In a message dated 12/10/2010 1:31:20 PM Pacific Standard Time, jamesmikedupont@googlemail.com writes:
If we prefer pages that can be cached and translated, and mark the others that cannot, then by natural selection we will in long term replaces the pages that are not allowed to be cached with ones that can be.
My suggestion is for a wikipedia project, something to be supported and run on the toolserver or similar.
I think if you were to propose that we should "prefer" pages that "can be cached and translated" you'd get a firestorm of opposition. The majority of our refs, imho, are still under copyright. This is because the majority of our refs are either web pages created by various authors who do not specify a free license (and therefore under U.S. law automatically enjoy copyright protection). Or they are refs to works which are relatively current, and are cited, for example in Google Books Preview mode, or at Amazon look-inside pages.
I still cannot see any reason why we would want to cache anything like this. You haven't addressed what benefit it gives us, to cache refs. My last question here is not about whether we can or how, but how does it help the project?
How?
W