[tex-live] license violation in tetex-texmf-3.0 fixed

[Robin Fairbairns]

george white wrote:

> An author can put any nonsense they want into a license -- it is up to
> the "users" to decide whether they can live with the license.  In
> practice, most will rely on compilations (teTeX, TeX Live, linux
> distros) to make such decisions.  At the risk of repeating myself,
> there is a duty on the part of packagers and testers to check for
> license problems before making a "release".  It is unfortunate that
> license issues take time away from more productive pursuits.

indeed the author can.  however, if he submits it to ctan, he's likely
to be encouraged to do something more sensible.  we try to get the
licence categorisation in the catalogue right ... but as people will
have seen from this thread, i had the licence of cstex wrong (since
cstex contains the curiously-licenced csplain).

the fact is that, with sufficiently capricious authors, and fallible
categorisers and packagers, there will always be licence problems.

> > As David pointed out, you are not allowed by GPL to add any further
> > restrictions. It is definitly unclear what license one has (and is
> > allowed to) use for a derived work. I think this problem is also at the
> > heart of RMS's comment on debian-legal where Thomas asked about this
> > license. It is ok for a free software license to carry a renaming
> > clause. But you cannot do this by adding restrictions on top of GPL. IMO
> > this makes csplain etc non-free.
> 
> More like "non-sense", and in fact more in the category of inviting
> misinterpretation by
> referring to GPL which, as you note, forbids additional restrictions.

it may not be sensible, but it caught both thomas and me, which was
presumably the intent -- to mislead people into thinking it was
actually gpl.

> I assume the intent of restricting the use of certain names is
> intended to reduce "forking", which does make life harder for package
> maintainers who get bug reports against 3rd party "enhancements".  A
> rule that said "no program of macro  can be invoked under the name
> 'epsf' or 'ps2pdf'." so we would have 'dvips-epsf.tex' and
> 'dvipsone-epsf.tex', 'gs-ps2pdf'[.bat], 'WinEDT-ps2pdf'[.bat] etc.
> would significantly reduce the traffic in c.t.tex and c.t.pdf.  In
> practice, licenses are very blunt tools and just as likely to injure
> the tool's user as to perform the task at hand.

the problem is that users quite regularly mistype package names when
reporting problems.

licences _are_ blunt tools, but they're the only tools we have against
the above problem, and against the problem of protecting distributors
against litigious package authors.

sigh.