Around 14 o'clock on Nov 3, Jim Gettys wrote:
> Or we could have Xlib generate the error: that is easy, I think... It > could still look as if it were generated as a protocol error. Having Xlib generate the error is actually somewhat hard -- it would have to generate a synthetic request (as a placeholder for the sequence number), then it would need to ensure that the error was inserted at the right place in the result stream. I've figured out how to generate a length error in the server; just send a request who's textElt length is longer than the request length. That's certainly the easiest way to generate a length error. [EMAIL PROTECTED] XFree86 Core Team SuSE, Inc. _______________________________________________ Xpert mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xpert
