> You don't want to depend on the symlinks, things should work without
> requiring the user to create any symlink.
> It should depend on what
> actual devices are present on the machine,
So, directly stat /dev/ttyS*, then check dosdevices for anything overriding
these?
A C Hurst <[EMAIL PROTECTED]> writes:
> Would using get_dos_device() be better, since that does fall through to
> default behaviour when the
> symlink isn't there?
No, it doesn't take into account existing devices either.
> I chose oldconfig.c since that's where the \DEVICEMAP\Scsi entries get
> > The code in oldconfig.c is called by process.c/__wine_kernel_init(), on
> > server startup.
> > The entries are volatile, and are dynamically created if you count scanning
> wineprefix/dosdevices/
> > and doing a stat() on anything named com[1-9] it finds there every time
> > wine starts up
A C Hurst <[EMAIL PROTECTED]> writes:
> The code in oldconfig.c is called by process.c/__wine_kernel_init(), on
> server startup.
> The entries are volatile, and are dynamically created if you count scanning
> wineprefix/dosdevices/
> and doing a stat() on anything named com[1-9] it finds there
> > Dear winehackers,
> > I sent this patch to wine-patches, with a few iterations of editing,
> > and wondered if I could have some feedback on it, before I just resend
> > it.
> I don't think hardcoding this(as this patch seems to do afaics, I barely
> know the code) is the right approach here. I
> Dear winehackers,
> I sent this patch to wine-patches, with a few iterations of editing,
> and wondered if I could have some feedback on it, before I just resend
> it.
I don't think hardcoding this(as this patch seems to do afaics, I barely
know the code) is the right approach here. I think it sh
