I saw this post about the Spam Cube, and I am not sure what to think. In order to get this to work, you connect it in-between your computer (or router for multiple computers) and your DSL/Cable modem.
It seems like it looks at all of the packets that go over ports 110 (POP3) and 143 (IMAP). Then it probably works in one of two ways.
- It looks at each message as it is downloaded, and modifies it if it is spam
- It implements a mail server, and then becomes the client to the specified server. So when you think that you are connecting with an external server, you actually are just connnecting witht he Spam Cube. Then the Spam Cube connects with the external mail server.
- Will this only work for unencrypted connections? Will this work with imaps or pops? (I personally use these so my passwords and the contents of my messages are not sent in clear text)
- They claim it will support up to 4 computers. Do they mean that, or do they mean that will protect 4 email accounts? How do they make sure that only 4 computers are connecting, especially when DHCP servers may change the ip addresses at any time. Also when the computers are behind a router, the Spam Cube will see all of the computers as the sam ip and MAC address.
- If the Spam Cube is connected in between a dsl modem and a router, does it have to be configured with the ip address from the isp, or does it just use some self assigned addresses?
- If this is inspecting all of the packets that go across the network interfaces, what performance impact does this have on other network traffic?
- How is it determining what is spam? Is spamassassin used on this box? Or is it using some dns blackist?
- How does the Spam Cube move spam messages for POP3 accounts? POP3 does not support the notion of server side folders. Maybe all that it is doing is putting a header on the messages, and the mail client needs to perform a rule to move this to the folder. (From the marketing messaging on the web site, it sounds like this shouldn't be necessary.)