my 2cts on Personalization, AOL, performance

Hi,
A few things I wanted to respond to. First of all, I agree that personalization is an important feature for certain lists. Especially if your target audience isn't the average computer geek. I need it for two reasons: easy unsubscription and bulletproof bounce detection (or better: reported-as-spam-by-AOL-user-detection). One could argue that users need to be educated better, but sometimes that's almost like striving for world peace. I can agree that mailman might be too slow for huge personalized lists, but my lists aren't bigger than 3k users... I have been running mailman for many years now. In the beginning it was running on an ancient Pentium 75 machine, so performance was a big issue for me (I couldn't even approve more than 5 posts at a time or else the system could crash ;) ) but now I have a 2Ghz machine with fast disks at my disposal so mailman won't have any trouble sending out those 3000 messages...
Speaking of personalization: what is the main reason it hasn't been implemented for digests yet? Simply a lack of time or are there some important design issues which make it difficult? If isn't too difficult I could try to invest some time it myself to try to get it to work (I'm not a python god, but I can manage)
about improving delivery speed: what about implementing LMTP? (http://www.networksorcery.com/enp/rfc/rfc2033.txt)
Regards,
Ricardo.

On Fri, 2004-01-30 at 15:06, Ricardo wrote:
For stupid reasons, Mailman decorates digests when it's composing them, so by the time SMTPDirect.py is sending out personalization, it's too late. Untangling all that is more work than I'd be comfortable adding to Mailman 2.1.
about improving delivery speed: what about implementing LMTP? (http://www.networksorcery.com/enp/rfc/rfc2033.txt)
I'm not sure LMTP is appropriate, though I haven't studied this protocol in detail, so I could be mistaken. Mailman doesn't manage a queue in the sense that LMTP requires; it relies on the MTA for most delivery guarantees.
-Barry

At 9:06 PM +0100 2004/01/30, Ricardo wrote:
about improving delivery speed: what about implementing LMTP? (http://www.networksorcery.com/enp/rfc/rfc2033.txt)
That's for local message delivery, from the MTA to the recipient.
That's not where we're hurting.
-- Brad Knowles, <brad.knowles@skynet.be>
"They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." -Benjamin Franklin, Historical Review of Pennsylvania.
GCS/IT d+(-) s:+(++)>: a C++(+++)$ UMBSHI++++$ P+>++ L+ !E-(---) W+++(--) N+ !w--- O- M++ V PS++(+++) PE- Y+(++) PGP>+++ t+(+++) 5++(+++) X++(+++) R+(+++) tv+(+++) b+(++++) DI+(++++) D+(++) G+(++++) e++>++++ h--- r---(+++)* z(+++)

On Fri, 2004-01-30 at 15:06, Ricardo wrote:
For stupid reasons, Mailman decorates digests when it's composing them, so by the time SMTPDirect.py is sending out personalization, it's too late. Untangling all that is more work than I'd be comfortable adding to Mailman 2.1.
about improving delivery speed: what about implementing LMTP? (http://www.networksorcery.com/enp/rfc/rfc2033.txt)
I'm not sure LMTP is appropriate, though I haven't studied this protocol in detail, so I could be mistaken. Mailman doesn't manage a queue in the sense that LMTP requires; it relies on the MTA for most delivery guarantees.
-Barry

At 9:06 PM +0100 2004/01/30, Ricardo wrote:
about improving delivery speed: what about implementing LMTP? (http://www.networksorcery.com/enp/rfc/rfc2033.txt)
That's for local message delivery, from the MTA to the recipient.
That's not where we're hurting.
-- Brad Knowles, <brad.knowles@skynet.be>
"They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." -Benjamin Franklin, Historical Review of Pennsylvania.
GCS/IT d+(-) s:+(++)>: a C++(+++)$ UMBSHI++++$ P+>++ L+ !E-(---) W+++(--) N+ !w--- O- M++ V PS++(+++) PE- Y+(++) PGP>+++ t+(+++) 5++(+++) X++(+++) R+(+++) tv+(+++) b+(++++) DI+(++++) D+(++) G+(++++) e++>++++ h--- r---(+++)* z(+++)
participants (3)
-
Barry Warsaw
-
Brad Knowles
-
Ricardo