On Mon, May 03, 2010 at 03:14:22PM +0000, Kai-Martin Knaak wrote: > You don't need a data base for this kind of indirection. Any download > script would do. However, it makes the process depend on stability of > external sources --sources that can change, or go away without any day. > Experience shows that this will happen for one reason or another.
Yup, and in some sense this is even true of free material. When I have put together download scripts, my checklist is: - rollover to a series of URLs - accept uncompressed, .gz, and .bz2 versions transparently - include an sha1sum to confirm you got what was intended - the last URL in the rollover list is a URL I control For the non-redistributable case, the last item is problematic. You still need it, but it has to be somehow not publicly accessible, so it can qualify as a legal backup copy. If the original becomes unavailable, the backup can become the reference copy for a clean-room reimplementation. So this trick becomes a way to defer and prioritize development of truly free models, not eliminate them. - Larry _______________________________________________ geda-user mailing list [email protected] http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
