there is a tiny more improvement coming to eliminate cross site scripting

[15:01:29 CDT(-0500)] <alexn1> but I still have few OER pull requests and more will be coming

[15:02:57 CDT(-0500)] <colinclark> Ok

[15:03:13 CDT(-0500)] <colinclark> Sounds like michelled is your gal, once you've wrapped up the ATutor work

[15:03:23 CDT(-0500)] <alexn1> ok thx

[15:03:35 CDT(-0500)] <colinclark> We don't want to keep those fixes waiting, since there's a deadline very soon

[15:05:16 CDT(-0500)] <alexn1> ok will keep it in mind. Andrey managed to update staging today so I need to update my pull requests which have been sitting and waiting for this moment. Those pull requests should address 3 out of 4 major issues.