[I've changed the subject line.]
2009/3/11 Lars Aronsson <lars(a)aronsson.se>se>:
If the content is free, people don't need to drink
from our
watertap. It's the water that's important, not the tap. We could
have a minimal webserver to receive new edits. Serving replication
feeds to a handful of media corporations (who might pay for it!)
should be far cheaper than to receive all this web traffic. Some
universities might serve up ad-free mirrors. We could be the
Associated Press instead of the New York Times, the producer
instead of the retailer.
Or is the fact that we spend so much to maintain the 7th most
visited website an admission to the fact that the space between
the copies actually has a great value to us? A value that will be
strengthened by cementing its URL and/or the name Wikipedia
(attributing the project) into the new license?
I'm not against that. I will go with whatever. I'm very flexible
and I still think this is a very fun technical experiment. But I
think the change is worth some consideration.
This is somewhat true. MediaWiki still needs a bloody huge central
database server (or three) and so it has them.
I suppose the place to ask your question is on wikitech-l.
Being able to duplicate the infrastructure is necessary for forking to
be meaningful:
http://davidgerard.co.uk/notes/2007/04/10/disaster-recovery-planning/
I'm not sure anything listed there has meaningfully changed in the
last two years.
- d.