Hi all, my $ 0.02: once I faced a simillar situation when implementing SSL for askemos.org.
Since the threading system of rscheme and chicken are simillar in a way, one should consider that user level threads are hardly scheduled to multiple processors, while kernel level processes do with ease. Thus delegating encryption into separate subprocesses could be beneficial. Besides a security consideration: the isolation might also prevent possible bugs in either chicken or opelssl/libcrypt to interfere in the other one. best regards /Jörg Am Donnerstag, den 29.09.2005, 10:00 +0000 schrieb Thomas Chust: > Am 29.09.2005, 01:21 Uhr, schrieb Alex Shinn <[EMAIL PROTECTED]>: > > > [...] > > However, the docs[1] are incomplete to the point of being almost > > non-existant, I haven't been able to find simple examples anywhere, > > and the source is huge and dense. But if you understand it I'd be > > happy to help! :) > > [...] > > Hello, > > this is precisely why I have been using cryptlib, which is very well > documented, instead. But cryptlib is infrequently updated, although it has > bugs as well of course, and it is not nearly as widespread as OpenSSL. > > I will just try to dig into OpenSSL deeper to find a nice way of making it > work with CHICKEN. The man-pages have become better since I last browsed > through them, but they still resemble reverse engeneering guidelines more > than real documentation ;) > > When I am programming this binding, I will probably try to make its > interface compatible to the SSL-API offered by PLT Scheme -- so if I'm > having trouble I can still consult the source of MzScheme. > > cu, > Thomas > > > _______________________________________________ > Chicken-users mailing list > [email protected] > http://lists.nongnu.org/mailman/listinfo/chicken-users _______________________________________________ Chicken-users mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/chicken-users
