[Lazarus] Source and Editor woes

[Hans-Peter Diettrich]

Martin schrieb:
> On 24/09/2011 12:13, Hans-Peter Diettrich wrote:
>> While editing the LCL docs I came across the following nasty things:
>>
>> After an commit the XML file(s) may be found different, so that the 
>> IDE asks to reload the file(s). Did somebody else observe this strange 
>> behaviour?
> 
> The IDE shows a diff, if you select one of the files in the dialog. So 
> you can check what changed.

That turned out to be not really useful. It contained 5600 lines, IIRC, 
impossible to find out what's different and what's the correct version. 
When I saved the editor file under a different name, fc did find only 
two different lines - with no visible difference (whitespace? line 
endings?).

> I have seen this with pascal files, if the contain a $ID, which svn 
> updates.

It also happens in other situations, where SVN is not involved.

BTW, the FPDoc Editor might get into the way, when the sources are 
inspected while editing the docs. In my case this cannot happen, because 
I redirected it to a common doc folder, but every other user may have 
this problem. I didn't dare right now to revert this feature, but what 
will happen when the FPDoc Editor is *also* used to update the docs, 
while the XML files are open in the source editor???


>> Because the lines in the XML files can be very long, I sometimes edit 
>> them in full-screen mode. Unfortunately the editor window is restored 
>> to normal size on many events:
>> - text completion (pops up whenever the cursor rests after an dot)
>> - block folding
>> - (don't remember more)
>>
>> Can this behaviour be cured, somehow?
> 
> Which revision?

Unknown - but multiple revisions show that behaviour.

> Any docking extension active in the ide?

In one case, yes.

> Maybe another package?

I've installed only the minimum set of standard packages.


> I recall having had that once, for a very short time (quite a while 
> ago), I do not think it was my config, it might have been some extra 
> package that was installed, or it got fixed (at least as far as my tests 
> go)

I.e. you can not reproduce the described behaviour?

DoDi