On Wed, May 09, 2001 at 02:06:33PM -0700, Ian Clarke wrote:
> On Wed, May 09, 2001 at 08:16:22PM +0100, toad wrote:
> > It is a real problem because Brandon and others want to use highly descriptive
> > human written metadata, for example the Dublin Core stuff includes several fuzzy
> > fields.
>
> Well then Dublin Core or whatever should be treated as a different kind
> of metadata, with a field pointing to the CHK of the data itself, or
> even encourage people to distribute two CHKs pointing to the Dublin Core
> stuff and the metadata.
>
> What it should *NOT* do is be treated as a *replacement* for the current
> metadata mechanism which is working fine thanks very much.
I agree with Ian in my disliking for these ideas. There will always be
different ways in which splitting the data up might help some parts
coalesce more efficiently then if we weren't - trying to make a special
system for "highly descriptive human written metadata" is kludgy and
annoying.
In reality what people are trying to do is simply make a linear splitfile
where the metadata is in the first part. The way we have been doing that
until now is via redirects - I understand that people think the network is
to slow for redirects ATM, but we are better off solving that problem then
instituting (and subsequently getting stuck with) these sorts of kludges.
OTOH, if people want to extends the URI scheme to allow for a splitfile to
encoded directly in a URI then that could be ok (but it would have to be
general, not just split between "metadata" and "data").
--
'DeCSS would be fine. Where is it?'
'Here,' Montag touched his head.
'Ah,' Granger smiled and nodded.
Oskar Sandberg
[EMAIL PROTECTED]
_______________________________________________
Devl mailing list
[EMAIL PROTECTED]
http://lists.freenetproject.org/mailman/listinfo/devl
