On Friday 16 January 2004 04:16, Alexandre Julliard wrote: > Robert Lunnon <[EMAIL PROTECTED]> writes: > > This is true, but it is not very time consuming. I had considered > > caching the cpuid results which would eliminate the multiple handling but > > didn't feel it was worth the effort. This function does not tend to get > > called often in windows programs (Usually once) but If you like I'll add > > the caching. > > All that extra work is simply because you are adding abstractions > where none are necessary. It looks like a bad case of Not Invented > Here syndrome: you have already submitted that code a number of times, > and I have already explained that it's way too complex for what we > need, so I didn't put it in. Now someone else has done a simple > implementation that works fine, and you are trying to replace it all > once again with your original code that I already rejected. If you > want to improve the cpuid support, please improve the existing code; > as long as you keep pushing your all singing all dancing new > implementation you won't get anywhere.
This is an incorrect charaterisation, but since you hold the keys to the cvs do as you wish. BTW: As a side note to the maintainers of rewind, the rewind project is welcome to use all my patches also. Bob
