naming catalogs

[Lauren Wood <lauren@softquad.com>]

On today's phone call, I promised to write something about why I think it's 
a good idea to have a default catalog name. Here goes - feel free to shoot 
holes.

The catalog resolver should, as a fallback when no other catalog is defined, 
look for a file named "xcatalog" (I don't really care about the name) in an 
implementation-defined directory (for example, with the document, in a 
directory where DTDs are commonly found, or some other choice). In 
circumstances where files were sent using MIME, the xcatalog would be 
expected to be one of the files.

Why is this good? Basically, it provides a fallback for the case when the 
system designer hasn't specified where a starting catalog is to be found and 
the user doesn't know enough to figure out how to set one up themselves. The 
document PI that we have incorporated is a help when the author has added 
it, but not if not. Each implementation can define a starting catalog file 
for itself; it seems to me that interoperability is helped if there is a 
defined fallback name for this. People tend to use fallback names as a 
starting point a lot, even if the implementation allows other names, so I 
think the tendency would be for many system designers and authors to use a 
file named "xcatalog" and send it along in email, where other systems would 
easily be able to use the file, rather than having the receiver have to 
figure out which of the sent files was the starting catalog file, and then 
rename it to match whatever their system uses (once they've figured that out).

We want to make it easy to use catalogs; I think having a default name helps 
with this and I can't see the downside to it.

Lauren