[01:00:45] *** tschindl has quit IRC
[01:42:31] *** buk has quit IRC
[01:42:38] *** buk has joined #eclipse-e4
[01:42:38] *** buk has joined #eclipse-e4
[01:48:35] *** briandealwis has quit IRC
[03:56:55] *** buk has quit IRC
[03:57:00] *** buk has joined #eclipse-e4
[03:57:15] *** buk has joined #eclipse-e4
[04:51:12] *** rcjsuen has quit IRC
[07:55:27] *** tschindl has joined #eclipse-e4
[08:52:28] *** tschindl has quit IRC
[09:15:13] *** tschindl has joined #eclipse-e4
[09:49:43] *** paulweb515_ has quit IRC
[09:50:42] *** paulweb515_ has joined #eclipse-e4
[11:22:15] *** tschindl has quit IRC
[11:23:46] *** tschindl has joined #eclipse-e4
[13:15:14] *** tschindl has quit IRC
[13:17:47] *** tschindl has joined #eclipse-e4
[14:29:44] *** yarthorn has joined #eclipse-e4
[15:06:59] <rcjsuen_> emoffatt: Not related to min/max addon, dunno if we want a rebuild for this ~357688
[15:07:00] <Arbalest> Bug 357688 - https://bugs.eclipse.org/bugs/show_bug.cgi?id=357688 - Platform / UI / 4.2 - PC / Linux - NEW / normal / - Assignee: gheorghe - Maximizing & restoring views results in missing titles
[15:08:49] <emoffatt> rcjsuen_: Not sure...I'll ask Bogdan once he gets in, depends on the fix I think
[15:09:35] <emoffatt> I'm just starting to verify defect fixes...
[15:10:31] <emoffatt> yarthorn: I have a N&N that I'm preparing for the fade transition on min/max...should be ready in an hour, is that OK ?
[15:11:10] <rcjsuen_> I seem to have a lot of problems.
[15:12:20] *** rcjsuen_ is now known as rcjsuen
[15:30:09] *** buk has left #eclipse-e4
[15:39:00] <rcjsuen> paulweb515_: Do you recall seeing cheese on Linux where the toolbar/tab folder is drawn bad?
[15:54:43] <magnet> hi guys
[15:57:13] <magnet> I could use some advice :). I am porting our 4.0 application to 4.1, and I'm hitting a few easy bugs but one outstanding. We had created a class that binds our widgets (buttons, menu items etc) to commands and used runAndTrack to refresh the enablement state. Since 4.1 any context lookup that happens in handlerService.canExecute() isn't recorded anymore in localComputations, so these are not added to the RAT dependencies and thus my runnable isn'
[15:57:13] <magnet> t called when they change anymore :(
[15:57:42] <magnet> any pointers, aside from introspecting my handlers myself in my class so that I can force the lookup to be "recorded"?
[16:02:33] <rcjsuen> magnet: Don't know. Maybe email the mailing list and maybe Oleg can help you.
[16:03:01] <magnet> rcjsuen, thanks. I'll do it if I can't solve it by myself :)
[16:03:18] <magnet> in my experience solving bugs is quicker than writing a mail though ;)
[16:06:25] *** briandealwis has joined #eclipse-e4
[16:24:32] <magnet> rcjsuen, for the record, I fixed it, but I had to change core.contexts code, so I'll send a mail in any case to know why computation recording is disabled when doing a context get through one-time injections (untracked injectors)
[16:31:55] <magnet> rcjsuen, another bug I hit that you may have heard of (new on 4.1.0), using GTK (linux), sometimes toolbar buttons presses don't do anything. Like 1/10th of the clicks go nowhere. It's kinda hard to debug, but from what I tested, SWT gets the event
[16:31:58] <magnet> heard of it? :)
[16:32:11] <rcjsuen> magnet: No, I don't kno that one.
[16:32:28] <rcjsuen> I think I might've heard it
[16:32:31] <rcjsuen> but no one's opened a bug I don't think
[16:33:25] <magnet> I didn't test with E4.1 SDK, and I'm not a big toolbar user anyway. But it's reproducible on my app, just click 20 times after changing focus and it's bound to happen once or twice
[16:33:57] <magnet> if I find 15 mins I'll open a bug
[16:34:19] <rcjsuen> magnet: What do you mean "changing focus"
[16:34:27] <rcjsuen> magnet: Alt+Tab a few times and then click a tool item 20-30 times?
[16:34:40] <magnet> rcjsuen, alt-tab between each click is more likely to work
[16:34:54] <rcjsuen> I wonder if the tool item might've been disabled
[16:34:56] <rcjsuen> in between the alt+tab
[16:34:58] <magnet> it's kinda hard to know if your click failed if you click too fast
[16:35:14] <magnet> it _does_ happen when my app is in focus though
[16:35:19] <magnet> when another viewpart is in focus
[16:35:22] <rcjsuen> magnet: What should I click on? Something like 'Open Type' that opens a dialog?
[16:35:33] <magnet> and those MToolItems are wired to commands whose handlers  return true
[16:35:37] <rcjsuen> is this a tool bar for a view or the workbench window?
[16:35:49] <magnet> rcjsuen, give focus to project explorer
[16:35:55] <magnet> then perspective switch
[16:35:56] <magnet> for instance
[16:36:04] <magnet> I haven't tested it on Windows
[16:36:09] <magnet> and i'm using a rather old GTK :)
[16:36:19] <rcjsuen> magnet: I have 2.20 here.
[16:36:25] <magnet> still it worked perfectly on 4.0
[16:36:37] <magnet> i'm on gnome 2.28.1
[16:36:41] <magnet> no idea about GTK version
[16:36:49] <rcjsuen> I don't use GNOME :)
[16:37:21] <magnet> I don't either, it's just there ;). I use Gnome Terminal, Xchat, Chromium & Eclipse :p
[16:37:42] <rcjsuen> well, you still have the desktop env :)
[16:38:03] <magnet> it's lying around.. ;)
[16:38:23] <magnet> anyway, that's not a big issue because it's only on the toolbar
[16:38:29] <magnet> (with MToolItems)
[16:38:54] <magnet> and it's only a matter of clicking twice. Most of the users of my app already confuse double & single clicks anyway
[16:38:56] <magnet> :D
[16:40:58] <emoffatt> rcjsuen: I'm going to move any unfixed M2 bugs to M3, any objections ?
[16:41:24] <magnet> hi eric
[16:41:29] <emoffatt> hi
[16:41:41] <rcjsuen> Makes sense to me.
[18:08:45] *** yarthorn has quit IRC
[18:11:16] *** yarthorn_ has joined #eclipse-e4
[18:11:23] *** yarthorn_ is now known as yarthorn
[18:24:31] *** tschindl has quit IRC
[18:38:42] *** yarthorn has quit IRC
[19:01:34] *** briandealwis has quit IRC
[19:46:15] *** yarthorn has joined #eclipse-e4
[19:51:37] <paulweb515_> rcjsuen_: I used to see it, especially if I use minimized views or forget my GDK_NATIVE_WINDOWS=1
[19:53:42] <rcjsuen> paulweb515_: It doesn't seem to fix my problem here. I brought my notebook in for some Linux testing, oh well.
[19:54:35] <paulweb515_> rcjsuen_: I haven't seen it, but I haven't done anything but code browsing recently :-)
[19:54:49] <rcjsuen> it seems to happen for views with 0 tool items
[19:54:52] <rcjsuen> like 'outline' with nothing
[19:54:54] <rcjsuen> or 'Problems' view
[19:55:01] <rcjsuen> i just resize left/right enough and then it'll draw it funny
[19:55:12] <rcjsuen> i don't have net access so can't up the screenshot i'll do it tonight/tmr
[19:55:16] <rcjsuen> @home
[19:57:07] <paulweb515_> rcjsuen_, emoffatt: Is someone talking to the SAP/Search console guys?
[19:57:23] <paulweb515_> they seem to be doing stuff (I'm hoping it's good stuff :-)
[19:57:41] <emoffatt> not that I'm aware of....I did see their email go by
[19:58:50] <emoffatt> perhaps next week after M2 goes out...I'll find the email and forward it to myself to move it back up the stack :)
[19:59:26] <emoffatt> but it's in a specific view as I understood the email...is this just a precursor to kick it around ?
[20:00:45] <rcjsuen> emoffatt: you're aware that saving perspectives is broken for the shared area, right
[20:00:49] <rcjsuen> I forgot if this bug was discussed or not
[20:01:00] <rcjsuen> or if it was discussed (and believed to have since been fixed)
[20:01:09] <paulweb515_> emoffatt: AFAIK it's their prototype implementation of a replacement for the Search view, using their tech
[20:01:17] <paulweb515_> emoffatt: so yes, a kick-around kinda thing
[20:01:57] <emoffatt> rcjsuen: how ? I tried it out a bit but didn't see any issues (except the multi-window one)
[20:02:09] <rcjsuen> yes, I am referring ot the multi window scenario
[20:02:31] <emoffatt> paulweb515_: so it's not a replacement for our 'Quick Access' then ?
[20:02:35] <rcjsuen> well this is more broken then i thot
[20:03:47] <emoffatt> rcjsuen: think we should wire it off unless the window is the same as the one you saved it in ?
[20:03:55] <paulweb515_> emoffatt: that's not what I got from they work, but I hadn't looked into it too deeply before I left.  Boris and Dani were working with them, and I just helped with git/e4 build stuff
[20:04:09] <paulweb515_> Now Boris is gone, I will have to pay more attention ... when I get back :-)
[20:04:17] <rcjsuen> emoffatt: Not sure I understand what you're suggesting here.
[20:04:25] <emoffatt> no prob dude (but I'll be happy when you do :) )
[20:04:28] *** briandealwis has joined #eclipse-e4
[20:05:27] <emoffatt> rcjsuen: the general issue is that the references from MPlaceholders get incorrectly cloned, causing us to 'kidnap' stuff from other windows
[20:06:26] <emoffatt> so one fix would be to disable the command if we can identify that the window we're cloning into isn't the one it was saved from
[20:07:19] <emoffatt> I think it's OK for M2 and there should already be a defect open to make the 'getRef' work correctly (i.e. using some sort of resolver to locate the correct shared element
[20:07:38] <emoffatt> not too many folks are using multiple windows AFAIK
[20:17:06] <rcjsuen> emoffatt: briandealwis uses multiple windows.
[20:17:29] <rcjsuen> But I don't know if he saves perspectives.
[20:17:34] <rcjsuen> Well, he couldn't before anyway ;o
[20:17:35] <briandealwis> I don't
[20:17:48] <briandealwis> ?because of that :)
[20:18:16] <rcjsuen> makes sense
[20:29:36] <emoffatt> briandealwis: well you can now but only if you re-open the perspective in the same window you saved it in... ;(
[20:30:14] <emoffatt> I know that DJ does as well
[20:30:28] <emoffatt> but I'm not sure if he uses save perspective
[20:36:29] <briandealwis> emoffat: I also only run a single perspective per window? so that doesn't help me much.  But it's no different from the current situation, so that's ok :)
[20:38:12] *** lpereira has joined #eclipse-e4
[22:02:16] *** magnet_ has joined #eclipse-e4
[22:14:43] <rcjsuen> emoffatt: dunno if Dean's new code will fix this or not ~357875
[22:14:43] <Arbalest> Bug 357875 - https://bugs.eclipse.org/bugs/show_bug.cgi?id=357875 - Platform / UI / 4.2 - PC / Windows XP - NEW / normal / - Assignee: Platform-UI-Inbox - Saving a perspective with a minimized part stack leaves it behind when closed
[22:22:15] <emoffatt> hmmm, likely not would be my guess....sheesh
[22:22:50] <rcjsuen> emoffatt: It has to do with your persp id tracking code
[22:22:58] <rcjsuen> emoffatt: because if you go back to Java and then switch persp it hides
[22:24:31] <emoffatt> it's even worse than that I think...there seem to be two issues competing...
[22:25:28] <emoffatt> we clone the 'visible = false' value for minimized stacks but not the MToolControl representing it (i.e. we have to grow one on the render)
[22:25:57] <emoffatt> ...or...we do *not* clone the visible false but then we have to remove the tags
[22:26:24] <emoffatt> otherwise the first minimize call will likely end up being a no-op
[22:54:01] *** rcjsuen is now known as rcjsuen_
[23:24:37] *** yarthorn has quit IRC





top