On Friday, 9. July 2010 18:06:48 Martin Zenzes wrote: > > The fixed size array will not be enough if git is unable to come up > > with a short version of the md5 (f.e. if two md5 sums are close to > > each other). > > Yeah, the fixed size itself if not fixed, depends on the output > generated by the script (which shoulb be a constant length string? At > least, the function could pad/cut the string to a constant length). But > 8^26 seems like a huge number to identify a commit out of a fixed number > of commits (which won't be more that 10000). Collisions won't likely > happen, according to linus...
"Won't likely" will happen ;) I think we could live with a collision as this is only for developer built releases. > > If the patch is for libftdi-1.x only, then yes. > > Ok, but against which repo should I base my patch? As I'm fairly new, > I'm still confused about which one is "the" libftdi... In that case I would suggest to add this new feature only to libftdi 1.x. If someone steps up and writes autoconf support for it, I'll also add it to libftdi 0.x. Cheers, Thomas -- libftdi - see http://www.intra2net.com/en/developer/libftdi for details. To unsubscribe send a mail to [email protected]
