On Friday 26 June 2009, Xiaofan Chen wrote:
> On the other hand, it may be easier to create a WinUSB backend for
> OpenOCD which covers the needs for OpenOCD (or libftdi) and
> OpenOCD (or libftdi) only. You may not need to be a Windows driver
> developer to do this.
Correct me if I'm wrong here: currently, the *ENTIRE* reason
to care about the D2XX library is to get simpler Windows support.
If that's so ... wouldn't it make the most sense just to rip out
all the D2XX code, and replace it with WinUSB calls?
On the plus side: no GPL worries, and simpler stories for
distributors and end users. No waiting for "someone" to
fix the holes in libusb on MS-Windows, which have already
been languishing for two years.
On the minus side: lose support for now-obsolete MS-Windows
versions (but why care about older-than-XP?); forgoes the
notion of a 100% identical programming interface (in just
one file, go cry me a river); and some "remove old driver"
stuff may need to be done.
Mmm?
- Dave
p.s. Another "plus". This is probably very doable by one
person, the classic "man on a mission". I've seen
harder things done in less than two weeks, even with
MS-Windows obstacles in the way.
_______________________________________________
Openocd-development mailing list
Openocd-development@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/openocd-development
