[Lazarus] Bug 28281

Michael Van Canneyt michael at freepascal.org
Tue Jan 26 16:15:07 CET 2016



On Tue, 26 Jan 2016, Mattias Gaertner wrote:

> On Mon, 25 Jan 2016 21:37:16 +0100 (CET)
> Michael Van Canneyt <michael at freepascal.org> wrote:
>
>>
>> Hi,
>>
>> Any news on bug 29281 ?
>>
>> http://bugs.freepascal.org/view.php?id=28281
>>
>> I can add the following extra info:
>>
>> if compilation fails, it quite often never detects this and fails to show the compiler error.
>> I need to do a recompile, "abort" the previous build, and then it sometimes shows the actual error.
>
> I wonder why it happens only on your platform. Perhaps it is only
> triggered by specific stdout/stderr content length and error IDs. This
> would explain, why a recompile works, where the output is shorter than
> a full compile.

I can reproduce it on 2 machines. One a desktop, one a laptop.
Both run the same Kubuntu version, have more or less the same configuration.

Since my report is marked as related to/duplicate of another, it's not just
me that has this...

>> Also, I get a 100% failure rate when I press ctrl-F9 when the project
>> inspector is the active dialog.
>
> I have no idea, how the project inspector could influence the external
> tools. I hope it is not some memory corruption.

Maybe the problem is keyboard (keyup/keydown/onkey events) related. 
Keydown caught in project inspector, transferred to mainform (or wherever) for processing, 
keyup caught somewhere else, command again executed.
Something like that.

It doesn't happen when I select the compile item from the run menu. 
That would support this theory.

Anyway, I will test tonight with your new define. 
We'll see where this gets us. Thanks for looking into this.

Michael.




More information about the Lazarus mailing list