Reaction to this anti-spam tool I found -NSPAM

[pasdetrois]

I came across this while researching ways to get rid of some persistent spam. Has anyone heard of this and more importantly whether the approach seems viable?

"Nspam signifies ‘No spam’. It is a free spam scanner introduced by the start-up called Arithme Software based in Ranchi, India. Unlike other existing anti-spam filters Nspam does not filter the body or other parts of a particular message. It just looks for Nspam header aka Numerical ID. We all know that spam exists because no restriction whatsoever exists in sending the emails. The problem of spam could probably be solved if the emails are charged but unfortunately it would not be possible because of the emails’ unique evolution. Most importantly, a bot sends spam for most of the times. Nspam has utilized these facts. It authenticates the sender before sending email/s. The user is required to answer a randomly generated short question. After the correct input the Nspam issues 64 character Numerical ID as a valid stamp. It signifies ham or spam free message. Nspam then uploads the Numerical ID to a Central Server maintained at Arithme’ end. Upon receiving the messages the Nspam JUST looks for the Numerical ID. If it finds the ID then after confirmation from the Central Sever it simply directs the messages to the Inbox. If it does not find the ID it directs the messages to in-built customized version of Spambayes Bayesian Filter. This filter has been added to Nspam to facilitate the user. The filter would allow the delivery of the messages without Numerical ID. Therefore, the Bayesian Filter of Nspam needs to be trained before any use. In the case of non-training the Nspam would direct the messages without Numerical ID to the folder of ‘Junk Suspects’. The messages with the Numerical ID would be sent to the Inbox even if the filter is non-trained. The registration of Nspam is mandatory. Nspam is in beta stage. Few functionalities of Nspam like ‘Tell Others’ and ‘Confirm Recipients’ would work in the final release. You may visit http://www.arithme.net for more information."

[Jeff G.]

Nspam appears to be a combination of bad ideas (CR and single point of failure in a foreign country (for those of us not in India)) with one good idea (Bayesian filtering). Sorry, but the bad ideas win.

[pasdetrois]

Nspam appears to be a combination of bad ideas (CR and single point of failure in a foreign country (for those of us not in India)) with one good idea (Bayesian filtering).  Sorry, but the bad ideas win.

34400[/snapback]

Thanks! I was afraid of that...

[Jeff G.]

You're welcome.