Re: Getting started

From: Andrew Kuchling <akuchlin_at_CNRI.Reston.Va.US_at_hypermail-project.org>
Date: Thu, 23 Apr 1998 14:02:42 -0400 (EDT)
Message-ID: <13631.32413.742424.444864_at_newcnri.cnri.reston.va.us>


Kent Landfield writes:
>Yep. Sounds like my approach to hypermail. ;)
>
>It is clear that the issues are the same, the only differences is in
>the implementation and selection of supported features.

        Perhaps the best course would be to go for the best of both approaches, then, and turn Hypermail into a C library with a well-defined API for archiving messages. The existing executable would still exist for high-traffic use, since it would just sit on top of the library, but you could also write an extension module (perhaps using SWIG or some similar tool) for your favorite scripting language, and build your own customized archives that way, if raw performance isn't a problem.

        Since you'd want to set mechanism and not policy, the lowest level wouldn't know about Unix mailboxes or parsing date formats or anything like that; instead, you'd have a data type for an archiver, and could pass it the date of this message, Message-ID to which it's a follow-up, etc., and could then get the previous/next message in this thread, previous/next message by date, and other relevant things that you might want to put in the HTML. After the code is finished generating the HTML, the text would be passed to the archiver and actually be written to disk and added to the indexes. Filing messages into multiple archives is then simple, and you can customize the display all you like.

-- 
A.M. Kuchling			http://starship.skyport.net/crew/amk/
<snf> At least it's not a moral. Worse than beginnings, morals. I've got no
time for them. No time at all.
	-- One of the three Fates, in SANDMAN #69: "The Kindly Ones:13"
Received on Thu 23 Apr 1998 08:05:32 PM GMT

This archive was generated by hypermail 2.2.0 : Thu 22 Feb 2007 07:33:49 PM GMT GMT