Jeff Licquia <[EMAIL PROTECTED]> writes:
> > I don't follow the allusion to cascading change requirements.
> >
> > Could someone pose a simple example? Or was the cascade a nightmare?
>
> OK, here's what I was thinking.
>
> Let's imagine something like LaTeX licensed under something like the
> LPPL, and let's also assume that I'm going to hack it.
>
> So, I edit "article.sty". OK, no problem; just rename it to
> "article-hacked.sty".
(You mean, I think, for LaTeX2E that you want to hack article.cls,
which is the LaTeX "article" class.)
I'm dubious about the sanity of bring up here a dependence of "book"
on "article", and, in any event what I describe next is probably not
what I would actually do, but, granted your scenario, TDS and Kpathsea
become relevant.
For my private use as a user, I would hack article.cls without renaming
it in my private texmf tree as ~/texmf/tex/latex/base/article.cls (and
not tell anybody :-) ).
Nothing else needs to be in my tree, and (with default setups) the
change will be automatically found for me as a user without having an
impact on any other user.
(I must then be prepared for breakage if I import a document from
elsewhere in source form.)
It only gets touchy when others become involved. I'm not sure how it
might play "legally" when the desired scope might be, say, a group on
my local platform -- for which a texmf tree could be provided -- and
all group members consent. It becomes enormously serious if I want to
burn a CD with this hacked article.cls in a new GNU/Linux system's
main texmf tree. LPPL certainly should say that I must not call it
article.cls in that context.
-- Bill
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
