[00:16:30] <nkomura>	TrevorParscal: all prototypes and deployments are down
[00:16:37] <TrevorParscal>	lookin
[00:19:28] <TrevorParscal>	strangeness
[00:19:30] <TrevorParscal>	poking
[00:37:43] <TrevorParscal>	because of this commit -> http://www.mediawiki.org/wiki/Special:Code/MediaWiki/62857
[00:38:05] <TrevorParscal>	which was a removal of "obsolete" extensions
[00:38:28] <TrevorParscal>	our settings, copied generally from the cluster, were requesting no longer existing extensions
[00:38:42] <TrevorParscal>	I just comented out the lines that asked for them
[17:48:57] <TrevorParscal>	hello people
[17:50:09] <TrevorParscal>	RoanKattouw_away: ping me when you are less away
[17:54:15] <RoanKattouw>	TrevorParscal: Back
[17:54:29] <RoanKattouw>	I tried setting up Saros with gmail.com as my Jabber server but failed
[17:54:29] <TrevorParscal>	hi
[17:54:39] <TrevorParscal>	interesting
[17:54:48] <RoanKattouw>	I also redid the merging of fixes to deployment prototype
[17:54:52] <TrevorParscal>	server: gmail.com
[17:55:11] <TrevorParscal>	username: [user]@gmail.com
[17:55:15] <TrevorParscal>	did you do it like that?
[17:55:46] <RoanKattouw>	No I left off the @gmail.com part, lemme try again
[17:58:02] <RoanKattouw>	Doesn't work either
[17:58:08] <RoanKattouw>	Saros is only offering me "Create new account" though
[17:58:37] <TrevorParscal>	RoanKattouw: hmm
[17:59:23] <TrevorParscal>	you seem to be logging in and out quickly
[17:59:55] <RoanKattouw>	?
[18:00:02] <RoanKattouw>	Oh
[18:00:11] *RoanKattouw 	disables Google chat in gmail window
[18:02:28] <RoanKattouw>	Still no joy
[18:02:36] <TrevorParscal>	well that's no fun
[18:03:06] <RoanKattouw>	And the link to a list of public Jabber servers is broken
[18:03:17] <RoanKattouw>	For me at least
[18:04:24] <TrevorParscal>	well, let's just play with it later
[18:04:39] <RoanKattouw>	OK
[18:04:39] <TrevorParscal>	how are you doing?
[18:04:44] <RoanKattouw>	I'm good
[18:05:09] <RoanKattouw>	I picked up my driver's license today so naturally my mother dragged me into her car and had me drive a little bit :)
[18:05:23] <TrevorParscal>	awesome
[18:05:24] <RoanKattouw>	Right now I'm poking at a signal processing assignment
[18:05:48] <TrevorParscal>	cool
[18:07:40] <TrevorParscal>	we need to make it so newlines don't get inserted at inline tags on paste
[18:07:44] <TrevorParscal>	like <a> tags
[18:07:55] <RoanKattouw>	On paste inside the wikiEditor you mean?
[18:07:57] <TrevorParscal>	if you copy test from a wikipedia paragraph, the text contains articles
[18:07:58] <TrevorParscal>	yes
[18:07:59] <RoanKattouw>	We also need to kill <script>s
[18:08:07] <TrevorParscal>	s/articles/links
[18:08:08] <TrevorParscal>	yes
[18:08:20] <RoanKattouw>	Why the heck would newlines get inserted for inline tags?
[18:08:35] <TrevorParscal>	they do for me in all browser's i've tested
[18:08:53] <RoanKattouw>	Maybe it happens while we strip said tags?
[18:09:06] <TrevorParscal>	I'm poking
[18:09:58] <TrevorParscal>	it only happens when you select part of the paragraph that contains links
[18:10:18] <TrevorParscal>	so what's coming in is not a series of paragraphs, it's just some text nodes with tags between them
[18:11:29] <RoanKattouw>	What happens for multiline selections containing links?
[18:11:39] <TrevorParscal>	they get stripped
[18:11:44] <RoanKattouw>	?
[18:12:00] <TrevorParscal>	it has to do with the way we strip tags
[18:12:18] <TrevorParscal>	we rely on the .text() function to operate on <p> tags
[18:12:47] <TrevorParscal>	but if there's not a surrounding (or series of surrounding) <p> tags for the pasted content, it breaks
[18:13:01] <RoanKattouw>	Aha
[18:13:24] <RoanKattouw>	So the fact that we get "foo<a>whee</a>bar" instead of <p>that stuff</p> kills us?
[18:13:41] <TrevorParscal>	yes
[18:14:26] <RoanKattouw>	Anything against detecting this and just wrapping it in a <p>?
[18:14:42] <RoanKattouw>	That's the lazy man's solution of course, you might find a better way by looking at the code
[18:14:45] <TrevorParscal>	and .text() seems to be stripping out some spaces - looking into it
[18:14:56] <TrevorParscal>	RoanKattouw: seems sane to me
[18:15:24] <TrevorParscal>	it also seems that if you select half of one <p> and several others (or even part of the next) you get a series of <p> tags
[18:15:46] <TrevorParscal>	so you should be able to just check if all root elements are <p>s (or <div>s too, just to be safe)
[18:15:51] <TrevorParscal>	basically block elements
[18:16:05] <RoanKattouw>	.children().not('p').length ?
[18:16:11] <RoanKattouw>	Oh block elements, blah
[18:16:27] <RoanKattouw>	Oh I know the trick for this
[18:16:37] <RoanKattouw>	.children().length < .childNodes.length
[18:16:46] <TrevorParscal>	oooh
[18:16:59] <RoanKattouw>	That'll detect top-level textnodes, because .children() doesn't return those
[18:17:08] <TrevorParscal>	very clever!
[18:26:09] <TrevorParscal>	RoanKattouw: we may need our own .text() function
[18:27:07] <RoanKattouw>	Ugh
[18:49:29] <TrevorParscal>	that's so strange
[18:49:38] <TrevorParscal>	on paste
[18:49:54] <TrevorParscal>	for each foreign element that's got no children but has textnodes
[18:50:09] <TrevorParscal>	I do a $(this).text( $(this).text() );
[18:50:27] <TrevorParscal>	and it fixes the spacing issues of $(selector).text()
[18:50:58] <TrevorParscal>	spacing issues being "something <a>here</a>" becoming "somethinghere"
[18:54:59] <RoanKattouw>	Wait WTF
[18:55:10] <RoanKattouw>	$(this).text( $(this).text() ); that's just ridiculous
[18:55:27] <TrevorParscal>	agreed
[18:56:05] <TrevorParscal>	almost as rediculous as chrome/safari showing a blank line in the iframe (from pasted content) that doesn't exist in the dom
[18:56:09] <TrevorParscal>	the iframe is haunted
[18:57:38] <TrevorParscal>	there's a \n being pasted between the <p> tags
[18:57:42] <TrevorParscal>	wow
[18:59:23] <TrevorParscal>	or, a space ratehr
[18:59:32] <TrevorParscal>	s/ratehr/rather
[19:02:57] <TrevorParscal>	it's a \n which our traverser thing turns into a space
[19:02:59] <TrevorParscal>	ah ha!
[19:19:53] <Ryan_Lane_>	Someone had mentioned looking at extensions that modified the sidebar based on user group. I've rewritten the SidebarEx extension to make it compatible with newer versions of MediaWiki, and work properly when sidebar caching is enabled. I'll check in the code, and do a write-up on mediawiki.org tonight.
[19:20:41] <Ryan_Lane_>	It likely isn't good enough for use without further optimization though. It also needs a hook added to core to work.
[19:29:46] <RoanKattouw>	Ryan_Lane_: Did you commit this to SVN yet?
[19:30:00] <Ryan_Lane_>	RoanKattouw: Not yet. I'm at work. I can't commit from here.
[19:30:06] <RoanKattouw>	OK
[19:30:19] <Ryan_Lane_>	RoanKattouw: I needed this for a work project though. So, I decided to knock it out
[19:30:52] <RoanKattouw>	Cool
[19:31:21] <RoanKattouw>	The reason I'm saying this is cause I'd like to review it, I'm like our code reviewer :)
[19:31:32] <Ryan_Lane_>	yeah. it definitely needs some review :)
[19:33:35] <TrevorParscal>	RoanKattouw: using the traverser, this should be ok right? -> $( t.node ).remove();
[19:33:42] <TrevorParscal>	where t is the traverser
[19:33:56] <TrevorParscal>	is there a better way to do that?
[19:34:35] <TrevorParscal>	i'm trying to remove text nodes with nothing but whitespace in them
[19:35:01] <RoanKattouw>	That's fine provided that you don't call t.next() after that, call it before
[19:35:07] <TrevorParscal>	k
[19:35:11] <RoanKattouw>	You probably understand why
[19:35:41] <TrevorParscal>	looking through the traverser, indeed
[19:36:35] <RoanKattouw>	Even not looking at the traverser you can probably see how getting the node after t.node is difficult when t.node is suddenly nowhere
[19:36:51] <TrevorParscal>	right
[19:47:03] <TrevorParscal>	the traverser is not traversing over the text nodes in the body in IE8
[19:47:10] <TrevorParscal>	any clue why that would happen?
[19:47:25] <RoanKattouw>	WTF
[19:47:39] <RoanKattouw>	How are you calling it?
[19:48:03] <TrevorParscal>	It's the code we added durring that skype call
[19:48:05] <TrevorParscal>	nothing new
[19:48:45] <TrevorParscal>	var t = context.fn.traverser( context.$content );
[19:48:46] <TrevorParscal>	while ( t ) { console.log( 'NodeValue: ' + t.node.nodeValue ) }
[19:49:12] <TrevorParscal>	in IE after paste there are text nodes that contain only spaces (with the code I'm about to commit)
[19:49:23] <TrevorParscal>	and that will not traverse over them
[19:49:30] <TrevorParscal>	so, the iframe is once again haunted :)
[19:54:01] <RoanKattouw>	Huh
[19:54:03] <RoanKattouw>	OK
[19:54:48] <TrevorParscal>	i'm going to commit this code in a min - did you actually write a patch for the <p> encapsulation thing?
[19:55:02] <RoanKattouw>	No
[19:55:06] <TrevorParscal>	i will
[19:55:15] <RoanKattouw>	I'm still working on that homework project thing and keeping an eye on the TV
[19:55:40] <RoanKattouw>	They're showing some stuff about parks and stuff in Vancouver now, apparently Vancouver is also a "bike-friendly" city
[19:55:54] *RoanKattouw 	sees a pattern and decides he likes the West Coast
[19:58:00] <TrevorParscal>	i need to get some food
[19:59:15] <TrevorParscal>	I'm committing some code
[19:59:22] <TrevorParscal>	I gotta get some food really quick
[19:59:24] <TrevorParscal>	brb
[19:59:32] <TrevorParscal>	i did not do the <p> wrap thing yet
[20:01:24] <RoanKattouw>	OK enjoy your lunch
[20:50:35] <TrevorParscal>	hi
[20:51:39] <TrevorParscal>	RoanKattouw: how's it going?
[20:52:20] <RoanKattouw>	I'm watching TV, chatting on IRC and submitting an assignment all at the same time
[20:52:24] <RoanKattouw>	Can't do much better than that :)
[21:02:35] <TrevorParscal>	got the wrapping thing solved a different way
[21:02:38] <TrevorParscal>	slightly
[21:03:37] <RoanKattouw>	OK
[21:08:17] <TrevorParscal>	RoanKattouw: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/62889#c5989
[21:08:21] <TrevorParscal>	fixed now
[21:09:16] <RoanKattouw>	Yeah saw it
[21:09:21] <TrevorParscal>	:)
[21:16:24] <TrevorParscal>	.text() is still being unreliable
[21:19:07] <TrevorParscal>	in chrome, spaces are being collapsed
[21:19:12] <TrevorParscal>	and in IE spaces are being imagined
[21:19:18] <TrevorParscal>	at tag boundries
[21:44:02] <TrevorParscal>	RoanKattouw: you still doing your assignment stuff?
[22:03:20] <RoanKattouw>	No I'm done now
[22:04:09] <RoanKattouw>	I just watched the most certain Dutch gold medal disappear because the guy didn't switch from the inside lane to the outside lane, making two inside turns too many so effectively cutting the corner
[22:04:30] <RoanKattouw>	Which would be bad enough if it weren't for the fact that he wanted to go into the right lane but his coach sent him back
[22:04:39] <RoanKattouw>	:(
[22:06:20] <RoanKattouw>	But I'm ready to do some work in 15 mins or so
[22:07:49] <TrevorParscal>	awesome
[22:08:06] <TrevorParscal>	could use some help when you are ready -  i know you are probably getting sleepy at this point though
[22:29:04] <RoanKattouw>	OK I'm ready
[22:30:48] <RoanKattouw>	TrevorParscal: Ready and waiting :)
[22:32:19] <TrevorParscal>	cool
[22:35:04] <RoanKattouw>	TrevorParscal: So, you needed help?
[22:39:48] <TrevorParscal>	yes
[22:39:56] <TrevorParscal>	I'm comiting something
[22:42:06] <TrevorParscal>	So these paste bugs
[22:42:17] <TrevorParscal>	the bug reports are very messy and confusing
[22:42:50] <TrevorParscal>	the report will start as one thing and evolve to several other things, so every time we close them, we are doing it legitimately but based on out understanding of the outstanding portions of the bug report
[22:43:06] <TrevorParscal>	then other strange sometimes unrelated cases are added - so it's a total mess
[22:44:29] <RoanKattouw>	Yeah
[22:44:59] <TrevorParscal>	So, what I want to do is map out the various places (browsers/other software) that text can be copied, and the various places it can be pasted
[22:45:13] <TrevorParscal>	then start filling in the combinations that work or do not work
[22:45:28] <RoanKattouw>	Rigth
[22:45:49] <RoanKattouw>	We're doing a copypaste compatibility matrix, unbelievable
[22:45:50] <TrevorParscal>	then we can systematically identify what's broken/working and decide what to fix or figure out what can't be fixed
[22:45:57] <TrevorParscal>	yep
[22:46:58] <TrevorParscal>	so, it's about midnight there
[22:47:07] <TrevorParscal>	how much longer are you intending on working
[22:47:43] <RoanKattouw>	Let's say about an hour
[22:47:51] <nkomura>	can we create the matrix directly into the test plan?
[22:47:54] <nkomura>	http://spreadsheets.google.com/ccc?key=0AsH79AXxmoh-dEhRZ0JFVmtaOTFEeVFvano2cTk5ZHc&hl=en
[22:48:15] <RoanKattouw>	TrevorParscal: If you could create that matrix, then split it into 4 parts and assign them to each of us
[22:48:15] <nkomura>	calcey actually can help creating the matrix
[22:48:16] <TrevorParscal>	sure
[22:48:21] <nkomura>	if we provide the framework
[22:48:32] <TrevorParscal>	maybe we should make it it's own doc
[22:48:38] <TrevorParscal>	this has like 15 sheets
[22:48:46] <TrevorParscal>	it's up to you - either way
[22:48:55] <TrevorParscal>	nkomura: you want it in there?
[22:49:09] <nkomura>	i prefer a single document over multiple
[22:49:18] <TrevorParscal>	sounds fine to me
[22:49:28] <TrevorParscal>	I'm setting the matrix up now
[23:01:53] <TrevorParscal>	does this look sane to everyone?
[23:02:28] *nkomura 	is looking
[23:04:34] <TrevorParscal>	RoanKattouw: are you OK with being our IE person for this map?
[23:04:46] <RoanKattouw>	Please no
[23:04:48] <nkomura>	do we care cross browser edit map?
[23:04:49] <TrevorParscal>	ha ha
[23:04:56] <RoanKattouw>	My VM makes my computer go into half-swapdeath
[23:04:59] <TrevorParscal>	nkomura: ???
[23:05:07] <TrevorParscal>	I will take the majority of IE stuff
[23:05:15] <RoanKattouw>	OK
[23:05:22] <nkomura>	e.g. IE8/EDIT  -> SA4/Edit
[23:05:26] <RoanKattouw>	That still leaves the cross entries
[23:05:37] <RoanKattouw>	I for one cannot possibly paste from IE to FF, my VM doesn't let me do that
[23:05:52] <RoanKattouw>	And I agree with Naoko that those entries are mostly useless
[23:06:37] <nkomura>	shall we color code the grid by priority
[23:06:43] <TrevorParscal>	useless?
[23:07:23] <TrevorParscal>	so you are comfortable assuming that they will always be using only one web browser?
[23:07:36] <TrevorParscal>	is that a safe assumption?
[23:07:49] <TrevorParscal>	Most people do not use multiple web browsers
[23:07:57] <TrevorParscal>	afaik
[23:08:23] <nkomura>	for the case for wikiEditor to wikiEditor, i think so
[23:09:41] <TrevorParscal>	The top is Paste _TO_
[23:10:31] <RoanKattouw>	Oh sorry
[23:10:40] <TrevorParscal>	:)
[23:10:44] <TrevorParscal>	we liking the chart now?
[23:11:30] <nkomura>	we still need browser specific matrix for "paste to" too
[23:11:30] <TrevorParscal>	What about the diff between Safari on WIndows and Safari on Mac?
[23:11:41] <TrevorParscal>	nkomura: we do
[23:12:05] <RoanKattouw>	TrevorParscal: Permission to flip the definition of the rows and columns?
[23:12:07] <TrevorParscal>	the chart is assuming - as we just discussed - that you always use the same browser
[23:12:14] <TrevorParscal>	Why would we do that?
[23:12:16] <RoanKattouw>	Oh right
[23:12:18] <RoanKattouw>	Sorry
[23:12:19] <RoanKattouw>	Yeah got it now
[23:12:31] <RoanKattouw>	But then the assignments don't strictly make sense
[23:12:44] <TrevorParscal>	you are testing with a browser
[23:12:46] <RoanKattouw>	OH wait they do
[23:12:49] <RoanKattouw>	Yes I get it now
[23:12:50] <TrevorParscal>	ha ah
[23:12:55] <TrevorParscal>	everyone get it?
[23:13:16] <TrevorParscal>	please ask before making changes to the structure - it's always possible I didin't think about this right, but I'm pretty sure I did
[23:14:29] <RoanKattouw>	I unassigned Chrome for me and gave myself OpenOffice and text editors
[23:14:29] <TrevorParscal>	May as well and keep nimish on webkit browsers
[23:14:45] <TrevorParscal>	I moved nimish to safari and chrome
[23:14:54] <nkomura>	the structure make sense
[23:14:59] <TrevorParscal>	RoanKattouw loves Opera
[23:15:03] <TrevorParscal>	ha ha ha
[23:15:07] <nimish_g>	TrevorParscal: why testing if HTML from web browsers copy and paste into 3rd party things correctly?
[23:15:12] <TrevorParscal>	will adam be back working tomorow?
[23:15:19] <nkomura>	so the column C is catch-all browser for WikiEditor
[23:15:45] <RoanKattouw>	nimish_g: Good point
[23:15:50] <RoanKattouw>	I Xed those cells out
[23:16:02] <TrevorParscal>	in the case of copying HTML from a browser and pasting it to some other thing, it's unimportant
[23:16:07] <TrevorParscal>	except
[23:16:16] <TrevorParscal>	well - that's beyond the scope
[23:16:38] <nkomura>	what do you want to see in Textarea column?
[23:17:03] <RoanKattouw>	I assume that's Trevor greying out cells now?
[23:17:24] <nimish_g>	ya he is. ok this makes sense
[23:17:31] <TrevorParscal>	yes
[23:17:39] <TrevorParscal>	RK = yellow
[23:18:01] <nkomura>	TrevorParscal: i hope adam will be back tomorrow
[23:18:02] <RoanKattouw>	I was gonna take blue but you took it
[23:18:13] <RoanKattouw>	Damn Eclipse is slow when closing
[23:18:22] <TrevorParscal>	TP = ??ber gree
[23:18:54] <TrevorParscal>	nimish_g: is the only one with word
[23:19:07] <nimish_g>	word
[23:19:40] <TrevorParscal>	NG?
[23:19:55] <nkomura>	question:
[23:19:59] <TrevorParscal>	oh.. hmm
[23:20:12] <TrevorParscal>	the color is better
[23:20:18] <TrevorParscal>	NG looks like a result
[23:20:27] <TrevorParscal>	nkomura: ??
[23:20:45] <TrevorParscal>	are we ready to start filling things in then?
[23:20:50] <nkomura>	so when copy paste from a web page, copy/pasting a fraction and copy/pasting an entire page by select all behaves differently
[23:20:56] <nkomura>	how do we want to capture that?
[23:21:02] <RoanKattouw>	Yes I just svn upped and am now refreshing
[23:21:06] <RoanKattouw>	We should just check all of those things
[23:21:11] <TrevorParscal>	nkomura: it shouldn't - that would be failure
[23:21:21] <nkomura>	k
[23:21:33] <TrevorParscal>	I will do a legend
[23:21:40] <nkomura>	i'll take a note in the test case
[23:22:13] <nkomura>	brilliant organization, TrevorParscal
[23:23:14] <RoanKattouw>	I have a problem in pasting HTML -> wikiEditor on FF3
[23:23:30] <RoanKattouw>	When I copy "header text\nsome more text" it pastes with an extra newline at the end
[23:23:58] <RoanKattouw>	And I just got a weird cursor
[23:24:45] <RoanKattouw>	TrevorParscal: Big bug: pasting literal "<p>" converts it to a newline
[23:25:18] <TrevorParscal>	hold up
[23:25:22] <TrevorParscal>	before we start testing
[23:25:26] <TrevorParscal>	can we all please svn up
[23:25:30] <RoanKattouw>	I did
[23:26:26] <TrevorParscal>	then what do you mean literal <p> converts, it doesn't do that for me at all....
[23:27:24] <RoanKattouw>	Insert a literal <p> in your article, preview it, copy the <p> in a multiline selection and paste it into the wikiEditor
[23:27:51] <RoanKattouw>	Does not happen on single-line select, granted
[23:28:16] <RoanKattouw>	Hm might need to be <p></p>
[23:28:31] <RoanKattouw>	WTF
[23:28:34] <RoanKattouw>	OK that doesn't happen after all
[23:28:37] *RoanKattouw 	wonders what he was smoking
[23:28:57] *guillom 	points at roan's carpet
[23:31:03] <RoanKattouw>	TrevorParscal: OK now Firefox eats a newline for me
[23:31:11] <RoanKattouw>	Selected "foo\nbar" and inserted it, newline disappeared
[23:32:36] <TrevorParscal>	hld - messing with structure a bit
[23:35:57] <RoanKattouw>	Newline eating also happens for wikiEditor->wikiEditor copying, but that makes sense
[23:37:50] <RoanKattouw>	TrevorParscal: outerParent.childNodes() is not a function --> it's an attribute called .childNodes
[23:37:59] <RoanKattouw>	wikiEditor.js line 441
[23:45:26] <TrevorParscal>	patch commmitted
[23:45:38] <TrevorParscal>	commmmmmmmmmmittttttteddddd
[23:45:44] <RoanKattouw>	TrevorParscal: We have general cursor jumping issues after pasting formatted text (e.g. links)
[23:45:54] <RoanKattouw>	Need to save/restore cursor position like undo does
[23:46:04] <TrevorParscal>	you mind poping that in?
[23:46:26] <TrevorParscal>	RoanKattouw: yes, as we collapse the elements the cursor seems to move
[23:46:50] <RoanKattouw>	To real weird places
[23:47:53] <TrevorParscal>	yeah, it seems to be in relation to the number of tags stripped
[23:51:20] <TrevorParscal>	Should we have a standard text for testing that includes all the possible pitfalls?
[23:52:41] <RoanKattouw>	Maybe
[23:56:46] <RoanKattouw>	Hm, storing the selection based on string offsets is obviously not going to work
[23:57:51] <RoanKattouw>	Or wait, maybe...
[23:58:06] <TrevorParscal>	http://pastebin.com/sAraqMvz
[23:59:51] <RoanKattouw>	TrevorParscal: What I've been using: http://pastebin.com/QK4ixY9d