Re: PBKDF2 addition to OpenSSL - Why Not Hash?

hi Stas, On Tue, Jun 12, 2012 at 8:14 PM, Stas Malyshev <smalyshev@sugarcrm.com> wrote: > I disagree. I've always stated small, self-contained features that do > not involve infrastructure changes may be OK, and the RFC explicitly > says so too. Yes, and trivial or self contained and can be discussed on a case by case basis. I do not see a discussion nor a proposal here, only a commit followed by a merge. Even the SSL maintainers were not involved at all. That's not an acceptable way to deal with such things, sorry. > If this particular feature is not having consensus, fine, > let's revert it, but in general saying "5.4 is bug fixes ONLY" is not > correct, self-contained function additions, especially ones that have > good usage potential, may be also approved. Approved if proposed, well thought, well implemented and well tested. With all respects to Scott's work, not the 1st time that it happened. We already have issues in this exact extension because of bad API choices. A little momentum is always necessary to get things done right. Cheers, -- Pierre @pierrejoye | http://blog.thepimp.net | http://www.libgd.org