fluid-work IRC Logs-2011-04-21

[08:45:28 CDT(-0500)] <heidi_> anastasiac i'll add in that demo fix for the fl-icon issue
[08:45:40 CDT(-0500)] <anastasiac> heidi_, great, thanks
[10:15:29 CDT(-0500)] <michelled> harriswong: did the jquery ui upgrade get into the project repo yet?
[10:17:14 CDT(-0500)] <harriswong> michelled: yes, justin_o committed it. https://github.com/fluid-project/infusion/pull/27
[10:22:45 CDT(-0500)] <michelled> thx
[12:59:22 CDT(-0500)] <colinclark> michelled: Is our FLUID-4140 branch necessary any more?
[12:59:34 CDT(-0500)] <colinclark> I was doing some branch tidy up in my Github fork
[12:59:42 CDT(-0500)] <colinclark> You did indeed merge changes from that branch
[13:00:15 CDT(-0500)] <michelled> if 38d0d93bb98bd17a7dc8b741d3606933f57080b1 is the last commit then it's been merged
[13:00:22 CDT(-0500)] <michelled> so you can get rid of the branch
[13:00:58 CDT(-0500)] <colinclark> I don't actually see that rev number in the log
[13:01:12 CDT(-0500)] <colinclark> ah, it's out of date
[13:01:13 CDT(-0500)] <colinclark> hang on
[13:01:17 CDT(-0500)] <colinclark> wow, crustiness
[13:01:49 CDT(-0500)] <colinclark> last commit on that branch seems to have been yours
[13:01:50 CDT(-0500)] <colinclark> 58773708c64eac3e4b8d5568dccba02658833b95
[13:02:22 CDT(-0500)] <michelled> maybe that commit hash I dropped in is the commit to the project repo
[13:02:38 CDT(-0500)] <colinclark> probably. Maybe the merge itself?
[13:02:58 CDT(-0500)] <michelled> yes
[13:03:14 CDT(-0500)] <michelled> 587737 is the last commit on the branch
[13:04:20 CDT(-0500)] <michelled> 38d0d9 is the merge
[13:49:14 CDT(-0500)] <colinclark> harriswong: So how's it going with the jQuery issues?
[13:50:20 CDT(-0500)] <harriswong> colinclark: I think it's now back to a jquery issue. It seems like trigger('focus') fires focus twice. Code here: http://jsfiddle.net/WVcy5/
[13:50:43 CDT(-0500)] <harriswong> colinclark: i have left a note on http://bugs.jquery.com/ticket/7340, and waiting for a reply. In the mean time, moving forward fixing other failed unit testcase
[13:51:34 CDT(-0500)] <colinclark> Any idea why it's firing twice for the case of the focus event?
[13:52:50 CDT(-0500)] <harriswong> colinclark: in 1.5.2+ (including 1.6beta and git versoin), the trigger fn invoke focus when focusin is called
[13:53:24 CDT(-0500)] <colinclark> So, I'm confused
[13:53:35 CDT(-0500)] <colinclark> Are you saying that trigger("focus") fires twice?
[13:53:50 CDT(-0500)] <colinclark> Or that trigger("focusin") fires both a focus and a focusin, harriswong?
[13:54:42 CDT(-0500)] <harriswong> colinclark: I think trigger('focus') calls 'focus', then 'focusin', then mysteriously calls 'focus' again
[13:55:42 CDT(-0500)] <colinclark> Okay
[13:55:55 CDT(-0500)] <colinclark> So let's get beyond thinking into knowing, if you know what I mean
[13:56:17 CDT(-0500)] <colinclark> Here's what I'd suggest
[13:56:21 CDT(-0500)] <colinclark> write three unit tests
[13:56:35 CDT(-0500)] <colinclark> One unit tests for trigger("focus")
[13:56:42 CDT(-0500)] <colinclark> One unit test for trigger("focusin")
[13:56:49 CDT(-0500)] <colinclark> and if necessary, one unit test that calls them both
[13:57:11 CDT(-0500)] <colinclark> In your unit test, start by creating a listener that will count the number of times an event is fired
[13:57:14 CDT(-0500)] <colinclark> and assert what you expect
[13:57:29 CDT(-0500)] <colinclark> When I look at your JSFiddle, all I see is what you tell me happens
[13:57:41 CDT(-0500)] <colinclark> it doesn't actually recreate the problem for me
[13:57:49 CDT(-0500)] <colinclark> Does that make sense, harriswong?
[13:58:06 CDT(-0500)] <harriswong> colinclark: digesting on "it doesn't actually recreate the problem for me"
[13:58:23 CDT(-0500)] <colinclark> Okay, let's look at your fiddle
[13:58:28 CDT(-0500)] <colinclark> What is it?
[13:58:35 CDT(-0500)] <harriswong> http://jsfiddle.net/WVcy5/
[13:58:40 CDT(-0500)] <colinclark> It's a line of code that calls trigger()
[13:58:48 CDT(-0500)] <colinclark> and a comment telling me what the behaviour you saw is
[13:59:04 CDT(-0500)] <colinclark> Put it into a form where you can let others see the behaviour themselves
[13:59:20 CDT(-0500)] <colinclark> rather than just reading a comment
[13:59:23 CDT(-0500)] <colinclark> do you see what I mean?
[13:59:34 CDT(-0500)] <colinclark> The only way to truly show that a bug exists is to have a unit test for it
[13:59:43 CDT(-0500)] <colinclark> Then I can run it, or John Resig could run it, or anyone else could run it
[13:59:49 CDT(-0500)] <colinclark> and see exactly the same issue you're seeing
[13:59:56 CDT(-0500)] <harriswong> colinclark: will do. moving ahead with making 3 unit tests as suggested.
[13:59:58 CDT(-0500)] <colinclark> This is just some console logs
[14:00:03 CDT(-0500)] <colinclark> put it into a unit test, and it will rock
[14:00:04 CDT(-0500)] <colinclark>
[14:26:47 CDT(-0500)] <colinclark> michelled: Can you merge on a per-file basis in Git?
[14:34:03 CDT(-0500)] <michelled> colinclark: I've never tried it but it seems like you want to use git co with a path
[14:34:05 CDT(-0500)] <michelled> http://www.kernel.org/pub/software/scm/git/docs/git-checkout.html
[14:37:22 CDT(-0500)] <fatima> Hello, I am Fatima and i am working on a screen recording tool for the Open VULab at YorkU
[14:37:28 CDT(-0500)] <colinclark> Hi fatima
[14:37:32 CDT(-0500)] <fatima> Came to know about this channel and thought to join it
[14:37:32 CDT(-0500)] <colinclark> How's it going?
[14:38:07 CDT(-0500)] <colinclark> Cool, welcome
[14:38:17 CDT(-0500)] <fatima> Well, i am trying to improve the screen recordign tool that is written in Java and get stuck many time
[14:38:26 CDT(-0500)] <fatima> many times
[14:39:19 CDT(-0500)] <fatima> I was wondering if there is anyone here who could give soem advice on how to improve it further
[14:39:51 CDT(-0500)] <colinclark> fatima: We can certainly try
[14:40:06 CDT(-0500)] <colinclark> I spent some time with the code, but it was quite awhile ago
[14:40:29 CDT(-0500)] <fatima> It is working well now (atleast when the code does not break )
[14:41:16 CDT(-0500)] <colinclark>
[14:41:27 CDT(-0500)] <colinclark> Sounds like you were having some threading problems not too long ago
[14:41:28 CDT(-0500)] <fatima> Basically Java Robot class is used to capture screen shots, which are saved as Jpeg Images on the client machine for some time, then these images are streamed back to the server and stitched together to form a video
[14:42:08 CDT(-0500)] <fatima> The Image compression and disk IO takes a lot of time. Moreover, this application becomes very slow on slow internet connections
[14:42:59 CDT(-0500)] <fatima> I am looking into ways to imrpove it - For example compressing the images with some video codec and then stream them , was one of the suggestions
[14:43:08 CDT(-0500)] <colinclark> I was just going to ask...
[14:43:31 CDT(-0500)] <fatima> Also, the Java code becomes relatively slow on Mac (this application runs as an applet in a browser)
[14:43:33 CDT(-0500)] <colinclark> Seems like the biggest problem is treating each frame as an image, using JPEG compression
[14:43:40 CDT(-0500)] <fatima> Yes
[14:44:03 CDT(-0500)] <fatima> So far we are treating each framse as an Image
[14:44:22 CDT(-0500)] <colinclark> I haven't looked recently at Java's support for video codecs
[14:44:42 CDT(-0500)] <colinclark> But it seems that you won't need a really modern codec
[14:45:02 CDT(-0500)] <fatima> We do not want to install anything on client side othewsie Xuggle could have been used and it supports many codecs
[14:45:47 CDT(-0500)] <fatima> I think any decent video codec should be fine
[14:46:30 CDT(-0500)] <fatima> I was looking into Red5 server as well as other applications like it to see if i can use them
[14:48:29 CDT(-0500)] <colinclark> So where does the code live now, fatima?
[14:49:37 CDT(-0500)] <fatima> I think my other team member is moving it to git
[14:49:47 CDT(-0500)] <colinclark> Oh, nice
[14:51:51 CDT(-0500)] <fatima> http://openvulab.org/
[14:52:10 CDT(-0500)] <fatima> There is a link in this page in the most recent post
[14:52:56 CDT(-0500)] <mlam> harriswong: do you know what version of jQuery is in john kremer's branch right now? is it sitll 1.5.2 or did you upgrade it to 1.6?
[14:53:19 CDT(-0500)] <harriswong> mlam: 1.5.1
[14:53:54 CDT(-0500)] <harriswong> mlam: on my branch that was cloned from his is on 1.5.2, but testing 1.6beta as well.
[14:54:09 CDT(-0500)] <mlam> ok, cool, thanks
[14:54:19 CDT(-0500)] <harriswong> np
[15:09:31 CDT(-0500)] <colinclark> justin and I were talking yesterday mlam and harriswong
[15:10:54 CDT(-0500)] <colinclark> about jquery 1.6
[15:11:17 CDT(-0500)] <colinclark> it'll probably be out before we release 1.4
[15:11:23 CDT(-0500)] <colinclark> so we might as well do our testing on it
[15:11:36 CDT(-0500)] <mlam> ok, makes sense
[15:11:40 CDT(-0500)] <colinclark> For sanity, doing what you're doing makes sense--testing alongside a stable release
[15:11:45 CDT(-0500)] <colinclark> 1.5.2, in that case
[15:12:18 CDT(-0500)] <mlam> so harriswong, do you have a separate branch with jQuery 1.6?
[15:14:09 CDT(-0500)] <harriswong> mlam: not yet. I will push 1.5.2 to github that was cloned from john; then another one that is with 1.6beta as a separated branch.
[15:14:31 CDT(-0500)] <harriswong> mlam: Do you need it now or i can give you the links later?
[15:15:04 CDT(-0500)] <mlam> whenever you're ready harriswong
[15:15:22 CDT(-0500)] <harriswong> mlam: ok
[15:36:55 CDT(-0500)] <fatima> It was very nice talking to you. I will be in touch.
[16:10:43 CDT(-0500)] <colinclark> Bosmon:
[16:10:45 CDT(-0500)] <colinclark> Bosmon:
[16:10:47 CDT(-0500)] <colinclark> okay
[16:11:01 CDT(-0500)] <colinclark> Bosmon's running off
[16:11:09 CDT(-0500)] <colinclark> but we were just talking off-channel about options backwards compatibility
[16:11:29 CDT(-0500)] <colinclark> He's made this notion of options transformation first-class in the framework now
[16:11:45 CDT(-0500)] <colinclark> We can see this in a new unit test for ModelTransformations
[16:13:23 CDT(-0500)] <colinclark> An options object can now contain a property called "transformOptions"
[16:13:49 CDT(-0500)] <colinclark> Which is an object containing two required properties:
[16:14:20 CDT(-0500)] <colinclark> 1. "transformer," which is an EL path to the transformation function
[16:14:47 CDT(-0500)] <colinclark> It could be any sort of transformation, really
[16:15:23 CDT(-0500)] <colinclark> but we've got the sneak peeky fluid.model.transformWithRules that might be typically used
[16:15:38 CDT(-0500)] <colinclark> 2. "config," which includes any options required by the transformation function
[16:15:58 CDT(-0500)] <colinclark> Again, in the case of fluid.model.transformWithRules, "config" would be the transformation rules