Mon Nov 24 10:16:00 2003
"Jared McIntyre" <firstname.lastname@example.org> wrote:
> Chicken of the VNC lately has had issues talking to RealVNC servers.
> I think the problem is in the COTVNC ZRLE decoder for subencoding
> greater than or equal to 130. When it completes running through this
> subencoding for 16 and 24 bit color the variable "data" is pointing
> at a location that is not correct. I've read through the
> documentation for the protocol, and can't see where its going wrong.
> Has something changed in the protocol in the last year or so as it
> seemed to work correctly at one point.
> Here is a slightly modified verion of their code for subencoding >=
> 130. If someone with more knowledge than me about ZRLE would look at
> it and see if they can see the problem, I'd appreciate it. (For
> those who don't know the language, the funny  is the object message
> call in Objective-C, but that shouldn't be the problem)
The ZRLE encoding is the same as it ever was. As far as I'm aware VNC 3.3.7
and 4.0b4 contain correct implementations of both encoding and decoding ZRLE
according to the protocol spec.
I can't see anything obviously wrong in your logic, but then without knowing
more about Objective-C it's difficult to tell.