fluid-work IRC Logs-2011-05-20
[08:25:30 CDT(-0500)] <heidi> Justin_o we should include ariaTabs for the linearization demo - jquery UI itself isn't aria <cindyli> however, u will run into other problems though, for instance, line 667: " .model.selections" <cindyli> replace with , heidi <cindyli> u actually should replace all s since it has been replaced by 3 sub-controls <heidi> cindyli what was before and what will changing it to do? <cindyli> as u've known, has been splitted into 3 little controls <heidi> cindyli let me know when your changes are up and i'll merge. but yeah the control panels are working now!
[08:25:54 CDT(-0500)] <heidi> Justin_o https://github.com/fnagel/jQuery-Accessible-RIA/tree/master/Tabs
[08:39:39 CDT(-0500)] <Justin_o> heidi: that's too bad..
[08:39:54 CDT(-0500)] <Justin_o> this may be a good opportunity for us to write a wrapper that adds in aria
[08:40:13 CDT(-0500)] <heidi> Justin_o the link i sent is the aria helper... i don't have it working yet
[08:41:45 CDT(-0500)] <Justin_o> heidi: yes.. we might want to be cautious about using 3rd party plugins though.. we've been burned by them in the past
[08:42:24 CDT(-0500)] <heidi> Justin_o i think a lot of felix's work has been brought into jquery ui itself (not sure) - i think it's a reliable source, but i hear ya
[08:43:04 CDT(-0500)] <Justin_o> heidi: okay.. i guess just make sure to check it out first
[08:43:05 CDT(-0500)] <heidi> Justin_o he's also the one that worked on the rich dropdown we were considering using as well
[08:43:16 CDT(-0500)] <Justin_o> heidi: i see, okay
[08:46:45 CDT(-0500)] <heidi> Justin_o i'm still fuzzy on how our wrapper will work, so maybe you can help me out with that in a bit
[08:46:45 CDT(-0500)] <heidi> ?
[08:47:29 CDT(-0500)] <Justin_o> heidi: sure..let me know when you want to chat
[08:47:47 CDT(-0500)] <Justin_o> colinclark: just sent a long winded reply to your e-mail about the graph flipping stuff
[08:47:54 CDT(-0500)] <colinclark> Justin_o: oh, great
[08:48:12 CDT(-0500)] <colinclark> I hope it didn't seem like I was being critical
[08:48:18 CDT(-0500)] <colinclark> you've been a total Git wizard
[08:48:27 CDT(-0500)] <colinclark> I think things were just starting to dawn on me
[08:48:56 CDT(-0500)] <Justin_o> colinclark: no i understand..
[08:49:19 CDT(-0500)] <Justin_o> i actually think that this type of thing should probably only be done for extreme cases anyways.. like what happened last friday..
[08:49:31 CDT(-0500)] <Justin_o> the one that happened the other day we could probably have lived with
[08:50:35 CDT(-0500)] <colinclark> I'll read your email
[08:50:42 CDT(-0500)] <colinclark> and maybe, ultimately, michelled will have some thoughts
[08:51:01 CDT(-0500)] <colinclark> but my gut feeling is that this is us trying to impose stability on a fleeting and illusory concept
[08:51:04 CDT(-0500)] <Justin_o> colinclark: yes.. i'm wondering what michelled will have to say..
[08:51:04 CDT(-0500)] <colinclark> right, fish?
[09:00:38 CDT(-0500)] <michelled> I agree with Justin_o that we should leave minor flips as they are
[09:00:51 CDT(-0500)] <michelled> I think we should try to get into the habit of using the --no-ff and --log flags
[09:01:01 CDT(-0500)] <michelled> and if we have a major flip we should discuss it
[09:01:31 CDT(-0500)] <michelled> I think the black line is valuable as a documentation and communication tool
[09:01:45 CDT(-0500)] <michelled> so keeping it showing the stable points in history would be wonderful
[09:02:04 CDT(-0500)] <michelled> but we should be so strict about it that we cause ourselves a lot of hassle
[09:02:29 CDT(-0500)] <michelled> shouldn't that is
[09:05:42 CDT(-0500)] <Justin_o> michelled: i would agree with all that
[09:09:02 CDT(-0500)] <colinclark> Yeah, I think the point is that we're saying "--no-ff is a standard community coding practice"
[09:09:11 CDT(-0500)] <colinclark> Not unlike, say, JSLint
[09:09:13 CDT(-0500)] <colinclark> or unit tests
[09:09:27 CDT(-0500)] <colinclark> The key difference, I think, is the cost of addressing a failure
[09:09:35 CDT(-0500)] <colinclark> meaning, some idiot like me forgets to include --no-ff
[09:09:44 CDT(-0500)] <colinclark> In the case of JSLint, it's pretty cheap
[09:09:55 CDT(-0500)] <colinclark> Either back out the change or commit on top with the lintification
[09:10:04 CDT(-0500)] <colinclark> In the case of --no-ff, the fix is pretty costly
[09:10:22 CDT(-0500)] <colinclark> As we've said, it's only suitable for really massive situations
[09:10:41 CDT(-0500)] <colinclark> So, that suggests to me that we have to be okay with the graph flipping periodically
[09:10:49 CDT(-0500)] <colinclark> and to some unknown extent
[09:11:03 CDT(-0500)] <colinclark> I don't know if you have thoughts, Justin_o, on how much flippage is too much flippage?
[09:13:10 CDT(-0500)] <Justin_o> colinclark: my initial thoughts would be 1) if puts a branch that we really don't want to appear as the black line up... for example one that is full of experimental commits or odd commit messages or something like that... 2) if the history gets put into a state where it is difficult to discern
[09:13:30 CDT(-0500)] <Justin_o> colinclark: also maybe if the branch extends back more than a few commits/days
[10:33:26 CDT(-0500)] <jessm> standup everyone?
[10:52:36 CDT(-0500)] <heidi> hey, sorry if my standup was funky. lag issues. i'm testing aria tabs plugin/will work on wrapper w/ justin, going to include cindy's UIO control templates into fat panel, and check out anastasia's new themes
[12:08:59 CDT(-0500)] <jessm> Justin_o: ping
[12:19:54 CDT(-0500)] <Justin_o> jessm: sorry.. .i'm here now
[12:21:03 CDT(-0500)] <jessm> no worries!
[12:21:06 CDT(-0500)] <jessm> and nevermind
[12:21:09 CDT(-0500)] <jessm> lol
[12:21:32 CDT(-0500)] <Justin_o> jessm: okay..
[12:37:51 CDT(-0500)] <heidi> cindyli i've added in the divs for the control templates and added the demands blocks for the templates in the demo scripts. were there extra js files i need to include in the header as well?
[12:38:16 CDT(-0500)] <cindyli> cindyli: no
[12:38:49 CDT(-0500)] <cindyli> what the ..., heidi: no
[12:39:11 CDT(-0500)] <heidi> okay. was there more i need to do?
[12:39:13 CDT(-0500)] <cindyli> heidi: problem? why do u ask?
[12:39:35 CDT(-0500)] <heidi> not showing up in my demo yet. i put in the divs... and added the demand blocks with url to templates
[12:39:56 CDT(-0500)] <cindyli> heidi: do u have a branch that i can look at?
[12:40:21 CDT(-0500)] <heidi> cindyli do those two things sound like enough? yep, i can update my branch, but is that right?
[12:40:32 CDT(-0500)] <cindyli> heidi: sounds right
[12:40:47 CDT(-0500)] <cindyli> heidi: 1. js -> add demands block
[12:42:06 CDT(-0500)] <cindyli> 2. modify say "components/uiOptions/html/FullPageNoPreview.html", replace the controls divs with 3 placeholder divs
[12:42:38 CDT(-0500)] <cindyli> that should be it, heidi
[12:45:56 CDT(-0500)] <heidi> cindyli hm, i'm missing something then. i'll push up to github, one sec
[12:47:32 CDT(-0500)] <heidi> cindyli https://github.com/heidiv/infusion/tree/FLUID-4230 – run the fat panel demo in /demos/uiOptions
[12:47:45 CDT(-0500)] <cindyli> ok
[12:57:20 CDT(-0500)] <heidi> thoughts cindyli?
[12:57:58 CDT(-0500)] <cindyli> heidi: same problem. the template looks good though. tracing
[12:58:09 CDT(-0500)] <heidi> thanks!
[12:58:18 CDT(-0500)] <cindyli> np
[13:03:19 CDT(-0500)] <cindyli> heidi: found the issue. if u look at components/uiOptions/js/UIOptions.js, line 323
[13:03:23 CDT(-0500)] <cindyli> var form = fluid.findForm(saveButton);
[13:03:31 CDT(-0500)] <cindyli> saveButton is no longer there
[13:03:45 CDT(-0500)] <heidi> cindyli yeah that error has been there awhile. i'm not sure it's related to this tho?
[13:04:06 CDT(-0500)] <heidi> it's an issue have to fix soon... there's a jira for auto-saving when controls change (no save button)
[13:04:19 CDT(-0500)] <cindyli> yes
[13:04:34 CDT(-0500)] <heidi> but i think this should work despite that error
[13:04:41 CDT(-0500)] <heidi> anything else look wrong?
[13:05:09 CDT(-0500)] <cindyli> in the meantime, if u wanna try out the fat panel, commented out the line where bindHandlers() gets called, heidi
[13:05:28 CDT(-0500)] <heidi> in UIOptions.js ?
[13:07:47 CDT(-0500)] <cindyli> heidi: yes
[13:08:48 CDT(-0500)] <heidi> cindyli that did it - thanks!
[13:09:17 CDT(-0500)]
[13:09:52 CDT(-0500)]
[13:10:26 CDT(-0500)]
[13:12:14 CDT(-0500)]
[13:13:06 CDT(-0500)]
[13:13:19 CDT(-0500)] <heidi> mmhm
[13:14:48 CDT(-0500)] <cindyli> heidi: so in order to collaborate these 3 contols, now it's their parent's change applier's job
[13:15:27 CDT(-0500)] <cindyli> same happens to model
[13:15:51 CDT(-0500)] <heidi> cindyli should this be changed in your branch instead?
[13:16:26 CDT(-0500)] <cindyli> ha, that makes sense since i didn't pay attention to livePreview, that you created
[13:16:48 CDT(-0500)] <cindyli> ok, heidi, i will change in my branch and u do a merge?
[13:17:19 CDT(-0500)] <heidi> cindyli sounds good! yeah the livepreview is from Justin_o - a lot of merges happening
[13:17:54 CDT(-0500)] <cindyli> ah ha. good we figured out
[13:23:05 CDT(-0500)] <yura_> Bosmon2: colinclark, Justin_o, harriswong, mlam, I noticed fluid.value fails if i pass empty jQuery object as nodeIn where jQuery val was just returning undefined. Is that intentional ?
[13:23:50 CDT(-0500)] <yura_> michelled: ^
[13:23:51 CDT(-0500)] <heidi> thanks cindyli the UIO work yr doing is awesome
[13:24:07 CDT(-0500)] <cindyli> super!!
[13:26:25 CDT(-0500)] <cindyli> heidi: oh, once i see thanks, i thought you get it working. is it ok now? UIO has lots of ur work. awesome to all us.
[13:27:01 CDT(-0500)]
[13:27:18 CDT(-0500)] <cindyli> heidi: sure. soon
[13:30:26 CDT(-0500)] <michelled> yura_: I don't know why we don't check for that - Bosmon2, is there a reason?
[13:32:03 CDT(-0500)] <colinclark> How does it fail, yura_?
[13:32:29 CDT(-0500)] <yura_> when trying to evaluate this : "input" !== node.nodeName.toLowerCase()
[13:32:35 CDT(-0500)] <yura_> nodeName is not there
[13:32:39 CDT(-0500)] <yura_> colinclark: ^
[13:32:41 CDT(-0500)] <colinclark> right
[13:32:50 CDT(-0500)] <colinclark> it needs to be unwrapped if it's a jquery
[13:32:58 CDT(-0500)] <colinclark> fish is right, it's probably a boz question
[13:33:47 CDT(-0500)] <yura_> well, there is unwrap on first line of fluid.value but it's still a jquery after that
[13:34:03 CDT(-0500)] <yura_> so maybe it's fluid.unwrap issue
[13:34:11 CDT(-0500)] <colinclark> hmm
[13:34:14 CDT(-0500)] <colinclark> seems unlikely
[13:34:16 CDT(-0500)] <colinclark> but who knows?
[13:34:37 CDT(-0500)] <yura_> well i mean i look at unwrapped and wrapped nodes and they look the same
[13:36:36 CDT(-0500)] <cindyli> heidi: it's in. sorry, i overlooked.
[13:38:00 CDT(-0500)] <heidi> cindyli cool no probs
[13:43:39 CDT(-0500)] <heidi> cindyli one thing about the templates: they shouldn't include the headers, the <h2>s
[13:44:00 CDT(-0500)] <cindyli> heidi: ok
[13:44:27 CDT(-0500)] <cindyli> which branch should we fix it, heidi
[13:44:37 CDT(-0500)] <heidi> cindyli in the control template branch
[13:44:46 CDT(-0500)] <heidi> (yours!)
[13:44:58 CDT(-0500)] <heidi> those <h2>s will have to go back into the fullpreview template
[13:46:13 CDT(-0500)] <cindyli> well, 4210 then
[13:46:38 CDT(-0500)] <cindyli> whoa, that's the branch that has been merged. michelled, is it ok that i put in a bit more work into 4210 and merge again?
[13:47:14 CDT(-0500)] <cindyli> heidi: michelled: i should say 4210 has been merged into project repo
[13:47:40 CDT(-0500)] <michelled> cindyli: you may want to recreate the branch given all the flipping that's been going on
[13:47:42 CDT(-0500)] <heidi> cindyli i thought 4216 was control templates?
[13:47:54 CDT(-0500)] <cindyli> yes, heidi
[13:48:00 CDT(-0500)] <michelled> cindyli: then you can make your changes and issue another pull request
[13:48:05 CDT(-0500)] <heidi> oh sorry, misread
[13:48:05 CDT(-0500)] <heidi>
[13:48:31 CDT(-0500)] <cindyli> michelled: ic. thx.
[13:59:10 CDT(-0500)] <cindyli> michelled: heidi, a pull request for 4210 has been sent
[14:00:02 CDT(-0500)] <heidi> ah, great
[14:09:08 CDT(-0500)] <heidi> cindyli did you take the <h2>s out for pull request?
[14:09:40 CDT(-0500)] <cindyli> heidi: yes
[14:09:43 CDT(-0500)] <heidi> great
[14:10:41 CDT(-0500)] <cindyli> heidi: is 4230 that u r working on base on my branch 4216?
[14:11:21 CDT(-0500)] <heidi> cindyli i've merged 4216 into my 4230 branch
[14:11:43 CDT(-0500)] <heidi> do i also need 4210?
[14:12:06 CDT(-0500)] <cindyli> great, heidi. the reason i ask is 4216 is pretty much done, but i cannot send a pull request since fatPanel and fullPageNoPreview are not working and u will take on. going to explain this on 4216 jira. yes, u do need 4210
[14:12:53 CDT(-0500)] <cindyli> 4210 had been merged into project repo, the only new thing there is moving <h2>s out of controls templates
[14:13:33 CDT(-0500)] <heidi> ok
[14:16:29 CDT(-0500)] <heidi> cindyli i tried to merge in 4210 and there are conflicts. yours/head look very different not sure what's happening
[14:16:43 CDT(-0500)] <heidi> for each of the controls templates
[14:17:18 CDT(-0500)] <heidi> so many merges tho, i could just be confused/mucked up
[14:17:30 CDT(-0500)] <cindyli> heidi: ya, reasonable since lots of changes in those templates.
[14:18:00 CDT(-0500)] <cindyli> heidi: let me try to merge 4210 into my 4216. 4216 is based on 4210 anyway, then u only need to merge 4216
[14:21:12 CDT(-0500)] <heidi> okay great
[14:24:35 CDT(-0500)] <cindyli> heidi: done. go ahead to merge 4216
[14:24:40 CDT(-0500)] <heidi> k!
[14:26:08 CDT(-0500)] <heidi> cindyli perfect thanks again
[14:26:30 CDT(-0500)] <cindyli> anytime
[14:43:03 CDT(-0500)] <Bosmon2> yura_ - what value() do you expect from an empty jQuery collection?
[14:43:17 CDT(-0500)] <yura_> undefined
[14:43:39 CDT(-0500)] <Bosmon2> I see
[14:43:46 CDT(-0500)] <Bosmon2> What value() do you expect from a jQuery collection with two elements?
[14:43:54 CDT(-0500)] <yura_> same as val
[14:43:57 CDT(-0500)] <yura_> the value of the first one
[14:44:02 CDT(-0500)] <Bosmon2> awful
[14:45:47 CDT(-0500)] <Bosmon2> This semantic suggests to me that the case of both 0 or more than 1 elements to this function have to be counted as a coding error...
[14:46:01 CDT(-0500)] <Bosmon2> How does it happen that the control that you want to take the value of doesn't exist?
[14:47:11 CDT(-0500)] <heidi> Justin_o the livepreview seems to apply to UIO itself in the fat panel (makes for a fun time, but just want the content below it)
[14:47:44 CDT(-0500)] <heidi> i think it applies to UIO itself once the panel is closed and reopened - is that right jameswy ?
[14:48:28 CDT(-0500)] <jameswy> heidi: That's correct. In the fat panel version, UIO applies to UIO on the next reopening.
[14:58:14 CDT(-0500)] <harriswong> Bosmon2: Hi, for FLUID-3953, it seems like IE9 is throwing an error with fluid.jById when an element contains both "id" and "rsf:id". Unit test added https://github.com/harriswong/infusion/tree/FLUID-3953.
[15:00:03 CDT(-0500)] <Bosmon2> harriswong - what error does it throw?
[15:00:12 CDT(-0500)] <harriswong> Bosmon2: Do you have any suggestions on how this should be fixed? This applies to normal jQuery calls ($('#boo)) too
[15:00:25 CDT(-0500)] <Bosmon2> harriswong - can you describe the problem in more detail?@
[15:01:22 CDT(-0500)] <harriswong> Bosmon2: I have attached 2 screenshots here: http://issues.fluidproject.org/browse/FLUID-3953, let me know if you need more details
[15:03:46 CDT(-0500)] <harriswong> Bosmon2: more details as follow:
[15:04:02 CDT(-0500)] <harriswong> Bosmon2: The error is thrown in Fluid.js line 1670.
[15:04:48 CDT(-0500)] <Bosmon2> harriswong - could you supply more details of the error?
[15:05:33 CDT(-0500)] <harriswong> Bosmon2: el.getAttribute("id") returns something different than id. On Firefox, if we do a el.attributes, we will see an array of 3 elements, which are 0:"class", 1:"id", 2:"rsf:id"; On IE9, el.attributes will give us 3 elements: 0:"class", 1:"rsf:id", 2:"rsf:id"
[15:06:25 CDT(-0500)] <harriswong> Bosmon2: on FF, it is correct to have "id" and "rsf:id" separated; in IE9, you can see that "id" got overwritten by "rsf:id".
[15:07:12 CDT(-0500)] <harriswong> Bosmon2: Because of that, line 1669: if (el.getAttribute("id") !== id) throws the error "Problem in document......"
[15:07:14 CDT(-0500)] <Bosmon2> harriswong - that assumes that there is some kind of relationship between the return of el.getAttribute and el.attributes
[15:07:17 CDT(-0500)] <Bosmon2> Which there may or may not be
[15:07:48 CDT(-0500)] <Bosmon2> But can you explain what happens in the case of the jQuery call ($('#boo)) above?
[15:07:52 CDT(-0500)] <Bosmon2> What failure is actually reported?
[15:11:35 CDT(-0500)] <harriswong> Bosmon2: For an element <div id="test" rsf:id="test-rsf"></div>, if i run jQuery("#test").attr('id') on IE9, it will give me "".
[15:11:57 CDT(-0500)] <harriswong> Bosmon2: If i run jQuery("#test").attr('rsf:id'), it will give me "test-rsf"
[15:12:01 CDT(-0500)] <Bosmon2> Ok, so it doesn't actually fail in that case - it just returns nothing
[15:12:34 CDT(-0500)] <Bosmon2> Amazing that IE comes from the "future" to plague us with yet more stupid bugs
[15:12:52 CDT(-0500)] <Bosmon2> Would you like to research into the matter and see if there is some discussion about it?
[15:13:16 CDT(-0500)] <harriswong> Bosmon2: on the IE9 console I just tested with, yes, it returns an empty string it seems. However, if i run the unit test, it gives me this "Died on test#n: Object doesn't support property or method 'fail'" ....."
[15:13:31 CDT(-0500)] <Bosmon2> harriswong - that is what the fluid.fail method always reports
[15:13:40 CDT(-0500)] <harriswong> Bosmon2: ah.
[15:13:42 CDT(-0500)] <Bosmon2> I can refer you to my earlier mail to the list on the subject of exceptions
[15:13:45 CDT(-0500)] <Bosmon2> Did you see that one?
[15:14:12 CDT(-0500)] <harriswong> Bosmon2: I might have but do not recall.... would you please give me the link to that if possible? thanks.
[15:14:33 CDT(-0500)] <harriswong> Bosmon2: Ok, i will look into it and see if there is discussion about this matter.
[15:15:10 CDT(-0500)] <Bosmon2> Thanks, harriswong
[15:15:14 CDT(-0500)] <Bosmon2> The mail is titled "Exception handling, "notrycatch", debugging in IE, and new diagnostics"
[15:16:09 CDT(-0500)] <harriswong> Bosmon2: Thanks!