Bill Taroli wrote: > I completely agree with your point. But taken from a different > perspective, this may be one reason to justify that such a product not > be used in production IT environments. The point should *not* be missed > that something so crucial to one's infrastructure -- that you would of > course want to keep up to date -- should *require* updating on a weekly > basis to solve *software* issues.
Oh well, someone had to give me an opening for diversifying :) Two points this brings forward. Firstly, and foremost, it does have to be accepted that Clam is still in pre version one state. Stability in any software can only be achieved after an extended period of updating and testing to make sure most avenues are covered. Things stabilise and level off eventually, but that cannot happen straight away from scratch. Secondly, if something is that crucial to your infrastucture, (and if I've said it once, I've said it a thousand times), you should never have a single point of failure within a system. If you are not running a backup, then whatever comes is only to be expected. This applies to anything, not just AV scanning. > Obviously, keeping signatures up to date is extremely important. But if > software is so buggy that regular code upgrades are required, one > really needs to start wondering why that's the case... is it for > functionality enhancements, or due to quality? Simple answer to that one. Inhouse testing and real world conditions are two completely different beasts. You can throw whatever you have at a system whilst you are testing it internally, but the testing is limited to the environments you can create. Once it hits the real world, theory goes out of the window. There are so many variations to everything, there is no way you can conceive of them all within a test environment. Matt _______________________________________________ http://lurker.clamav.net/list/clamav-users.html
