RE: [Deinterlace-discuss] Problems fixed

[John A. <Jo...@ad...>]

> Rob Muller <rob...@gm...> wrote:
> > When trying to find out what code uses the flawed functions=20
> I stumbled=20
> > upon InitializeSleep(). This code amazes me, what happens if the=20
> > computer is busy for just a few ms when this code is called?
>=20
> There're a few crazy (not-so-scientific) code like this in=20
> DScaler. :) I presume the code was initially put in place as=20
> an=20
> I-don't-know-a-better-way-but-this-will-probably-do-for-now=20
> solution and was never revised.  The SAA713x and CX2388x=20
> counterparts, perhaps followed BT8x8's lead without following up.
Hmmm, haven't noticed this one before...

It does look pretty odd, probably should be a straight conversion to use
the high frequency timer,  I think I recall some issues with the
accuracy of Sleep and i2c but this doesn't look like an ideal way of
solving the problem.

> Maybe Sleep(...) can be changed to just poll=20
> GetTickCount(...) the way it's done in InitializeSleep(...)?
I think the high frequency timer may be better, there can be ups to a
+/- 1 error on each end of the GetTickCount

> (In a similar failure of foresight (rather, from laziness),=20
> I've put in few code for SAA713x that takes certain state=20
> changes by the hardware for granted and polls for them=20
> without a timeout. :))
Probably worth putting at least a TODO: near them.

Cheers

JOhn