Chad Seibert wrote:
Well, as I'm writing the proposal, I'm more emphasizing the structure of generic implementations of crypto-like stuff, not the actual implementations of MD5, AES, RSA, etc. The second half of my proposal was on actual implementation, where I discuss using either Crypto++ or Botan, with author permission. I'm in the midst of getting that permission.
As you wish, though I'm not sold on the benefit of reinventing the wheel. I thought Botan and Crypto++ both have interfaces that hide details. Perhaps if it has selectable implementations via strategy (like ASIO) and hides the differences between OpenSSL, GNUTLS, Botan, Crypto++, SChannel etc then it would have value. I'd be concerned about something that's driven by 'because I can' rather than by solving a problem, particularly if you can't necessarily hit tough performance targets. I do know that Jack mentioned he'd like some help fixing the Ajisai code that went into Botan so it doesn't try to do its own IO, and can be used as a passive state machine. That would make it easier to integrate with other software (eg ASIO), and it would be nice if a result could be that ASIO got a Botan SSL integration alongside the OpenSSL one. Extra points for a SASL layer for ASIO too. (It would also be nice if ASIO got support for Solaris event ports, but that's another matter.
_________________________________________________________________ The New Busy is not the too busy. Combine all your e-mail accounts with Hotmail. http://www.windowslive.com/campaign/thenewbusy?tile=multiaccount&ocid=PID28326::T:WLMTAGL:ON:WL:en-US:WM_HMP:042010_4 _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost