"Kelvin D. Olson" <[EMAIL PROTECTED]> writes: > Yeah, it's heady, difficult stuff to work with (I'm a web developer, > not a Linux admin)... but still, it's pretty cool.
Comfort kills. Lack of comfort creates. > Second, an idea (if it hasn't already been proposed): > > In the ~/.tmda/templates/confirm_request.txt file, I can put the > TMDA-tagged confirmation address anywhere I wish to. Reply-to, > Subject, Body... wherever. How terrible would it be to have options > for tmda-filter that instruct it to look beyond the recipient > address for the cookie? > > FILTER_PARSE_BODY = "true" > > FILTER_PARSE_SUBJECT = "true" It's been proposed, and I did consider adding an option to include the confirmation ``cookie'' in the Subject header to parse out instead of in the actual Reply-To address. This would for example, make TMDA easy to use for dialup/POP users who don't have control of their mail server's configuration. The potential increase in users was an attractive incentive. However, I decided against it for a few reasons. First, because it makes the confirmation process much more unreliable and difficult to diagnose. Also because I feel that without tagged addresses, we are back to the typical whitelist system with challenges, which IMO has too many disadvantages/flaws to be usable. This is discussed further at http://tmda.net/faq.cgi?req=show&file=faq01.012.htp So in summary, I decided to sacrifice a potentially larger user base for a more complete and reliable system. That said, there are some similar systems such as Active Spam Killer that do use header parsing instead of tagged addresses (see http://tmda.net/faq.cgi?req=show&file=faq01.009.htp). Though biased of course, I feel TMDA is superior to all of these, and is worth the extra initial effort. _____________________________________________ tmda-users mailing list ([EMAIL PROTECTED]) http://tmda.net/lists/listinfo/tmda-users
