View Issue Details
|ID||Project||Category||View Status||Date Submitted||Last Update|
|0000679||GeoSetter||User Interface||public||2010-11-07 09:47||2011-02-06 02:43|
|Product Version||3.3.60 Release|
|Target Version||Fixed in Version|
|Summary||0000679: Ctrl-M doe not show the map|
|Description||Ctrl-M hides the map, but does not bring it back on - the map shows for a second and then disappears.|
Turning on the map by using the menu works fine.
|Tags||No tags attached.|
||Unfortunately I can not reproduce this by now...|
That's strange. I just installed the most recent version on my another computer and it showed the same symptoms. No changes in settings, just installed, lanched, pressed Ctrl-M (map disappeared), Ctrl-M again - the map showed for a fraction of a second and then disappeared again.
The first computer is Win7 64-bit, the second one is XP 32 bit.
||Can you please send me your settings? You can export them to a zip file by using "File | Backup Settings..."|
geosetter_settings.zip (3,792 bytes)
||As I mentioned before, it reproes for me with the default settings. Anyway, the file with settings is attached.|
I still can not reproduce it :-/
What happens after pressing Strg+1 and then using Strg+M? Do you have this problem also with Strg+T, Strg+P, Strg+I?
Ctrl-T, Ctrl-P, Ctrl-I work fine.
Ctrl-1 restores the default layout, so it shows the map. Ctrl-M after that hides the map as expected. Second Ctrl-M brings the map back for a fraction of a second, then the map hides again.
Funny thing, the repro seems to be sporadic. Now one of my computers still shows the symptoms, while another one works as it is supposed to (i.e., the map shows up and does not disappear).
I have no idea why the symptoms disappeared on one of my computers. I will try to play a little in attempt to find out what condition/sequence of events brings GeoSetter to this state.
Sorry, it took me awhile to figure this out. Looks like this issue happens if the mouse cursor is within the map area when the program starts.
Repro steps that worked on both my computers:
1. Start GeoSetter. Press Ctrl-1 to switch to the default layout.
2. Move mouse cursor somewhere in the middle of the map area. Memprize this position.
3. Exit GeoSetter (I use Alt-F4 to make sure that I do not move the mouse cursor).
4. Start GeoSetter again. I use keyboard to make sure that mouse cursor stays where the map area will be. However starting the app and quickly moving mouse cursor to the memorized position also works.
5. Press Ctrl-M twice. Second Ctrl-M should show the problem.
As a side note I noticed that Alt-F4 sometimes does not close the application. I will try to find out the exact repro and open another issue.
||Thank you very much, I can reproduce it now :-) It's the Internet Explorer (the map) which seems to process the key commands also and forwards them to the main windows. So the main window receives the commands twice... I will see what I can do...|
||I hope this is fixed now. Can you perhaps try out http://www.geosetter.de/geosetter_beta.exe|
||Unfortunately the Ctrl-M problem is still there. Alt-F4 problem was fixed though.|
||Is it perhaps possible that you didn't update to the beta 3.3.62? Perhaps you are using 3.3.61? Please check this in the help menu... Otherwise I have no idea at the moment why you still have this problem...|
Nope, just double-checked. 3.3.62 (build 1933).
I noticed though that the repro that I used before does not show the symptoms 100% of the time now. Sometimes things indeed work fine. However 5-10 Ctrl-M's usually bring the problem back.
One time I tried about 20 Ctrl-M's and the problem still did not show up. I clicked on the map and the very next Ctrl-M-Ctrl-M brought the problem back.
It looks like your change partially fixed the problem, but there is still some time window when the issue can happen.
Ok, many thanks for your efforts... As I said it's very strange. The pressed key will be processed twice, by the main form and by the map. I added already a small time window (0,5 second) in which the key will be ignored when it should be processed again.
I have to think about it again ;-)
Without processing the key commands of the map in the main window, Ctrl+A for example would select the map and would make it unusable if the map have the focus. Also the application would not react to other hotkeys anymore...
Do I understand correctly, you process keyboard commands twice - one time because your app gets them in the normal way, and another time - because the hosted IE somehow gets them as well and then forwards them to the host (your app)?
Out of curiosity, can you distinguish these two sources (direct commands from the keyboard vs forwarded commands) and just ignore all or some forwarded commands?
Alternatively, can you completely prevent IE from receiving keyboard commands? It does not look like the map should see any keyboard events at all...
> Do I understand correctly, you process keyboard commands twice
yes, it seems so ;-) I didn't know it before you've created this ticket.
> and just ignore all or some forwarded commands?
yes, that's what I want to do. But it's not so easy to know which commands have to be ignored. Right now, I put each processed action in a list and do not process it again when it will be triggered from the map in a time window < 0,5 seconds and already exists as the last item in the list.
> It does not look like the map should see any keyboard events at all...
But unfortunately then nothing will happen sometimes when pressing Ctrl+M or pressing Ctrl+A in the map...
If you want to, you can try it again: http://www.geosetter.de/geosetter_beta.exe (3.3.63)
I know, the solution is not very good, I've set the time window in which an action can't be processed again when it has been performed by the map window to 1 second now.
I do not see the "Ctrl-M does not show the map" problem anymore. However I see couple of new problems - one minor and another one kinda serious.
The minor problem is that now if the map is not shown, I press Ctrl-M followed by another Ctrl-M and the second one does not hide the map. I guess this is expected considering your 1 sec ignore interval.
The serious problem is as follows. I have a folder with 20-30 big .dng files (each around 15-20 MB), so it takes awhile to generate thumbnails for all of them.
If I manage to press Ctrl-M twice on the program start before all thumbnails get generated, I end up with some thumbnails stuck in the "rolling wheel" state and they never be generated.
|2010-11-07 09:47||arkb||New Issue|
|2010-11-07 14:29||Friedemann||Status||new => assigned|
|2010-11-07 14:29||Friedemann||Assigned To||=> Friedemann|
|2010-11-07 22:24||Friedemann||Note Added: 0001323|
|2010-11-07 23:07||arkb||Note Added: 0001325|
|2010-11-07 23:10||Friedemann||Note Added: 0001326|
|2010-11-07 23:20||Friedemann||Note Added: 0001327|
|2010-11-07 23:43||arkb||File Added: geosetter_settings.zip|
|2010-11-07 23:45||arkb||Note Added: 0001328|
|2010-11-07 23:52||Friedemann||Note Added: 0001329|
|2010-11-08 00:13||arkb||Note Added: 0001330|
|2010-11-08 08:07||arkb||Note Added: 0001334|
|2010-11-08 20:39||Friedemann||Relationship added||has duplicate 0000684|
|2010-11-08 20:41||Friedemann||Note Added: 0001335|
|2010-11-08 22:09||Friedemann||Status||assigned => resolved|
|2010-11-08 22:09||Friedemann||Fixed in Version||=> 3.4.51 beta|
|2010-11-08 22:09||Friedemann||Resolution||open => fixed|
|2010-11-08 23:40||Friedemann||Status||resolved => feedback|
|2010-11-08 23:40||Friedemann||Resolution||fixed => reopened|
|2010-11-09 00:32||Friedemann||Note Added: 0001338|
|2010-11-09 00:32||Friedemann||Status||feedback => resolved|
|2010-11-09 00:32||Friedemann||Resolution||reopened => fixed|
|2010-11-09 02:13||arkb||Note Added: 0001340|
|2010-11-09 02:13||arkb||Status||resolved => feedback|
|2010-11-09 02:13||arkb||Resolution||fixed => reopened|
|2010-11-09 16:17||Friedemann||Note Added: 0001342|
|2010-11-09 17:24||arkb||Note Added: 0001343|
|2010-11-09 17:24||arkb||Note Edited: 0001343|
|2010-11-09 17:24||arkb||Note Edited: 0001343|
|2010-11-09 17:45||Friedemann||Note Added: 0001344|
|2010-11-09 18:14||arkb||Note Added: 0001345|
|2010-11-09 18:25||Friedemann||Note Added: 0001346|
|2010-11-09 23:27||Friedemann||Note Added: 0001347|
|2010-11-10 08:28||arkb||Note Added: 0001349|
|2011-02-06 02:43||Friedemann||Status||feedback => resolved|
|2011-02-06 02:43||Friedemann||Fixed in Version||3.3.62 beta =>|
|2011-02-06 02:43||Friedemann||Resolution||reopened => fixed|