Am I reading this correctly? "manually update..."??? You say it's
possible to send multiple attachments to the feeder. But everytime I or
any one of my hundreds of users choose to do so, I somehow have to know
a) that it occurred, and b) know precisely how many attachments were
sent in, and then go manually update a field in a database table to
reflect that number? I'm obviously missing something. If this is
indeed the case, the answer is, "no... multiple attachments are not
supported".
What's going to happen if I don't keep up with it? There's no way I can
monitor and manually update this with as many users as I have. How
would I know to begin with that someone had sent multiple attachments?
There is no way I'm going to be able to prevent them from sending
multiple attachments. I understand you said small descrepancies in the
numbers are ok. This is quickly going to become NOT small descrepancies.
Just out of curiosity... why can't the feeder simply update the counter
itself on multiple attachments? It does it on single attachments,
doesn't it?
Jerry
Vincenzo Gianferrari Pini wrote:
Hi Bud,
Bud Bach wrote:
This was asked before but I didn't see a response. Is it possible to
send a
message with multiple spam/ham attachments to the spam/ham feeder or
is it
strictly one attachment per message? [james 2.3]
Yes. But in addition to that you would need to manually update the
bayesiananalysis_messagecounts table counters: as the process is to
automatically add 1 to the ham or spam counter, if you feed say 10
attachments you should manually add 9 to the counter.
If the current value of the counters are already high, a small
percentage of "error" in the counters can be ignored (say if you have
already 1000, ending up with 1001 instead of 1010 doen't matter).
Also, It seems like spammers could easily feed their email to the ham
feeder
if steps aren't taken to prevent them from doing so. Is there a
simple way
to prevent people from feeding spam/ham by requiring them to be smtp
authorized or something like that? Thanks. -- Bud
Your concern is perfectly correct. To avoid this I have SMTP AUTH
activated and use a config line like the following:
<mailet match="[EMAIL PROTECTED]"
class="BayesianAnalysisFeeder">
where "internal.use" is any not-existing domain (or anyway a domain
different from my servernames). The same applies to ham.
This way only SMTP AUTHenticated users will be able to feed spam/ham
messages.
Another way is to add a match="SMTPAuthSuccessful" somewhere in the
config.xml logic, but the previous solution IMHO is simpler.
Ciao,
Vincenzo
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
