[Fwd: Debian TEI delay]

[Syd Bauman]

> > I only want to be paranoid about variants that claim or seem to
> > be from or approved by TEI poisoning the world. Other variants
> > should at least be permitted, if not outright encouraged.
> that [DocBook] copyright would cover this, surely?

While I'm confident something like this could work, no, I'm not sure
the DocBook one would do the trick. What does "label your DTD" mean?
And the permissions paragraph does not require that the modifications
paragraph be retained. (Not necessarily a stumbling block, unless
someone interpreted "modify the DocBook DTD in any way" to mean
modify the declarations, not the comments.)

<p>> > Furthermore, I am worried that including only a copyright notice
> > like the DocBook one fails to warn the user of the negative
> > consequences of direct modification of the schema, as opposed to
> > extending it properly. Of course, perhaps either the consequences
> > are no longer as problematic as they were, and perhaps we no
> > longer care.
> not sure of the worry. If the punter takes tei2.dtd and changes
> it to foo.dtd, removes the "TEI name" from it, and butchers
> it, do we care?

No, not in any copyright or legal sense do we care; as long as they
don't attribute their bad encoding to us, people can use the worst
DTDs in the world all they want.

(Of course, in the altruistic "we want you to have good data" sense
we care, but empathy has no place in this discussion :-)

<p>> what are the bad consequences? that they think they are using the
> TEI?

No, I meant the consequences for the user that come from having an
modified TEI(-like) DTD that was not extended using the indirect
method of a user extensions file, i.e., real difficulty in
ascertaining what, exactly, is different from vanilla TEI and total
pain in the @$$ to upgrade.

I know the various problems direct modification of a flattened DTD
can cause, and conversely the advantages of using proper user
extension files in the P4 DTD world. What I don't understand is
whether or not the same is true in the RelaxNG world.

<p>> but we are the copyright owners. we can reissue the stuff with
> new copyrights as often as we like.

Absolutely. But if there's 1 version out there with a too-permissive
license, that version will be the source of trouble, not the new,
corrected one.

<p>> I hate the idea on principle of paying a lawyer (:-}).

No comment from me, it just bears repeating.

<p>In the big picture I think the GFDL may be the way to go. (That is, I
think perhaps the Board should charge the editors to split the
chapter on conformance as recommended by the FSF, and put the
Guidelines under the GFDL.) However, there is something to be said
for a variant of the DocBook notice, not the least of which we could
probably do it a lot faster.

And where, btw, do you (Sebastian) imagine this DocBook-like notice
going? I'm presuming in every TEI DTD fragment file, and in each
flattened TEI DTD created by the Pizza Chef, no?