Google Kreme writes:
On 10 Apr 2005, at 17:55 :12, David R. Linn wrote:
Snicker - nobody uses FTP anymore?
If you want *really* old procmail, I recommend the archives of
comp.sources.misc where you can find, among other things,
"The Procmail mail processing package. (v1.10 1991-02-13)"
Ah, yes, the first version of procmail I used, back on netcom.com.
Actually, I think there was an earlier use on ucscb, but someone
simply wrote a recipe for me to reject posts from an annoying luser.
That was 1988 though, so it wasn't exactly procmail. Or it was
pre-1.0 at least.
According to BuGLess, he started coding procmail "around October,
1990," (he told me that in August 2001 in an e-mail after he stopped
developing procmail,) but I think that may have been the original C
version that he first called "procmail."
Prior to that, /usenet/comp.sources.unix, had tons of shar files
containing sed/egrep script submissions for parsing e-mail headers and
filing e-mail in mbox folders, (maybe still searchable from google.)
John
BTW, BuGLess was quite proud of the scoring/weighting he added in
1994/06/14: v3.00; it added Bayesian Chaining to procmail
scripts. See:
http://www.johncon.com/john/StochasticUCEDetection/
for examples/particulars, which hasn't changed much since
1994. Bayesian Chaining is a very powerful concept, as recently
suggested by Paul Graham:
http://www.paulgraham.com/spam.html
There is a lot of stuff like that in procmail. The multigram program
in SmartList works by comparing n-grams of an e-mail address against a
list of addresses. (Bayesian Chaining and n-grams are the
methodological cornerstones of modern informatics, and there are
rumblings of patent issues developing on both.)
--
John Conover, conover(_at_)rahul(_dot_)net, http://www.johncon.com/
____________________________________________________________
procmail mailing list Procmail homepage: http://www.procmail.org/
procmail(_at_)lists(_dot_)RWTH-Aachen(_dot_)DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/procmail