Re: Suggestion for dual-licensed LaTeX (was Re: Encoding the name in the file contents (was Re: Towards a new LPPL draft))
Thomas Bushnell, BSG writes:
> Frank Mittelbach <email@example.com> writes:
> > I would suggest for (nearly) all typesetting systems to use a license like
> > LPPL, simply because (nearly) all of them have as one of their purposes the
> > goal to allow interchange of documents.
> Here's the big mistake.
> You think you are matching the license to the purpose.
yes indeed we do.
> Actually, you are doing the opposite. You are simple declaring that
> any use of the software which does not advance that purpose is going
> to be disregarded.
absolutely not. it makes me really sad to hear that claim repeated over and
over again --- please substantiate it instead of just (re)stating it!
> All those people who might want to do something really wacky with the
> code other than typesetting and interchange of documents are being
> written off.
- anybody is free and invited to do whatever she likes with the code if there
is no distribution
- anybody is free and invited to do whatever she likes with the code as long
as the resulting work is not pretending to be the "original" part of the
ULL. For this we offer three alternatives which can be used individually
- if the remapping feature is used anything, any scrap of the code, can be
individually replaced to do something else without any limitations. And
replacement here is not cascading or anything that makes it difficult to
use and it allows reinterpretation of data without any need to modify the
- if renaming happens (or more exactly if the identification towards the
original interpreter (eg LaTeX, pdfLaTeX, Omega,...) changes) then the
derived work an be put under any licenses whatsoever, can become a new
part of the ULL (or can not) depending on the author of the derived codes
whim, can be used within or outside the scope of the original interpreter
- if the register thing as suggested by Jeff is used then anything can even
be modified without any name change requirement. As in selfmodifying
languages such as TeX, any command is as much data as it is a command,
and any bit of data can be turned into a command. Thus (a) there isn't
something like an API the is identifiable for the TeX language as a whole
(any bit of code part of the document loaded at some point when
processing a document can change the API) and thus (b) different
interpreters or different starting kernels can interpret the register
command completely differently up to ignoring it, in addition (c) even
within the original interpreter and/or the original kernel one can
provide a separate package which changes the interpretation of the
register command string at will.
so you can be as wacky as you wish and and produce whatever results you want
to obtain and it leaves nobody out who wishes to produce anything new or
So what is it that you feel deprived of?
The license has a goal that it tries to protect, true, which is that the
interpretation of the ULL when starting from one particular kernel is
identical between machines. But if you are interested in any different
interpretation of the ULL then all you need to do is start from a different
kernel (and we even offer to provide that alternate kernel that allows
arbitrary interpretations by default).
Other licensesd also have goals including all licenses that are listed in DSFG
10 and all of them technically reduce an absolute freedom (without any rules
to follow) a little bit, to protect those goals, eg
GPL - prohibits distribution without source, requires you to license any
derivation under exactly the same license anda few other things
BSD - has nearly none but it prohibits to promote derived software by telling
that it is large written by the original author
Artistic - fources you to obey certain rules or follow certain procedures
when you do modification (they leave you a choice which to follow, just as
LPPL does) but they don't grant you the right to modify and disobey all of
point 3. They also prohibit to earn money from any distribution of the
package (which I think would included modified version, but perhaps that is
All such rules and restrictions seem to me fine enough and most importantly
they do not hinder anybody to reuse technically the code to perform anything
she wishes. Nevertheless they are rules that restrict absolute freedom
(and after all absolute freedom means no rules at all, not even those of DSFG)
The same is however true for LPPL (in conjunction with the corrections we over
and over suggested to make to it)
To UNSUBSCRIBE, email to firstname.lastname@example.org
with a subject of "unsubscribe". Trouble? Contact email@example.com