Raul, Berin, about 8-10 weeks ago I did a openSSL crypto lib binding / interface to BouncyCastle, then I moved that code over to Juice, currently in my own branch here at my system. Currently I have several Signature and encryption/decryption algos up and running. The speed gains, in particular for Signatures, are very good.
There is no difference from the Java point of view - its all standard JCE interfaces. I took some code from BC to deal with the standard JCE interfaces, keys, key parameters and alike. About 6/8 weeks ago Dims and I started the process to get a certificate from Sun to produce a signed JCE provider Jar. Just yesterday I received the cert and the root cert from Sun. Now I can build a signed Jar and work on the tests, the openSSL thread callbacks (it's not complicated - Just a SMOP (Small Matter Of Programming) ). As soon as I have a somewhat stable version that supports the most important Signature algos I would like to check in the code. Regards, Werner Raul Benito wrote: > I was benchmarking against the C version of Aleksey Sanin > http://www.aleksey.com/xmlsec/, > I have yet test against C++ one, but I will do as soon as possible, > And I will tell you. > > Raul > On 1/15/06, Berin Lautenbach <[EMAIL PROTECTED]> wrote: > >>Raul Benito wrote: >> >>>I'm just hacking arround juice to see how we fare against C libxml sec >>>library. And I can tell you that all in all very well. I invite you to >>>read more in my blog at http://r-bg.com/apache >> >>Hey cool! Is that using latest SVN version, or 1.2.1 of the C++ >>library? If the latter - can you try with latest SVN? I'd be most >>interested. I did a whole lot of optimisation stuff a few months back >>that might change the equation. I'm particularly interested because the >>4K case is a borderline case for me between different implementations of >>node lists. >> >>Cheers, >> Berin >> >> > > > > -- > http://r-bg.com >
