fluid-work IRC Logs-2011-04-26

[08:42:30 CDT(-0500)] <colinclark> Hey Justin_o
[08:43:00 CDT(-0500)] <Justin_o> colinclark: hellow
[08:43:03 CDT(-0500)] <Justin_o> hello
[08:43:18 CDT(-0500)] <colinclark> I'm slightly confused about one Test Swarm thing I've noticed
[08:43:24 CDT(-0500)] <Justin_o> okay
[08:43:37 CDT(-0500)] <colinclark> It seemed to run all day nicely on Firefox 4
[08:43:56 CDT(-0500)] <colinclark> Bosmon and I both seem to have noticed that it stalls after a test or two in Chrome
[08:44:19 CDT(-0500)] <colinclark> For me, it's as if it gets stuck on some kind of non-commit
[08:44:35 CDT(-0500)] <colinclark> "Status: Running Infusion Commit #d0d730e D0d730e tests..."
[08:44:53 CDT(-0500)] <colinclark> I can see the UI Options preview iFrame at the bottom of the page, and it seems never to bother to fetch more tests
[08:44:58 CDT(-0500)] <colinclark> Can you reproduce that at all?
[08:45:45 CDT(-0500)] <Justin_o> colinclark: i'll take a look and see
[08:46:04 CDT(-0500)] <colinclark> You know, I'm actually seeing the same in Firefox 4 now
[08:46:11 CDT(-0500)] <colinclark> Somehow I managed to get it to rock yesterday
[08:46:43 CDT(-0500)] <Justin_o> colinclark: It looks like it stalled for me in safari now
[08:46:49 CDT(-0500)] <Justin_o> oh it's going again
[08:46:55 CDT(-0500)] <Justin_o> there's a test there that isn't a test
[08:47:09 CDT(-0500)] <Justin_o> or I should say there's an html page that isn't a unit test file
[08:47:20 CDT(-0500)] <Justin_o> not sure if that's where you saw there error
[08:47:58 CDT(-0500)] <Justin_o> so the test will stall until it times out
[08:48:22 CDT(-0500)] <jamon> the Hello UI options one Justin_o?
[08:48:30 CDT(-0500)] <Justin_o> jamon: yep
[08:48:42 CDT(-0500)] <jamon> saw that too
[08:49:12 CDT(-0500)] <colinclark> That's quite a drag
[08:49:26 CDT(-0500)] <colinclark> So Test Swarm will freak if it encounters any HTML pages that aren't test pages?
[08:49:54 CDT(-0500)] <Justin_o> colinclark: well.. it's looking for the results
[08:50:04 CDT(-0500)] <Justin_o> which never occur so it waits till it times out
[08:51:19 CDT(-0500)] <Justin_o> jamon: did you change the regex that was used to find the test pages
[08:51:33 CDT(-0500)] <Justin_o> i think before it would look for files ending in "-test.html"
[08:51:36 CDT(-0500)] <jamon> Justin_o: just copied and pasted the svn one
[08:51:48 CDT(-0500)] <Justin_o> hmm..
[08:52:00 CDT(-0500)] <Justin_o> i would have thought it would have ignored that file
[08:56:43 CDT(-0500)] <michelled> colinclark, cindyli: I forgot to schedule our meeting for today - have either of you seen heidi today?
[08:57:10 CDT(-0500)] <cindyli> michelled: not yet
[08:57:57 CDT(-0500)] <michelled> shall we tentatively say 3:00?
[08:58:19 CDT(-0500)] <cindyli> michelled: fine with me
[08:59:59 CDT(-0500)] <cindyli> michelled: ha, heidi makes into the office. she is here
[09:04:59 CDT(-0500)] <michelled> ok, colinclark, heidi, cindyli: when would you like to meet to talk about UI Enhancer?
[09:05:20 CDT(-0500)] <cindyli> anytime is good to me
[09:06:01 CDT(-0500)] <Justin_o> jamon: can you point me at the perl script that adds the tests for testswarm
[09:07:26 CDT(-0500)] <jamon> Justin_o: ~/swarm.fluidproject.org/scripts/testswarm-infusion-git.pl
[09:07:53 CDT(-0500)] <Justin_o> jamon: thanks
[09:08:32 CDT(-0500)] <jhung> colinclark, jessm: who is managing the drumbeat project site at the moment? Is managing the feeds as easy as creating an account?
[09:08:50 CDT(-0500)]

[09:08:52 CDT(-0500)] <Justin_o> to
[09:08:55 CDT(-0500)]

[09:09:05 CDT(-0500)] <Justin_o> sorry..
[09:09:06 CDT(-0500)]

[09:09:25 CDT(-0500)] <Justin_o> basically that just adds -test in between the * and the .html
[09:11:22 CDT(-0500)] <michelled> heidi: does 3:00 work for you?
[09:11:26 CDT(-0500)] <jamon> Justin_o: ah
[09:11:57 CDT(-0500)] <heidi> michelled 3pm yep
[09:12:13 CDT(-0500)] <michelled> ok, I'll send a message to the list
[09:13:02 CDT(-0500)] <heidi> hey cindyli , michelled should jameswy and i start styling the UIO in the trunk, or in cindy's branch or? thoughts?
[09:13:39 CDT(-0500)] <michelled> cindyli's branch is more up to date then trunk but isn't fully working yet
[09:13:54 CDT(-0500)] <michelled> you should be fine to work off trunk
[09:14:01 CDT(-0500)] <heidi> okay cool, thanks
[09:14:53 CDT(-0500)] <jamon> Justin_o: made that change, will see when someone makes a commit
[09:15:53 CDT(-0500)] <Justin_o> jamon: thanks... hopefully it works
[09:16:11 CDT(-0500)] <colinclark> jhung: jessm can make you an admin once you set up an account on the Drumbeat site, I believe
[09:18:06 CDT(-0500)] <michelled> hi gsoc_pulkit, how's it going?
[09:18:08 CDT(-0500)] <colinclark> cindyli, michelled: What's left to fix before pushing the latest of cindyli's fork into the project repo?
[09:18:16 CDT(-0500)] <michelled> tests
[09:18:20 CDT(-0500)] <michelled> they are all broken
[09:18:25 CDT(-0500)] <colinclark> aha
[09:18:31 CDT(-0500)] <gsoc_pulkit> Hi michelled
[09:19:22 CDT(-0500)] <michelled> gsoc_pulkit: I'm actually going on vacation today and I'll be gone for two weeks.
[09:19:39 CDT(-0500)] <michelled> but there are lots of people around who can help you get started in the community
[09:19:52 CDT(-0500)] <gsoc_pulkit> ohk.
[09:19:59 CDT(-0500)] <michelled> gsoc_pulkit: Justin_o is a good person to start with
[09:20:16 CDT(-0500)] <gsoc_pulkit> I will get started on the documentation
[09:20:22 CDT(-0500)] <gsoc_pulkit> Hi Justin_o
[09:20:29 CDT(-0500)] <michelled> colinclark: we also need to integrate my branch with cindyli's branch in order to get the textfield sliders working again
[09:20:42 CDT(-0500)] <colinclark> michelled: Ah, interesting
[09:20:48 CDT(-0500)] <michelled> gsoc_pulkit: anastasiac has been leading a push with our documentation
[09:20:49 CDT(-0500)] <gsoc_pulkit> michelled: Yeah, I have seen that he is very active in the irc channel
[09:20:51 CDT(-0500)] <colinclark> your branch can't go in independently from cindyli's?
[09:21:00 CDT(-0500)] <michelled> anastasiac: if you have a starting point can you point gsoc_pulkit to it?
[09:21:28 CDT(-0500)] <anastasiac> michelled, gsoc-pulkit, I'll catch up on this thread and put together some useful links
[09:21:49 CDT(-0500)] <gsoc_pulkit> anastasiac: thanks.
[09:22:18 CDT(-0500)] <michelled> colinclark: with UIO becoming a renderer component textfield sliders need to be incorporated into the renderer tree
[09:22:21 CDT(-0500)] <michelled> I'm working on that now
[09:22:22 CDT(-0500)] <heidi> michelled accordion's gone in UIO in trunk now, right?
[09:22:41 CDT(-0500)] <michelled> colinclark: in the past, we handled the creation of the textfield sliders manually
[09:22:50 CDT(-0500)] <michelled> heidi: yep, been gone for a while
[09:22:52 CDT(-0500)] <colinclark> yep, i remember
[09:22:53 CDT(-0500)] <colinclark> sounds good
[09:22:55 CDT(-0500)] <heidi> cool
[09:23:15 CDT(-0500)] <colinclark> cindyli: I can't wait to see those green bars... your changes will make UI Options a lot better
[09:23:52 CDT(-0500)] <cindyli> colinclark: hopefully
[09:24:36 CDT(-0500)] <cindyli> colinclark: michelled, i'm currently working on uiEnhacer to ioc it. should i work on the tests first
[09:24:47 CDT(-0500)] <michelled> colinclark: cindyli might need some help with determining whether the tests are even useful anymore - at the time of UIO's creation a lot of the framework wasn't really built so quite a few tests were written to test things that are now tested in the framework tests.
[09:24:59 CDT(-0500)] <colinclark> ok
[09:25:08 CDT(-0500)] <colinclark> I can try to take a look later today, before the UIE meeting
[09:25:24 CDT(-0500)] <colinclark> I'm on grant proposal patrol today, so it's all going to be a bit of a stretch
[09:27:06 CDT(-0500)] <colinclark> cindyli: Do you figure that you and michelled can get to a point where your code is integrated by the end of today?
[09:27:49 CDT(-0500)] <michelled> colinclark: once I get the renderer tree going we're ready for integration
[09:28:08 CDT(-0500)] <michelled> cindyli will do the integration in her branch because it's unlikely we can get her branch into trunk today
[09:28:16 CDT(-0500)] <colinclark> how come?
[09:28:32 CDT(-0500)] <cindyli> because of the tests i assume
[09:28:37 CDT(-0500)] <colinclark> ok
[09:28:38 CDT(-0500)] <michelled> well, maybe we can - cindyli thoughts on getting those tests working?
[09:28:39 CDT(-0500)] <colinclark> I'm sure it would make heidi and jameswy really happy if they were working off the latest and greatest
[09:28:48 CDT(-0500)] <michelled> ya
[09:28:52 CDT(-0500)] <cindyli> agree
[09:29:04 CDT(-0500)] <colinclark> ok
[09:29:05 CDT(-0500)] <cindyli> ok, i will look into the tests first
[09:29:12 CDT(-0500)] <colinclark> I'll try to look at some of the unit tests for relevancy
[09:29:13 CDT(-0500)] <cindyli> see if i can get them working
[09:29:16 CDT(-0500)] <colinclark> in case that helps
[09:29:30 CDT(-0500)] <colinclark> In the meantime, cindyli, I think the key will be to ask questions about each test
[09:29:35 CDT(-0500)] <colinclark> "What is this testing?"
[09:29:49 CDT(-0500)] <colinclark> Is it testing the component itself, or some service offered by the framework?
[09:29:59 CDT(-0500)] <colinclark> Does it also cover the component working in harmony with all its subcomponents?
[09:30:14 CDT(-0500)] <colinclark> The key thing you've done is to start to break apart UI Options into smaller pieces
[09:30:28 CDT(-0500)] <colinclark> We want to make sure that each of the smaller pieces is independently tested
[09:30:39 CDT(-0500)] <colinclark> but on the other hand, we want to have tests like the sort that mlam has been writing for Uploader...
[09:30:49 CDT(-0500)] <colinclark> tests that cover the integration of all the little pieces into a working whole
[09:30:59 CDT(-0500)] <colinclark> I don't know if that's helpful or too abstract
[09:31:04 CDT(-0500)] <cindyli> understand.
[09:31:17 CDT(-0500)] <colinclark> cool
[09:31:32 CDT(-0500)] <colinclark> mlam: How are those Uploader tests going, based on your conversation yesterday with Bosmon in the channel?
[09:31:35 CDT(-0500)] <cindyli> just have to figure out how to do those
[09:31:40 CDT(-0500)] <colinclark>
[09:33:18 CDT(-0500)] <jhung> colinclark: thanks (re: drumbeat).
[09:33:51 CDT(-0500)] <jhung> jessm: can you make me an admin for the drumbeat site so I can investigate? username: jhung.
[09:35:10 CDT(-0500)] <jessm> jhung: let me check
[09:35:18 CDT(-0500)] <mlam> colinclark: they're going well, just doing some refactoring to get it into tip top shape
[09:35:25 CDT(-0500)] <colinclark> awesome
[09:35:35 CDT(-0500)] <colinclark> mlam: Any sense of when they'll land in the project repo?
[09:35:55 CDT(-0500)] <mlam> I'm hoping to finish everything off by 1 or 2...then ping Bosmon to take a final look
[09:36:11 CDT(-0500)] <mlam> so hopefully they'll be in by the end of today
[09:38:23 CDT(-0500)] <colinclark> great!
[09:42:28 CDT(-0500)] <jessm> jhung: there's no such option
[09:42:29 CDT(-0500)] <jessm> urgh
[09:43:29 CDT(-0500)] <jhung> jessm: uugh. That sucks. But I'm assuming there may not be anything interested for Smriti and I to look at from the admin side anyway? I assume the drumbeat is just aggregating the feeds from those URLs, right?
[09:43:57 CDT(-0500)] <jessm> jhung: exactly – there is absolutely nothing interesting; just a text box for inputing links
[09:44:26 CDT(-0500)] <jhung> jessm: okay. Thanks!
[10:03:51 CDT(-0500)] <jameswy> cindyli: Have you changed much of the original UIOptions.html?
[10:04:06 CDT(-0500)] <cindyli> no
[10:04:16 CDT(-0500)] <jameswy> Thanks!
[10:04:19 CDT(-0500)] <cindyli> np
[10:04:24 CDT(-0500)] <jhung> heidi: do you have admin privs for the inclusive learning handbook site?
[10:06:02 CDT(-0500)] <cindyli> jameswy: after looking at it, i think i did, at least all the rsf:id have been replaced by classes
[10:08:10 CDT(-0500)] <jameswy> cindyli: Ah, good to know. Anything else? We're ripping things up right now.
[10:08:24 CDT(-0500)] <heidi> jhung gonna do password reminder, one sec
[10:08:43 CDT(-0500)] <cindyli> jameswy: u only meant UIOptions.html, right?
[10:08:55 CDT(-0500)] <jameswy> cindyli: Yep.
[10:11:19 CDT(-0500)] <cindyli> jameswy: did a comparison, line 17, 29 regarding text slider, the classes are modified, don't recall i did that. did you, michelled?
[10:11:41 CDT(-0500)] <cindyli> but anyway, they are and that's all modified in UIOptions.html
[10:11:59 CDT(-0500)] <michelled> cindyli: no, I didn't touch the html
[10:14:00 CDT(-0500)] <cindyli> interesting
[10:14:58 CDT(-0500)] <Justin_o> anastasiac: did you push this one up the project repo https://github.com/fluid-project/infusion/pull/34
[10:16:37 CDT(-0500)] <colinclark> Maybe I'm missing something...
[10:16:46 CDT(-0500)] <colinclark> but it seems like heidi and jameswy are heading for a conflict nightmare
[10:17:03 CDT(-0500)] <colinclark> cindyli: if you replaced all rsf:ids with class names...
[10:17:12 CDT(-0500)] <michelled> colinclark: you're right - maybe heidi and jameswy should work on cindyli's branch after all
[10:17:16 CDT(-0500)] <colinclark> I can imagine a very large portion of the file will have changed from the version they are working on
[10:17:18 CDT(-0500)] <anastasiac> Justin_o, no - not yet. I was going to do it yesterday, but...
[10:17:25 CDT(-0500)] <jameswy> colinclark, cindyli: it's not too bad-- there were only a half dozen rsf:ids
[10:18:55 CDT(-0500)] <Justin_o> anastasiac: okay, so it's still on your radar?
[10:19:04 CDT(-0500)] <anastasiac> yes, Justin_o, on my list for today
[10:19:14 CDT(-0500)] <Justin_o> anastasiac: thanks
[10:21:27 CDT(-0500)] <cindyli> jameswy: michelled, it was me who moved the classes used by text slider around. moved flc- to be in front of -fl. no classes' names are modified. should be ok
[10:23:56 CDT(-0500)] <anastasiac> gsoc_pulkit, a nice place to start learning about Infusion is heidi's Developer Intro: http://wiki.fluidproject.org/display/fluid/Developer+Introduction+to+Infusion+Framework
[10:24:06 CDT(-0500)] <anastasiac> the starting point for our docs is http://wiki.fluidproject.org/display/fluid/Infusion+Documentation
[10:25:18 CDT(-0500)] <Justin_o> gsoc_pulkit: hello, sorry... i think i missed a message from you earlier
[10:26:05 CDT(-0500)] <heidi> cindyli michelled i think our changes our significant enough a merge prob won't happen. so we'll manually add cindy's changes in as we go. does that sound ok?
[10:26:24 CDT(-0500)] <gsoc_pulkit> anastasiac: Thanks. I will look at it.
[10:26:27 CDT(-0500)] <cindyli> osunds good
[10:26:44 CDT(-0500)] <cindyli> oh, u can work off the html in my branch
[10:27:07 CDT(-0500)] <gsoc_pulkit> Justin_o: no problem, I just wanted to say hi!
[10:28:09 CDT(-0500)] <Justin_o> gsoc_pulkit: thanks... if you have any questions, feel free to ping me...
[10:29:01 CDT(-0500)] <gsoc_pulkit> Justin_o: sure. thanks.
[10:43:53 CDT(-0500)] <colinclark> anastasiac: I'm still trying to get my head around your post to fluid-work about script blocks. Do you have any code in a repo somewhere?
[10:45:48 CDT(-0500)] <anastasiac> colinclark, not committed yet. I can commit what I have this afternoon. It's just experimental so far, so it's not necessarily representative of what we'll end up going with
[10:46:07 CDT(-0500)] <colinclark> I find it's really hard to give advice in the abstract
[10:46:11 CDT(-0500)] <colinclark> code helps a lot
[10:46:43 CDT(-0500)] <colinclark> I think that's probably the case for anyone
[10:47:36 CDT(-0500)] <anastasiac> of course!
[11:00:29 CDT(-0500)] <Justin_o> harriswong, mlam : http://jqueryui.com/docs/Changelog/1.8.12
[11:02:07 CDT(-0500)] <mlam> cool, thanks Justin_o
[11:02:48 CDT(-0500)] <Justin_o> np hopefully jquery and jquery ui will stop updating long enough for us to get a release out
[11:04:59 CDT(-0500)] <mlam> yah, no kidding
[11:20:28 CDT(-0500)] <heidi> cindyli some of the UIO controls will be switching from radio buttons to a checkbox. should we do that before you or after you? should we change/break the UI and you can fix later or?
[11:21:45 CDT(-0500)] <cindyli> heidi: go ahead with the change i will fix later
[11:22:24 CDT(-0500)] <heidi> k
[11:33:33 CDT(-0500)] <Justin_o> colinclark: at this weeks dev meeting can we talk a bit about the contributor gallery?
[11:43:08 CDT(-0500)] <colinclark> sure
[11:43:15 CDT(-0500)] <Justin_o> thanks
[12:16:02 CDT(-0500)] <anastasiac> colinclark, Bosmon, regarding the "instructional demo builder": I've pushed my code to my github repo (FLUID-4145 branch) and sent an email to the list with specifics of what's what
[12:21:06 CDT(-0500)] <colinclark> cool, anastasiac
[12:33:18 CDT(-0500)] <Justin_o> fluid-everyone: You can now directly merge pull requests through the github ui https://github.com/blog/843-the-merge-button
[12:33:39 CDT(-0500)] <michelled> have you tried it yet Justin_o?
[12:33:50 CDT(-0500)] <Justin_o> one thing to note is that currently they only use the --no-ff flag, and not the --log flag which we like
[12:33:56 CDT(-0500)] <jameswy> fluid-everyone: we're having a mock standup at http://connect.ocad.ca/fluid-work right now. Join if you have a minute or two,
[12:34:04 CDT(-0500)] <Justin_o> michelled: not yet
[12:34:18 CDT(-0500)] <Justin_o> but we can play around with it if you like
[12:35:11 CDT(-0500)] <Justin_o> jameswy: why are you having a mock standup
[12:35:13 CDT(-0500)] <Justin_o> ?
[12:49:25 CDT(-0500)] <jameswy> colinclark, michelled: It doesn't look like UIO ships with a preview by default--why is that?
[12:53:39 CDT(-0500)] <michelled> jameswy: actually, that is a new development - the preview is now in the standalone example.
[12:54:06 CDT(-0500)] <michelled> we broke it out because we found that we kept telling people to create their own and it really didn't feel like part of the component
[12:54:17 CDT(-0500)] <michelled> also, the default for 1.4 will be whole page preview I think
[12:54:30 CDT(-0500)] <jameswy> michelled: Hm... what do you mean by "whole page preview"?
[12:54:39 CDT(-0500)] <michelled> sliding panel
[12:54:47 CDT(-0500)] <michelled> I'm using the wrong terminology
[12:54:52 CDT(-0500)] <jameswy> Haha.
[12:54:57 CDT(-0500)] <jameswy> But what about for the full-page UIO?
[12:55:16 CDT(-0500)] <jameswy> I feel like it should ship with a default preview of some sort, much like we ship with a default style for UIO.
[12:55:31 CDT(-0500)] <jameswy> Basic elements for an immediate out-of-box experience.
[12:55:45 CDT(-0500)] <jameswy> As opposed to having to do further configuration and development.
[12:55:46 CDT(-0500)] <michelled> well, perhaps the standalone example as it is now will become the full-page UIO
[12:55:57 CDT(-0500)] <michelled> we need to fine tune how we are going to organize all of this
[12:56:08 CDT(-0500)] <jameswy> <nods>
[12:57:08 CDT(-0500)] <Justin_o> heidi, jameswy : what do you think of this as a start for the FSS contributor gallery http://wiki.fluidproject.org/display/fluid/FSS+Contributor+Gallery
[12:57:41 CDT(-0500)] <michelled> hmmm.. jameswy I just remembered that the demo also has it's own preview
[12:57:41 CDT(-0500)] <jameswy> Justin_o: Works for me!
[12:57:46 CDT(-0500)] <heidi> c'est bon
[12:58:31 CDT(-0500)] <Justin_o> thanks... i'm planning on talking about the contributor space at the dev meeting tomorrow.. i think we need to talk about requirements and etc.. like licenses
[12:58:58 CDT(-0500)] <jameswy> Justin_o: How about adding description and date added?
[12:59:22 CDT(-0500)] <Justin_o> jameswy: i was thinking about the date, didn't think of a description but that makes sense... i'll add those
[13:00:31 CDT(-0500)] <michelled> Justin_o: can you send a note to the list with what you decide at dev meeting tomorrow? I'd like to be able to catch up when I'm back
[13:04:05 CDT(-0500)] <colinclark> anastasiac: You there?
[13:13:21 CDT(-0500)] <anastasiac> colinclark, sorry - I'm here now
[13:13:48 CDT(-0500)] <colinclark> hey
[13:14:05 CDT(-0500)] <colinclark> so, can you explain in really simple terms what your goal is for the instructional demos stuff?
[13:14:24 CDT(-0500)] <colinclark> I gather that you have a bunch of pages--full HTML pages
[13:14:34 CDT(-0500)] <colinclark> and you want to render them into another page
[13:14:38 CDT(-0500)] <colinclark> Is that right so far?
[13:15:12 CDT(-0500)] <anastasiac> basically, yes: the ultimate goal is to have instructional demos in the docs that include both the working demo and a visual display of the code used for it
[13:15:47 CDT(-0500)] <anastasiac> we don't want to just put the demos right into the page manually, since that would require duplicating the same code/markup in both the 'working' demo and the display version
[13:16:27 CDT(-0500)] <anastasiac> by starting with demos in the distribution, we get a) distributed demos and b) a single source of the material for the docs
[13:17:26 CDT(-0500)] <colinclark> right
[13:17:50 CDT(-0500)] <colinclark> So, I'll try to use some terminology... hopefully it's consistence
[13:17:54 CDT(-0500)] <colinclark> consistent
[13:18:07 CDT(-0500)] <colinclark> There's a "source" document, which is the HTML page containing the demo itself
[13:18:18 CDT(-0500)] <colinclark> and a "target" document which is the "chrome" of the page
[13:18:23 CDT(-0500)] <colinclark> right?
[13:18:42 CDT(-0500)] <anastasiac> right
[13:19:42 CDT(-0500)] <colinclark> Okay, so you have some container, say a div, in the target document
[13:20:10 CDT(-0500)] <colinclark> And, is it the case that you want to put the contents of the source's <body> tag into that container?
[13:20:56 CDT(-0500)] <anastasiac> well, no, that's not quite an adequate description
[13:21:28 CDT(-0500)] <anastasiac> It will depend, in part, on a few things:
[13:21:53 CDT(-0500)] <anastasiac> whether or not the demo script is in a script tag in the html or in a seperate file, for one
[13:22:11 CDT(-0500)] <colinclark> okay
[13:22:12 CDT(-0500)] <colinclark> you
[13:22:13 CDT(-0500)] <anastasiac> and whether or not we have each small demo in its own html (which was my original thinking)
[13:22:17 CDT(-0500)] <colinclark> you're getting ahead of me here
[13:22:45 CDT(-0500)] <colinclark> fuzzing over the question of how the demo's code is linked in, was my previous question correct?
[13:23:00 CDT(-0500)] <colinclark> In other words, you want to take the body of one page and put it into a container in the other?
[13:23:24 CDT(-0500)] <anastasiac> ok, in simple terms, partly true: we want to put the body into another page; but into two places in the other page
[13:23:38 CDT(-0500)] <colinclark> okay
[13:23:45 CDT(-0500)] <colinclark> elaborate the "two places" part
[13:23:58 CDT(-0500)] <colinclark> that's your point about the working and display versions?
[13:24:05 CDT(-0500)] <anastasiac> right
[13:24:22 CDT(-0500)] <anastasiac> we need the html for the working demo, and we also want the user to see a display of the html
[13:24:54 CDT(-0500)] <colinclark> okay
[13:25:05 CDT(-0500)] <colinclark> so a source view and a "see the component and play around with it" view?
[13:25:13 CDT(-0500)] <anastasiac> right
[13:25:39 CDT(-0500)] <colinclark> okay
[13:25:50 CDT(-0500)] <colinclark> am i right in thinking that the source view part is pretty straightforward
[13:25:59 CDT(-0500)] <colinclark> ?
[13:26:25 CDT(-0500)] <colinclark> i.e. it's just a question of getting the source page as a string and putting it into the target document in a way that escapes the HTML tags appropriate?
[13:26:26 CDT(-0500)] <anastasiac> yes: if you have the original strings, you just insert them into the right elements
[13:26:35 CDT(-0500)] <colinclark> ok
[13:26:47 CDT(-0500)] <colinclark> So the hard part is putting a live demo into some other page, right?
[13:26:54 CDT(-0500)] <anastasiac> right
[13:27:09 CDT(-0500)] <colinclark> Okay, so that brings us back to your point about the inadequacy of my previous statement
[13:27:22 CDT(-0500)] <colinclark> about just putting the source <body> in to a target <div>
[13:27:36 CDT(-0500)] <colinclark> So, the next question is, do we care about the stuff in the source's <head> at all?
[13:27:55 CDT(-0500)] <colinclark> In other words, do we also need to link in any style sheets and scripts into the target page?
[13:28:55 CDT(-0500)] <anastasiac> if we assume that a) the target page is including infusion already, b) the script and css are embedded in the body in <script> and <style> tags, no: the head is not required
[13:29:03 CDT(-0500)] <colinclark> Okay
[13:29:09 CDT(-0500)] <colinclark> b) is the key point, then
[13:29:13 CDT(-0500)] <anastasiac> right
[13:29:18 CDT(-0500)] <colinclark> I am of two minds about b), how about you?
[13:29:29 CDT(-0500)] <anastasiac> well, yes, definitely
[13:29:41 CDT(-0500)] <colinclark> On one hand, it's often really nice to see everything all in one spot
[13:29:46 CDT(-0500)] <anastasiac> exactly
[13:29:50 CDT(-0500)] <colinclark> I'm thinking, for example of Paul Rouget's style (from Mozilla)
[13:30:01 CDT(-0500)] <colinclark> where he makes demos specifically designed for "View Source learning"
[13:30:10 CDT(-0500)] <colinclark> All his styles and scripts are sitting right there in the HTML page
[13:30:26 CDT(-0500)] <colinclark> On the other hand, it has two weaknesses, I think
[13:30:35 CDT(-0500)] <colinclark> 1. No one should ever do that in a real web app
[13:30:48 CDT(-0500)] <colinclark> 2. For long demos, it can be really hard to read--lots and lots of scrolling
[13:31:08 CDT(-0500)] <Justin_o> jameswy: is this better? http://wiki.fluidproject.org/display/fluid/FSS+Contributor+Gallery
[13:31:16 CDT(-0500)] <Justin_o> i didn't know what to put for the descriptions though..
[13:31:28 CDT(-0500)] <Justin_o> i'll make an appeal to the contributors for that I guess
[13:32:30 CDT(-0500)] <colinclark> anastasiac: Does that seem in line with your sense of the pros and cons of the "view source learning" style?
[13:32:41 CDT(-0500)] <anastasiac> yep, that sums it up very well
[13:33:34 CDT(-0500)] <colinclark> Okay, so if we went with the "view source learning" style, i.e. all demo assets are actually inline in the document...
[13:33:49 CDT(-0500)] <colinclark> Is the solution to getting the source into the target fairly straightforward?
[13:34:31 CDT(-0500)] <anastasiac> getting the source into the target in display mode is straightforward - I've got that working. it's the working demo that's the issue
[13:35:03 CDT(-0500)] <colinclark> how so?
[13:36:08 CDT(-0500)] <anastasiac> well, simply injecting it in works just fine, as I did in the jQTest.js version. But given that all of these demo pages will have a remarkably similar form, it's well suited to using a template with the Renderer - hence the start at creating a renderer component
[13:36:23 CDT(-0500)] <anastasiac> but the rendering of the script tags doesn't result in the script executing - hence, no working demo
[13:36:53 CDT(-0500)] <Justin_o> anastasiac: can this be closed now http://issues.fluidproject.org/browse/FLUID-4182
[13:37:10 CDT(-0500)] <colinclark> anastasiac: The jQTest.js file is fascinating to me
[13:37:14 CDT(-0500)] <colinclark> There's a lot of code there
[13:37:35 CDT(-0500)] <anastasiac> Justin_o, yes, heidi should be able to resolve it
[13:37:37 CDT(-0500)] <colinclark> a real RegEx pro must have written that
[13:37:43 CDT(-0500)] <Justin_o> anastasiac: thanks
[13:38:07 CDT(-0500)] <anastasiac> colinclark, that was taken from the jQuery UI demos - I wanted to understand how they were doing it
[13:38:11 CDT(-0500)] <colinclark> What does it do, exactly?
[13:38:30 CDT(-0500)] <anastasiac> it strips off unwanted markup from the source HTML string
[13:39:27 CDT(-0500)] <colinclark> Seems like it, yes
[13:39:36 CDT(-0500)] <colinclark> How is it different from doing something like this? http://pastie.org/1836321
[13:41:16 CDT(-0500)] <anastasiac> well, the code you've pasted injects the entire fetched file into the target, for one; I'm guessing that what it loads is, already, a fragment
[13:42:50 CDT(-0500)] <colinclark> Here's my test source page: http://pastie.org/1836329
[13:43:12 CDT(-0500)] <colinclark> I was just trying to remind myself of all these issues, which I guess we last confronted back in the days of Engage and the ScreenNavigator
[13:43:16 CDT(-0500)] <colinclark> So I whipped up those two pages
[13:43:52 CDT(-0500)] <colinclark> I guess, alternatively, I could have used $("#fragmentContainer").load("fragment.html body") or something to that effect
[13:44:13 CDT(-0500)] <anastasiac> yes, I did play with load as well.
[13:44:39 CDT(-0500)] <anastasiac> so your example would inject the full 'fragment' into the target, including the head, the body tag, the html tag, right?
[13:44:48 CDT(-0500)] <colinclark> No, it doesn't work that way
[13:44:59 CDT(-0500)] <colinclark> Two things are at work here
[13:45:19 CDT(-0500)] <colinclark> jQuery's ability to correctly load injected <script> tags
[13:45:45 CDT(-0500)] <colinclark> and the fact that the browser will actually trim off all the extras that "don't make sense" in the context of injecting one chunk of markup into the other
[13:46:13 CDT(-0500)] <colinclark> So any browser we've ever met just won't let you create a chunk of DOM containing tags like <head> and <body> in the context of some other page's DOM
[13:46:21 CDT(-0500)] <colinclark> Is that something you observed in your tests as well?
[13:46:28 CDT(-0500)] <anastasiac> ok, so that makes the working example simple
[13:46:38 CDT(-0500)] <anastasiac> yes, I had noticed that
[13:47:07 CDT(-0500)] <colinclark> So I'm trying to get a sense of the advantages of the RegEx approach [13:47:19 CDT(-0500)] <colinclark> I guess it's much less reliant on "implementation details" [13:47:23 CDT(-0500)] <colinclark> What else? [13:47:58 CDT(-0500)] <anastasiac> well, the approach in your pastie essentially does the same thing as the regexps do: strips off the unwanted html [13:48:24 CDT(-0500)] <anastasiac> one difference is that we need to further process the string, to split the markup from the script, for display purposes [13:48:42 CDT(-0500)] <colinclark> for the source view, you mean? [13:48:50 CDT(-0500)] <anastasiac> with your approach, the stripped string could be extracted from the DOM, I suppose [13:48:51 CDT(-0500)] <anastasiac> yes [13:49:11 CDT(-0500)] <colinclark> or, alternatively, we could change the dataType from "html" to "text" and then process that however we need [13:49:24 CDT(-0500)] <colinclark> So why did you settle on the regexs instead of another approach? [13:49:54 CDT(-0500)] <anastasiac> well, I haven't settled on any approach yet I'm still at the exploration phase [13:49:58 CDT(-0500)] <colinclark> ok [13:50:07 CDT(-0500)] <colinclark> So that's half of it [13:50:29 CDT(-0500)] <colinclark> We'll still have to decide on an approach... "view source learning" or "lead by example" [13:50:40 CDT(-0500)] <colinclark> That can be a separate discussion, and one that everyone is involved with [13:50:57 CDT(-0500)] <colinclark> I think I'm on the side of the latter [13:51:05 CDT(-0500)] <colinclark> So, we should look at the other half of the problem... [13:51:23 CDT(-0500)] <anastasiac> interesting - it was you who first recommended the 'view source learning' approach to me [13:51:37 CDT(-0500)] <colinclark> Perhaps I recommended it as "interesting" [13:51:38 CDT(-0500)] <colinclark> Reme [13:52:03 CDT(-0500)] <colinclark> Remember, my goal is to point out all the factors we should weight [13:52:07 CDT(-0500)] <colinclark> weigh [13:52:14 CDT(-0500)] <colinclark> stupid typing [13:52:32 CDT(-0500)] <jameswy> Justin_o: Looks good. [13:52:48 CDT(-0500)] <colinclark> So, walk me through some of the issues if we were to try to solve the problem for this type of demo, anastasiac [13:53:01 CDT(-0500)] <colinclark> I guess the source view might be even easier in many ways [13:53:10 CDT(-0500)] <Justin_o> jameswy: thanks [13:53:13 CDT(-0500)] <colinclark> no need to extract the code from some markup, just load the .js file and display it? [13:53:50 CDT(-0500)] <anastasiac> do you mean the 'lead by example' approach? [13:54:18 CDT(-0500)] <colinclark> yeah [13:54:25 CDT(-0500)] <colinclark> separate js, css, and html files [13:57:56 CDT(-0500)] <Justin_o> jamon: hello, anastasiac pushed some changes up to the project repo about 45 minutes ago.. i don't see it in testswarm though [13:58:55 CDT(-0500)] <jamon> Justin_o: hmm, will check in a moment [13:59:02 CDT(-0500)] <Justin_o> jamon: thanks [13:59:09 CDT(-0500)] <anastasiac> sorry, colinclark, distracted by a conversation [13:59:22 CDT(-0500)] <anastasiac> so issues surrounding these demos: [14:00:08 CDT(-0500)] <anastasiac> the format of the source files, as we've been discussing: "view source learning" vs "lead by example." The two require differences in the solution [14:00:52 CDT(-0500)] <jamon> Justin_o: what was the commit hash? [14:01:18 CDT(-0500)] <Justin_o> jamon: i think it was efe30ec57ff1a3be3400b6dff986b683678c4bdb [14:01:31 CDT(-0500)] <anastasiac> we're expecting there to be many of these little demos, so separate files results in many more files. Not a major problem, just another consideration [14:02:36 CDT(-0500)] <colinclark> I guess the biggest difference is maybe that we'll have to care about the <head> of the source document, too? [14:02:45 CDT(-0500)] <anastasiac> I'm not quite sure what you meant above, colinclark, when you said "just load the .js file and display it" - were you thinking "load" via ajax, or through the head? [14:03:18 CDT(-0500)] <jamon> Justin_o: job there now? [14:03:22 CDT(-0500)] <colinclark> Well, I was talking about the source view [14:03:22 CDT(-0500)] <colinclark> S [14:03:26 CDT(-0500)] <colinclark> So I guess neither [14:04:17 CDT(-0500)] <colinclark> I was saying that handling the source view would be easier in the case where all the code was in a separate file, since you'll just have to fetch the file and display it in the source view, rather than having to do some kind of string manipulation on a document containing a mix of stuff [14:04:21 CDT(-0500)] <colinclark> Does that make sense? [14:04:35 CDT(-0500)] <anastasiac> yes [14:05:05 CDT(-0500)] <Justin_o> jamon: yep it's there now.. thanks.. but it looks like that file snuck in again [14:05:13 CDT(-0500)] <michelled> colinclark, cindyli, heidi: do you prefer skype or connect ? [14:06:28 CDT(-0500)] <Justin_o> jamon: yah... need some more digging into that one... since it's causing all the jobs to show up as having an error [14:06:31 CDT(-0500)] <anastasiac> well, that's one of the things I was pondering [14:06:50 CDT(-0500)] <michelled> let's do skype then - I was having issues with connect today. let's hope this works better [14:06:54 CDT(-0500)] <anastasiac> we could identify the "source" script through the head, or it could possibly be specified through a configuration option [14:07:18 CDT(-0500)] <michelled> fluid-everyone: does anyone else want to join the uienhancer meeting? [14:07:20 CDT(-0500)] <michelled> we are talking code strategy [14:07:26 CDT(-0500)] <colinclark> So, correct me if I'm mistaken, but this seems to be squarely into ScreenNavigator territory, anastasiac [14:07:54 CDT(-0500)] <colinclark> What it did was to take a source document, and carefully inject only the unique head material into the target page [14:08:00 CDT(-0500)] <anastasiac> I haven only started to look into the screen navigator, so I'm certainly not going to correct you [14:08:17 CDT(-0500)] <colinclark> Meaning, only the script and link tags that were unique to the source document [14:08:28 CDT(-0500)] <colinclark> Does that seem roughly like what you were thinking of, except no configuration options required? [14:09:04 CDT(-0500)] <anastasiac> that sounds right, for the 'lead by example' version [14:09:11 CDT(-0500)] * anastasiac is checking the screen nav code [14:09:14 CDT(-0500)] <colinclark> Are there any other requirements we are missing? [14:09:28 CDT(-0500)] <colinclark> The Screen Navigator, as it shipped in Engage, has other concerns [14:09:31 CDT(-0500)] <colinclark> which would need to be factored out [14:09:39 CDT(-0500)] <colinclark> It's the "Navigator" part of the equation [14:09:52 CDT(-0500)] <colinclark> it assumes that there is a sequences of pages [14:10:09 CDT(-0500)] <colinclark> and uses the old-fashioned URL hash hack to make back-button navigation work [14:10:15 CDT(-0500)] <colinclark> which is something you don't need in this context [14:10:28 CDT(-0500)] <colinclark> What you need is roughly the same thing you encountered back in the days of CSpace, too, it sounds like [14:10:29 CDT(-0500)] <colinclark> T [14:10:41 CDT(-0500)] <colinclark> The ability to load unique head material into the target page [14:10:54 CDT(-0500)] <colinclark> Unless there are other requirements we're missing [14:11:28 CDT(-0500)] <harriswong> justin_o: ping [14:11:37 CDT(-0500)] <anastasiac> colinclark, does the screennav copy the relevant link and script element into the target header, or does it actually load the js and css files? [14:11:52 CDT(-0500)] <colinclark> hmm [14:11:56 CDT(-0500)] <colinclark> I'm not sure I understand the question [14:12:24 CDT(-0500)] <anastasiac> sorry, I'm still becoming familiar with how the screen nav works [14:12:45 CDT(-0500)] <anastasiac> so maybe i'm misunderstanding [14:12:55 CDT(-0500)] <Justin_o> harriswong: hello [14:13:18 CDT(-0500)] <anastasiac> the screen nav modifies the <head> of the target page by adding a <script> tag linking to the js - is this correct, colinclark? [14:13:49 CDT(-0500)] <heidi> michelled are we meeting now? [14:14:09 CDT(-0500)] <michelled> yes, heidi - I'll call you in one minute [14:14:29 CDT(-0500)] <harriswong> justin_o: hi, how does fluid wrap special characters? In ImageReordererTestsCOnstants.js line 47, i see "gallery:::gallery....", I want to escape those colons to "\\:\\::" [14:15:25 CDT(-0500)] <harriswong> justin_o: it seems like jquery now needs all special chars escaped, for http://bugs.jquery.com/ticket/5241 as of 1.5+ [14:15:31 CDT(-0500)] <Justin_o> harriswong: that's what you'd use fluid.jById and fluid.byId for [14:16:05 CDT(-0500)] <harriswong> justin_o: thanks. will look into that. [14:16:34 CDT(-0500)] <colinclark> anastasiac: It injects string material from the head of one document into the other [14:17:29 CDT(-0500)] <anastasiac> ok, so the screen nav approach could be used to handle the working demo part, but we still need to actually load the script and css for display purposes [14:17:35 CDT(-0500)] <colinclark> To be clear, in itself, it won't do what you need as-is [14:17:42 CDT(-0500)] <colinclark> But a refactoring of it could [14:17:42 CDT(-0500)] <Justin_o> harriswong: no problem.. [14:17:43 CDT(-0500)] <colinclark> and should [14:17:49 CDT(-0500)] <anastasiac> Justin_o, do we have a working engage demo anywhere? [14:17:56 CDT(-0500)] <heidi> erin justin's going to create a contributor template for fss contribs. that'll help [14:17:57 CDT(-0500)] <Justin_o> anastasiac: yes.. [14:17:59 CDT(-0500)] <colinclark> anastasiac: fluidengage.org [14:18:12 CDT(-0500)] <Justin_o> colinclark: you beat me to it [14:18:15 CDT(-0500)] <erin> heidi: excellent. thanks Justin_o [14:18:17 CDT(-0500)] <colinclark> [14:18:37 CDT(-0500)] <anastasiac> ah, right - requires Safari, doesn't it, Justin_o? thanks [14:18:57 CDT(-0500)] <Justin_o> anastasiac: yep... some webkit browser anyways [14:21:15 CDT(-0500)] <Justin_o> jamon: can you try to make one more change to the perl script [14:21:29 CDT(-0500)]