fluid-work IRC Logs-2011-09-01
[11:08:06 CDT(-0500)] <cindyli> michelled: regarding 4370, there's another styling issue in IE6 in the absense of the new component, i will try to fix it
[11:08:37 CDT(-0500)] <michelled> cindyli: is it something that doesn't exist already in master?
[11:08:51 CDT(-0500)] <cindyli> michelled: it doesn't exist in master
[11:09:02 CDT(-0500)] <michelled> ok, let me know how it goes
[11:09:06 CDT(-0500)] <cindyli> sure
[13:03:01 CDT(-0500)] <Bosmon> I can see everyone is filtering back
[14:04:11 CDT(-0500)] <michelled> Bosmon: cindyli was just showing me her changes to the 4370 branch
[14:04:33 CDT(-0500)] <michelled> it seems that to get it working in ie6 we need to move the creation of the page enhancer into document ready
[14:05:07 CDT(-0500)] <michelled> I know that's not ideal but it seems like the only time when the styling changes we require will actually happen
[14:05:33 CDT(-0500)] <michelled> otherwise we end up with yellow and black sections when we are supposed to have the default theme
[14:06:12 CDT(-0500)] <michelled> Bosmon: any chance you can review and push the pull request when she makes it?
[14:12:59 CDT(-0500)] <michelled> cindyli: I closed justin_o's pull request for 4370 - please open a new one when you are ready
[14:13:11 CDT(-0500)] <cindyli> sure, michelled
[14:25:14 CDT(-0500)] <Bosmon> michelled - ok
[14:25:23 CDT(-0500)] <Bosmon> So currently it initialises too early on IE6?
[14:49:40 CDT(-0500)] <cindyli> yes, Bosmon, it's not only on ie6. it affects all browsers. moreover, the effect is beyond the styling issue in ie6. you may have noticed that sometimes, even tho the settings of "make inputs larger" and "emphasize links" are turned on, when you refresh the page, these settting effects on the main page are gone
[14:51:12 CDT(-0500)] <cindyli> it's because pageEnhancer looks inside <body> for <a>, <input> tags, but as the page hasn't been fully loaded when pageEnhancer gets initialized. These link and input elements are not enhanced
[14:51:46 CDT(-0500)] <cindyli> calling pageEnhancer at document.ready solves this problem
[14:52:34 CDT(-0500)] <Bosmon> I am trying to do a diff of your branch, but it has such a complex history that it is tricky....
[14:52:47 CDT(-0500)] <Bosmon> Do we know how to do a diff just along a particular lineage?
[14:55:02 CDT(-0500)] <Bosmon> Ok, michelled has helped me
[14:55:03 CDT(-0500)] <cindyli> Bosmon: i haven't pushed in the recent change. need probably another 10 mins
[14:55:19 CDT(-0500)] <Bosmon> It would be better if the ie6Inversion component were added via a demands block
[14:55:26 CDT(-0500)] <Bosmon> But I suspect the current framework wouldn't handle it
[14:55:32 CDT(-0500)] <cindyli> i've removed that component
[14:55:43 CDT(-0500)] <Bosmon> Oh, ok - what is your new strategy?
[14:55:44 CDT(-0500)] <cindyli> ie6Inversion, from my working copy
[14:56:04 CDT(-0500)] <cindyli> a function called in refreshView
[14:56:18 CDT(-0500)] <cindyli> same as setStyles()
[14:59:45 CDT(-0500)] <michelled> fyi to get to the compare view in github you need to click the 'switch branches' link and then the 'compare' button
[15:00:15 CDT(-0500)] <michelled> if you want to compare between repositories then you need to add the user name and a colon before the branch name
[15:00:20 CDT(-0500)] <michelled> for example:
[15:00:31 CDT(-0500)] <michelled> https://github.com/cindyli/infusion/compare/fluid-project:master...FLUID-4370
[15:00:49 CDT(-0500)] <michelled> that link compares cindyli's branch to the master of the project repo
[15:01:11 CDT(-0500)] <Bosmon> That's pretty remarkable
[15:23:06 CDT(-0500)] <cindyli> Bosmon, michelled, i've sent a pull request on 4370. code review please. thanks.
[15:26:06 CDT(-0500)] <michelled> thx cindyli
[15:26:14 CDT(-0500)] <cindyli> NP
[15:34:46 CDT(-0500)] <cindyli> michelled: just noticed the new push made onto project repo this afternoon. there are conflicts at merging into my branch. working on it
[15:35:04 CDT(-0500)] <michelled> thx cindyli - appreciated
[15:37:46 CDT(-0500)] <michelled> anastasiac: this is the mac git completion tool I mentioned: http://repo.or.cz/w/git.git/blob/HEAD:/contrib/completion/git-completion.bash
[15:38:05 CDT(-0500)] <anastasiac> cool, thanks michelled
[15:39:12 CDT(-0500)] <cindyli> michelled: solved and pushed
[15:39:24 CDT(-0500)] <michelled> thx cindyli did you retest?
[15:44:30 CDT(-0500)] <cindyli> yes, michelled, tested the page that i resolved the conflict on
[15:44:44 CDT(-0500)] <michelled> cool, thx
[15:44:48 CDT(-0500)] <cindyli> np
[15:56:28 CDT(-0500)] <Bosmon> hi cindyli - I have been looking over your patch
[15:56:54 CDT(-0500)] <Bosmon> The main issue with it is something that we have in Uploader too - you have some platform-specific behaviour, but the tests are all driven by the same self-same platform detection
[15:57:15 CDT(-0500)] <Bosmon> This means that the tests are themselves platform-specific, which isn't really desirable
[15:57:40 CDT(-0500)] <Bosmon> What I suggest is that you follow the same strategy we use with progressive enhancement with Uploader - which is to put a layer of indirection in the platform detection
[15:57:47 CDT(-0500)] <Bosmon> And place some tags in the static environment to represent them
[15:58:01 CDT(-0500)] <Bosmon> For example, "fluid.browser.ie" and "fluid.browser.version.6" or something like that
[15:58:23 CDT(-0500)] <Bosmon> Then if you write your tests in terms of these tags rather than jquery's own detection, you can actually test the platform-specific adaptation on all platforms
[15:58:37 CDT(-0500)] <Bosmon> Does that makes sense?
[15:59:08 CDT(-0500)] <Bosmon> Also, having the tags prepares some of the way for the "proper" version of dealing with these kinds of tasks in 1.5
[15:59:20 CDT(-0500)] <Bosmon> Where we will just issue demands blocks in the expected way.....
[15:59:30 CDT(-0500)] <Bosmon> sorry, by tags I mean "type tags"
[15:59:40 CDT(-0500)] <Bosmon> Of the kind you can see if you look in ProgressiveEnhancement.js
[16:05:10 CDT(-0500)] <cindyli> i see, Bosmon