[00:01:51] *** tschindl has joined #eclipse-e4
[00:10:50] *** tschindl has quit IRC
[01:04:30] *** rcjsuen has quit IRC
[01:05:07] *** rcjsuen has joined #eclipse-e4
[01:32:42] *** rcjsuen__ has joined #eclipse-e4
[01:36:00] *** rcjsuen has quit IRC
[01:53:14] *** briandealwis has quit IRC
[01:57:35] *** rcjsuen__ has quit IRC
[02:28:41] *** rcjsuen has joined #eclipse-e4
[03:24:13] *** rcjsuen has quit IRC
[03:26:24] *** rcjsuen has joined #eclipse-e4
[03:56:35] *** rcjsuen has quit IRC
[08:33:41] *** tschindl has joined #eclipse-e4
[09:23:17] *** tschindl has quit IRC
[09:43:25] *** tschindl has joined #eclipse-e4
[11:38:59] <magnet> hi there, anyone around? :)
[13:34:24] <rcjsuen_> magnet: hi
[13:34:26] *** rcjsuen_ is now known as rcjsuen
[14:34:45] <magnet> hi rcjsuen
[14:35:17] <magnet> do you know why the PartStackRenderer checks for a NoClose tag on MPlaceHolders instead of just checking isCloseable()?
[14:35:35] <magnet> (the closeable attribute is ignored)
[14:37:49] <magnet> (StackRenderer:840)
[14:48:07] <rcjsuen> magnet: Do you have this method?
[14:48:08] <rcjsuen> 	private boolean isClosable(MPart part) {
[14:48:11] <rcjsuen> 		// if it's a shared part check its current ref
[14:48:13] <rcjsuen> 		if (part.getCurSharedRef() != null) {
[14:48:15] <rcjsuen> 			return !(part.getCurSharedRef().getTags()
[14:48:16] <rcjsuen> 					.contains(IPresentationEngine.NO_CLOSE));
[14:48:19] <rcjsuen> 		}
[14:48:21] <rcjsuen> 		return part.isCloseable();
[14:48:23] <rcjsuen> 	}
[14:48:24] <rcjsuen> My 840 doesn't match your 840.
[14:49:24] <magnet> yeah this one -- i'm using 0.10.0 release
[14:49:48] <rcjsuen> Let me try to roll back time
[14:49:52] <magnet> part.getCursharedRef() is MPlaceholder and has an isCloseable() method
[14:50:03] <magnet> but instead we just check for NoClose tag
[14:50:26] <magnet> in fact isCloseable() is never called on placeholder elements
[14:51:20] <rcjsuen> Looks like my code isn't that far back.
[14:52:14] <magnet> rcjsuen, i have exactly the same method as you
[14:52:17] <rcjsuen> my history is from 2010-06-04
[14:52:34] <magnet> only the line number differs )
[14:52:35] <magnet> :)
[14:52:43] <rcjsuen> oh okay I see what you mean now
[14:52:48] <rcjsuen> yes you're right its eems the placeholder is not being checked
[14:53:06] <rcjsuen> I would say this is a bug.
[14:53:19] <magnet> so would I, but I wanted your opinion on the matter before opening one :)
[14:53:36] <magnet> it's confusing because that makes the default to closeable when it's a placeholder
[14:53:42] <magnet> even if you didn't set it closeable in your model
[14:56:32] <rcjsuen> magnet: Do you have a case where you want a part to not be closeable in one perspective but closeable in another one?
[14:58:08] <magnet> rcjsuen, currently it's not on the part that you have to put the tag, but on the placeholder anyway
[14:58:21] <magnet> so with tags you can already do that :)
[14:58:34] <magnet> I don't need it, no, but that doesn't mean someone wouldnt
[14:58:40] <rcjsuen> oic
[14:58:48] <rcjsuen> It seems like a strange use case but we supported it in 3.x.
[14:58:55] <rcjsuen> So I was wondering whose use case this is :)
[14:59:22] <magnet> i have no problem with it, my sole problem is that we have a model attribute that does something very specific and a tag with exactly the same semantics
[14:59:27] <magnet> i'd rather use the model attribute :)
[15:01:13] <rcjsuen> ah
[15:05:11] <magnet> btw, how do you guys use the context debug bundle? I mean, how do you open it? Show view command?
[15:06:05] <rcjsuen> magnet: Ctrl+3
[15:06:17] <magnet> ok
[15:06:23] <magnet> unfortunately that relies on the compatibility layer :)
[15:06:52] <rcjsuen> I guess you can try just using EPartService to show it.
[15:06:59] <magnet> I added an handler based on Tom's live model editor to make it open when pressing Shift+Alt+F10 even on pure apps
[15:07:00] <rcjsuen> We don't have a command/handler for showing it right nowl
[15:07:12] <magnet> yeah, I could share my code
[15:07:15] <magnet> it's a few lines
[15:08:45] <magnet> oh. well see you in a few minutes maybe in case we're netsplit'd
[15:10:26] *** paulweb515_ has quit IRC
[15:10:27] *** magnet has quit IRC
[15:11:36] *** paulweb515_ has joined #eclipse-e4
[15:22:09] *** magnet has joined #eclipse-e4
[16:00:31] *** briandealwis has joined #eclipse-e4
[16:03:13] <magnet> rcjsuen, btw I also tried adding a disposal listener on the Display in swt.E4Application that removes UiSynchronize from the application context, but the problem still happens and some services aren't happy
[16:03:20] <magnet> !ENTRY org.eclipse.e4.core 4 0 2011-09-08 16:02:25.093
[16:03:20] <magnet> !MESSAGE Unable to process "SelectionServiceImpl.synchService": no actual value was found for the argument "UISynchronize".
[16:03:28] <magnet> when closing.
[16:04:11] <magnet> unless these services are made more tolerant for the absence of an UiSynchronize instance in the context, i don't see other solutions than my patch
[16:56:36] <rcjsuen> ah
[16:57:06] <rcjsuen> magnet: My main concern is just that someone actually asked the synchronizer to do something. But if it's disposed nothing happens, seems kind of weird.
[16:57:34] <magnet> rcjsuen, it's not going to do anything if the display is disposed anyway
[16:57:40] <rcjsuen> Perhaps the async case not so much but if I say "have my thread wait and syncExec this runnable", when syncExec(Runnable) returns the guy would expect that the runnable was run, right?
[17:05:20] <magnet> rcjsuen, true, but it's UI stuff, if UI is gone...
[17:05:27] <magnet> he might expect it :)
[17:05:33] <rcjsuen> yeah, I suppose
[17:06:53] <rcjsuen> let me check something
[17:19:45] *** briandealwis has quit IRC
[17:25:15] *** briandealwis has joined #eclipse-e4
[17:29:15] <tschindl> emoffatt: ping
[17:30:37] <rcjsuen> tschindl: He is talking to Dean right now.
[17:30:39] <rcjsuen> What's up?
[17:31:23] <tschindl> no hurry he only wanted me to tell him once i know his talk of eclipsecon europe is accepted
[17:31:43] <rcjsuen> oic
[17:33:52] *** briandealwis has quit IRC
[17:48:29] <emoffatt> tschindl: pong...
[17:48:54] <tschindl> so we have 3 of our 5 talks accepted
[17:49:04] <tschindl> the 2 of mine and your 2nd talk
[17:49:19] <tschindl> the tutorial and your other talk will not make it into the conference
[17:49:24] <emoffatt> beauty...so I have the 'where are we going and I'm co with one of yours ?
[17:49:33] <tschindl> yep
[17:50:00] <tschindl> i think you'll receive the accept mail in a few days
[17:50:20] <emoffatt> I have to check with the foundation just to *make sure* that that's enough...no free pass...no go
[17:50:42] <tschindl> hm - i'm not sure i short talk is a free pass
[17:50:53] <emoffatt> I think I'm OK and I can likely wrangle it anyways but thanks for the pre-notice, it'll help
[17:51:14] <emoffatt> but it's one short talk and co-presenter on another correct ?
[17:51:21] <tschindl> yes
[17:51:30] <emoffatt> it's whether the 'co' one counts :)
[17:51:56] <emoffatt> in any case knowing where I am will at least allow me to start the conversation...
[17:52:28] <emoffatt> thanks a lot...when are the selections finished (everybody here has to hold off on booking flights.. until we knwo)
[17:53:10] <tschindl> i fact they are but some of the talks have to be shortened from extended to standard
[17:53:29] <tschindl> so the schedule can still change
[17:53:35] <emoffatt> so there's a general email coming soon(ish) ?
[17:53:50] <tschindl> yes for those who are in in any case
[17:54:14] <emoffatt> excellent...do you know if Kim Moir is in ?
[17:54:25] <tschindl> let me see
[17:55:25] <tschindl> Migrating to git is in
[17:56:12] <tschindl> is there another talk from her?
[17:56:24] <emoffatt> and Simon Kaegi ? they're the only other two from here
[17:56:47] <emoffatt> not sure, how long is it ?
[17:57:03] <tschindl> Orion: Coding on the Web is in
[17:57:40] <tschindl> and javascript at Eclipse from simon is also in
[17:58:48] <tschindl> a standard talk will only give you 50% discount according to http://www.eclipsecon.org/europe2011/content/speakerfaqs#terms
[17:59:28] <emoffatt> the question is whether I count as 1 talk or two
[17:59:45] <emoffatt> or are you thinking about Kim ?
[18:00:01] <tschindl> if my 2nd talk would not got changed from a extended to standard could give you my 2nd discount
[18:00:08] <tschindl> no about us
[18:00:28] <tschindl> we have together 3 standard talks == 1.5 discounts
[18:00:57] <emoffatt> don't you get one for being on the selection committee ? :)
[18:01:06] <tschindl> i'm not sure about that
[18:01:12] <tschindl> i'd need to ask
[18:01:24] <tschindl> let me check
[18:01:38] <tschindl> if i get one you are free to use the one from the shared talk :-)
[18:01:59] <emoffatt> sheesh...the amount of work you have to do to review the papers... should be worth something
[18:02:13] <emoffatt> but I guess that's up to the organizers / foundation...
[18:04:51] <emoffatt> let me know what happens...I'll start the conversation at this end with the eclipse foundation folks / McQ (I just need to make sure that the trip is funded before buying the tickets :))
[18:07:14] *** briandealwis has joined #eclipse-e4
[18:15:10] *** briandealwis1 has joined #eclipse-e4
[18:15:10] *** briandealwis has quit IRC
[18:20:30] *** briandealwis has joined #eclipse-e4
[18:20:30] *** briandealwis1 has quit IRC
[18:22:16] *** briandealwis1 has joined #eclipse-e4
[18:22:16] *** briandealwis has quit IRC
[18:23:52] *** briandealwis1 has quit IRC
[18:24:05] *** briandealwis has joined #eclipse-e4
[18:24:56] *** briandealwis has quit IRC
[18:24:58] *** briandealwis1 has joined #eclipse-e4
[18:54:51] <briandealwis1> tschindl: Hi Tom.  I was just trying to load an older App.e4xmi with the model tooling editor, and it fails as my binding table doesn't seem to have a "bindingContextId".  I wondered if the editor could instead have an error bar to display such errors
[18:55:23] <tschindl> you mean similar to the generic emf editor?
[18:55:48] <tschindl> file me a bug and I'll try to implement such a feature
[18:56:38] <briandealwis1> ok
[18:57:04] <briandealwis1> tschindl: was there a bug prefix you preferred?
[18:57:57] <briandealwis1> oh, and the generic EMF editor doesn't like unknown attributes either
[19:00:08] *** tschindl has quit IRC
[19:06:47] <briandealwis1> It's too bad we didn't bump the version number when we changed bindingContextId -> bindingContext
[19:59:33] <rcjsuen> paulweb515_: We squash all visibleWhen expressions for a given menu's id into one large composite AndExpression, correct?
[20:00:17] <paulweb515_> rcjsuen: no ... we do group MMCs for the same menu id with the identical expression
[20:00:27] <paulweb515_> sorry, not same menu id, same locationURI
[20:00:40] <paulweb515_> at least, contiguous ones :-)
[20:00:57] <rcjsuen> hm
[20:01:25] <paulweb515_> when we get a programmatic contribution, we AND all of the service expressions to make one big one at the end
[20:01:57] <paulweb515_> like contributing through a part site will give you activePart & activeWindow
[20:02:09] <paulweb515_> rcjsuen: what part of the code are you looking at?
[20:02:16] <rcjsuen> paulweb515_: https://bugs.eclipse.org/bugs/show_bug.cgi?id=349423#c4
[20:02:28] <rcjsuen> since jdt uses action sets and photran uses its own custom visibleWhen
[20:02:40] <rcjsuen> then i expect the expressions to be diff, and based on what you say should not be merged
[20:03:51] <paulweb515_> rcjsuen: when we build up that, don't we build up a bunch of ContributionRecords?
[20:04:13] <paulweb515_> rcjsuen: I thought we had code in the ContributionRecord to say "that menu is visible if any of the CRs says it's visible"
[20:04:19] <paulweb515_> rcjsuen: in effect, a giant OR
[20:06:20] <rcjsuen> Okay I'll try that paper trail and post my findings on the bug later
[20:06:20] <rcjsuen> thx
[20:43:31] <rcjsuen> paulweb515_: But a CR is a 1-1 mapping to an MMC.
[20:43:43] <rcjsuen> so if one MMC is bad it would alrdy just set the menu to false
[20:48:50] <paulweb515_> rcjsuen: right, but a specific contribution should be visible if any contributing CR says true
[20:51:43] <rcjsuen> I think that part of the logic has problems.
[21:57:41] *** lpereira has joined #eclipse-e4
[22:16:41] <rcjsuen> paulweb515_: Can I programmatically create an OrExp
[22:17:01] <rcjsuen> mm doesn't look like
[22:32:33] <rcjsuen> paulweb515_: I posted an illustrative patch (not meant for delivery) with a brief explanation of the pb https://bugs.eclipse.org/bugs/show_bug.cgi?id=349423#c5
[22:37:42] *** rcjsuen is now known as rcjsuen_
[22:53:48] <briandealwis1> emoffatt: the shell toolbar bug is ~337485
[22:53:48] <Arbalest> Bug 337485 - https://bugs.eclipse.org/bugs/show_bug.cgi?id=337485 - e4 / UI / 0.9 - Macintosh / Mac OS X - NEW / normal / - Assignee: e4.ui-inbox - e4 should use shell toolbar where possible
[23:08:23] *** briandealwis1 has quit IRC
[23:10:10] *** rcjsuen has joined #eclipse-e4
[23:31:30] *** lpereira has quit IRC





top