Make test drill, next steps, call for help with Winetest

[Jakob Eriksson]

Francois Gouget wrote:
> While I prefer the autorun approach because it has fewer dependencies on 
> the Windows side and thus allows me to test in as clean a Windows as 
> desired, your approach could be pretty useful for testing on a real 
> Windows machine. Maybe if you post your script with some instructions it 
> will inspire some people to set up automated testing on their real 
> Windows machines. I'm sure it would make the Direct3D guys happy (if the 
> testers have Nvidia ;-) ).
>
>   
I forgot... I did a C hack. cc -o runstuff.exe runstuff

Then copied runstuff.exe to c:\win95\Start-Menu\Program\Autostart

#include <stdlib.h>
#include <unistd.h>


int main ()
{
    chdir ("/tmp");
    unlink ("winetest-latest.exe");
    system ("wget
http://www.astro.gla.ac.uk/users/paulm/WRT/CrossBuilt/winetest-latest.exe");

    system ("./winetest-latest.exe -q -c -t JakobAuto");

    system ("shutdown -h now");

    return 0;
}


About virtual machines starting in a clean state, nothing stops anyone
from booting linux
in between and "dd if=windows_part.bin of=/dev/hdb1" and then reboot
into Windows.

This way you could get consistent Windows regression testing on real
hardware.


regards,
Jakob