fluid-work IRC Logs-2012-05-23

Owned by Fluid IRC Log Bot
Last updated: May 23, 2012Version comment
29 min read

[07:58:03 CDT(-0500)] <michelled> cindyli: are you working on the font issues in chrome and FF on windows?

[07:59:22 CDT(-0500)] <cindyli> michelled: yes

[08:00:17 CDT(-0500)] <michelled> cool - let me know how it goes

[09:21:48 CDT(-0500)] <cindyli> michelled: one thing to clarify, we want to use another font for those browsers only on windows without affecting other OS, right?

[09:22:48 CDT(-0500)] <michelled> if that's possible cindyli, yes

[09:22:54 CDT(-0500)] <cindyli> ok

[09:25:00 CDT(-0500)] <colinclark> hey michelled, question for you

[09:25:06 CDT(-0500)] <colinclark> about my favourite topic

[09:25:07 CDT(-0500)] <colinclark> if you have a sec

[09:25:31 CDT(-0500)] <michelled> hi colinclark: I'm actually looking at dealing with focus in ie6 right now (smile)

[09:25:38 CDT(-0500)] <colinclark> lol

[09:25:49 CDT(-0500)] <colinclark> I wonder--could we make the focus indicator a bit more subtle?

[09:25:58 CDT(-0500)] <colinclark> Thinner--perhaps it would reflow the content less

[09:26:03 CDT(-0500)] <colinclark> and a little less blazing

[09:26:18 CDT(-0500)] <colinclark> I think it could probably still be done in a way that made it very clear

[09:26:43 CDT(-0500)] <colinclark> On all browsers, I mean

[09:27:38 CDT(-0500)] <michelled> colinclark: yep, it's certainly easy to change it. what I can't do easily is move it to using outline which is, in my opinion, the best way to style keyboard focus

[09:27:53 CDT(-0500)] <michelled> colinclark: but changing up the border to another style is simple.

[09:28:16 CDT(-0500)] <colinclark> right

[09:30:35 CDT(-0500)] <colinclark> I think we should consider it with jvass when we get a chance

[09:33:34 CDT(-0500)] <michelled> colinclark: I don't know what to do about keyboard focus on IE6

[09:33:56 CDT(-0500)] <michelled> I really can't use the border because of the way it renders things - if you think jumping is bad on FF you should see this

[09:33:59 CDT(-0500)] <michelled> or maybe you shouldn't

[09:34:05 CDT(-0500)] <colinclark> yeah, probably not

[09:34:13 CDT(-0500)] <colinclark> and the default focus outline is just totally unworkable?

[09:34:32 CDT(-0500)] <colinclark> I'd be tempted just to turn our custom focus indication off in the junior IEs

[09:34:51 CDT(-0500)] <colinclark> Next time we make a site, we'll make it Progressively

[09:34:57 CDT(-0500)] <michelled> well, that's interesting, because that's what we did for our UIO demo

[09:35:02 CDT(-0500)] <michelled> meaning left the default styling

[09:35:04 CDT(-0500)] <colinclark> and so we can just leave the junior IEs with a mobile-style site

[09:35:13 CDT(-0500)] <michelled> and in both the UIO demo and the site it's a little hard to notice

[09:35:32 CDT(-0500)] <colinclark> I'm not worried as much about "hard to notice" as "not working" if you see what I mean

[09:35:46 CDT(-0500)] <michelled> but I guess I'd argue that ie6 users will probably notice it because :focus doesn't work on ie6 so they probably always see the default styling

[09:36:30 CDT(-0500)] <colinclark> right

[09:36:46 CDT(-0500)] <colinclark> and IE6 is fortunately on our periphery in terms of browser support

[09:36:54 CDT(-0500)] <colinclark> So, some principles are becoming more clear to us

[09:37:09 CDT(-0500)] <colinclark> 1. never look broken, no matter the browser

[09:37:13 CDT(-0500)] <colinclark> the corollary to that is

[09:37:35 CDT(-0500)] <colinclark> 2. design progressively, so we can provide a minimal experience for older devices and browsers without the pain of last minute fixes

[09:38:23 CDT(-0500)] <colinclark> in the meantime, i think you're good to just go back to the default style, michelled

[09:38:30 CDT(-0500)] <michelled> 3. testing IE 6 is a lesson in patience

[09:38:35 CDT(-0500)] <colinclark> lol

[09:38:43 CDT(-0500)] <colinclark> If we do #2, #3 will be largely moot

[09:38:47 CDT(-0500)] <colinclark> or at least, a lot freaking simpler

[09:39:22 CDT(-0500)] <colinclark> I'd find it preferable to go back to the default focus indicator on all browsers, but I gather it's actively broken in a number of spots

[09:39:55 CDT(-0500)] <michelled> yep, the whole outline issue

[09:40:03 CDT(-0500)] <colinclark> right

[09:40:04 CDT(-0500)] <colinclark> okay

[09:40:46 CDT(-0500)] <colinclark> Next time Johnny is in, we'll work with him to create a "Making an Awesome Website" methodology and checklist

[09:40:57 CDT(-0500)] <colinclark> With all our lessons learned here

[09:50:50 CDT(-0500)] <jessm> anastasiac: content date with me?

[09:51:12 CDT(-0500)] <anastasiac> sure, I'll get on skype

[10:02:00 CDT(-0500)] <anastasiac> alexn2, I've commented on your scrollbars pull request: Unfortunately, we can't modify the original FSS files. you'll need to find a way to override that style setting inside the idi theme style.css file

[10:02:15 CDT(-0500)] <alexn2> hmmm interesting

[10:04:13 CDT(-0500)] <anastasiac> cindyli, I'm seeing two pull requests from you, both for the chrome font issue - are they the same, or does one override the other?

[10:04:40 CDT(-0500)] <cindyli> anastasiac: they are not the same, https://github.com/inclusive-design/idi-theme/pull/52 is to fix the font issue with chrome/firefox on windows, and "51" to replace in the right yellow arrow icons for IE, can you have a look

[10:05:00 CDT(-0500)] <cindyli> one sec, perhaps one was from yesterday

[10:05:07 CDT(-0500)] <cindyli> i will close that one

[10:05:26 CDT(-0500)] <anastasiac> cindyli, I'm comparing #48 with #50

[10:06:24 CDT(-0500)] <cindyli> right, anastasiac, ignore 48, closing that one

[10:06:34 CDT(-0500)] <anastasiac> k, thanks for checking cindyli

[10:06:39 CDT(-0500)] <cindyli> np, anastasiac

[10:12:40 CDT(-0500)] <alexn2> anastasiac: I moved my change into idi-theme into pull request IDI-33

[10:12:57 CDT(-0500)] <anastasiac> thanks, alexn2, I'll have another look

[11:07:25 CDT(-0500)] <anastasiac> michelled, cindyli, alexn2: most of this morning's pull requests have been merged into development; please update

[11:07:39 CDT(-0500)] <cindyli> ok, thanks, anastasiac

[11:37:54 CDT(-0500)] <anastasiac> cindyli, did you see my comment on your IDI-38 pull request, about the new yellow arrows having bad transparency?

[11:42:07 CDT(-0500)] <cindyli> ana

[11:42:12 CDT(-0500)] <cindyli> anastasiac: not yet, checking

[11:51:18 CDT(-0500)] <Justin_o> jhung: just realized that i set the file size limit to 400mb not 40mb… can't remember if that was intentional or not..

[11:51:52 CDT(-0500)] <Justin_o> jhung: what do you think about the limit.. I think we were trying to make it big enough that they could upload large tiff files without having to muck around the options

[11:52:56 CDT(-0500)] <Bosmon> yura, colinclark - https://github.com/substack/yarnify

[11:53:09 CDT(-0500)] <Bosmon> Could be useful in time as some infrastructure for kettle

[11:53:27 CDT(-0500)] <colinclark> what a great name

[11:53:29 CDT(-0500)] <colinclark> and metaphor

[11:53:35 CDT(-0500)] <colinclark> michelled will love it, too

[11:54:03 CDT(-0500)] <Bosmon> Based on "Browserify generates a single static bundle that you can drop into your application with a single <script> tag. You can use browserify with any kind of web stack that can host up static files."

[11:54:13 CDT(-0500)] <colinclark> wow, interesting

[11:55:08 CDT(-0500)] <Bosmon> It looks sufficiently agnostic that we could probably make an automatic "yarnify bridge" to the set of all Fluid components, somewhat similar to our "thatist bridge"

[11:55:34 CDT(-0500)] <Bosmon> We still have to fix the renderer so we can actually render them, of course : P

[12:07:31 CDT(-0500)] <Justin_o> jhung: okay.. one more thing too.. the server is currently capped at 100MB.. so there is an error here some where.. the server shouldn't be capped lower than the client.. i suppose the other way around is fine

[12:09:24 CDT(-0500)] <jhung> okay justin_o: that sounds fine. I guess we'll need to document that?

[12:16:29 CDT(-0500)] <alexn2> michelled, anastasiac: IE6 white on black theme when you hover on a home icon it has a blueish highlight

[12:47:34 CDT(-0500)] <anastasiac> alexn2, we've found a problem that's a result of the overflow changes

[12:47:45 CDT(-0500)] <alexn2> uh oh

[12:47:47 CDT(-0500)] <alexn2> what is it ?

[12:48:18 CDT(-0500)] <anastasiac> the prefs button at the top of the screen is pushed down - there's space above it. This is in modern browsers

[12:48:37 CDT(-0500)] <anastasiac> apparently, 'overflow:auto" on the body will fix it, but I don't know what that might break again in IE7

[12:49:39 CDT(-0500)] <Bosmon> anastasiac - please also think about overflow:none on the UIOptions body itself

[12:50:34 CDT(-0500)] <Bosmon> I am not quite seeing all the instances of the awful font problem fixed on Chrome...

[12:50:43 CDT(-0500)] <Bosmon> For example, the links at the top of the page still suffer from it

[12:51:16 CDT(-0500)] <Bosmon> As do the "Stay updated" and "IDI institutions" text

[12:56:10 CDT(-0500)] <cindyli> anastasiac: the transparency issue with the pull request for IDI-38 has been fixed. Can you give it a try? Thanks.

[12:56:38 CDT(-0500)] <anastasiac> will do, cindyli

[12:56:47 CDT(-0500)] <cindyli> thanks

[12:57:00 CDT(-0500)] <anastasiac> alexn2, see Bosmon's comments on overflow above

[13:00:58 CDT(-0500)] <cindyli> anastasiac: i just made another commit into IDI-38 pull request simply to merge in latest development branch. it was quite out of date.

[13:01:22 CDT(-0500)] <anastasiac> thanks, cindyli

[13:01:28 CDT(-0500)] <cindyli> np

[13:01:50 CDT(-0500)] <anastasiac> yes, cindyli and alexn2: Lots of stuff is being pushed into the dev branch, so please merge the latest into your branches before issuing pull requests

[13:02:08 CDT(-0500)] <alexn2> Bosmon: I'm not quiet following what is the problem with overlow (smile) Could you give some more details about it

[13:02:13 CDT(-0500)] <anastasiac> saves me merge conflicts later, and ensures your changes work with others' changes

[13:05:09 CDT(-0500)] <Bosmon> alexn2 - as you open up the UIOptions UI using the preferences button, you will see a vertical scrollbar at the right momentarily appear and disappear

[13:05:09 CDT(-0500)] <Justin_o> jhung: sorry… back now.. what do you think the cap should be.. i'll have to make a change somewhere so we might as well set it to the preferred value

[13:05:20 CDT(-0500)] <Bosmon> In some cases this will even cause the content to shift to the left during the opening phase

[13:05:32 CDT(-0500)] <Bosmon> We were discussing this issue with cindyli and anastasiac last Friday

[13:05:49 CDT(-0500)] <alexn2> Bosmon: which OS and browser should I use to replicate this ?

[13:05:51 CDT(-0500)] <Bosmon> btw, the team in general, I have reported the awful Chrome font rendering issue here https://code.google.com/p/chromium/issues/detail?id=129399

[13:05:55 CDT(-0500)] <Bosmon> alexn2 - any browser, any OS

[13:06:00 CDT(-0500)] <Bosmon> All you need is your eyes (smile)

[13:06:42 CDT(-0500)] <alexn2> well I do have eyes but I do not have this bug (smile) I can even send you a screenshot

[13:06:53 CDT(-0500)] <alexn2> used Mac OS Chrome

[13:07:08 CDT(-0500)] <alexn2> and Firefox

[13:07:18 CDT(-0500)] <alexn2> can someone else try it as well

[13:08:56 CDT(-0500)] <alexn2> actually Cindy showed me on her IE6 this behaviour. I guess it is not visible in my case only because my browsers are very fast and this scroll bar tend to disappear once UIO panel reached its maximum height

[13:13:44 CDT(-0500)] <jhung> justin_o: so this is the cap on the server per file?

[13:16:20 CDT(-0500)] <Justin_o> jhung: yes.. so in the client it's capped at 400mb per file, on the server it's capped at 100mb..

[13:16:33 CDT(-0500)] <Justin_o> jhung: basically on the sever requests have a 100mb cap on their body siz

[13:16:34 CDT(-0500)] <Justin_o> size

[13:16:48 CDT(-0500)] <jhung> Make it the same then. I can't think of a reason why they would be different. Right?

[13:17:48 CDT(-0500)] <Justin_o> jhung: sure.. which one should we take

[13:17:57 CDT(-0500)] <Justin_o> i figure 100mb is a reasonably large size.. what do you think?

[13:18:09 CDT(-0500)] <jhung> let's set 100MB for now.

[13:18:21 CDT(-0500)] <Justin_o> jhung: great.. sounds good.. i'll file a jira and fix that up

[13:18:57 CDT(-0500)] <jhung> k

[13:20:58 CDT(-0500)] <jhung> justin_o: can you give me some example server configuration options the user can adjust? File size is one. Is output directory another?

[13:22:26 CDT(-0500)] <Justin_o> jhung: they could change any of cherrypy's settings.. file size being one.. what do you mean by output directory though

[13:22:31 CDT(-0500)] <Justin_o> you mean the working directory?

[13:23:26 CDT(-0500)] <jhung> yeah - where the PDF gets written to before the download prompt.

[13:24:22 CDT(-0500)] <Bosmon> alexn2 - yes, you will need to be very fast to take a screenshot

[13:24:29 CDT(-0500)] <Bosmon> Given the opening action only takes 400ms

[13:25:50 CDT(-0500)] <alexn2> Bosmon: It is a good find

[13:26:14 CDT(-0500)] <alexn2> Bosmon: but we won't be able to fix it for the idi website in time. it is only 4 minutes left before the code freeze

[13:26:21 CDT(-0500)] <Bosmon> ok, cool (smile)

[13:26:27 CDT(-0500)] <Justin_o> jhung: here's the blocker http://issues.fluidproject.org/browse/DECA-265

[13:26:28 CDT(-0500)] <alexn2> but we could fix it in the next iteration of fixes though

[13:26:32 CDT(-0500)] <Justin_o> I'll go look into the directory thing now

[13:26:34 CDT(-0500)] <Bosmon> 4 minutes to fix a 400ms bug (smile)

[13:27:08 CDT(-0500)] <alexn2> it is more like overflow one. and it should be tested in multiple browsers. rebuild infusion and submit a pull request to idi to include it

[13:27:20 CDT(-0500)] <alexn2> but yes I see lots of 4s

[13:27:41 CDT(-0500)] <Bosmon> I put this overflow rule in myself, in the early days when I was not confident that the UIOptions height calculation would always work

[13:28:14 CDT(-0500)] <Bosmon> The idea was that if the height matching failed, the user would at least be able to scroll to the lower part of the panel rather than be "permanently stuck"

[13:28:16 CDT(-0500)] <alexn2> If I were a superstitious Chinese I would say that those numbers tell us about death and we should not play with higher powers.

[13:28:24 CDT(-0500)] <Bosmon> But it seems that the calculation has improved reliable in practice

[13:28:25 CDT(-0500)] <alexn2> and that the fix should go another time

[13:28:26 CDT(-0500)] <Bosmon> Yes, it's quite true

[13:28:34 CDT(-0500)] <Bosmon> It would be hugely better if we had 8 minutes to fix an 800ms bug

[13:28:48 CDT(-0500)] <alexn2> yes significantly

[13:29:14 CDT(-0500)] <Bosmon> I had a Chinese friend who said he felt positively dizzy when seeing a phone number on the side of a skyscraper.... 888-88888

[13:29:21 CDT(-0500)] <Bosmon> He said he simply couldn't conceive of anything more lucky

[13:29:26 CDT(-0500)] <Justin_o> jhung: here's a look at the cherrypy config http://code.google.com/p/decapod/source/browse/config/decapod.conf?repo=server

[13:29:47 CDT(-0500)] <Justin_o> jhung: so they could chance the path to the library to change the location of the working directory

[13:30:07 CDT(-0500)] <Justin_o> although it will create the specific subdirectories internally, that part isn't configurable

[13:34:42 CDT(-0500)] <alexn2> Bosmon: I have a friend who travelled into China as a bible smuggler. He told me about his experience in one of the small restaurants. He told me that you have to order an even number of dishes since odd numbers is a bad luck. So you left with 2,4,6 and 8. Then 2 is too little and not enough food (because of the portion sizes). 4 is a bad luck. So you left with 6 and 8… which is too much food and too expensive. (smile)

[13:34:47 CDT(-0500)] <jessm> anastasiac: ping

[13:35:25 CDT(-0500)] <Bosmon> alexn2 - it's a plot to inflate the economy (smile)

[13:36:09 CDT(-0500)] <anastasiac> jessm, polo

[13:36:36 CDT(-0500)] <jessm> anastasiac: let's do this thing – Skype me when you're ready

[13:39:21 CDT(-0500)] <cindyli> anastasiac: the last pull request is out to ur hands.

[13:39:53 CDT(-0500)] <anastasiac> thanks! I'll get to it once I'm off my call with jessm

[14:00:12 CDT(-0500)] <Justin_o> colinclark: uploader question.. for the file size limit.. is that the maximum size a file can be or must the file be smaller than that

[14:00:14 CDT(-0500)] <Justin_o> ?

[14:00:36 CDT(-0500)] <colinclark> Justin_o: hmm

[14:00:49 CDT(-0500)] <colinclark> I fear the answer to that is probably in code

[14:01:02 CDT(-0500)] <colinclark> I guess it's a subtle distinction (tongue)

[14:01:07 CDT(-0500)] <jessm> anastasiac: deb fels' bio has a http in it w/ no link

[14:01:46 CDT(-0500)] <Justin_o> colinclark: yes.. it seems at least in practice that the file must be smaller.. but I had expected that it was the max file size..

[14:02:01 CDT(-0500)] <Justin_o> colinclark: i'm wondering what the intention was though.. to know if this is a bug or not

[14:02:16 CDT(-0500)] <colinclark> Honestly, I'm not sure

[14:02:24 CDT(-0500)] <colinclark> a lot of these contract decisions weren't really made

[14:02:29 CDT(-0500)] <colinclark> they were inherited from SWFUpload

[14:02:34 CDT(-0500)] <colinclark> which, as you know, scares me

[14:05:42 CDT(-0500)] <Justin_o> colinclark: okay.. so i'm looking at the uploader api in the wiki now http://wiki.fluidproject.org/display/fluid/Uploader+API#UploaderAPI-%7B%7BqueueSettings%7D%7DOptions

[14:06:07 CDT(-0500)] <Justin_o> and it seems like it should have been the max size.. so i guess i'll file this as a bug.. and we can decide later to change the code or the docs

[14:06:34 CDT(-0500)] <colinclark> sure, that makes sense

[14:15:21 CDT(-0500)] <anastasiac> alexn2, cindyli, colinclark, jessm, Justin_o, yura: The dev IDI site is now up-to-date with everything we've done today. We'd like to freeze the site and test it on the various browser, checking for regressions (i.e. things that are worse than they are on the production site) or anything that looks broken. IF we pass QA, we can ask avtar to push the changes to the production site and call it "ready" for the launch tomorrow. Are you all

[14:15:21 CDT(-0500)] <anastasiac> available to help test?

[14:15:43 CDT(-0500)] <cindyli> sure, anastasiac

[14:15:48 CDT(-0500)] <alexn2> I get IE7 in Windows

[14:15:51 CDT(-0500)] <alexn2> (smile)

[14:15:55 CDT(-0500)] <alexn2> since my VM is open with it

[14:16:23 CDT(-0500)] <alexn2> I mean - I will test with IE7 so you can pick other browser

[14:16:53 CDT(-0500)] <jessm> anastasiac: sure

[14:17:06 CDT(-0500)] <jessm> anastasiac: i'm on mac FF, Safari and Chrome – tell me where to be

[14:17:16 CDT(-0500)] <jessm> anastasiac: also, did you get my edit above ^ ?

[14:17:51 CDT(-0500)] <anastasiac> so we're checking the dev site: http://dev.inclusivedesign.ca and first of all making sure we haven't regressed from what's already on the production site http://inclusivedesign.ca

[14:17:51 CDT(-0500)] <Justin_o> jhung: do you have an actual image file that is 100mb that we can test with

[14:17:59 CDT(-0500)] <anastasiac> jessm, yes, got it

[14:18:09 CDT(-0500)] <anastasiac> jessm, could you take Chrome?

[14:18:11 CDT(-0500)] <cindyli> anastasiac: i will do IE6 on windows first, let me know of other broswer/platform you want me to work on

[14:18:19 CDT(-0500)] <Justin_o> anastasiac: what configuration would you like me to test in

[14:18:20 CDT(-0500)] <Justin_o> i'm open

[14:18:32 CDT(-0500)] <anastasiac> Justin_o, why don't you take IE8, then? thanks

[14:18:39 CDT(-0500)] <Justin_o> anastasiac: will do

[14:18:58 CDT(-0500)] <jessm> anastasiac: i got chrome!

[14:19:30 CDT(-0500)] <anastasiac> I'll start with IE9

[14:20:59 CDT(-0500)] <anastasiac> Justin_o, jessm, cindyli, alexn2: We're looking for regressions, and anything we think we really don't want people to see tomorrow. Please also try reading the text, check for spelling, typos, broken links, etc

[14:21:18 CDT(-0500)] <jessm> anastasiac: clarify, we should be checking the dev site now?

[14:21:34 CDT(-0500)] <anastasiac> jessm, yes: dev site is up-to-date, so go for it

[14:22:19 CDT(-0500)] <anastasiac> alexn2, jessm, Justin_o, cindyli, remember to clear your cache, if you don't have ti turned off (smile)

[14:22:59 CDT(-0500)] <jessm> anastasiac: marcia rioux space after PhD

[14:23:13 CDT(-0500)] <anastasiac> alexn2, jessm, Justin_o, cindyli: play with thinks like focus styling (tab through the links), hover styling, resizing the window. And try out the UIO themes and other options

[14:24:25 CDT(-0500)] <jessm> anastasiac: brian cantwell smith space after Professor

[14:26:19 CDT(-0500)] <alexn2> ok I found a regression. In IE7 if you compare this http://dev.inclusivedesign.ca/infrastructure/ to http://inclusivedesign.ca/infrastructure/

[14:26:41 CDT(-0500)] <alexn2> the icon which looks like a water drop close to locations shifts headers to the right a bit

[14:27:55 CDT(-0500)] <alexn2> it looks like a bug with Equipment page

[14:27:58 CDT(-0500)] <alexn2> which we had before

[14:30:07 CDT(-0500)] <alexn2> ok this problem on all browsers I think

[14:30:14 CDT(-0500)] <alexn2> it does not look good even on my MAC

[14:30:22 CDT(-0500)] <alexn2> but it looks like a new column was added

[14:30:31 CDT(-0500)] <alexn2> so not sure if it is related to the old bug anymore

[14:30:34 CDT(-0500)] <jessm> do others see the "people" nav item in the live site?

[14:30:41 CDT(-0500)] <jessm> anastasiac: am i supposed to?

[14:31:05 CDT(-0500)] <anastasiac> jessm, you should see it on dev, but not on production

[14:31:15 CDT(-0500)] <jessm> roger that

[14:33:44 CDT(-0500)] <Justin_o> anastasiac: is inclusive design institute supposed to be all lower case or have the words start with capitals.. or maybe it doesn't matter.. i just notice that the name at the top is all lower case, but at the bottom in the contact us they start with caps

[14:34:07 CDT(-0500)] <anastasiac> Justin_o: logo is lower case, otherwise capitalized

[14:34:10 CDT(-0500)] <alexn2> another regression in IE7 (not sure about other browsers). If you switch to high contrast theme (yellow on black) all headers will have a border

[14:34:20 CDT(-0500)] <alexn2> while they did not have before

[14:34:33 CDT(-0500)] <anastasiac> alexn2, I see that in IE9, too

[14:37:08 CDT(-0500)] <alexn2> visit project page links are tabbable http://dev.inclusivedesign.ca/research/ocadu/ not a regression but it is an accessibility inconvenience for screen readers.

[14:37:41 CDT(-0500)] <Justin_o> alexn2, anastasiac: i'm only seeing that on the home page, the other pages seem fine ( for the border issue)

[14:37:44 CDT(-0500)] <anastasiac> alexn2, yes, a known issue - I just haven't fixed it yet (not a github change)

[14:37:59 CDT(-0500)] <anastasiac> Justin_o, do you see it on the news page?

[14:38:21 CDT(-0500)] <Justin_o> anastasiac: ah yes.. it's there too

[14:38:30 CDT(-0500)] <anastasiac> k

[14:38:50 CDT(-0500)] <Justin_o> anastasiac: and infrastructure.. so maybe i should revise and say.. some pages don't have that problem

[14:40:36 CDT(-0500)] <anastasiac> Justin_o, cindyli, alexn2, yura: if you feel you've tested thoroughly, let me know - there are more browsers to test (smile)

[14:41:16 CDT(-0500)] <alexn2> not sure if it is regression or a fix. When you tab through links they do not have a yellow border anymore. IE7

[14:41:30 CDT(-0500)] <Justin_o> anastasiac: don't feel like that yet, but will let you know.. I have to leave at 4 so i don't know if i'll get to another browser

[14:41:57 CDT(-0500)] <anastasiac> np, Justin_o, there are only a couple left. thanks for your help!

[14:42:46 CDT(-0500)] <alexn2> ok I think I can take another browser to test anastasiac

[14:43:09 CDT(-0500)] <anastasiac> alexn2, how about FF on Mac or Windows

[14:43:17 CDT(-0500)] <alexn2> sure

[14:43:19 CDT(-0500)] <alexn2> Windows

[14:43:59 CDT(-0500)] <Justin_o> anastasiac: are the "read article" links not supposed to be tabable?

[14:44:21 CDT(-0500)] <anastasiac> Justin_o, wherever there are more than one link to something, only one would be tabbable

[14:44:31 CDT(-0500)] <Justin_o> anastasiac: okay.. thanks.

[14:45:06 CDT(-0500)] <Justin_o> anastasiac: minor issue the focus styling on links within the twitter feeds causes the layout to shift

[14:45:15 CDT(-0500)] <Justin_o> the text below shifts

[14:45:20 CDT(-0500)] <cindyli> anastasiac: Links for leaders on research cluster page left side menu point to, for example,

[14:45:20 CDT(-0500)] <cindyli> http://dev.inclusivedesign.ca/about/#Treviranus which no longer exist. They should point to people/#Treviranus

[14:45:32 CDT(-0500)] <anastasiac> cindyli, good - yura just caught that

[14:45:38 CDT(-0500)] <yura> FIRST (smile)

[14:45:50 CDT(-0500)] <cindyli> haha, i will shoot some more to be the first

[14:46:12 CDT(-0500)] <cindyli> anastasiac: IE6 Regression:

[14:46:12 CDT(-0500)] <cindyli> 1. http://dev.inclusivedesign.ca/research/policies/

[14:46:12 CDT(-0500)] <cindyli> The project desc is too far from the logo

[14:46:13 CDT(-0500)] <cindyli> 2. all the left side menu has primary content wrapping under it.

[14:46:26 CDT(-0500)] <cindyli> IE6 Bug but not regression:

[14:46:26 CDT(-0500)] <cindyli> 1. "back to news" left arrow hover icon is not swapped in.

[14:47:04 CDT(-0500)] <anastasiac> cindyli, for the project description issue: is that only on the policies page, or all clusters?

[14:47:17 CDT(-0500)] <cindyli> all clusters, anastasiac

[14:47:22 CDT(-0500)] <alexn2> not sure if it s valid but before Mozarella Firefox had unreadable text but readable headings (http://inclusivedesign.ca/about/) in dev branch it has readable text but not so smooth headings (http://dev.inclusivedesign.ca/people/)

[14:49:23 CDT(-0500)] <jessm> alexn2: OMG i love it – mozzarella firefox!

[14:49:43 CDT(-0500)] <alexn2> I was wondering if anyone notices (big grin)

[14:49:55 CDT(-0500)] <jessm> hi-larious

[14:50:10 CDT(-0500)] <alexn2> *if anyone would notice (sorry my typing has a major regression since morning)

[14:50:14 CDT(-0500)] <Justin_o> anastasiac: not a regression, but if a header is hidden under the navigation bar, you can tab to it, but it won't' slide below the navigation

[14:50:24 CDT(-0500)] <cindyli> The cause is the firefox/windows is using ariel now rather than lato

[14:50:40 CDT(-0500)] <anastasiac> Justin_o, yes, a known issue

[14:50:58 CDT(-0500)] <yura> some background info - http://answers.yahoo.com/question/index?qid=20080317162900AAQeM89

[14:52:28 CDT(-0500)] <Justin_o> anastasiac: also not a regression, but the columns on the infrastructure page are too narrow so the W. is on it's own line for the addresses

[14:53:19 CDT(-0500)] <Justin_o> anastasiac: also the focus styling for UIO is a different colour in the default theme from the rest of the site

[14:54:06 CDT(-0500)] <anastasiac> Justin_o, good catch

[14:54:10 CDT(-0500)] <jessm> wow, i totally suck at testing, y'all are catching everything

[14:55:30 CDT(-0500)] <jessm> where does focus go after "UIO"?

[14:55:33 CDT(-0500)] <Justin_o> anastasiac: if you use verdana and increase the linespacing.. the word "institue" will slide below the navigation

[14:55:35 CDT(-0500)] <jessm> there's another click before the next link

[14:56:01 CDT(-0500)] <alexn2> FF in Windows. http://dev.inclusivedesign.ca/research/ my window occupies ~50% of my total screen. When you hover over cluster icon Learn More will be split into 2 lines. Before the phrase was a one liner.

[14:56:08 CDT(-0500)] <Justin_o> jessm: i think that's a skip to content link

[14:56:14 CDT(-0500)] <jessm> ah

[14:57:07 CDT(-0500)] <jessm> wow, I'm so impressed with the testing

[14:57:14 CDT(-0500)] <jessm> anyone more skilled want to take chrome for mac?

[14:58:17 CDT(-0500)] <alexn2> well Chrome and FF for MAC should be almost flawless. I would give a chocolate to anyone who finds an issue there

[15:00:23 CDT(-0500)] <jessm> incentive...

[15:00:53 CDT(-0500)] <Justin_o> anastasiac: on the infrastructure page for the address when setting a contrast theme, their focus styling does not change the whole background, only a strip at the bottom

[15:01:10 CDT(-0500)] <jessm> anastasiac: has someone tested the join the mailing list link lately?

[15:01:30 CDT(-0500)] <jessm> anastasiac: also, can someone just drop all the page text systematically into word.doc and check for spelling?

[15:01:51 CDT(-0500)] <alexn2> so...

[15:02:02 CDT(-0500)] <alexn2> contentwise http://dev.inclusivedesign.ca/people/ for Evan Weaver

[15:02:06 CDT(-0500)] <alexn2> it says he is a Chair

[15:02:10 CDT(-0500)] <anastasiac> I've spell checked the about page and the workshops page

[15:02:26 CDT(-0500)] <alexn2> I googled and found only http://en.wikipedia.org/wiki/Chair

[15:02:42 CDT(-0500)] <alexn2> some info might be missing about Evan

[15:02:53 CDT(-0500)] <alexn2> it does not say organization where he works

[15:03:27 CDT(-0500)] <alexn2> well it is like this in FF Windows

[15:03:31 CDT(-0500)] <anastasiac> alexn2, I'm seeing "Chair, School of Information and Communications Technology

[15:03:31 CDT(-0500)] <anastasiac> Seneca College"

[15:03:46 CDT(-0500)] <anastasiac> which is the info Seneca gave to us - do you not see that?

[15:03:47 CDT(-0500)] <alexn2> oh my bad

[15:03:54 CDT(-0500)] <alexn2> I think I have some cache issues

[15:04:02 CDT(-0500)] <jessm> some have a <br> after their title and it goes to the next line

[15:04:08 CDT(-0500)] <alexn2> I restarted browser, it looked like it never finished loading or something

[15:04:08 CDT(-0500)] <jessm> evan weaver is one of them

[15:04:30 CDT(-0500)] <jessm> we are inconsistent on this page with format where alexn2 is looking

[15:04:41 CDT(-0500)] <jessm> we do title, institution sometimes and others we do

[15:04:42 CDT(-0500)] <jessm> title

[15:04:45 CDT(-0500)] <jessm> institution

[15:04:51 CDT(-0500)] <jessm> no comma and a break

[15:05:01 CDT(-0500)] <jessm> not a blocker in my mind

[15:05:16 CDT(-0500)] <alexn2> jessm: could you give us an example

[15:05:40 CDT(-0500)] <jessm> evan weaver has a comma and a break

[15:05:57 CDT(-0500)] <jessm> kevin stolarick comma and break

[15:06:08 CDT(-0500)] <jessm> brian cantwell smith no comma and no break

[15:06:20 CDT(-0500)] <jessm> marlene scardamalia no comma and break

[15:16:40 CDT(-0500)] <jessm> anastasiac: is this the right spelling? Paediatric on tom chau's bio?

[15:17:23 CDT(-0500)] <anastasiac> jessm, yes, it's acceptable

[15:18:01 CDT(-0500)] <jessm> canadian?

[15:18:12 CDT(-0500)] <jessm> anastasiac: under peter pennefather's bio formatting should have two t's

[15:24:29 CDT(-0500)] <jessm> anastasiac: UOIT page has a typeo

[15:24:35 CDT(-0500)] <jessm> i spelled that wrong

[15:24:48 CDT(-0500)] <jessm> in the paragraph starting The Faculty of Education...

[15:24:54 CDT(-0500)] <jessm> Finally, sone – should be some

[15:25:51 CDT(-0500)] <jessm> are folks still there?

[15:26:38 CDT(-0500)] <anastasiac> jessm, I'm still here, others are too but not necessarily testing. I think we're tested out

[15:26:55 CDT(-0500)] <jessm> i'm not done!

[15:27:00 CDT(-0500)] <anastasiac> jessm, I've put the issues we've found on the iteration plan - next step is to decide which are blockers

[15:27:07 CDT(-0500)] <cindyli> anastasiac: 2 more things with IE6 regression:

[15:27:08 CDT(-0500)] <cindyli> 3. the gap on the front page in btw the area with cirles and the text below is too wide.

[15:27:08 CDT(-0500)] <cindyli> 4. the front page footer is mixed up with actual content.

[15:27:09 CDT(-0500)] <anastasiac> no, jessm, you get special consideration (smile)

[15:27:52 CDT(-0500)] <jessm> ok, i'm done

[15:27:57 CDT(-0500)] <jessm> i dropped it all into word

[15:28:02 CDT(-0500)] <jessm> spell-checked

[15:28:37 CDT(-0500)] <anastasiac> thanks, jessm, that's great!

[15:28:52 CDT(-0500)] <jessm> anastasiac: now, let's move on this – what is our next step?

[15:29:26 CDT(-0500)] <jessm> are you fixing blockers?

[15:29:39 CDT(-0500)] <anastasiac> jessm, next step is to go through the issues and decide what's a blocker - then we fix them (smile)

[15:29:49 CDT(-0500)] <anastasiac> jessm, I've put everything on the iteration page

[15:33:47 CDT(-0500)] <anastasiac> cindyli, are you still here?

[15:35:02 CDT(-0500)] <cindyli> yes, anastasiac

[15:35:20 CDT(-0500)] <anastasiac> are you available to help fix a few bugs before leaving?

[15:35:36 CDT(-0500)] <cindyli> sure, what are those? anastasiac

[15:35:46 CDT(-0500)] <anastasiac> we're just deciding that: http://wiki.fluidproject.org/display/fluid/Floe+Iteration+Plan

[15:36:13 CDT(-0500)] <jessm> cindyli: we're just all figuring out next steps

[15:36:17 CDT(-0500)] <jessm> but some of these are tiny tiny

[15:36:32 CDT(-0500)] <cindyli> okay

[15:37:35 CDT(-0500)] <anastasiac> cindyli, for your layout problems in ie6 on the cluster pages, was that in the default theme, or high contrast

[15:37:36 CDT(-0500)] <anastasiac> ?

[15:37:52 CDT(-0500)] <cindyli> anastasiac: both

[15:41:22 CDT(-0500)] <anastasiac> cindyli, regarding "2. all the left side menu has primary content wrapping under it.", which page is that?

[15:41:56 CDT(-0500)] <cindyli> about, people, anastasiac

[15:56:53 CDT(-0500)] <anastasiac> cindyli, the iteration page lists a few blockers at the top that are really just spelling, typos, broken links. Could you start on those?

[15:56:57 CDT(-0500)] <anastasiac> shouldn't take long

[15:57:46 CDT(-0500)] <cindyli> anastasiac: i'm working on the IE6 cluster page layout issue. should I finish it first or switch to spelling stuff?

[15:58:20 CDT(-0500)] <anastasiac> cindly, I think that's not a blocker

[15:58:32 CDT(-0500)] <anastasiac> unless it's super quick to fix

[16:00:19 CDT(-0500)] <cindyli> ok, it's apparantly not super quick. will give another 2 mins

[16:00:23 CDT(-0500)] <anastasiac> cindyli, are you close to fixing the layout, or still working on it?

[16:00:34 CDT(-0500)] <cindyli> working on it for another 2 mins

[16:14:49 CDT(-0500)] <travis_84> hey colin and bosmon, did you get my link to the blog?

[17:33:48 CDT(-0500)] <avtar> anastasiac, colinclark: i'm going to head over if everything looks good on your end

[17:34:06 CDT(-0500)] <anastasiac> avtar, looks good to me, thanks!

[17:34:25 CDT(-0500)] <avtar> perfect

[17:34:33 CDT(-0500)] <colinclark> +1

[17:36:01 CDT(-0500)] <travis_84> colinclark: I posted the latest on my blog when you get a chance

[17:38:07 CDT(-0500)] <colinclark> hey travis_84

[17:38:15 CDT(-0500)] <colinclark> Bosmon sent it along last night

[17:38:21 CDT(-0500)] <colinclark> I had a quick look and your wireframes looked really nice

[17:38:30 CDT(-0500)] <colinclark> great classic Mac OS sort of aesthetic

[17:39:24 CDT(-0500)] <travis_84> hah just basic layout but hey that's cool

[17:40:12 CDT(-0500)] <travis_84> thanks for the update

[17:41:44 CDT(-0500)] <colinclark> and the "non-modal" refactoring also looks good

[17:41:49 CDT(-0500)] <colinclark> i'd like to spend a bit more time with your blog post

[17:41:58 CDT(-0500)] <colinclark> i was intrigued by how you were thinking of simplifying CSS in some way

[17:42:01 CDT(-0500)] <colinclark> which is quite interesting

[17:43:17 CDT(-0500)] <colinclark> travis_84: What do you think your next steps will be?

[17:46:09 CDT(-0500)] <travis_84> right now I am placing buttons for the center editing module, for the basics... add new obj., save, undo, redo, etc.

[17:50:06 CDT(-0500)] <travis_84> as far as CSS much of the problem of multi-params. like that described. I know I don't want to add a long list of CSS objects for them to "check" the ones they need

[17:52:48 CDT(-0500)] <travis_84> I guess the question is how much should be done while editing an html doc. or whether all css should be delegated to the css doc editing mode

[17:53:17 CDT(-0500)] <Bosmon> travis_84 - visual designs look good, yes

[17:53:38 CDT(-0500)] <Bosmon> I'd probably avoid going into too many details related to particular modes until you have the basic editing idiom sorted out for a few simple cases

[17:54:01 CDT(-0500)] <Bosmon> And as your current question is highlighting, you may want to have easy navigation between things that you are currently calling "modes"

[17:54:24 CDT(-0500)] <Bosmon> Frequently producing "a work" will involve simultaneous work on a number of documents of different types

[17:54:43 CDT(-0500)] <Bosmon> Which might have various kinds of relationships between them

[17:54:59 CDT(-0500)] <Bosmon> For example, the fact that CSS documents are organised around selectors which "hit" various parts of an HTML document

[18:02:37 CDT(-0500)] <travis_84> as shown (sort of) in the wireframes, when a new project is made, it will always come stock with an attached style.css. selecting the .html file or the .css file will automatically save the current state and switch modes

[18:04:09 CDT(-0500)] <Bosmon> Probably there should be no "extra state saving" - most modern apps (at least Apple-inspired ones) will ensure that all relevant state is kept saved at all times

[18:04:26 CDT(-0500)] <Bosmon> colinclark is fond of this afforance too (smile)

[18:04:58 CDT(-0500)] <travis_84> ahh ok

[18:05:38 CDT(-0500)] <Bosmon> And rather than "switching modes" it might be better to think of it as a case of "seeing something else" : P similarly to our avoidance of "modality" in the interface to date

[18:06:02 CDT(-0500)] <Bosmon> There may not be much difference in the visual interface from this difference in thinking, but it might lead to a few things, who knows....

[18:06:49 CDT(-0500)] <Bosmon> But in general, to think that there is "one thing" that you are seeing different parts of at different times, rather than a "thing" which changes state depending on what you are editing

[18:07:13 CDT(-0500)] <Bosmon> That view will probably make it easier to think about what the behaviour of the non-visual editing modes should be

[18:09:06 CDT(-0500)] <travis_84> yes

[18:09:17 CDT(-0500)] <travis_84> nothing actually will change

[18:10:30 CDT(-0500)] <travis_84> except the contents displayed in the center, much like happens in browser tabs

[18:10:37 CDT(-0500)] <travis_84> i guess

[18:11:17 CDT(-0500)] <Bosmon> There's a choice there - between the Mac-like "there is only one app" model, and the Eclipse-like "there is one tab for each document model"

[18:11:32 CDT(-0500)] <Bosmon> We can ignore the dreadful Microsoft MDI-like "there is one window for each document" model : P

[18:12:06 CDT(-0500)] <travis_84> I prefer the Mac in this cas

[18:12:11 CDT(-0500)] <travis_84> case*

[18:12:35 CDT(-0500)] <Bosmon> Yes - probably the "tab space" will get very crowded... in Eclipse, I usually have so many documents open that it is worthless anyway

[18:12:47 CDT(-0500)] <Bosmon> I am always in the "99+" case that it reserves for when it gives up : P

[18:13:20 CDT(-0500)] <Bosmon> Better to save the tab space, and invest in some better primitives for getting between documents quickly

[18:13:21 CDT(-0500)] <travis_84> haha, the top left doc list is enough for me

[18:14:07 CDT(-0500)] <Bosmon> The one I like the best is the "LRU electric mode" that I saw first in emacs and the OS/2 EPM editor ... you hit a key, and a list of documents comes up in LRU order, with the current one omitted

[18:14:31 CDT(-0500)] <Bosmon> So in my case it is F12... if I am working on 2 documents, I can hit F12, then enter, and that switches between the two of them

[18:14:39 CDT(-0500)] <Bosmon> If I am working on 3, I can hit F12, then down arrow, then enter

[18:15:23 CDT(-0500)] <Bosmon> But you could imagine those combinations could easily be mapped to dedicated voice commands

[18:15:34 CDT(-0500)] <Bosmon> "next-document", or "next-document 1" or so on

[18:15:49 CDT(-0500)] <Bosmon> I think Eclipse calls it "Next Editor"

[18:16:01 CDT(-0500)] <Bosmon> But it's not a well-known feature

[18:16:09 CDT(-0500)] <Bosmon> Given its rather bland name

[18:16:22 CDT(-0500)] <Bosmon> emacs, "electic-buffer-menu" is much more exciting : P

[18:16:28 CDT(-0500)] <Bosmon> electric

[18:18:06 CDT(-0500)] <travis_84> mapping to voice is a good idea, also "list all html documents"

[18:18:54 CDT(-0500)] <Bosmon> A key thing it would be nice to see something of, during this project, even if it is very basic, is exploring the way the user might be able to dynamically augment the interface by creating new queries and voice commands

[18:19:08 CDT(-0500)] <Bosmon> Since I think that is one of the key goals behind the project as originally formed

[18:19:25 CDT(-0500)] <Bosmon> That is, "helping users craft their own queries to match their style of working"

[18:20:04 CDT(-0500)] <travis_84> can you give an example?

[18:20:14 CDT(-0500)] <Bosmon> As in the example of T.V. Raman playing his bejeweled-like game using voice - at the outset, when he is not a very good player, he may not actually know what queries he wants

[18:20:34 CDT(-0500)] <Bosmon> Perhaps we haven't talked about this example yet, but it was one of the key examples in Clayton's mind motivating the "non-visual" idea

[18:20:59 CDT(-0500)] <Bosmon> So Raman is a high-ranking Google engineer who has a small amount of vision, but mostly uses his laptop with the screen turned off

[18:21:23 CDT(-0500)] <Bosmon> And his example of an interaction was playing a game like bejeweled, which is traditionally considered highly visually oriented

[18:21:50 CDT(-0500)] <Bosmon> His idea was that you could consider that what vision really enables you do to is quickly find the results of certain "queries"

[18:22:13 CDT(-0500)] <Bosmon> But the kinds of queries you are interested in is highly task-dependent, as well as highly dependent on the state and goals of the person

[18:22:42 CDT(-0500)] <Bosmon> So he might, for example, start out with queries like, "tell me the location of the first <red> stone in the column to the left of the cursor"

[18:22:51 CDT(-0500)] <Bosmon> Where, perhaps <red> is a substitutable element in the query

[18:23:07 CDT(-0500)] <Bosmon> But might eventually work up to more sophisticated ones, depending on what he understands about how to play the game

[18:23:39 CDT(-0500)] <travis_84> like WoW Macros

[18:23:40 CDT(-0500)] <Bosmon> But ideally, he has a "UI" which can adapt to his understanding and capabilities of the task

[18:23:43 CDT(-0500)] <Bosmon> Yes, exactly like that

[18:24:01 CDT(-0500)] <Bosmon> But preferably with a less malign syntax and semantic (smile)

[18:24:20 CDT(-0500)] <Bosmon> Of course, the great challenge is that the macro system itself is something that needs to be operated with the voice

[18:24:21 CDT(-0500)] <Bosmon> etc. etc.

[18:24:35 CDT(-0500)] <Bosmon> Which is why it is key that the system is oriented around "editing structured documents in general"

[18:24:42 CDT(-0500)] <travis_84> streamlining repeatative tasks to a single command

[18:24:45 CDT(-0500)] <Bosmon> Whether they are HTML, CSS, JSON or any other tree-oriented material

[18:25:10 CDT(-0500)] <Bosmon> Ideally the macros themselves would be a thing which the system could be turned to editing - and we would probably to start with have a simple JSON-like syntax for that

[18:27:40 CDT(-0500)] <travis_84> how do we plan to handle errors in the initial recognition, like Siri says "did you mean <such and such>"? if we are avoiding dialogs?

[18:28:05 CDT(-0500)] <Bosmon> Perhaps as some kind of "autocomplete" or "suggestions" system

[18:28:17 CDT(-0500)] <Bosmon> Think about how Google deals with possibly mistyped search queries

[18:28:30 CDT(-0500)] <Bosmon> It does its best to "do something sensible immediately" but at the same time highlights any problems that there might have been

[18:28:50 CDT(-0500)] <Bosmon> In the case of a query, "doing something sensible immediately" involves finding some query that returns more than 0 results

[18:29:14 CDT(-0500)] <Bosmon> But may have different effects in the case of different kinds of commands

[18:29:33 CDT(-0500)] <Bosmon> CLearly if the command might have some destructive effects, and you aren't sure about it, it might be best not to actually do anything immediately : P

[18:30:16 CDT(-0500)] <travis_84> my thoughts exactly

[18:30:30 CDT(-0500)] <travis_84> autocomplete sounds dangerous

[18:30:59 CDT(-0500)] <Bosmon> From my experience, the key thing about autocomplete is to "never do anything that changes the meaning of what the user has entered already without their consent"

[18:31:03 CDT(-0500)] <travis_84> but offering suggestions without a popup...

[18:31:18 CDT(-0500)] <Bosmon> I've seen countless absolutely horrible autocomplete systems that make it basically impossible to type normally

[18:31:22 CDT(-0500)] <Bosmon> Well.... there are popups, and popups (smile)

[18:31:37 CDT(-0500)] <Bosmon> Just because "something appears" doesn't mean its necessarily a modal, or intrusive interaction

[18:32:04 CDT(-0500)] <Bosmon> In the visual case, it may just be highlighting that pressing certain special keys are temporarily bound to various available effects

[18:32:14 CDT(-0500)] <Bosmon> But if you don't invite the special effects, you can just carry on as normal

[18:32:52 CDT(-0500)] <Bosmon> In terms of speech interactions, this might correspond to the difference between the various levels of "politeness" in ARIA live region support

[18:33:31 CDT(-0500)] <Bosmon> For example, a "polite" interaction will wait until the system is idle (the user doesn't seem to be interacting for a short while) before reporting info or alternatives

[18:33:45 CDT(-0500)] <Bosmon> A "rude" interaction might butt in, interrupting the current voice stream

[18:34:12 CDT(-0500)] <Bosmon> For certain extremely urgent interactions, "rudeness" might be appropriate, but one way of casting the interaction problem might be in terms of trying to find ways to get work done in as "polite" a way as possible

[18:35:00 CDT(-0500)] <travis_84> oops

[18:35:28 CDT(-0500)] <travis_84> got typing so fast that I hit a bad key combo

[18:35:44 CDT(-0500)] <Bosmon> (smile)

[18:36:33 CDT(-0500)] <Bosmon> I pasted a catch-up into Skype...

[18:38:09 CDT(-0500)] <travis_84> so as I was trying to type...having a slideout like in the case of the properties panel could work?

[18:38:30 CDT(-0500)] <Bosmon> yes.... things can appear, so long as they don't forcibly disrupt the user's current flow

[18:38:42 CDT(-0500)] <Bosmon> Or cut off interaction operations that are normally available

[18:39:05 CDT(-0500)] <Bosmon> This is the real pernicious aspect of modal dialogues - that the user is forced to become aware of them, and interrupt something that they might ordinarily be doing

[18:39:39 CDT(-0500)] <Bosmon> And this is also the kind of problem with "bad" autocomplete implementations - ones that insert stuff into the text stream by default, rather than waiting to be invited

[18:41:11 CDT(-0500)] <travis_84> in the case of the blind, it would speak and they could either say 'yes' or w/e or just move on?

[18:41:40 CDT(-0500)] <Bosmon> Yes, something like that - and in terms of "politeness" it would only speak if they weren't already trying to do something already

[18:42:17 CDT(-0500)] <travis_84> right like a

[18:42:45 CDT(-0500)] <travis_84> don't output when a recognition is in affect

[18:43:21 CDT(-0500)] <Bosmon> Given that our initial implementation is going to be "as far as possible a normal webapp", it might well be that we just use standard ARIA mechanisms for all of our output speech

[18:43:41 CDT(-0500)] <Bosmon> Which implies actually using the ARIA politeness setting itself, rather than just thinking up something that is a bit like it : P

[18:44:05 CDT(-0500)] <Bosmon> But that's a decision that I guess depends on how good an interaction we can get whilst "playing by the rules"

[18:44:23 CDT(-0500)] <Bosmon> It would also save a lot of implementation work, if it proved workable

[18:44:41 CDT(-0500)] <Bosmon> Have you looked much at the ARIA descriptions and "live regions" system?

[18:44:58 CDT(-0500)] <Bosmon> It looks very daunting as a spec at first sight, but we found out in practice that almost none of it is implemented (smile)

[18:45:05 CDT(-0500)] <Bosmon> Which simplifies working with it quite a lot (smile)

[18:47:32 CDT(-0500)] <travis_84> yes I have, I think I mentioned it in passing in the blog

[18:47:56 CDT(-0500)] <Bosmon> cool

[18:49:28 CDT(-0500)] <travis_84> when rolling over an option, another "live" region continualy updates according to what's focused

[18:49:43 CDT(-0500)] <travis_84> is that what you mean?

[18:49:53 CDT(-0500)] <Bosmon> Yes, that's right

[18:50:07 CDT(-0500)] <Bosmon> And it may or may not interrupt the voice flow, depending on the "politeness" setting of the update

[18:50:52 CDT(-0500)] <travis_84> the w3c talked about its use as a stockmarket type widget

[18:51:48 CDT(-0500)] <travis_84> or # of characters remaining counter

[18:51:56 CDT(-0500)] <travis_84> another question regarding navigation, say I set each module in a tab order, would the arrow keys be the secondary navigation method when reaching the desired module?

[18:52:34 CDT(-0500)] <Bosmon> That sounds like a reasonable choice

[18:52:51 CDT(-0500)] <Bosmon> In our "reorderable" widget we allow arrow keys within a particular domain of markup

[18:53:02 CDT(-0500)] <Bosmon> Whilst retaining the tab key for larger-scale navigation between domains

[18:53:39 CDT(-0500)] <Bosmon> My CATTT is now pushing his empty can of CATT FFOOD round and round the floor

[18:53:45 CDT(-0500)] <Bosmon> A clear form of communication (smile)

[18:54:15 CDT(-0500)] <travis_84> haha indeed

[18:54:59 CDT(-0500)] <travis_84> thanks, this has made me feel more comfortable with my progress

[18:55:05 CDT(-0500)] <Bosmon> I'm glad

[18:55:16 CDT(-0500)] <Bosmon> I hope you will shout if you are feeling uncomfortable about something, or neglected : P

[18:55:31 CDT(-0500)] <Bosmon> I know it's a pretty daunting project in its scope, but all we need to do is hack off some definable lump of it

[18:55:41 CDT(-0500)] <Bosmon> Noone expects you to transform the entire world overnight (smile)

[18:58:24 CDT(-0500)] <travis_84> right, soon enough we will definitely need to sit and write this into order and i'll need to plan everything out

[18:58:41 CDT(-0500)] <Bosmon> Definitely

[19:00:15 CDT(-0500)] <travis_84> right now I am a mad mathmatician with papers EVERYWHERE and all of it is organized in my head

[19:00:22 CDT(-0500)] <travis_84> bad idea : P

[19:00:38 CDT(-0500)] <Bosmon> Well... everything has to start in the HEAD : P