[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Revised LaTeX Project Public License (LPPL)

Scripsit Walter Landry <wlandry@ucsd.edu>
> Jeff Licquia <licquia@debian.org> wrote:

> >      2. If the file is used directly by the Base Format when run, and
> >         the Base Format provides a facility for such files to be
> >         validated as being original parts of The Work, then the file
> >         does not represent itself as being the unmodified original
> >         Work.  This does not imply that the Base Format must provide
> >         such a facility; only that, if such a facility is available,
> >         it must be used in the normal way and it must enable the Base
> >         Format to validate as being modified.  If the Base Format does
> >         not provide such a file validation facility, then the file may
> >         be modified without reference to such a facility.

> I think that this is not good enough.  This sounds a lot like "trusted
> computing".  There are valid reasons to want to run untrusted
> versions.

The point is not to prevent the running of untrusted versions, but to
make sure that the Base Format would emit a warning to the user if it
happens inadvertantly. I belive the plan was that the LaTeX people
would amend their Base Format such that the validation code could be
turned off in an easy and general manner for people who don't care
about whether the code they run is the official original.

(Such turning-off need not entail changes to document source or other
LaTeX source files than the ones that are functionally changed; it
could be achieved quite simply with a short wrapper shell script that
turns validation off before inputting the user's document source).

Also, the entire Debian package with LaTeX in it could be forked
simply by changing, in one place, the Base Format code for validating
(LaTeX) packages to be a no-op. After that is done, the insertion of
formal this-is-not-an-original notices in files that one change is not
a functional change anymore, but rather just a particular kind of
prominent notice that the file has been changed.

> >   b. You must change any identification string in any modified file of
> >      the Derived Work to indicate clearly that the modified file is
> >      not part of The Work in its original form.

> Strings for other programs (think browser id-strings) must be
> modifiable to anything at all.  Strings strictly for human consumption
> can be required to indicate that it is different.

Knowing just a wee bit of what LaTeX uses "identification strings"
for, I think this refers to the "Hi, I'm package X, written by N.N"
notes that go into the human-readable output (and log file). It would
render the whole 5.a.2 business void if the package name in the
\ProvidesPackage header were consideret an "identification string".

I agree that the license draft could conceivable be *applied* in a way
that makes it non-free (for example, if the Base Format's validation
code defaulted to simply refusing to run tampered-with code instead
of emitting warnings and offering the option to squelch these). I do
am not convinced that this is the case for the intended application.

Henning Makholm          "Gå ud i solen eller regnen, smil, køb en ny trøje,
                   slå en sludder af med købmanden, puds dine støvler. Lev!"

Reply to: