fluid-work IRC Logs-2008-07-18

Owned by Fluid IRC Log Bot
Last updated: Jul 18, 2008Version comment
9 min read

[06:59:38 EDT(-0400)] * phiggins (n=dante@c-68-34-199-67.hsd1.tn.comcast.net) has joined #fluid-work
[08:13:05 EDT(-0400)] * Justin_o (n=Justin@142.150.154.235) has joined #fluid-work
[08:40:49 EDT(-0400)] * colinclark (n=colin@bas1-toronto09-1279534421.dsl.bell.ca) has joined #fluid-work
[08:41:49 EDT(-0400)] * colinclark (n=colin@bas1-toronto09-1279534421.dsl.bell.ca) has joined #fluid-work
[09:38:55 EDT(-0400)] * theclown (n=theclown@142.150.154.101) has joined #fluid-work
[10:07:45 EDT(-0400)] * anastasiac (n=team@142.150.154.160) has joined #fluid-work
[10:08:29 EDT(-0400)] * michelled (n=team@142.150.154.197) has joined #fluid-work
[10:34:47 EDT(-0400)] * Jacob2 (n=Fluid@142.150.154.106) has joined #fluid-work
[10:35:00 EDT(-0400)] * Bosmon (n=Antranig@ginger.caret.cam.ac.uk) has joined #fluid-work
[10:35:26 EDT(-0400)] <Bosmon> The focusing events seem extremely inconsistent, even down to being affected by whether the Firebug console is open or not...
[10:35:43 EDT(-0400)] <Bosmon> But the key issue seems to be that the "highlight" listener is bound to the text field, whereas the blur is only received on the editable field
[10:36:09 EDT(-0400)] <Bosmon> I would suggest that a blur received on either of them should be considered a valid trigger for the highlight...
[10:36:20 EDT(-0400)] <Bosmon> There is this really bizarre "second click" phenomenon though
[10:36:54 EDT(-0400)] <Bosmon> Clicking a second time "away" from any editable control causes a further blur event to be delivered, this time for the text control as the highlight listener is expecting
[10:37:57 EDT(-0400)] <Bosmon> The first click only causes a blur for the edit control, which means the control state toggles (partially) but the focus indicator is left behind
[10:38:04 EDT(-0400)] <Bosmon> This is all on FF2 on Windows...
[10:38:33 EDT(-0400)] * colinclark (n=colin@142.150.154.101) has joined #fluid-work
[10:39:33 EDT(-0400)] <michelled> Bosmon: I agree that both the text and the editable field should listen to blur. Would that fix the focus style issues?
[10:40:36 EDT(-0400)] <Jacob2> are losing the cursor in FF?
[10:40:49 EDT(-0400)] <Jacob2> or the caret specifically?
[10:41:20 EDT(-0400)] <Jacob2> ....never mind....
[10:41:22 EDT(-0400)] * Jacob2 (n=Fluid@142.150.154.106) has left #fluid-work
[10:41:38 EDT(-0400)] <anastasiac> by the way, Bosmon, the Firebug console does tend to mess up focus testing sometimes...
[10:42:59 EDT(-0400)] <anastasiac> iirc, the click causing the first blur on the edit causes focus to be places on the test - this is probably why it takes a second click to actually move focus away
[11:52:48 EDT(-0400)] * jessm (n=Jess@rrcs-70-63-141-143.midsouth.biz.rr.com) has joined #fluid-work
[12:30:19 EDT(-0400)] * Justin_o (n=Justin@142.150.154.235) has left #fluid-work
[12:48:54 EDT(-0400)] * Justin_o (n=Justin@142.150.154.235) has joined #fluid-work
[12:50:27 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined #fluid-work
[12:53:02 EDT(-0400)] * Jacob1 (n=Fluid@142.150.154.106) has joined #fluid-work
[12:53:08 EDT(-0400)] * Jacob1 (n=Fluid@142.150.154.106) has left #fluid-work
[12:59:47 EDT(-0400)] * Justin_o (n=Justin@142.150.154.235) has left #fluid-work
[13:04:51 EDT(-0400)] * colinclark (n=colin@142.150.154.101) has joined #fluid-work
[13:22:25 EDT(-0400)] * EricDalquist (n=EricDalq@dyn-0-111.uwnet.wisc.edu) has joined #fluid-work
[13:24:21 EDT(-0400)] <anastasiac> Bosmon, thanks for 'that'-ifying the Inline Edit! Just a head's up that the sample code and tests will have to be updated with any changes you make to the API.
[13:24:32 EDT(-0400)] <anastasiac> In particular, there are:
[13:25:01 EDT(-0400)] <anastasiac> the sample file: sample-code/inline-edit/announcements/announcements.html
[13:25:26 EDT(-0400)] <anastasiac> the unit tests: tests/fluid-tests/js/InlineEditTests.js
[13:25:45 EDT(-0400)] <anastasiac> and the manual test file: tests/fluid-tests/manual/inline-edit/InlineEdit.html
[13:26:07 EDT(-0400)] <anastasiac> Once you've committed, I'll make sure that the API documentation on the wiki is updated.
[13:31:24 EDT(-0400)] <Bosmon> Yes, thanks for that list
[13:35:04 EDT(-0400)] * colinclark (n=colin@142.150.154.101) has joined #fluid-work
[14:07:07 EDT(-0400)] <colinclark> clown: Off the top of your head, do you know where Dijit's keyboard accessibility code lives?
[15:11:27 EDT(-0400)] <colinclark> Bosmon: So what's the tooltip plugin you're looking at?
[15:11:41 EDT(-0400)] <Bosmon> OK, so we are talking about how best to achieve FLUID-779
[15:12:13 EDT(-0400)] <Bosmon> In theory we could roll our own, but perhaps it would be desirable to piggyback on some others' efforts in creating a generalised tooltip plugin
[15:12:18 EDT(-0400)] <Bosmon> http://bassistance.de/jquery-plugins/jquery-plugin-tooltip/
[15:12:21 EDT(-0400)] <Bosmon> I have found this one
[15:12:41 EDT(-0400)] <Bosmon> It looks generally competent, and handles some issues that would well have occured to us over time and had to deal with ourselves
[15:13:04 EDT(-0400)] <Bosmon> (20:10:42) Antranig: edge positioning, native tooltip suppression, and select z-order are 3 important ones
[15:13:21 EDT(-0400)] <colinclark> Bosmon: I'll take a look.
[15:13:24 EDT(-0400)] <colinclark> Sounds cool.
[15:13:34 EDT(-0400)] <Bosmon> The question is, if we "adopted" it, how might we integrate our development lifecycle with theirs
[15:13:54 EDT(-0400)] <anastasiac> theclown: colinclark asked a question above about dijit keyboard a11y, but used the wrong nick...
[15:14:22 EDT(-0400)] <colinclark> oops
[15:15:01 EDT(-0400)] <Bosmon> The comments stretch up till February 2008, so it is moderately if not "actively" maintained
[15:15:11 EDT(-0400)] <Bosmon> Would we want to just copy/paste this into the Fluid repo?
[15:15:35 EDT(-0400)] <Bosmon> It depends on two further plugins, dimensions (hard) and bgiframe (soft)
[15:15:54 EDT(-0400)] <colinclark> Bosmon: dimensions is now baked into jquery ui core
[15:15:59 EDT(-0400)] <Bosmon> oh, ok cool
[15:16:01 EDT(-0400)] <colinclark> So we already satisfy that dependency.
[15:16:10 EDT(-0400)] <colinclark> Assuming it's upward compatible with jQuery UI 1.5 (smile)
[15:16:33 EDT(-0400)] <colinclark> We have a directory structure for our third-party depedencies. You can troll through the repository and see where we're putting things.
[15:17:48 EDT(-0400)] <Bosmon> oh, ok
[15:18:06 EDT(-0400)] <Bosmon> It does actually seem to be fairly well attested, since it is copied directly into main jQuery plugins:
[15:18:06 EDT(-0400)] <Bosmon> http://docs.jquery.com/Plugins/Tooltip
[15:20:29 EDT(-0400)] <Bosmon> colinclark - not quite sure - wouldn't it go into fluid-components itself?
[15:20:42 EDT(-0400)] <Bosmon> I assume it would go alongside something like "json"
[15:20:58 EDT(-0400)] <colinclark> Bosmon: Yes, exactly.
[15:20:58 EDT(-0400)] <Bosmon> Or would it not even go into "jquery" itself?
[15:21:05 EDT(-0400)] <colinclark> in jquery makes sense
[15:21:14 EDT(-0400)] <Bosmon> ok then
[15:21:45 EDT(-0400)] <Bosmon> Well, maybe not, it does have a whole "package" of its own
[15:21:46 EDT(-0400)] <colinclark> Bosmon: This appears to give us a lot for free.
[15:22:04 EDT(-0400)] <Bosmon> Perhaps the most sensible placee would just be its own package of "jquery-tooltip"
[15:22:13 EDT(-0400)] <colinclark> Sure.
[15:22:14 EDT(-0400)] <Bosmon> Yes, and though it is a little fascistic, it is not enormously so
[15:22:28 EDT(-0400)] <Bosmon> I mean, someone has to create the markup for the physical tooltip from somewhere...
[15:22:42 EDT(-0400)] <Bosmon> Yes, last release is from Feb 2008
[15:22:45 EDT(-0400)] <colinclark> We'll want to use the unminified version, presumably.
[15:22:54 EDT(-0400)] <Bosmon> I will test it out and see if it is compatible with all our current stuff
[15:23:13 EDT(-0400)] <Bosmon> Yes, but I guess it would be nice to have its demo package and everything available for local reference?
[15:24:20 EDT(-0400)] <colinclark> Bosmon: Do you think?
[15:24:26 EDT(-0400)] <colinclark> I tend to favour just the barest dependencies.
[15:24:26 EDT(-0400)] <Bosmon> I dunno
[15:24:43 EDT(-0400)] <Bosmon> You think I should just take its one active file, and then simply the "latest" version of bgiframe or whatever?
[15:24:56 EDT(-0400)] <colinclark> I'll have to check with anastasiac, but if I remember correctly, any JavaScript dependencies do currently have to be added to the Ant build script.
[15:25:08 EDT(-0400)] <colinclark> Since order matters when building, the list is hardcoded at the moment.
[15:25:16 EDT(-0400)] <Bosmon> To be sure
[15:25:44 EDT(-0400)] <Bosmon> It's just an issue whether we want its demo "usage" stuff to be available in our own repo, or just ask people to follow the links that are at the head of the file
[15:25:52 EDT(-0400)] <anastasiac> regarding dependencies: If the file is in the jQuery folder, it will automatically be included by the build script
[15:25:55 EDT(-0400)] <colinclark> Bosmon: That seems reasonable. Does bgiframe have any depedencies?
[15:26:04 EDT(-0400)] <anastasiac> it doesn't have to be explicitly named in the script
[15:26:08 EDT(-0400)] <colinclark> asnastasiac: But doesn't the order have to be hardcoded?
[15:26:08 EDT(-0400)] <Bosmon> bgiframe looks like the end of its line
[15:26:28 EDT(-0400)] <colinclark> anastasiac: How would it know to concat jQuery.js before, say, something that depends on it?
[15:26:42 EDT(-0400)] * colinclark should know this, since he wrote the build scripts.
[15:26:54 EDT(-0400)] <colinclark> To the code!
[15:26:54 EDT(-0400)] <Bosmon> bgiframe has not itself been updated since July 2007
[15:26:54 EDT(-0400)] <theclown> colinclark, missed your question for various reasons. "clown: Off the top of your head, do you know where Dijit's keyboard accessibility code lives?" Not sure I even know where it is off the bottom of my head. Then again, I'm not sure what is meant by "Dijit's keyboard accessibiltiy code".
[15:26:58 EDT(-0400)] <anastasiac> I'm not sure, but I'm looking at the scripts, and I can't see anything that specifies an order
[15:27:03 EDT(-0400)] <Bosmon> So there is no danger of mismatches...
[15:27:04 EDT(-0400)] * anastasiac keeps looking
[15:27:22 EDT(-0400)] <colinclark> theclown: The code that Simon and others wrote to support Enter/Space activation...
[15:27:32 EDT(-0400)] <anastasiac> aha! found it
[15:27:33 EDT(-0400)] <colinclark> I think it used to be called onDijitClick or something along those lines.
[15:27:48 EDT(-0400)] <anastasiac> there's a property in the build.properties file that lists the files in the right order
[15:27:49 EDT(-0400)] <colinclark> anastasiac: Did you update that list when you upgraded to jQuery UI 1.5?
[15:27:55 EDT(-0400)] <colinclark> Cool.
[15:28:09 EDT(-0400)] * anastasiac slaps herself in the forehead
[15:28:26 EDT(-0400)] <colinclark> anastasiac: Don't sweat it. I never liked that requirement.
[15:28:38 EDT(-0400)] <colinclark> But I haven't been able to figure out any other strategy.
[15:28:39 EDT(-0400)] <anastasiac> well, I'm fixing it now
[15:28:46 EDT(-0400)] <colinclark> wicked
[15:28:56 EDT(-0400)] <theclown> colinclark: don't recall that explicitly, although 'dijiclick' rings a bell. There is a plugin for container keyboard navigation, but that's not space/enter.
[15:29:18 EDT(-0400)] <theclown> s/plugin/mixin/
[15:29:30 EDT(-0400)] <colinclark> theclown: Hmm. Okay, thanks. I know I've seen the code, but it was a long time ago. I always find digit's source code hard to navigate through.
[15:29:49 EDT(-0400)] <theclown> not doubt dijit too.
[15:29:51 EDT(-0400)] <colinclark> I will have to do some trolling around.
[15:29:57 EDT(-0400)] <theclown> sorry
[15:29:59 EDT(-0400)] <colinclark> yes, sorry
[15:30:00 EDT(-0400)] <colinclark> dijit
[15:30:03 EDT(-0400)] <colinclark> That thingy
[15:30:10 EDT(-0400)] <theclown> without finger nails
[15:30:17 EDT(-0400)] * colinclark laughs.
[15:30:59 EDT(-0400)] <Bosmon> So, any votes for putting in just jquery.tooltip.js into SVN or its entire distro package?
[15:34:08 EDT(-0400)] <Bosmon> anastasiac: I think the clock is set incorrectly on the box serving JIRA....
[15:34:20 EDT(-0400)] <Bosmon> Looks like somehow it is 2 hours in the future
[15:34:56 EDT(-0400)] <anastasiac> yes, I think that's a known issue. It also affects the wiki timestamps.
[15:35:02 EDT(-0400)] <colinclark> Evil.
[15:35:16 EDT(-0400)] <anastasiac> I've mentioned it to our sysadmin - I'll ping him again next week, when he's back from vacation
[15:36:25 EDT(-0400)] <colinclark> Bosmon: I vote for just the' file and its dependencies.
[15:36:36 EDT(-0400)] <colinclark> michelled: Do you have any preferences?
[15:36:55 EDT(-0400)] <michelled> I haven't been reading this - give me a moment
[15:37:01 EDT(-0400)] <colinclark> (tongue)
[15:43:05 EDT(-0400)] <michelled> colinclark: I'm happy with the file and its dependencies.
[15:43:11 EDT(-0400)] <colinclark> Cool
[15:44:07 EDT(-0400)] <michelled> colinclark: I think I remember why we used 'keydown' instead of 'keypress'. I think some browsers only throw 'keypress' for characters not other keys
[15:44:30 EDT(-0400)] * Justin_o (n=Justin@142.150.154.235) has joined #fluid-work
[15:44:32 EDT(-0400)] <colinclark> michelled: I seem to remember a distinction, in fact, for the spacebar
[15:44:36 EDT(-0400)] * michelled wishes she had a better memory
[16:03:45 EDT(-0400)] * Justin_o (n=Justin@142.150.154.235) has left #fluid-work
[16:07:39 EDT(-0400)] <Bosmon> OK
[16:07:58 EDT(-0400)] <Bosmon> A sort of global issue - should we not be namespacing our default CSS classes better?
[16:09:18 EDT(-0400)] * Bosmon (n=Antranig@ginger.caret.cam.ac.uk) has joined #fluid-work
[16:09:25 EDT(-0400)] <Bosmon> Let's try that again...
[16:09:33 EDT(-0400)] <Bosmon> styles: {
[16:09:34 EDT(-0400)] <Bosmon> invitation: "invitation",
[16:09:34 EDT(-0400)] <Bosmon> defaultViewText: "invitation-text",
[16:09:34 EDT(-0400)] <Bosmon> focus: "focus"
[16:09:51 EDT(-0400)] * Jacob1 (n=Fluid@142.150.154.106) has joined #fluid-work
[16:09:54 EDT(-0400)] * michelled (n=team@142.150.154.197) has left #fluid-work
[16:09:54 EDT(-0400)] <Bosmon> A default style of "focus" seems a bit minimal - should we not want to namespace these as inlineEdit-invitation, inlineEdit-focus etc.?
[16:11:37 EDT(-0400)] <Jacob1> I spoke with Colin about namespace css
[16:11:49 EDT(-0400)] <Jacob1> we probably should be, especially with such generic names
[16:12:58 EDT(-0400)] <Bosmon> OK
[16:13:00 EDT(-0400)] <Bosmon> I will start doing that
[16:13:27 EDT(-0400)] <Bosmon> Also, this "mixDefaults" seems a bit unsatisfying, from various angles
[16:13:47 EDT(-0400)] <Jacob1> i was talking with Colin about specificaly using a single option for prefixes, then just taking over from that point with our own conventions
[16:13:51 EDT(-0400)] <Bosmon> i) are we sure we really want to blend all of our options slap onto the instance itself? It seems we run the risk of worrying collisions
[16:15:04 EDT(-0400)] <Jacob1> do you mean as opposed to each instantiation having its own unique prefixes/options?
[16:15:21 EDT(-0400)] <Bosmon> ii) surely we would want an automated solution to this, along the lines of deepCopy?
[16:15:40 EDT(-0400)] <Bosmon> Well, I was just thinking of assigning the options structure one level lower down
[16:15:43 EDT(-0400)] <Jacob1> i've no experience with deepCopy -what is it?
[16:15:45 EDT(-0400)] <Bosmon> So that it would become stuck onto that.options
[16:15:58 EDT(-0400)] <Bosmon> Well, just anything that would deepCopy something (tongue)
[16:16:20 EDT(-0400)] <Bosmon> I don't seem to be able to lay my hands on one right now, but it would work along the lines of the deepEquiv thing that is in the renderer tests
[16:16:23 EDT(-0400)] <Bosmon> Only somewhat simpler...
[16:16:44 EDT(-0400)] <Jacob1> I think we're probably thinking along the same lines
[16:17:05 EDT(-0400)] <Bosmon> ok
[16:17:10 EDT(-0400)] <Bosmon> I will do that also then (tongue)
[16:17:27 EDT(-0400)] <Bosmon> Really, this is what comes of "trying to write a Fluid Component in 50 minutes" (smile)
[16:17:34 EDT(-0400)] <Jacob1> (smile)
[16:19:41 EDT(-0400)] <Bosmon> Ah
[16:19:48 EDT(-0400)] <Bosmon> It looks like jQuery.extend can already handle this
[16:19:54 EDT(-0400)] <Bosmon> So we can smash that function down into 1 line..
[16:31:28 EDT(-0400)] * Jacob1 (n=Fluid@142.150.154.106) has left #fluid-work
[17:00:13 EDT(-0400)] * apetro-_ (n=apetro@71-11.racc2000.com) has joined #fluid-work
[17:43:41 EDT(-0400)] * theclown (n=theclown@142.150.154.101) has left #fluid-work
[18:33:35 EDT(-0400)] * EricDalquist (n=EricDalq@adsl-76-208-70-140.dsl.mdsnwi.sbcglobal.net) has joined #fluid-work