Hello!
I am glad to notice the renewed activity around MODS/RDF! I think it is
important to consider these questions carefully.
In this e-mail, I won't list my comments about the MODSRDF XSLT. Instead,
I consider that a few things still need to be clarified when it comes to
the question of hash vs. slash namespaces and possible redirections. The
namespace related decisions we take today may have important consequences
later on.
I am aware of the advantages of slash namespaces, but if we want to avoid
overloaded URIs and CN, we may need to use redirections. This is where a
couple of questions arise.
The number of new or modified HTTP header fields (whether formally
standard or not) has been (very roughly) a couple per year during the last
twenty years or so. If we want our data to be stored for decades and
centuries, we may wish to try to avoid (or at least be aware of) possible
tight dependencies on specific HTTP headers. If we choose to recommend the
use of redirections, I consider that we should clarify which kinds of
redirections (problably with a "303 See Other" status code if we reach
consensus about that) and how we plan to use them in some detail.
> The disadvantage does not seem to apply since the file is not very large.
I agree. The size will probably never be a big problem. But if we choose
to recommend redirections, what about considering the use of
slash-to-fragment redirections to have a chance to minimize the number of
HTTP requests?
Regards!
SaaĊĦha,
|