MBRD ARCHIVING - TECHNOLOGY
the main idea is:
dload ALL the mb list of headers (page by page, as many as they are)
chew the info into straright forward form in a text file , e.g. level;originator;email;subject;date/time
then compare this to the available fail (from previous update), and save it.
dump new-only messages into proper form into a html 
, and fire the html to teleport pro, or any other webdownloader (webcopy, or deep's bot, or gnu's wget)
after that, zip everything into proper file(s)
and you'll need another script to recreate the html-main-list from the database you have
plus i have a script to filter out all the crap from single posts (e.g. reply form, ad's etc)

and that's it!

so.. how-to-use-the-scripts:

dload the main pages
save as + join into one big "html" (simply copy a+b+c+d... res)
extract the prev-ver of the tree
** msgbrd.pl that_file
//produces getonly.htm & updated tree in the upper directory
** web-copier getonly.htm
** msg2.pl mbs*.*   (all the files)
// will filter out the crap and save them as number-of-the-mssg.htm
** zip properly
to recreate, 
** tre2msg.pl mbxxxx.tre,
 and put the result mbxxxx.htm & all the single messages from the zip into one directory. this last script have sort of help, if run without args

all the rest is utilities or previous versions
tre2msg.php is the online php2 ver (altern.org/ephemerides)

the best way to figure out how the above works is to use it ;) 