On Thu, Oct 7, 2010 at 4:59 AM, Alex Brollo <alex.brollo(a)gmail.com> wrote:
Special pages, if I understand all their features, are
special why:
# they come from a live API query;
# they cannot be managed/created/edited by users;
# they have no chronology (it would be nonsense).
Special pages are special because they don't have any contents stored
at all, as such. When you view a special page, the software just
executes a particular PHP script, which can be hardcoded to do
absolutely anything. For instance, Special:MyPage just returns an
HTTP redirect, not a page with contents. They don't have a history in
general because they don't have any fixed contents at all -- the
contents are generated on the fly in an arbitrary fashion.
It.source uses many "list pages", daily
updated by a bot, containing other
project-specific queries. They are "normal" pages, and their chronology is
bot useless and heavy. DynamicPageList extension could solve in part such a
useless overload of web space, but its output can't be finely tuned.
Storing many revisions in history is not expensive. Don't worry about
it. There's no way any users, even admins, will ever be permitted to
update pages without leaving any history -- it violates the principle
of reversibility that underlies how wikis work. Is there some problem
you have with these list pages other than the fact that they have lots
of history that no one cares about? If not, this is useful to read:
http://en.wikipedia.org/wiki/Wikipedia:Don't_worry_about_performance