| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0028096 | Lazarus | Packages | public | 2015-05-14 22:51 | 2016-01-20 17:55 |
| Reporter | Takeda Matsuki | Assigned To | Ondrej Pokorny | ||
| Priority | normal | Severity | minor | Reproducibility | always |
| Status | resolved | Resolution | fixed | ||
| Platform | x86_64 | OS | Windows | OS Version | 6.3 |
| Product Version | 1.5 (SVN) | Product Build | 49022 | ||
| Target Version | 1.6 | Fixed in Version | 1.7 (SVN) | ||
| Summary | 0028096: AnchorDockingdsgn drive IDE into WM_Size Loop Detected | ||||
| Description | Over a week I was tried to catch where is the spot which make Lazarus got WM_Size Loop detected on my Lazarus.. I check one by one.. And I found the "AnchorDockingdsgn" package was not compatible with some changes on IDE itself.. Stepping to produce : 1. Install the anchordocking package 2. Install the AnchordockingDsgn package 3. After Restart the IDE, Start to dock the Window.. 4. Save the configuration. Close the IDE and Re Run the IDE.. 5. Here we go.. The bug would come.. | ||||
| Tags | No tags attached. | ||||
| Fixed in Revision | 51337,51338,51346 | ||||
| LazTarget | 1.6 | ||||
| Widgetset | Win32/Win64 | ||||
| Attached Files |
| ||||
|
|
Laz Error 20150511.png (17,641 bytes) |
|
|
Dock what window? Please attach your anchordocklayout.xml. |
|
|
I don't understand this error message. Both bounds and new bounds are same , why error message ? |
|
|
anchordocklayout.xml (2,506 bytes) |
|
|
anchordockoptions.xml (117 bytes) |
|
|
@Matias : Done. I already attaching the files, Sir. Please compile clean Lazarus then install anchordocking and anchordockingdsgn. Have you tried it? Edit: After install the packages above, Docking just like what I did, then save the layout, then close the Lazarus. Then Run the Lazarus, booommm... You got it.. |
|
|
The issue persists. (Lazarus: trunk, SVN rev 51304; fpc 2.6.4 or fpc 3.0, OS: Windows 7, 32bit) As described above, compile lazarus with anchordockingdsgn.lp, open lazarus, maximize application, close lazarus, open again: Error. Closing lazarus in an not maximized state will not produce the error. |
|
|
Lazarus error ID28096 2016-01-16.png (13,836 bytes) |
|
|
No booommm here, maximized or not maximized. Try with a clean configuration, for example by using the --pcp option. The configuration format changed when the desktops feature was introduced. Your configuration may be corrupted by some intermediate broken revision. If you still can reproduce with a clean config then explain the exact steps how to reproduce. "After Restart the IDE, Start to dock the Window.." is not accurate enough. [Edit] Actually I get some errors with the Windows version after playing with docking for a while. GTK2 and QT versions under Linux work well without errors always. |
|
|
This one will probably be hard to solve. I once experienced the same issue with clean config. On the good side, I closed the WMSize loop errors (there were about 10) and then Lazarus started OK, although with a corrupted (default) layout. I then ordered the windows the way I wanted, saved the desktop and the errors haven't appeared ever since. |
|
|
Sorry Juha for taking over an issue assigned to you but I wanted to save you time since it seemed you tried to find the source of the problem at the wrong place (0029194). FYI, the loop was caused by CoolBar and Palette initialization in TMainIDEBar.DoShow. Please test. Should it be merged to 1.6? - I don't know, maybe there will be some regressions I am not aware of. There's not much time to test for the 1.6. What do you think? |
|
|
After r51338 lazarus won't start. Just to make sure I deleted everything, even the config folder. You can test it with a plain BIGIDE(r51337 it's ok, I'm on windows). |
|
|
>> After r51338 lazarus won't start. Ohhh yes :) Please test r51346 |
|
|
Now it's ok(the lazarus won't start part). To see if the bug it's gone, I need more time, since it's very sporadic, quite annoying and long lasting bug. :) PS: I will also post it on the forum, so other people can test it. |
|
|
Wow, thanks! I was indeed looking for the problem in a wrong place. > Should it be merged to 1.6? I think it should. It does not look an awfully dangerous change to me. If it caused a regression I think it would show up quickly. In my tests both the undocked and docked IDEs work well. I cannot say about the bug because I could not reproduce it recently. |
|
|
Ok, I added it to the merge list. |
|
|
Wow, I was having the same issue and I just browsed through the issue tracker. I thought it was related to different dual monitor layouts. When my primary monitor was on the right, it happened regularly, when it was on the left, it didn't happen. While looking through issues with "WMSize" in the subject, I came accross issue 0022747, which looks like a duplicate to me. You might want to close it too. |
|
|
>> In my tests both the undocked and docked IDEs work well. I cannot say about the bug because I could not reproduce it recently. It was a Windows-only issue, if I am not mistaken. |
|
|
> It was a Windows-only issue, if I am not mistaken. Yes and I got the error messages under Wine at some point but not any more later. Other users had similar experiences, the error did not happen always and not to everybody. |
|
|
>> Other users had similar experiences, the error did not happen always and not to everybody. The WM_SIZE loop can have different sources. This was only one of them. I think a corrupted layout (e.g. like in 0022747) can cause a WM_SIZE loop as well. Also some IDE extensions can cause it as well (packagetabs had this problem too). Therefore I think this issue is resolved. Other WM_SIZE loops should be reported separately. |
|
|
I wasn't able to reproduce it anymore. Let's hope it's gone for good! |
