On 11/12/2010 12:15 AM, Daniel Kahn Gillmor wrote: > Or am i misunderstanding your suggestion?
Not really, no. I am not trying to tell you what your problem really is or how it ought be solved -- you're the guy who knows the ins and outs of it, after all. :) That said, I will just repeat three well-worn bits of generally applicable advice: 1. Don't optimize code that isn't a bottleneck -- there's nothing wrong with making GnuPG do unnecessary work so long as that part isn't the bottleneck. 2. Don't make assumptions about where your code bottlenecks. Profile it. 3. GnuPG is a very mature project that's had a lot of people hammering on it. Your own code is probably much newer with far fewer people hammering on it. "Potentially falling prey to ... bugs in gpg's code" by making GnuPG be clever about the data pipeline may be safer than making your code be clever about the data pipeline. ... I don't have any answers for how you might approach this, other than what I've already mentioned. Sorry!
smime.p7s
Description: S/MIME Cryptographic Signature
_______________________________________________ Gnupg-users mailing list Gnupg-users@gnupg.org http://lists.gnupg.org/mailman/listinfo/gnupg-users
