[08:52:45 EST(-0500)] * jayshao (n=jayshao@jayshao.oirt.rutgers.edu) has joined #fluid-work
[09:22:37 EST(-0500)] * anastasiac (n=chatzill@142.150.154.149) has joined #fluid-work
[09:29:18 EST(-0500)] * clown (n=clown@user383.megabit.utoronto.ca) has joined #fluid-work
[09:35:49 EST(-0500)] <anastasiac> good morning, clown. Nice work on horizontal portlet moving. Can we consider FLUID-202 resolved now?
[09:36:02 EST(-0500)] <clown> not quite.
[09:36:31 EST(-0500)] <clown> instead of configuring my laptop to run sakai for testing in the image gallery, i'm using the daily build server.
[09:36:45 EST(-0500)] <anastasiac> ok...
[09:37:09 EST(-0500)] <clown> i'm just testing a few things now. i expect to be resolving the bug in a couple of minutes.
[09:37:16 EST(-0500)] <clown> you make it in?
[09:37:29 EST(-0500)] <anastasiac> of course
[09:37:35 EST(-0500)] <anastasiac> no one else is here yet
[09:37:41 EST(-0500)] <anastasiac> I was wondering about Simon's flight, though
[09:37:52 EST(-0500)] <anastasiac> I've heard hundreds of flights are cancelled
[09:41:27 EST(-0500)] <clown> oh yeah. alistair has another snow day.
[09:43:05 EST(-0500)] <anastasiac> yay!
[09:44:45 EST(-0500)] <clown> yes, he is in video game heaven right now.
[09:45:00 EST(-0500)] <clown> FLUID-202 is resolved. yipee!
[09:53:01 EST(-0500)] <clown> I think there is a weakness in JSLint. Given a long file (e.g., Reorderer.js), it doesn't appear to scan all of it. The error log stops at line 678 (implied global 'jQuery'), but there are 1020 lines in there. The last appearance of jQuery is on line 830.
[09:58:47 EST(-0500)] * EricDalquist (n=EricDalq@adsl-76-204-102-88.dsl.mdsnwi.sbcglobal.net) has joined #fluid-work
[09:59:49 EST(-0500)] <anastasiac> interesting - and not good
[10:06:01 EST(-0500)] <clown> makes my wonder if there is a command line version, or an eclipse plug-in that might have greater tolerance for big files.
[10:09:53 EST(-0500)] <anastasiac> yes, that would be nice
[10:11:10 EST(-0500)] <anastasiac> michelle thinks that jslint might stop if it finds a problem that it needs you to fix before it can continue?
[10:30:36 EST(-0500)] <clown> don't think so. if you look at the details it says "Problem at line 681 character 17: Too many errors. (66% scanned)."
[10:31:09 EST(-0500)] <clown> all of the errors thus far are 'undefined global'
[10:31:13 EST(-0500)] <anastasiac> looks like it only stopped because there were too many errors
[10:31:28 EST(-0500)] <clown> there's an echo in this chat room...
[10:31:56 EST(-0500)] <anastasiac> my point is, it will likely scan the whole document we get rid of the errors
[10:33:59 EST(-0500)] <clown> good point. we have to get rid of the undefined globals.
[10:34:21 EST(-0500)] * EricDalquist (n=EricDalq@adsl-76-204-102-88.dsl.mdsnwi.sbcglobal.net) has joined #fluid-work
[11:13:07 EST(-0500)] <clown> got to go off line for a bit.
[11:13:13 EST(-0500)] * clown (n=clown@user383.megabit.utoronto.ca) has left #fluid-work
[11:28:52 EST(-0500)] * clown (n=clown@user383.megabit.utoronto.ca) has joined #fluid-work
[12:16:39 EST(-0500)] * colinclark (n=colin@142.150.154.101) has joined #fluid-work
[13:26:48 EST(-0500)] * colinclark (n=colin@142.150.154.101) has joined #fluid-work
[13:40:02 EST(-0500)] * michelled (n=team@142.150.154.199) has joined #fluid-work
[15:06:51 EST(-0500)] <colinclark> clown: Did you work on findFirstOrderableSiblingInColumn ()?
[15:07:06 EST(-0500)] <clown> yes.
[15:07:21 EST(-0500)] <colinclark> I was just doing a bit of code review. I noticed something that I wanted to ask you about.
[15:07:28 EST(-0500)] <clown> okay.
[15:07:39 EST(-0500)] <colinclark> topMostOrderableSibling = jQuery ("#" + id)[0];
[15:07:42 EST(-0500)] <colinclark> Do you think this is safe?
[15:08:09 EST(-0500)] <colinclark> It assumes that the id doesn't contain any characters that are also used in CSS selectors.
[15:08:15 EST(-0500)] <colinclark> eg. dots or colons, etc.
[15:09:02 EST(-0500)] <colinclark> I've seen us use another approach as well: jQuery ("[id=" + someId + "]")
[15:09:11 EST(-0500)] <clown> how does it assume that. i thought that id's could contain dots and colons. maybe my memory is wrong.
[15:09:27 EST(-0500)] <colinclark> They can contain dots and colons, that's the risk.
[15:09:36 EST(-0500)] <clown> what's the risk?
[15:09:50 EST(-0500)] <colinclark> It will interpret those characters as part of the selector.
[15:10:02 EST(-0500)] <colinclark> So for example, a query like this will fail:
[15:10:11 EST(-0500)] <colinclark> <div id="foo.bar" />
[15:10:22 EST(-0500)] <colinclark> jQuery ("#foo.bar")
[15:10:47 EST(-0500)] <colinclark> You have to escape the id.
[15:10:52 EST(-0500)] <clown> why will that fail? or, what is it about jQuery() that will make it fail?
[15:11:02 EST(-0500)] <colinclark> Because . is a valid selector character.
[15:11:12 EST(-0500)] <colinclark> So jQuery interprets that "#foo.bar" selector as saying:
[15:11:36 EST(-0500)] <colinclark> (one sec, just typing.) 
[15:12:07 EST(-0500)] <clown> i thought that the presence of the # was enough to tell jQuery that this is an id. but, i haven't studied it.
[15:12:15 EST(-0500)] <colinclark> "Get me all the descendents of the element with an id of 'foo' which have a class of 'bar'"
[15:12:25 EST(-0500)] <colinclark> Make sense?
[15:12:37 EST(-0500)] <clown> no...
[15:12:48 EST(-0500)] <colinclark> The latter form, jQuery ("[id=" + "foo.bar" + "]") will not fail.
[15:12:48 EST(-0500)] <clown> oh, wait...
[15:13:40 EST(-0500)] <clown> okay. i've seen both forms used in the code, especially the unit test code
[15:13:40 EST(-0500)] * jhung (n=Fluid@142.150.154.178) has joined #fluid-work
[15:14:03 EST(-0500)] <clown> just so i'm clear on it would this work (assumes '\' acts as an escape character):
[15:14:09 EST(-0500)] <colinclark> Yes, exactly.
[15:14:23 EST(-0500)] <colinclark> So unless we can guarantee that an id has already been escaped, it's dangerous to use jQuery ("#foo.bar").
[15:14:24 EST(-0500)] <clown> jQuery ("#foo\.bar")
[15:14:28 EST(-0500)] <colinclark> Yep, that's it.
[15:15:29 EST(-0500)] <colinclark> Though I will admit that the # selector syntax is far prettier than using the full [id=... form.
[15:15:30 EST(-0500)] <colinclark>
[15:15:50 EST(-0500)] <clown> i'll say...
[15:16:31 EST(-0500)] <colinclark> The #id form is used in several markup examples, and a huge amount in the unit tests.
[15:16:38 EST(-0500)] <clown> and the fact that it's between quotes suggests that it's a literal. that is, don't interpret the dot as denoting a sub-type.
[15:17:06 EST(-0500)] <colinclark> Aside from providing a misleading example, it should be benign in unit tests because we create our own id scheme for test markup.
[15:17:20 EST(-0500)] <colinclark> But in real Reorderer code, it will break on any ids that container characters used in CSS selectors.
[15:17:48 EST(-0500)] <clown> i take it we've been lucky so far...
[15:17:54 EST(-0500)] <colinclark> Yep, I think so.
[15:18:13 EST(-0500)] <colinclark> It seems to me that the cost of doing a bit of string concatenation is cheaper than escaping every id that is passed to us.
[15:18:25 EST(-0500)] <colinclark> And less error-prone.
[15:19:11 EST(-0500)] <colinclark> Though not ideal.
[15:20:16 EST(-0500)] <clown> there was string concatenation in the erroneous form, albeit a bit less. not ideal, but what can you do?
[15:24:45 EST(-0500)] <clown> so, are you opening another JIRA for this (replace all "#"id" with "[id="+id"]")?
[15:31:29 EST(-0500)] <clown> anastasiac: have you had any problems with subversion for eclipse and os x leopard?
[15:32:57 EST(-0500)] <clown> i've just discovered that since i upgraded, parts of subversion are missing – the javahl package.
[15:33:36 EST(-0500)] <clown> i've found something at http://downloads.open.collab.net/binaries.html, but it says it's for tiger.
[15:35:49 EST(-0500)] <colinclark> Sorry, we were distracted talking about where to put patches.
[15:35:57 EST(-0500)] <colinclark> Like our recent patch to jQuery.ui.droppable
[15:36:11 EST(-0500)] <anastasiac> clown: no problems, except that the highlighting of words didn't seem to be working.
[15:36:20 EST(-0500)] <anastasiac> I actually just applied updates, and am relaunching
[15:36:58 EST(-0500)] <colinclark> clown: I vaguely remember hearing about Jen Bourey having problems yesterday.
[15:37:16 EST(-0500)] <colinclark> Similar problems with the JavaHL package, I mean. 
[15:37:30 EST(-0500)] <anastasiac> how would problems with javahl manifest themselves?
[15:37:48 EST(-0500)] <colinclark> I think she was having problems with commits not completing, which sounds horrible.
[15:38:01 EST(-0500)] <colinclark> Awhile ago, anastasiac asked why I never commit with Eclipse.
[15:38:09 EST(-0500)] <colinclark> That is my slightly paranoid answer.
[15:39:13 EST(-0500)] <colinclark> clown: Regarding the jQuery id issue, I can file a bug about it. Would you like me to go ahead and fix the four occurrences in the Reorderer.
[15:39:57 EST(-0500)] <clown> colinclark: re fixing the Reorderer – go ahead.
[15:40:42 EST(-0500)] <colinclark> k
[15:41:01 EST(-0500)] <clown> anastasiac: my problems with the JavaHL packages is that I don't got them. The "copy files and settings from existing accounts" did grab them when migrating from Tiger to Leopard.
[15:41:25 EST(-0500)] <clown> that should be DIDN"T grab them
[15:41:26 EST(-0500)] <anastasiac> what do they do? what are you not able to do without them?
[15:41:46 EST(-0500)] <clown> subclispe does not work at all.
[15:41:57 EST(-0500)] <clown> it complains that it cannot find the javahl libraries.
[15:42:05 EST(-0500)] <anastasiac> oh - no, I haven't had such problems.
[15:42:26 EST(-0500)] <anastasiac> It's a clean install here, but I upgraded my machine at home, and I don't recall problems there, either
[15:42:37 EST(-0500)] <anastasiac> (hm... I don't recall having tried, either
[15:42:51 EST(-0500)] <clown> command line subversion is fine. but colinclark has made me nervous now; so perhaps i'll stick with the command line.
[15:43:33 EST(-0500)] <clown> the funny thing is that Eclipse says that the subversion installation is properly configured...until you try to use it.
[15:43:39 EST(-0500)] <colinclark> 
[15:44:00 EST(-0500)] <clown> you tongue looks infected.
[15:44:10 EST(-0500)] <clown> so red...
[15:44:10 EST(-0500)] <colinclark> It is particularly swollen when on IRC.
[15:44:13 EST(-0500)] <colinclark> Strange, eh?
[16:14:55 EST(-0500)] <clown> for posterity: one needs to add a symoblic link into /Library/Extensions/Java from /usr/local/lib as follows:
[16:15:30 EST(-0500)] <clown> sudo ln -s /usr/local/lib/libsvnjavahl-1.jnilib /usr/loacl/lib/libsvnjavahl-1.jnilib
[16:15:47 EST(-0500)] <clown> and:
[16:15:56 EST(-0500)] <clown> sudo ln -s /usr/local/lib/libsvnjavahl.jnilib /usr/local/lib/libsvnjavahl.jnilib
[16:16:24 EST(-0500)] <clown> the subversion installer puts those libraries into /usr/local/lib by default.
[16:16:34 EST(-0500)] <anastasiac> thanks for posting those instructions, clown.
[16:16:47 EST(-0500)] <clown> no probllem.
[16:29:38 EST(-0500)] * michelled (n=michelle@CPE001310472ade-CM0011aefd3ca8.cpe.net.cable.rogers.com) has joined #fluid-work
[16:44:33 EST(-0500)] <michelled> jhung: haven't started the trip home yet?
[16:53:19 EST(-0500)] <jhung> just about to. Leaving here at 5p
[16:53:34 EST(-0500)] <jhung> sorry I have the speakers turned off.
[16:58:37 EST(-0500)] * anastasiac (n=chatzill@142.150.154.149) has left #fluid-work
[17:15:58 EST(-0500)] <colinclark> clown: another minor point that came up when I was code reviewing.
[17:16:09 EST(-0500)] <clown> yes?
[17:16:10 EST(-0500)] <colinclark> I'm wondering if we can standardize on a spelling of "movable."
[17:16:28 EST(-0500)] <colinclark> At least according to Google, it doesn't seem to be a real word.
[17:16:30 EST(-0500)] <clown> how about "orderable"? ;-P
[17:16:36 EST(-0500)] <colinclark> No!
[17:16:38 EST(-0500)] <colinclark>
[17:16:43 EST(-0500)] <clown> why not?
[17:16:57 EST(-0500)] <colinclark> Because it's now loaded terminology.
[17:17:08 EST(-0500)] <colinclark> Plus, in the case of the portlet layout work, it's misleading.
[17:17:23 EST(-0500)] <colinclark> I doubt the user is thinking of this in terms of "reordering."
[17:17:49 EST(-0500)] <colinclark> Movable is very exact.
[17:17:58 EST(-0500)] <colinclark> The question I had was one of spelling.
[17:17:59 EST(-0500)] <clown> how is it misleading? oh... well i doubt the user is all that concerned about the spelling of moveable either.
[17:18:12 EST(-0500)] <colinclark> Yes, but we are.
[17:18:59 EST(-0500)] <colinclark> I'd like to have consistent spelling of movable. Movable without the "e" returned more hits in Google with the define: keyword.
[17:19:23 EST(-0500)] <colinclark> Hardly good English, but it used as a jargon word in several fields without the "e."
[17:19:32 EST(-0500)] <colinclark> Are you averse to spelling it as "movable."
[17:19:33 EST(-0500)] <colinclark> ?
[17:20:25 EST(-0500)] <clown> as opposed to moveable? don't really care. "moveable" seems to be more natural typing, sort of.
[17:20:40 EST(-0500)] <clown> since there is no word "mov".
[17:21:02 EST(-0500)] <colinclark> So you're happy with "movable?"
[17:21:25 EST(-0500)] <clown> i think that's why my fingers tend to type the e. pure speculation. movable is fine (still prefer orderable).
[17:22:15 EST(-0500)] <colinclark> Dictionary.com says they are equivalent. So at least they think it is a real word.
[17:22:32 EST(-0500)] <clown> so, when are we renaming the Reorderer to the Removerer?
[17:22:34 EST(-0500)] <colinclark> Let's let orderable go. I really don't agree. Movable is much more explicit.
[17:22:36 EST(-0500)] <clown>
[17:22:39 EST(-0500)] <colinclark> LOL
[17:22:40 EST(-0500)] <colinclark>
[17:23:10 EST(-0500)] <colinclark> Webster's also concurs. I don't have my Oxford here, which is the real dictionary.
[17:25:42 EST(-0500)] <clown> oddly. the dashboard dictionary gives both.
[17:27:28 EST(-0500)] <colinclark> Yes. I think it's pulling from dictionary.com
[17:27:46 EST(-0500)] <colinclark> So I'm confused about itemAboveBelow (). Does it return the item above or below?
[17:27:57 EST(-0500)] <colinclark> Does it depend on the direction parameter?
[17:28:12 EST(-0500)] <clown> yes. it does both depending on the direction given.
[17:28:34 EST(-0500)] <clown> rather it does one or the other depending on the direction.
[17:28:57 EST(-0500)] <colinclark> Okay, so how do I know which one it does? The directions I pass in seem to be either PREVIOUS or NEXT
[17:29:35 EST(-0500)] <clown> you mentally map PREVIOUS to above, and NEXT to below?
[17:29:49 EST(-0500)] <clown> i toyed with 'adjacent'.
[17:29:58 EST(-0500)] <colinclark> Okay, I see.
[17:30:36 EST(-0500)] <colinclark> I will need more time to think about it, but it looks like the portletLayout object should actually be two separate units.
[17:30:38 EST(-0500)] <clown> as in itemAdjacent(), but that didn't seem to be clear as to which direction the adjacent thing was.
[17:31:11 EST(-0500)] <clown> it could be left, right, above, or below.
[17:31:30 EST(-0500)] <colinclark> Right.
[17:32:07 EST(-0500)] <colinclark> So itemAboveBelow() is only called by one method, get verticalSibling()?
[17:33:42 EST(-0500)] <colinclark> Looks to me like we could break up the layout along the lines of concern with indices.
[17:34:22 EST(-0500)] <colinclark> As I say, I need to look a bit more closely, but I get the feeling that there are a number of primitive functions that are designed to dig into the underlying layout matrix sent by the portal.
[17:34:40 EST(-0500)] <colinclark> And then there are number of more meaningful functions for actually getting stuff.
[17:35:38 EST(-0500)] <michelled> Yes, I think that's correct.
[17:40:17 EST(-0500)] <clown> colinclark: sorry, had to step away for a second. re: itemAboveBelow() is only called from getVerticalSibling(). Until this morning, that code was actually part of getVerticalStibling(). I just moved it into portletLayout. Anyhow, I have to call it day, pretty much now.
[17:42:01 EST(-0500)] <michelled> clown: are you working on getting the portletLayout to generate a findItems?
[17:42:28 EST(-0500)] <colinclark> clown: cool, thanks for the clarification. Let's talk more about portletLayout items tomorrow.
Page Comparison
Manage space
Manage content
Integrations