Re: [linux-minidisc] saving window geometry

["Thomas Arp" <manner.moe@gmx.de>]

-------- Original-Nachricht --------
> Datum: Sat, 5 Jun 2010 11:11:02 +0200
> Von: Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
> An: Thomas Arp <manner.moe@gmx.de>
> CC: linux-minidisc@lists.fu-berlin.de
> Betreff: Re: [linux-minidisc] saving window geometry

> On Sat, Jun 05, 2010 at 10:44:09AM +0200, Thomas Arp wrote:
> > Hi,
> > i added a small feature to save the window geometry and
> > column width of the himd browser when closed and restores
> > them at next startup.
> 
> Looks good to me.
> 
> Could you re-format the patch? There are some places where you "patch"
> whitespaces like lines 19 and 20 in the patch file:
> 
> @@ -154,7 +154,7 @@ QString QHiMDMainWindow::dumppcm(const QHiMDTrack
> & track, QString file)
>  
>      while(himd_nonmp3stream_read_block(&str, &data, &len, NULL,
>      &status) >= 0)                                          
>      {
> -
> +
> 

O.K., i think we have to check the code for correct line endings.
Maybe it is a bug in msysgit.
These "patches" are present just by creating a new branch.
I checked the code with Textpad showing whitespaces and line 
endings.
There are some blancs in the master branch at these lines which 
are not present on other branches.  

> Also:
> 
> void QHiMDMainWindow::save_window_settings()
> {
>     int i = 0;
> 
>     settings.setValue("geometry", QMainWindow::saveGeometry());
>     settings.setValue("windowState", QMainWindow::saveState());
>     for(;i == 6; i++)
> 
> Shouldn't we rather read the number of columns in the list view rather
> than assuming it's 6? I think using a constant here is more prone to
> errors when changing something.
> 
Good point, i will change this.

> Otherwise I think the patch is fine and I would apply it after you
> fixed the suggested changes or come up with some pretty good arguments
> ;).
> 
> Adrian

Thomas