Re: TeX Licenses & teTeX (Was: Re: forwarded message from Jeff Licquia)
> Date: Thu, 8 Aug 2002 00:55:52 -0400
> From: Glenn Maynard <firstname.lastname@example.org>
> > I am not a lawyer, so I cannot claim understanding of intricacies of
> > licenses. However, I think I understand Knuth's lucid writings about
> > his intentions with respect to TeX. He many times said that he wants
> > that after his death TeX version number is frozen at pi, and MF number
> > frozen at e, and absolutely no change is made to them thereafter. It
> If this is done, then the software is non-free. I don't know if the TeX
> community agrees, but I think most people--and not just people sharing
> Debian's views--would agree that software whose license forbids any
> changes to be made is unambiguously non-free.
> However, it's certainly not clear that this is the case. You make
> claims, but you don't support them. From my reading of
> he doesn't want a modified TeX to be called "TeX".
Ok. let us see whether our differences are real or imaginary. In the
latter case there is a hope we can have a common ground.
There are two ways to make modifications to a work "foo". First, you
can change some code in the work and distribute the modified verison
as "foo". Second, you can create a work "bar", which might be
different from "foo" by just one letter -- or by many lines of code --
and distribute it as "bar". For me a work is free if either of these
methods is allowed. Prior to the discussion on this list I thought
that DFSG-4 means, that while Debian community prefers the first
method, the work *may* allow only the second one to be considered
Now it seems that Thomas does not agree with this understanding and
says that I do not interpret DFSG correctly. It may be so. I am a
Debian *user*, not a Debian developer. However, you seem to
accept the second way to be valid.
Do we both agree that TeX and friends do not allow the first way of
modification, but allow the second one? Does it makes them free or
Now let us look at the Computer Modern fonts. You cannot modify them
in the first way, but you definitely can in the second way. There are
derivatives of CM released under other names, which are accepted by
the community. Do you consider CM to be non-free then? Why?
Well. you can say, I can accept the requirement of change the name
Computer Modern to some other name, but I cannot accept the
requirement of change the *filename* cmr10.tfm to some other
filename. If you consider cmr10.tfm to be a separate work, you might
be right. The trick is, it is not exactly a separate work. Rather, it
is a distinguishable part of larger entity called "The TeX". If you
make a drop-in replacement of the file cmr10.tfm, you might argue that
you did not modify Computer Modern fonts (especially if you take care
to change the banners). However, you definitely changed the larger
work, TeX. Therefore a change in cmr10.tfm inside TeX search tree
*requires* you to change the name of TeX in your distribution. That is
why you cannot distribute a replacement of cmr10.tfm AND call the
Where is the fault with my logic?
> > is evident for me that he does not want TeX to be gradually improved;
> > rather he visions completely new systems based on TeX ideas and
> > code.
> Completely new systems based on TeX code? Huh?
Glenn, if you do not know about such systems, this does not mean that
they do not exist, right?
There are several systems purporting to extend and modify TeX. They do
not call themselves TeX precisely because TeX is free for
modifications of the second kind only. teTeX includes three such
systems: pdfTeX, e-TeX and Omega (or four if you count
pdf-e-TeX). There are fonts based on CM. This is the way Knuth seems
to want people to modify his work. I think the possibility of these
modifications makes TeX, CM etc. free. Prior to this discussion I
thought this makes them DFSG-free too, I still cannot understand why
do you think they are not.
People are going to scream bloody murder about that.
-- Seen on linux-kernel