[tex-implementors] Re: [tex-live] LM as the default outline font?

Frank Küster frank at kuesterei.ch
Tue Mar 29 22:23:06 CEST 2005

Philip TAYLOR <P.Taylor at Rhul.Ac.Uk> wrote:

> Frank Küster wrote:
> [snip]
>> It doesn't matter much what they say or do:  By convention, software
>> versions use integers (and even letters), not real numbers.
> /Who's/ convention ?  If Staszek et all want to use a real-number
> based numbering convention, then that is their right, as package
> authors.

Of course it is their right.  Which is why I said

| It's not impossible to handle them, but it would be much more convenient
| if lmodern would follow the convention.

> If other software has problems as a result, then that
> software is probably considerably less flexible than it need to be.

I doubt that it would be possible to write such a flexible software.  It
would somehow have to be instructed to use the usual integer/char
convention for software A, but the real number convention for software
B.  This only works for a predefined number of software packages; it
would be doable for a linux distribution (as long as everybody creating
rpms/debs is aware of the problem), but, for example, hardly for a
script for keeping a CTAN copy up-to-date, because there is a quasi
unlimited number of packages there, with unknown versioning scheme.

Or rather it is not, because I have never found any software
deliberately using a real number scheme.  Perhaps the lmodern authors
intended to do so; I don't think it would be a good idea (which is why
we are sending copies of this thread to Stazek).

> It seems to me (and now we're getting /really/ off-topic)
> that what we are seeing here is exactly the same type of
> coercive pressure as is used to "persuade" package authors to
> include some quasi-legal "Free software" licence in any package
> for which they are responsible.  I for one resent that pressure,
> and I similarly resent these attempts to enforce a particular
> versioning scheme on the authors of what is potentially a very
> significant package.

There is certainly such pressure if you care for the group that asks for
the license, or for the versioning scheme.  If you don't, there's no
pressure, and the software might still be useful or not.  

The pressure for the version number, however, is of course much less
than the pressure for a license: If I don't like the version numbering
of a package on CTAN, I can simply invent my own one for distributing
purposes (using the release date, for example, which gives a reliable,
unambiguous version number - maybe adding upstreams version number after
it for reference).  If a package doesn't have a license (or a non-free
one), it simply cannot be included in the commercially distributed CTAN
derivatives, like TeX-live, Debian, or whatever. (I do *not* want to
discuss here what exactly "a non-free one" might mean).

Regards, Frank
Frank Küster
Inst. f. Biochemie der Univ. Zürich
Debian Developer

More information about the tex-live mailing list