I screwed up my white list?

[markallen]

What is the best practice if you have been deleting a bunch of spam without training, and the whitelist is now full of 10,000 spam addresses? Should I…

a) delete stuff from the white list
b) reset the corpus
c) do nothing and train all incoming spam from now on

I’m getting a high volume of spam as a result of not understanding it was important to train every spam

[Michael_Tsai]

@markallen

What is the best practice if you have been deleting a bunch of spam without training, and the whitelist is now full of 10,000 spam addresses?

Both (a) and (b), because some of those spam messages are probably recorded as good in the corpus.

@markallen

I’m getting a high volume of spam as a result of not understanding it was important to train every spam

Just to be clear, after the initial training it’s only necessary to train SpamSieve with the spam messages that it didn’t catch.

[markallen]

thanks for the quick reply

So, to be clear, should I clear the entire whitelist, or some portion of it? If some portion, what portion? Just addresses that show 0 hits?

[Michael_Tsai]

@markallen

So, to be clear, should I clear the entire whitelist, or some portion of it? If some portion, what portion?

Either clear the whole thing or go through it and delete all the spammy addresses (many of which will probably have 0 hits, but some of which probably won’t). It would probably be easiest to sort the whitelist by hits and save some of the useful rules that it’s easy to see are valid, but not worry about accidentally deleting some valid rules that don’t have very many hits.