[tei-council] directory layout again

[Syd Bauman]

> Sorry, but I disagree. We've suffered from years from having a tag
> called tei.2 -- let's make the naming reflect the reality of the
> situation.

I'm not sure I see that the name of the root element and the name of
the directory that holds the Guidelines/schemas/whatever that deal
with that root element are such that they'd cause the same suffering.

E.g., the language XHTML has both a version 1.0 and a version 1.1 in
separate directories in /usr/share/xml/xhtml/schema/dtd/, but in both
cases the root element is "html".


> There is P4 which we have agreed to freeze and to maintain in
> parallel for a few years, and there is the current version of the
> TEI, which is (as it happens) P5.

While I understand the sentiment, I think it is premature to call
that which will be released as P5 "current".


> It might become P6 one day, but it *is* currently the TEI
> proposals.

I think this, mostly through our sluggishness, is wishful thinking.
After all, that which will eventually be released as P5 still has a
chapter on the structure of its DTDs; still refers to base,
additional, and axillary tagsets all over; still shows users how to
invoke tagsets using DOCTYPE and parameter entity declarations; still
has source data in attributes; does not mention XInclude; still has
chapters on conformance and modification that make no sense anymore;
etc. While I think there are a few things (character set issues,
feature structures, manuscript description, and tagset documentation
jump to mind) where we can point to a chapter of P5 as a reasonable
current recommendation, P5 as a whole cannot, IMHO, enjoy that
privilege.


> So I think that P4 is the "marked case" as linguists say and should
> have a different name. I'm not married to "teiclassic" but it does
> convey what I think we want to say about this other version rather
> well --despite the associations with a well known marketing fiasco!

While this may well be the case in some number of months when P5 does
become the current version, I still don't think naming directories
like this is necessarily a good idea. (Although setting up symlinks
like this might be.) A user's documents that conform TEI P7 still
conform to TEI P7 even after TEI P8 comes out and P7 is considered
obsolete. As a user, I don't want to do a system upgrade and find
that my files are no longer valid and no longer can be properly
processed because TEI changed what they consider current. In some
sense, what's current to me (the user) is what *I'm using*, not what
TEI is recommending.

That said, I realize that such a system if created without symlinks
(or aliases or whatever), in which only the numbered versions are
accessible is a bit unusual. The current version of Emacs is in
"emacs21/"; but "emacs" is symlinked to it. When the installed
version becomes 22, the symlink is updated such that if I do nothing,
I get the new version.

This is not the case with /usr/share/xml/, though, is it? I see
different versions of DocBook, SVG, and XHTML in that directory, but
no symlinks so that there is a static "current" version.


Gotta go ...