Brion Vibber wrote:
On Apr 20, 2004, at 08:49, Ray Saintonge wrote:
The features that I consider important remain the
same as they have
been at other times when this subject has been raised. An article
must be capable of accepting multiple categorizations in multiple
systems.
A page may be assigned to arbitrarily many categories. Category pages
themselves may be assigned to categories, giving rise to a
hierarchical categorization system (though right now it's a little
awkward to navigate.) Parallel hierarchies shouldn't be a problem.
Perfect.
Plain text
categories are essential, but so too should various
coding systems. My simple suggestion remains that a category which
begins with two capital letters would indicate a code; exactly which
code would be indicated by the specific choice of those two letters.
The category names are opaque to the software; it doesn't have any
interest in what naming conventions are used. Codes are fine, but they
should be sufficiently legible that having a bunch of them appear at
the top of an article won't be confusing.
I think that we agree. I suppose that the issue that I raised may be
more editorial than software related. It is conceivable that at some
distant future time some very popular coding systems should be
recognizable by the software, but that should not become an impediment
to getting a functioning category system activated.
If codes are to be allowed there must be a way of distinguishing them
from plain text, and there must be a way of distinguishing them from
each other. Some people may find the Dewey Decimal System appropriate,
but I would suggest that something in category 537 be coded as "DD:537"
rather than simply "537" . Any coding system would need to be defined
and described
Ec