On Wednesday, February 8, 2006, 11:00:05 AM, Ali wrote: AR> I for one would like to see it incorporated. What Peter's take?
The SNF v2 engine is open source for precisely that purpose, and in general so that the SNF engine can be incorporated into any similar projects. I can't spare the cycles right now to participate, but I'm happy to do what I can to support anyone interested in doing that work. On a similar note, for those who have commercial interests which preclude incorporating SNF in open source projects, we are also prepared to provide commercial licensing for the code. _M PS: I have a prototype multi-threaded engine handler for the SNF2 engine if that will be helpful. The code is targeted to linux (or any gnu) pthread implementations and can be used as either a c++ object or as a c shared library (libSNFMulti.so). I could use a few testers for it. -- Before you ask... the code is not out there yet because I've not yet tested it fully (enough for my liking), I have some new features to implement, and it is intended to be released as part of the upcoming SDK. However, the code does work well in all of the tests I've run so far: On Fedora Core, Intel P4 2.4Ghz, running 4 scanning threads in a tight loop on 1 clean and 1 junk message, the engine scans 349 messages/second while reloading the rulebase once every 10 seconds (or so) in a separate thread. Rulebase reloads are internally verified and do not interrupt scanning processes. This E-Mail came from the Message Sniffer mailing list. For information and (un)subscription instructions go to http://www.sortmonster.com/MessageSniffer/Help/Help.html
