Console jcr:data download links should not be cached

The jcr:data download links are generated using standard Wicket ResourceLink, the default cache response headers are set (meaning: 1 year cache expiration).
This can result in the browser re-using previously cached content when the link url happens to be the same.
And the latter can easily happen because the backing wicket model hierarchy in the Console (especially) is very generic and depends on which nodes are clicked when...

This became a real annoyance since the introduction of webfiles, so we will need to backport to 3.00 (v10).

Fix is trivial by explicitly setting the cache duration to NONE.
And to ensure existing browser cached content no longer to be used, the link id (as encoded in the url) is trivially changed from "binary-download-link" to "binary-download-lnk".