On Wed, 2006-08-30 at 15:32 +0100, Matt Davey wrote: > > I guess we could use 'usb:' under the hood automatically if we get a > usb wakeup and detect that there isn't any visor module loaded. That will probably fail the very first time someone tries to sync on a cold-booted machine, since visor will not have been loaded the very first time. libusb should not be enabled by default, until it is *WELL* tested, and that is the point of making it the secondary option in 0.12.0. This is the first time the users will have ever seen or used libusb in production (other than those who were testing the pre-releases), and since we have no active test cases that will help us fix any obvious issues, we shouldn't just blindly make it the global default. Too many sweeping changes in one launch will royally irritate the users (upstream kernel breakage, visor configuration, new p-l, new g-p, new devices, new application support, new binary names) and now we're going to throw a whole new sync possibility at them? I vote no, until we get some real-world feedback from users who have opted to try it out, but I strongly oppose making it the default in this .0 release. Let's wait until it cools down a little bit through a few point releases before we're sure its solid enough on ALL machines to make it the default. That also gives you a little time to work out a migration path from visor to libusb for those users who wish to go from one to the other (and lets not forget, back again if libusb proves not to work for them. There ARE devices which simply do not sync using libusb, and visor is the ONLY option). -- David A. Desrosiers desrod gnu-designs com http://gnu-designs.com
Attachment:
signature.asc
Description: This is a digitally signed message part