Basically active-xmlrpc is an implementation of "active" as a set of independent software components, conforming to a set of API's, with all of the components accessible via XML-RPC interface, and located using a component registry.
The front-end can then use the components to gather information and render it. There is a patch for mainstream active to enable it to use active-xmlrpc as a back-end in part of the features.
Front-end written using Zope, which is in a much more advantage stage.
No ratings system yet
Static HTML content is not generated. instead, the existence of httpd-accel (like squid) with ignore-reload option is assumed. long Expires are used, and explicit purging is done using the PURGE method
One of the ideas of Active-xmlrpc was to "choose the right tool for the right job". Currently there are components written in Java, and components written in python
Active-xmlrpc handles file uploads in a non-orthodox way. uploads are not done when publishing. instead, they are uploaded via a small HTTP/FTP server written in python, before or after the publish
Mirroring itself is done by a "mirrorHandler" component(multiple mirrorHandlers can exist, each dealing with different mirror).
The user has ability to trace the file mirroring process
The database tables have support for threaded comments, using two additional fields:- threadid and depth. threadid is the id of the highest parent in the thread (unlike parentid which is the direct parent). depth is the comment thread depth.
Highlights (at CVS)
Please contact idan at idanso.dyndns.org If you wish to get the latest code tarballs
Sourceforge page: http://sourceforge.net/projects/sstorm
All code, except the ZODB exported object(which CVS does not fit to), is available via CVS.
cvs co sstorm-trunk
cvs co active-xmlrpc
cvs co active-xmlrpc-java cvs co active-front-ruby (Yeah, I started a new frontend in ruby).
(Yes, currently part of the code is at sourceforge, and other portion at savannah).-- PseudoPunk - 11 Jun 2002