From: Marcus Sundberg <mackan@stacken.kth.se>
  To  : ggi-develop@eskimo.com
  Date: Sat, 27 Feb 1999 17:33:14 +0100

Re: Inconsistency in ggiCopyBox

Andreas Oman wrote:
> According to the man page ggiCopyBox() will transfer an area between
> frames when the read frame is not the same as the write frame, but if
> you read the source for the linear targets (or what you're supposed to
> call them) like for instance libggi/default/linear_8/copybox.c it says:
> 
>     obox=(uint8 *)(LIBGGI_CURWRITE(vis))+y*linew+x;
>     nbox=(uint8 *)(LIBGGI_CURWRITE(vis))+ny*linew+nx;
>     for (line=0;line<h;line++,obox+=linew,nbox+=linew)
>         memmove(nbox,obox,w);
> 
> So it always copies from write frame to write frame. Which one is the way
> it's supposed to be? Either way will work for me, but it would be nice
> to get my programs right from the beginning. Copying from read to write
> frame seems better to me though...

IIRC we decided that it would copy from the read frame to the write
frame (and it also seems right to me), and apparently it got
documented but never implemented.

Thanks for pointing this out, I'll take care of fixing it.

Btw, how should we handle this in KGIcon? Recalc the values before
calling the ioctl(), or make KGIcon aware of the read/write frame
concept?

//Marcus
-- 
-------------------------------+------------------------------------
        Marcus Sundberg        | http://www.stacken.kth.se/~mackan
 Royal Institute of Technology |       Phone: +46 707 295404
       Stockholm, Sweden       |   E-Mail: mackan@stacken.kth.se