Possible Bug in video.pp (GetVideoMode(aMode);) delivers zero values.

[jc99]

the video.pp from rtl-console is heavily used in fv and it's descendents.

the DetectVideo-Procedure in drivers should detect the current -text-console mode, without changing it.
therefore it's declared as

Code: Pascal  [Select][+][-]

1. procedure DetectVideo;
2. VAR
3.   CurrMode : TVideoMode;
4. begin
5.   { Video.InitVideo; Incompatible with BP
6.     and forces a screen clear which is often a bad thing PM }
7.   GetVideoMode(CurrMode);
8.   ScreenMode:=CurrMode;
9. end;
10.  

But since video isn't initialized, it returns zero-values.
But on video.initVideo the screen is changed.
Is there a way to init video, another (less invasive) way ?

[jc99]

I found a possible workaround:

Code: Pascal  [Select][+][-]

1. procedure DetectVideo;
2. VAR
3.   CurrMode : TVideoMode;
4.   Driver: TVideoDriver;
5. begin
6.   { Video.InitVideo; Incompatible with BP
7.     and forces a screen clear which is often a bad thing PM }
8.   if video.ScreenWidth = 0 then
9.     begin
10.       GetVideoDriver (Driver);
11.       if Assigned(Driver.InitDriver) then
12.         driver.InitDriver;
13.     end;
14.   GetVideoMode(CurrMode);
15.   ScreenMode:=CurrMode;
16. end;
17.  

Now it detects the color and width but still there is a minor issue about the height, (I allways get 63 instead of 25/43 or 50 )

[Edit:]
Maybe it's a bug in FV and the initialization should be done in the initialization-section of the unit.

The best way to solve it, would be that initVideo has a (DoClearScreen=true) parameter. So you have the option whether to clear the screen, or not.