fluid-work IRC Logs-2013-02-19

Owned by Fluid IRC Log Bot
Last updated: Feb 19, 2013Version comment
6 min read

[07:53:49 CST(-0600)] <colinclark> hey Justin_o

[07:55:00 CST(-0600)] <Justin_o> colinclark: hello

[07:55:22 CST(-0600)] <colinclark> Here's the DOM I see in Firebug for the Text Style control in UI Options

[07:55:38 CST(-0600)] <colinclark> http://pastie.org/6224547

[07:56:10 CST(-0600)] <colinclark> You can see that the label is referring to an id of "text-font-selection" but the select element actually has the id "textFont"

[07:56:21 CST(-0600)] <colinclark> it does have a name of "textFont-selection", though

[07:56:59 CST(-0600)] <Justin_o> colinclark: i thought it was enough for the name to match

[07:57:52 CST(-0600)] <colinclark> A tester on Cloud4all with JAWS 14 seemed to think it wasn't working

[07:57:54 CST(-0600)] <colinclark> let me check the spec

[07:58:46 CST(-0600)] <colinclark> Justin_o: http://www.w3.org/TR/html-markup/label.html

[07:58:57 CST(-0600)] <colinclark> and http://www.w3.org/TR/html-markup/datatypes.html#common.data.idref

[07:59:09 CST(-0600)] <colinclark> It looks, unfortunately, like it must be an exact match to the ID attribute

[08:03:20 CST(-0600)] <Justin_o> colinclark: okay.. i guess we'll have to fix that then

[08:03:29 CST(-0600)] <colinclark> yup

[08:03:48 CST(-0600)] <Justin_o> colinclark: i guess that means we'll need to put out a maintenance release for the 1.4.x line

[08:03:56 CST(-0600)] <colinclark> it's possible

[08:04:08 CST(-0600)] <colinclark> let's see what happens once we get a fix in place

[08:04:25 CST(-0600)] <colinclark> Can you coordinate that work with michelled and yzen this morning?

[08:04:37 CST(-0600)] <colinclark> I'm in meetings but will be available to help in about two hours

[08:06:31 CST(-0600)] <Justin_o> colinclark: sure

[08:06:34 CST(-0600)] <colinclark> thanks!

[09:25:42 CST(-0600)] <colinclark> yzen: We're going to have our work cut out for us supporting UI Options users over the next few weeks

[09:25:58 CST(-0600)] <colinclark> I'm going to try to write some more documentation on a) how to extend it now and b) where we're going with it

[09:26:28 CST(-0600)] <colinclark> But we'll need to describe how to support multiple languages sooner rather than later

[09:26:51 CST(-0600)] <yzen> colinclark: sounds good, perhaps, i can take a look at it

[09:26:55 CST(-0600)] <colinclark> thanks so much

[09:27:24 CST(-0600)] <colinclark> let's chat this afternoon in the office--you, me, Justin_o, and michelled, just to do a bit more coordination of our development tasks

[09:27:40 CST(-0600)] <colinclark> are you in today, yzen, or are you working from home?

[09:28:01 CST(-0600)] <yzen> i m home right now, ill head to the office after the call

[09:28:37 CST(-0600)] <colinclark> no rush; i am too (smile)

[12:15:14 CST(-0600)] <Bosmon> colinclark, Justin_o - that use case that we decried for ProgressiveEnhancement last week has already arrived (smile)

[12:15:28 CST(-0600)] <Bosmon> Specifically, the need to support "grouping" for selective forgetting of tags

[12:15:50 CST(-0600)] <Bosmon> Here are the new "Uploader Compatibility Tests"

[12:15:51 CST(-0600)] <Bosmon> https://github.com/amb26/infusion/blob/FLUID-4916/src/webapp/tests/component-tests/uploader/js/UploaderCompatibilityTests.js#L78-L93

[12:16:17 CST(-0600)] <Bosmon> In particular they have the need to "forget" tags related to ensuring version compatibility, whilst "remembering" tags related to context capabilities

[12:17:07 CST(-0600)] <Bosmon> The central test loop is on lines 139-142

[13:27:24 CST(-0600)] <michelled> heidiv: hi

[13:27:30 CST(-0600)] <heidiv> hey michelled

[13:27:50 CST(-0600)] <michelled> I was just looking at the VP pull requests and it looks ilke you have two outstanding ones

[13:28:03 CST(-0600)] <michelled> are they ready to be reviewed at this point?

[13:28:08 CST(-0600)] <michelled> or are you still working on them?

[13:28:17 CST(-0600)] <heidiv> michelled i just updated the -overlay one, to fix the caption layout

[13:28:26 CST(-0600)] <heidiv> please take a look.

[13:28:30 CST(-0600)] <michelled> ok

[13:28:35 CST(-0600)] <heidiv> going to update the theme one now, it will be quick

[13:28:47 CST(-0600)] <michelled> ok, sounds good

[13:28:53 CST(-0600)] <heidiv> thanks

[13:51:24 CST(-0600)] <Justin_o> Bosmon: sorry, missed your comments here

[13:51:30 CST(-0600)] <Justin_o> Bosmon: that's still a case only for tests

[13:51:32 CST(-0600)] <Justin_o> ?

[13:52:17 CST(-0600)] <Justin_o> Bosmon: is it not safe to just forget all and re-add?

[13:55:56 CST(-0600)] <Bosmon> Justin_o - of course it is safe, but it is not helpful (smile)

[13:56:04 CST(-0600)] <Justin_o> Bosmon: (smile)

[13:56:21 CST(-0600)] <Bosmon> Our original use case was also primarily for tests

[13:56:26 CST(-0600)] <Justin_o> Bosmon: also, in regards to the UIO issue around labels and id's for the drop downs that colin mentioned ealier

[13:56:28 CST(-0600)] <Justin_o> r

[13:57:19 CST(-0600)] <Justin_o> Bosmon: it seems that commenting out the conditional on line 384 - 386 of fluidRenderer.js fixes the issue, but we're not sure why it's there as all the tests still passed after we commented it out

[13:58:10 CST(-0600)] <Bosmon> The tests may pass, but this implementation is now at a risk of creating an id collision in the document

[13:58:12 CST(-0600)] <Justin_o> https://github.com/fluid-project/infusion/blob/master/src/webapp/framework/renderer/js/fluidRenderer.js#L384-L386

[13:58:30 CST(-0600)] <Bosmon> We probably already have a framework JIRA open against the renderer for this bug, and if we don't, we should (smile)

[13:58:30 CST(-0600)] <Justin_o> Bosmon: how does it prevent an ID collision

[13:58:52 CST(-0600)] <Bosmon> Because the id allocated to the "parent component" may already be in use

[13:59:17 CST(-0600)] <Bosmon> Or if it doesn't generate a collision, we should verify why it doesn't (smile)

[14:03:00 CST(-0600)] <Justin_o> Bosmon: the checkboxes in the uio work, although they may not be in the structure you are thinking of.. i tried the renderer demo. You can see the issue there http://build.fluidproject.org/infusion/demos/renderer/html/renderer.html

[14:03:16 CST(-0600)] <Justin_o> the radio buttons work, but the select doesn't. The checkboxes sadly do not appear to have labels

[14:03:18 CST(-0600)] <Bosmon> Justin_o - yes, we don't have enough test cases for this issue, which is a bug either way

[14:09:58 CST(-0600)] <Justin_o> yzen, Bosmon: looks like michelled had filed this one a while ago http://issues.fluidproject.org/browse/FLUID-4718

[14:10:56 CST(-0600)] <michelled> unfortunate that I have no memory of this (sad)

[14:11:18 CST(-0600)] <colinclark> quite

[14:11:20 CST(-0600)] <Justin_o> michelled: at least you filed a jira, that's a good thing

[14:11:24 CST(-0600)] <colinclark> do we have a way to fix it?

[14:13:10 CST(-0600)] <Justin_o> colinclark: not yet.. yzen spotted some code that you could comment out to get rid of it, but Bosmon pointed out above that this could have other repercussions. Also there aren't enough tests right now to verify all the paths through the renderer that might be affected by this.

[14:14:05 CST(-0600)] <colinclark> I'll catch up on the IRC logs to see what you've discussed

[14:16:22 CST(-0600)] <Justin_o> colinclark, yzen, michelled: when would you like to meet for our UIO talk?

[14:17:07 CST(-0600)] <yzen> colinclark: anytime works for me

[14:18:29 CST(-0600)] <colinclark> I need about 15 minutes or so

[14:18:31 CST(-0600)] <colinclark> and then I'm ready

[14:23:04 CST(-0600)] <yzen> so Bosmon in case that -section postfix is necessary, then I would assume we should have a but that rewrites select's id just like the label's id is overridden ?

[14:36:26 CST(-0600)] <colinclark> I'm ready now, yzen, Justin_o michelled and whoever else

[14:37:04 CST(-0600)] <yzen> sounds good

[14:37:11 CST(-0600)] <Justin_o> colinclark: ready

[14:37:17 CST(-0600)] <michelled> me too

[14:37:27 CST(-0600)] <colinclark> can someone start the call?

[14:37:39 CST(-0600)] <Justin_o> i probably shouldn't

[14:49:41 CST(-0600)] <colinclark> https://160.40.60.185:8443/cloud4allSettings/preferenceSet.jsf

[16:21:41 CST(-0600)] <Bosmon> yzen - there is indeed such a "bit" already

[16:21:59 CST(-0600)] <Bosmon> id rewriting is handled on line 1044

[16:23:10 CST(-0600)] <Bosmon> We even have some test cases for this, but it seems only for the radio button case

[16:23:28 CST(-0600)] <Bosmon> Around line 734 of RendererTests.js

[16:23:47 CST(-0600)] <Bosmon> Once we make the fix, we should extent the test case on line 686 to deal with it

[16:27:04 CST(-0600)] <yzen> Bosmon: when i was stepping through only labels were being rewritten

[16:27:12 CST(-0600)] <yzen> which is what we already know, but the id for the select never does

[16:27:24 CST(-0600)] <Bosmon> yzen - what do you mean

[16:27:25 CST(-0600)] <yzen> the only thing that changes it seems like is the selection.fullId change

[16:29:25 CST(-0600)] <Bosmon> Yes, that should be what happens

[16:30:09 CST(-0600)] <Bosmon> But this change should end up in the "rewritemap"...

[16:30:11 CST(-0600)] <yzen> Bosmon: right but don't we need an updated id for the select itself , not selectiion ?

[16:31:51 CST(-0600)] <yzen> Bosmon: do you have time to skype maybe ?

[16:31:55 CST(-0600)] <Bosmon> yes

[16:32:01 CST(-0600)] <yzen> now?

[16:32:06 CST(-0600)] <Bosmon> yes

[16:32:12 CST(-0600)] <yzen> k ill dial you

[16:40:39 CST(-0600)] <yzen> function getRewriteKey(template, parent, id) {

[16:40:39 CST(-0600)] <yzen> return template.resourceKey + parent.fullID + id;

[16:40:39 CST(-0600)] <yzen> }

[16:42:25 CST(-0600)] <yzen> function getRewriteKey(template, parent, id) {

[16:42:25 CST(-0600)] <yzen> var fullID = parent.componentType === "UISelect" ?

[16:42:25 CST(-0600)] <yzen> parent.selection.fullID : parent.fullID;

[16:42:25 CST(-0600)] <yzen> return template.resourceKey + fullID + id;

[16:42:25 CST(-0600)] <yzen> }