fluid-work IRC Logs-2011-04-11
- Fluid IRC Log Bot
[08:57:19 CDT(-0500)] <heidi_> Justin_o are you fss'ing today <cindyli> optionnames: "$ ", <cindyli> optionlist: "$ ", <cindyli> selection: "$ "
[08:58:26 CDT(-0500)] <Justin_o> heidi_: hello
[08:59:03 CDT(-0500)] <Justin_o> i'm not sure how much FSS I'll get to today.. i did pull in your updates for FLUID-3880 and push it to my branch on github...
[08:59:23 CDT(-0500)] <Justin_o> I have to finish up the wiki page I was working on for the doc sprint, and get harriswong
[08:59:31 CDT(-0500)] <heidi_> cool. i think linear just needs testing right? and we're gonna ask johnny if he wants to help with that?
[08:59:32 CDT(-0500)] <Justin_o> 's pull request in
[08:59:40 CDT(-0500)] <Justin_o> heidi_: yep
[09:00:00 CDT(-0500)] <heidi_> Justin_o cool, i'm gonna work on a couple new issues then... the list item scrollbar and maybe fl-icon.
[09:00:04 CDT(-0500)] <Justin_o> i did mention it to him last week... on wednesday I'll be going over some git and github stuff with johnny too
[09:00:09 CDT(-0500)] <heidi_> cool
[09:00:14 CDT(-0500)] <Justin_o> heidi_: okay thanks
[09:00:54 CDT(-0500)] <colinclark> He mentioned that he managed to get linearization working in his WordPress theme, which is cool
[09:01:04 CDT(-0500)] <heidi_> oh great
[09:01:24 CDT(-0500)] <heidi_> yeah i think he'd provide good input on what we've done so far
[09:15:28 CDT(-0500)] <cindyli> Bosmon2: have time to answer some questions of "rendererComponent"?
[09:15:29 CDT(-0500)] <Justin_o> colinclark: should we all be adding our names here http://wiki.jqueryui.com/w/page/38817541/ARIA-Hackathon or is this meant more for the direct jquery ui folks?
[09:15:51 CDT(-0500)] <colinclark> cindyli: I imagine you're a few hours too early for Bosmon2 
[09:15:56 CDT(-0500)] <colinclark> What's your question?
[09:16:05 CDT(-0500)] <colinclark> Justin_o: Yes, go for it
[09:16:10 CDT(-0500)] <Justin_o> colinclark: thanks
[09:16:23 CDT(-0500)] <colinclark> Reminds me that I have a half-written announcement about the hackathon to send
[09:18:24 CDT(-0500)] <cindyli> colinclark: oh, ok. it's about rendering radio button item. my expander is like:
[09:18:24 CDT(-0500)] <cindyli> expander: {
[09:18:24 CDT(-0500)] <cindyli> type: "fluid.renderer.selection.inputs",
[09:18:24 CDT(-0500)] <cindyli> inputID: "inputIDbackgroundImages",
[09:18:24 CDT(-0500)] <cindyli> tree: {
[09:18:25 CDT(-0500)]
[09:18:25 CDT(-0500)]
[09:18:26 CDT(-0500)]
[09:18:26 CDT(-0500)] <cindyli> },
[09:18:27 CDT(-0500)] <cindyli> rowID: "rowIDbackgroundImages",
[09:18:27 CDT(-0500)] <cindyli> selectID: "backgroundImages", <--------------
[09:18:28 CDT(-0500)] <cindyli> labelID: "labelIDbackgroundImages"
[09:18:28 CDT(-0500)] <cindyli> }
[09:18:29 CDT(-0500)] <cindyli> I'm wondering what the "selectID" is. Does it need to be defined in template? as i know, doesn't need to, however, after rendering, i got error "Component UISelect with full ID backgroundImages could not be found within template at lump line 2 column 74 index 0 in file .", which seems it needs to be in template or somewhere?
[09:19:35 CDT(-0500)] <colinclark> Been awhile since I rendered a radio button... let me remind myself.
[09:22:00 CDT(-0500)] <cindyli> colinclark: i traced into "RendererUtility.js, line 115 before fluid.render() is called. the "tree" argument contains
[09:22:00 CDT(-0500)] <cindyli> tree.children[3] –
[09:22:00 CDT(-0500)] <cindyli> id: backgroundImages
[09:22:00 CDT(-0500)] <cindyli> componentType: UISelect
[09:22:00 CDT(-0500)] <cindyli> ...
[09:22:00 CDT(-0500)] <cindyli> seems the error is complaining about this element which corresponds to expander -> "selectID". so i wonder where i should get it defined
[09:25:14 CDT(-0500)] <colinclark> cindyli: So is it not just the mismatch between your two ids here?
[09:25:22 CDT(-0500)] <colinclark> inputID: "inputIDbackgroundImages",
[09:25:31 CDT(-0500)] <colinclark> selectID: "backgroundImages"
[09:25:31 CDT(-0500)] <colinclark> ?
[09:25:58 CDT(-0500)] <cindyli> colinclark: they need to be same?
[09:26:15 CDT(-0500)] <colinclark> The way radio buttons and checkboxes work in the Renderer is that there as a Select element in the tree that represents the user's selection.
[09:26:25 CDT(-0500)] <colinclark> It typically isn't bound to anything in the tree
[09:26:44 CDT(-0500)] <colinclark> and then you create a set of SelectChoice elements that represent each choice
[09:26:54 CDT(-0500)] <cindyli> yes
[09:26:56 CDT(-0500)] <colinclark> and the choices need to be associated with the Select element
[09:27:02 CDT(-0500)] <colinclark> So, I've never used this expander
[09:27:05 CDT(-0500)] <colinclark> I'm sure yura_ has
[09:27:14 CDT(-0500)] <colinclark> but the way I read it, without having looked at the code
[09:27:35 CDT(-0500)] <colinclark> The select.inputs expander will create both the Select and Select choices for you
[09:27:50 CDT(-0500)] <colinclark> But I'll have to look at the code to see what's going on
[09:27:53 CDT(-0500)] <colinclark> it doesn't quite make sense to me
[09:29:07 CDT(-0500)] <cindyli> i see the choices are expanded correctly on the tree.
[09:29:30 CDT(-0500)] <colinclark> cool
[09:30:03 CDT(-0500)] <cindyli> other than the choices, the other element on the tree is about tis UISelect. according to yura_, it might be the id to group the choices.
[09:30:13 CDT(-0500)] <cindyli> oh, yura_ had a look, also puzzled
[09:32:02 CDT(-0500)] <yura_> cindyli: yes I think selectID is the id to group you radio buttons by
[09:37:34 CDT(-0500)] <mlam> colinclark: anastasiac: about the argumentMap stuff from the doc sprint on Friday, I'm going to have to confirm the options value with Bosmon2. What we think is the wrong value for the options in the argumentMap is the only value that works. Then I can write the appropriate unit test for it in the uploader
[09:38:06 CDT(-0500)] <colinclark> odd
[09:38:13 CDT(-0500)] <anastasiac> yes, odd
[09:38:56 CDT(-0500)] <mlam> yes, the uploader does nothing when I changed the options value to the value we expect
[09:49:28 CDT(-0500)] <colinclark> cindyli: Is this code sitting in a repo somewhere I can take a look at?
[09:50:12 CDT(-0500)] <cindyli> colinclark: no. it's in a local test script.
[09:50:28 CDT(-0500)] <colinclark> Make a repo for it on github and push it up
[09:50:31 CDT(-0500)] <colinclark> I'm curious to take a look
[09:50:33 CDT(-0500)] <colinclark> if you've got the time
[09:51:00 CDT(-0500)] <cindyli> colinclark: i'm looking into another radio button renderer that yura created. i think i have some ideas. trying
[09:51:06 CDT(-0500)] <colinclark> ok
[09:52:06 CDT(-0500)] <cindyli> thx, colinclark, will let u know if i figure it out, or not
[09:52:20 CDT(-0500)] <colinclark> ok
[09:52:23 CDT(-0500)] <colinclark> have fun with it
[10:04:46 CDT(-0500)] <colinclark> anastasiac, harriswong: Since my goal is to catch up on some documentation sprinting today, I thought I'd start with a little bit of editing
[10:05:11 CDT(-0500)] <colinclark> harriswong: I tweaked your FAQ entry for fluid.demands(). Let me know if you think it's still in the spirit you intended: http://wiki.fluidproject.org/pages/viewpage.action?pageId=24936994
[10:07:17 CDT(-0500)] <colinclark> ok, off to the dentist
[10:13:19 CDT(-0500)] <Justin_o> anastasiac: i've updated this page http://wiki.fluidproject.org/display/docs/Event+injection+and+boiling
[10:13:28 CDT(-0500)] <anastasiac> Justin_o, thanks
[10:13:43 CDT(-0500)] <Justin_o> the event boiling section may be a bit light on text and description... feel free to suggest changes
[10:26:20 CDT(-0500)] <mlam> Bosmon2: are you free for a couple framework questions?
[10:49:36 CDT(-0500)] <mlam> harriswong: are the reorderer tests failing because of .val() as well?
[10:50:15 CDT(-0500)] <harriswong> mlam: so far, I don't think so. I could be wrong. I am experiencing events firing twice where it should be once.
[10:50:25 CDT(-0500)] <harriswong> mlam: this is only one of the problems though.
[10:50:41 CDT(-0500)] <harriswong> mlam: i haven't looked at the others yet.
[10:52:32 CDT(-0500)] <mlam> ah ok, was just curious
[11:05:50 CDT(-0500)] <jhung> anastasiac, jameswy: Do you guys have time to go over some of the results from tech doc user testing? Is 1p good?
[11:06:07 CDT(-0500)] <anastasiac> jhung, 1p works for me
[11:06:38 CDT(-0500)] <jameswy> jhung, anastasiac: 1p is kind of tight for me. Could we try for 3p?
[11:06:53 CDT(-0500)] <anastasiac> that works for me, too
[11:07:18 CDT(-0500)] <jhung> anastasiac, jameswy: 3p works. I'll ping y'all then.
[11:11:07 CDT(-0500)] <arun197> Hi all!
[11:12:33 CDT(-0500)] <colinclark> Hi arun197
[11:13:44 CDT(-0500)] <arun197> colinclark: Hello. Can I start working on some of the bugs with Video Player components? To get familiarity
[11:13:59 CDT(-0500)] <arun197> any specific links?
[11:14:40 CDT(-0500)] <colinclark> arun197: Sure
[11:14:41 CDT(-0500)] <colinclark> http://issues.fluidproject.org/secure/IssueNavigator.jspa?reset=true&jqlQuery=project+%3D+FLUID+AND+component+%3D+%22Video+Player%22+AND+resolution+%3D+Unresolved+ORDER+BY+priority+DESC%2C+key+DESC&mode=hide
[11:14:58 CDT(-0500)] <colinclark> There's a list of the currently-open JIRAs for the Video Player
[11:16:00 CDT(-0500)] <arun197> awesome. Thanks colinclark . I ll get on with it!
[11:17:34 CDT(-0500)] <arun197> colinclark: Will ping you as and when I get stuck.
[11:33:09 CDT(-0500)] <Justin_o_> colinclark, michelled_: did you either of you add any new unit tests for ui options.. specifically in regards to the iFrame
[11:33:27 CDT(-0500)] <colinclark> Justin_o_: New as in when?
[11:33:38 CDT(-0500)] <Justin_o_> colinclark: as in after 1.3.1 i guess
[11:36:50 CDT(-0500)] <colinclark> hmm
[11:36:59 CDT(-0500)] <colinclark> Yes, I did add one or two, Justin_o
[11:37:45 CDT(-0500)] <Justin_o> colinclark: ah okay... i think it doesn't work on IE when run from the local file system
[11:38:02 CDT(-0500)] <colinclark> But it does work in TestSwarm?
[11:38:05 CDT(-0500)] <colinclark> What's the error in IE?
[11:38:12 CDT(-0500)] <colinclark> and which versions are affected?
[11:38:17 CDT(-0500)] <Justin_o> SCRIPT5: Access is denied
[11:38:35 CDT(-0500)] <Justin_o> if it's run through a server it works fine..
[11:38:39 CDT(-0500)] <colinclark> fascinating
[11:38:40 CDT(-0500)] <colinclark> which test?
[11:38:42 CDT(-0500)] <Justin_o> it's like a cross domain issue
[11:38:51 CDT(-0500)] <Justin_o> the UIOptions unit test
[11:38:55 CDT(-0500)] <Justin_o> the page won't even finish loading
[11:39:18 CDT(-0500)] <Justin_o> colinclark: http://stackoverflow.com/questions/5087549/access-denied-to-jquery-script-on-ie
[11:39:58 CDT(-0500)] <Justin_o> http://forum.jquery.com/topic/cross-domain-ajax-and-ie
[11:48:25 CDT(-0500)] <jhung> ping athena.
[11:48:51 CDT(-0500)] <athena> hey
[11:49:06 CDT(-0500)] <jhung> did you have a chance to look at the design?
[11:49:11 CDT(-0500)] <athena> no 
[11:49:14 CDT(-0500)] <athena> really sorry
[11:49:19 CDT(-0500)] <athena> hoping to do that this afternoon
[11:49:24 CDT(-0500)] <jhung> don't sweat it.
[11:49:35 CDT(-0500)] <athena> definitely still meaning to get back to you on that though!
[11:50:21 CDT(-0500)] <jhung> No prob! Your feedback is always appreciated, so it's worth the wait IMO.
[13:40:05 CDT(-0500)] <jessm> hey y'all is jameswy around?
[13:40:09 CDT(-0500)] <jessm> is he busy?
[13:40:26 CDT(-0500)] <michelled_> jessm: Jutta asked him to provide some insight to the students in her class
[13:40:40 CDT(-0500)] <jessm> oh, ok, that explains that – thanks!
[13:40:42 CDT(-0500)] <michelled_> at least I think that's what I half heard
[14:00:42 CDT(-0500)] <jhung> jameswy, anastasiac: are either of you free to chat in 5mins?
[14:00:52 CDT(-0500)] <jameswy> jhung: 5 minutes sounds good
[14:00:57 CDT(-0500)] <jhung> cool
[14:01:02 CDT(-0500)] <anastasiac> jhung, yep. I'll get skype running
[15:19:19 CDT(-0500)] <heidi_> Justin_o i added http://issues.fluidproject.org/browse/FLUID-4184 but not sure 1.4 or not - thoughts?
[15:47:38 CDT(-0500)] <JonathanD> Howdy.
[15:48:12 CDT(-0500)] <jamon> hi JonathanD, colinclark will contact you shortly, he's on a call at the moment
[15:49:58 CDT(-0500)] <mlam> Bosmon2: free to chat a bit?
[15:51:30 CDT(-0500)] <colinclark> Hey JonathanD
[15:51:33 CDT(-0500)] <colinclark> How's it going?
[15:51:34 CDT(-0500)] <Bosmon2> Sure, mlam
[15:51:39 CDT(-0500)] <JonathanD> Howdy colinclark
[15:51:59 CDT(-0500)] <Bosmon2> So, yes, colinclark told me about the anomalous argumentMap for HTML5.remote
[15:52:04 CDT(-0500)] <Bosmon2> It certainly seems like some kind of bug
[15:52:21 CDT(-0500)] <Bosmon2> I guess I still can't test this without trying to run the live image gallery thing?
[15:52:44 CDT(-0500)] <mlam> Bosmon2: You can change the argument map in the local component as it has an argumentMap as well
[15:52:48 CDT(-0500)] <Bosmon2> What happens if you correct it to options: 1
[15:52:52 CDT(-0500)] <mlam> Nothing happens.
[15:53:09 CDT(-0500)] <Bosmon2> Oh, that's good
[15:53:10 CDT(-0500)] <mlam> I tried changing the value to 1 in the local component, and no files get added to the queue
[15:53:16 CDT(-0500)] <Bosmon2> Oh, I see
[15:53:19 CDT(-0500)] <Bosmon2> You don't mean nothing 
[15:53:27 CDT(-0500)] <Bosmon2> Yes, but the argumetnMap for the local component is correct
[15:53:30 CDT(-0500)] <mlam> haha, the other kind of nothing.
[15:53:33 CDT(-0500)] <Bosmon2> 2 is the correct position for options
[15:53:38 CDT(-0500)] <Bosmon2> It is only the remote one that is wrong
[15:53:49 CDT(-0500)] <Bosmon2> What happens if you correct the remote one to 1?
[15:53:56 CDT(-0500)] <mlam> let me check for you really quickly
[15:54:01 CDT(-0500)] <Bosmon2> cheers
[15:54:12 CDT(-0500)] <Bosmon2> Hopefully what happens in that case is the "good kind of nothing"
[15:54:48 CDT(-0500)] <jamon> colinclark: JonathanD has pointed out http://freenode.net/group_registration_form.php to get us registered as a group
[15:55:01 CDT(-0500)] <colinclark> jamon: Yep, I'm filling it out now
[15:55:05 CDT(-0500)] <jamon> k
[15:55:52 CDT(-0500)] <mlam> Bosmon2: the good nothing happened. Why is this the case?
[15:57:24 CDT(-0500)] <Bosmon2> mlam - it isn't essential to get the argumentMap right, depending on the material that is incoming in the options...
[15:57:34 CDT(-0500)] <Bosmon2> It is a much worse error to claim that an argument is the options when it is not
[15:57:41 CDT(-0500)] <Bosmon2> Since that will result it its being uninterpretable
[15:57:48 CDT(-0500)] <Bosmon2> if you make an error of the other kind, you may get away with it in some cases
[15:58:10 CDT(-0500)] <mlam> so it was by chance that the options didn't need any merging?
[15:58:12 CDT(-0500)] <Bosmon2> So in this case claiming that the nonexistent argument "2" is the options just means that the system thinks that no argument is the options
[15:58:16 CDT(-0500)] <Bosmon2> Yes, it is just by chance
[15:58:26 CDT(-0500)] <Bosmon2> There are some kinds of expansion that will not occur correctly if you get the map wrong
[15:58:35 CDT(-0500)] <mlam> I see I see.
[15:58:39 CDT(-0500)] <Bosmon2> And it looks like the layout was sufficiently simple and concrete in this case that they weren't required
[15:59:18 CDT(-0500)] <Bosmon2> However, if you remove the argumentMap entry completely for the remote, you will see it breaks completely
[15:59:38 CDT(-0500)] <Bosmon2> Since the "queue" argument becomes uninterpretable, since the system believes it is the options
[16:00:47 CDT(-0500)] <Bosmon2> Ok, that seems clear - Justin_o was mentioning you had some questions about the jQuery upgrade?
[16:01:07 CDT(-0500)] <mlam> Ah i see. Yes, thanks for clarifying
[16:01:12 CDT(-0500)] <mlam> yes, i had a question about fluid.value()
[16:01:19 CDT(-0500)] <mlam> I'm looking into the 1.5.2 jquery upgrade
[16:01:26 CDT(-0500)] <Bosmon2> Yes
[16:01:33 CDT(-0500)] <Bosmon2> I made an "opportunistic fix" to it in trunk a few weeks ago
[16:01:38 CDT(-0500)] <Bosmon2> But I didn't test it with the latest jQuery
[16:01:42 CDT(-0500)] <mlam> and i'm noticing that the fluid.value doesn't really do anything to shelter us from the negative effects of passing in an undefined value.
[16:01:54 CDT(-0500)] <Bosmon2> How so?
[16:02:32 CDT(-0500)] <mlam> There's an if statement in the function near the top, where it states, if it's not an input element or not a radio/checkbox , just set the val
[16:02:41 CDT(-0500)] <Bosmon2> On line 69 there is a branch which should be protecting us from this now... or did I misunderstand the issue?
[16:02:59 CDT(-0500)] <Bosmon2> Well, I guess you are working in jkkremer's branch which hasn't been updated for a very long while?
[16:03:14 CDT(-0500)] <Bosmon2> Do merge in the new implementation and try it out
[16:03:54 CDT(-0500)] <mlam> Ah. Ok, I'll give it another go.
[16:04:22 CDT(-0500)] <mlam> Bosmon2: and one more
[16:04:48 CDT(-0500)] <mlam> could you take a look at my FLUID-4163 branch tomorrow afternoon ? I made a pull request last week....they're the uploader integration tests
[16:05:02 CDT(-0500)] <Bosmon2> Sure, yes, I will ook at it this evening
[16:05:05 CDT(-0500)] <mlam> I want to make the argumentMap fix tomorrow and then write a proper unit test for it
[16:05:12 CDT(-0500)] <Bosmon2> That's great
[16:05:16 CDT(-0500)] <Bosmon2> How will you test the remote strategy?
[16:05:31 CDT(-0500)] <Bosmon2> I guess we are still planning to make proper XHR mocks, colinclark?
[16:05:41 CDT(-0500)] <mlam> I haven't tested the actual remote strategy yet.
[16:05:50 CDT(-0500)] <colinclark> Bosmon2: I think that is blocking on us, yes
[16:05:56 CDT(-0500)] <Bosmon2> Ok
[16:06:01 CDT(-0500)] <Bosmon2> I'll try and look at that this evening too
[16:06:16 CDT(-0500)] <Bosmon2> Is our plan to start by patching our local version of mockJax?
[16:06:27 CDT(-0500)] <Bosmon2> Assuming we decided that it is still the best framework to use
[16:06:43 CDT(-0500)] <mlam> Bosmon2: I've tested around the actual xhr calls to test the remote strategies...by ensuring proper events are fired, etc. But no actual xhr interaction
[16:07:04 CDT(-0500)] <Bosmon2> mlam - so how is the XHR itself mocked out?
[16:09:00 CDT(-0500)] <mlam> Tell me if it is bad practice, but for instance, to stop an upload, i re-wrote the stop function in the tests with the the code in the actual component minus the xhr call
[16:09:17 CDT(-0500)] <Bosmon2> Well sure, it is a bad practice
[16:09:25 CDT(-0500)] <Bosmon2> You have performed "testing through cut'n'paste"?
[16:09:32 CDT(-0500)] <colinclark> mlam: Can you show us an example
[16:09:32 CDT(-0500)] <colinclark> ?
[16:09:35 CDT(-0500)] <mlam> that way I can test that the proper events are triggered and that the uploader states are properly updated
[16:09:49 CDT(-0500)] <mlam> sure...let me maybe put some code in a pasty
[16:10:34 CDT(-0500)] <colinclark> or a link to Github?
[16:10:51 CDT(-0500)] <mlam> ok
[16:11:41 CDT(-0500)] <Bosmon2> I suggest the first thing to do here is to fix up the implementation that you use in your test cases, and supply a form of configuration to the Uploader that allows you to inject a mock XHR implementation
[16:11:48 CDT(-0500)] <Bosmon2> And produce a simple mock XHR that just does nothing
[16:12:16 CDT(-0500)] <mlam> https://github.com/mlam/infusion/blob/FLUID-4163/src/webapp/tests/component-tests/uploader/js/UploaderTests.js
[16:12:23 CDT(-0500)] <Bosmon2> cut'n paste code isn't a solution to anything
[16:12:47 CDT(-0500)] <mlam> line 301 is an example of what i'm talking about.
[16:15:09 CDT(-0500)] <mlam> Ok, I can remote the start/stop tests until the mock xhr implementation is in place
[16:15:14 CDT(-0500)] <mlam> **remove
[16:15:47 CDT(-0500)] <Bosmon2> Well, we don't need any good kind of mock in order for you to improve the tests
[16:16:14 CDT(-0500)] <Bosmon2> You can retain your start/stop tests together with a simple mock that just does a no-op
[16:19:59 CDT(-0500)] <mlam> Bosmon2: I guess let me clean up the tests to remove some code before you take a look at it
[16:21:42 CDT(-0500)] <Bosmon2> Which code will you remove?
[16:23:43 CDT(-0500)] <Bosmon2> Please remove all the cut and pasted material, but do retain the tests themselves
[16:24:11 CDT(-0500)] <mlam> just looking now....but looks like the stopUpload, triggerUpload functions
[16:24:29 CDT(-0500)] <colinclark> mlam: Bosmon2's got it right, here
[16:24:37 CDT(-0500)] <colinclark> You're not really testing the Uploader with these tests
[16:24:39 CDT(-0500)] <colinclark> you're testing the test
[16:24:43 CDT(-0500)] <colinclark> if you know what I mean
[16:25:37 CDT(-0500)] <mlam> I do. I was trying to test the events and uploader state of the remote without actually interacting with ajax
[16:26:43 CDT(-0500)] <colinclark> So, there's basically one line of code that is foiling an easy first step
[16:26:58 CDT(-0500)] <colinclark> I'm looking in the project repo, and it might be out of date...
[16:27:09 CDT(-0500)] <colinclark> but line 111 of HTML5UploaderSupport.js is it
[16:27:17 CDT(-0500)] <colinclark> Where we directly create an XHR
[16:28:14 CDT(-0500)] <mlam> Oh, I pulled that out into a separate function in my branch
[16:28:53 CDT(-0500)] <mlam> So the function just creates the XHR and returns it
[16:30:40 CDT(-0500)] <colinclark> mlam: Cool
[16:30:42 CDT(-0500)] <colinclark> So you should be all set
[16:30:47 CDT(-0500)] <colinclark> just create a little mocky XHR
[16:31:46 CDT(-0500)] <mlam> Bosmon2: so tomorrow, I'll clean up the tests, make a fix for the argumentMap, add a test for the argumentMap, and then I'll update John's branch tomorrow and test the outcome of the fluid.value() function
[16:36:32 CDT(-0500)] <colinclark> ok
[16:37:21 CDT(-0500)] <colinclark> Bosmon2: So, mlam can at least reproduce his current tests in a way that tests the implementation with a no-op XHR stub
[16:37:28 CDT(-0500)] <colinclark> That'll get him a little ways along
[16:37:51 CDT(-0500)] <colinclark> Then we'll need a real mock xhr soon
[16:38:30 CDT(-0500)] <colinclark> I guess the issue with mockjax vs. other is that we're not using jQuery's ajax() for the real implementation
[16:38:55 CDT(-0500)] <colinclark> so we'd have to also write a file upload-friendly fork or plugin for jQuery
[16:49:18 CDT(-0500)] <Bosmon2> Yes
[16:49:34 CDT(-0500)] <Bosmon2> Well, it seems to me we need to "fork" both projects, yes?
[16:49:35 CDT(-0500)] <colinclark> I don't really like the thought of having two mock XHRs lying around
[16:49:35 CDT(-0500)] <colinclark> I don't really relish the thought of extending jQuery, either
[16:49:35 CDT(-0500)] <colinclark> I mean, it's always been on a to do list
[16:49:38 CDT(-0500)] <colinclark> but not really one for 1.4
[16:49:47 CDT(-0500)] <colinclark> jQuery and MockJax
[16:49:47 CDT(-0500)] <colinclark> yes
[16:49:52 CDT(-0500)] <Bosmon2> ok
[16:50:00 CDT(-0500)] <Bosmon2> And hopefully encourage both of them to take our patches
[16:50:20 CDT(-0500)] <Bosmon2> As I remember from last week, we had divided it up that I would deal with mockJax and you with jQuery
[16:50:26 CDT(-0500)] <colinclark> Yes, that was the idea
[16:50:30 CDT(-0500)] <Bosmon2> ok, cool
[16:50:33 CDT(-0500)] <colinclark> I'm still cool with that
[16:50:48 CDT(-0500)] <colinclark> So, we use a few things that are only in XHR Level 2
[16:51:12 CDT(-0500)] <colinclark> 1. The upload.onprogress Event
[16:51:29 CDT(-0500)] <colinclark> Which seems to just fire periodically with progress information
[16:51:52 CDT(-0500)] <colinclark> Thinking about jQuery.ajax(), that's probably just another callback
[16:52:11 CDT(-0500)] <colinclark> Although everything in jQuery 1.5 is now a "promise" instead of a callback
[16:52:18 CDT(-0500)] <Bosmon2> So I saw....
[16:52:45 CDT(-0500)] <colinclark> 2. The ability to .send() a FormData object, instead of just plain old key/value pairs
[16:57:46 CDT(-0500)] <colinclark> 3. For Firefox 3.6, the sendAsBinary() method
[16:58:15 CDT(-0500)] <colinclark> That's all we'll need to contend with, as far as I know