Re: Re: Auth_HMAC problems

Jesus, I'm using fully RFC compliant HMAC now, using PEAR::Crypt_HMAC (which is why I want that to be seperate from PEAR::Message, I only need the HMAC implmentation). ob_start() is called in Auth_HMAC::protect() which is not called by the constructor. Reading the code (which has apidoc comments) and the example script should give you enough idea to see what I'm doing. Any questions, please don't hesitate to ask. btw, Jesus, theres now a #PEAR on freenode, come join me (and others) there, no more EFNet woes :D - Davey Jesus M. Castagnetto wrote:

One sugestion. If you are going to use your algorithm, instead of the standard HMAC defined in the RFC, perhaps you should use a different name than 'Auth_HMAC'. IIRC correctly you were using a divide and rearrange algorithm to generate the signature/digest, which was dissimilar to the one described for the HMAC digest algorithm. Had not looked at your code yet, but hopefully you are not calling that ob_start() in the constructor, when the object has not really finish initializing yet. ===== --- Jesus M. Castagnetto (jcastagnetto@yahoo.com) Research: http://metallo.scripps.edu/ Personal: http://www.castagnetto.org/ __________________________________ Do you Yahoo!? The New Yahoo! Search - Faster. Easier. Bingo. http://search.yahoo.com