> > On January 9, 2006 11:06 am, Jeremy Kitchen wrote: > > just reject viruses at the front door, and you'll be fine. > > 'client-side' scanning (squirrelmail IS a client, even though it's run > > on a server) is not a 'feature'. Don't think you should do it that way > > just because thunderbird does it. The only reason thunderbird or kmail > > have client-side virus scanning support is because some providers don't > > do their own scanning. > > Re-read your last sentence, then compare how Thunderbird accesses messages > from a POP server compared to how SquirrelMail accesses messages from a > POP server using the built-in Mail Fetch plugin (that completely > by-passes any and all mail servers at the site using SquirrelMail). > There is no functional difference, so why should one client be allowed to > scan messages while another isn't? > > While it's not the most optimal setup, having the option to scan messages > in the mail client should not be frowned upon. If your mail provider > does not scan your incoming messages, then the mail client is a good > place to scan messages. After-all, it's the only place *you*, the > recipient, fully control access to the e-mail message.
One difference is the T-bird client uses client cpu clicks whereas squirrel mail uses server clicks. Unless you can come up with a browser based scanner. 10,000 users all clicking and scanning at the same time seems like a potential problem for the average server. Personally I don't think there's such a thing as being too late to scan for viruses, but I do think if it's going to happen on my servers it's also going to be my processes with customer policy input that does it. dp _______________________________________________ http://lurker.clamav.net/list/clamav-users.html
