Re: setupx.dll16: Handle the error case for VCPM_VSTATWRITE in vcpUICallbackProc16. (RESEND)

[Alexandre Julliard]

Gerald Pfeifer <gerald(a)pfeifer.com> writes:
> Just realized I did not see a response to this patch.  Based on
> feedback I received for a different patch, I looked at this patch
> again and believe it indeed is easier.
>
> 1. We only set, but never used cbres, so removing it loses nothing.
>  2. With this patch we now set the return value to VCPN_FAIL in this
>     case.  However, this is not the only case in vcpUICallbackProc16;
>     for example, in another case in this function we pass the return
>     value of VCP_UI_CopyStart and that returns VCPN_FAIL all over the
>     place.

That doesn't mean it's appropriate in this case. Unless there is
evidence that a zero return from VCPM_DISKPREPINFO is supposed to cause
an error, it's better not to change the current behavior.

-- 
Alexandre Julliard
julliard(a)winehq.org

setupx.dll16: Handle the error case for VCPM_VSTATWRITE in vcpUICallbackProc16. (RESEND)

[Gerald Pfeifer]

On Mon, 6 Sep 2010, Alexandre Julliard wrote:
> > 1. We only set, but never used cbres, so removing it loses nothing.
> >  2. With this patch we now set the return value to VCPN_FAIL in this
> >     case.  However, this is not the only case in vcpUICallbackProc16;
> >     for example, in another case in this function we pass the return
> >     value of VCP_UI_CopyStart and that returns VCPN_FAIL all over the
> >     place.
> > That doesn't mean it's appropriate in this case. Unless there is
> > evidence that a zero return from VCPM_DISKPREPINFO is supposed to
> > cause an error, it's better not to change the current behavior.

Thanks, Alexandre.  Based on your feedback I just submitted an updated
patch that just simplifies this code, but does not change behavior.

Gerald