stvrtg wrote:
On 2/22/07, David Gerard <dgerard(a)gmail.com>
wrote:
The main problem with any form of archiving is that it's the busiest
pages that need it, and the busiest pages tend to
be controversial. So
picking how to archive them is not so much a technical matter as
politics. Date-based archiving is generally tolerable to most.
This doesnt replace date based Ordered archives. Im hoping that redundancy
and plenty of cross linking between Topical archives (TA) and OA will keep
people from disagreeing too much with the subjectivity that will go into
forming readable TA.
Refactoring was encouraged when Wikipedia was much smaller. I remember
trying it on a few pages, and it was a time consuming job. No wonder it
fell into disfavour. Is there a Google-like way in which every click to
or from a particular section of a talk page, and the recency of its
edits could be used as a measure of its importance. The topical archive
could then be ordered automatically according to those measures, with
the most important ones at the top. When an talk page exceeds a
predetermined number of kilobytes the program would put the less
important excess on archive page(s).
As long as this idea depends heavily on manpower input I can't see it
going very far.
Ec