User's thoughts about LPPL

[Boris Veytsman]

Greetings:

I apologize for butting in in the ongoing discussion. Moreover, I am
neither a lawyer nor a LaTeX3 team member (a couple of my programs are
in the distribution,  both under GPL and LPPL). Nevertheless I hope
that my thoughts might be of use.

I am a Debian and LaTeX user, so the present misunderstanding between
Debian and LaTeX3 concerns me a lot.

I think Debian team overlooks a couple of points.

1. Debian already uses software other than LaTeX under the "no changes
   unless the files are renamed" clause. This is Don Knuth's TeX and
   MF suite *and* the relevant fonts. Let me remind you that the
   licensing of TeX is rather peculiar:
   
   A. The program itself is in the public domain -- you can do
   whatever you want with the code or its parts

   B. The *name* TeX is reserved for Knuth's program. If you program
   is called TeX, it must satisfy triptest. You can NOT correct bugs
   in this program, you cannot do Debian QA for it -- you either take
   it as is or rename it.

   The same is going for Knuth's Computer Modern fonts. You can do
   whatever you want with the lettershapes -- as long as you do not
   call your product CM.

   If Debian wants to declare (and presumably delete from the main
   distribution) the software under this license, it would be
   hypocrisy to keep TeX and fonts. I wonder whether people realize
   that this means a complete disaster for the GNU info system? GNU
   info is prepared with a program called texinfo, which is basically
   a special TeX format. 



2. Debian people seem not to realize that LaTeX (and TeX) is BOTH a
   program and a language -- and a language requires
   standardization. The nightmare of incompatible HTML dialects proves
   this point well. Yes, standards limit freedom in some way. However,
   do you really want your grocer to have a freedom to call 800g a
   kilogram? 

   As a LaTeX user I have two requirements:

   A. Standardization. I want a LaTeX document to be compiled and
   printed exactly in the same way at my desk, at my publisher's desk,
   at my student's computers etc UNLESS I or students or publishers
   want otherwise.

   B. Flexibility. I want a possibility to completely change
   appearance of any document I received -- IF I WANT IT.

   The present state of LaTeX satisfies these requirements. Due to
   LPPL I am assured that my documents will look exactly the same if
   their source code is unchanged. Due to the fact that LaTeX is a
   macro language, I can redefine ANY command in any document. 

   Suppose a user is near blind and wants all documents to be printed
   in a big fontsize. He can create a program (in latexese called
   style) bigsize.sty and add to all his documents a line
   \usepackage{bigsize}. By doing this he makes a decision about
   document formatting. He is free to do this under LPPL. On the other
   hand, the authors of the documents know that the formatting of
   their works is exactly same UNLESS a user made an explicit decision
   to change.

To summarize: I think LPPL strikes a necessary balance between
standardization and flexibility. This balance was tested by 20+ years
of TeX, which is licensed under exactly same conditions.

-- 
Good luck

-Boris

These download files are in Microsoft Word 6.0 format. After
unzipping, these files can be viewed in any text editor, including
all versions of Microsoft Word, WordPad, and Microsoft Word Viewer
-- From Micro$oft


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Chris Lawrence]

On Jul 16, Boris Veytsman wrote:
> To summarize: I think LPPL strikes a necessary balance between
> standardization and flexibility. This balance was tested by 20+ years
> of TeX, which is licensed under exactly same conditions.

I don't think anyone here has a problem with a license that says "If
your LaTeX doesn't pass such and such a validation suite, you can't
call it LaTeX, but you can do whatever else you want to do with it."

I think the real issue is that the LaTeX project is trying to use its
license to enforce a norm of good behavior by distributors that would
be much better left to certification marks or a Knuthian statement
that "if you break it, both pieces are yours and you are the one who
will get bitched at, not us."  I think that's being obfuscated behind
Branden's establishment of hypotheticals that can be dealt with
through \renewcommand and the like.

I think Frank et al's concerns could be addressed fairly easily by
requiring distributors of modified versions of the entire LaTeX suite
to document the changes and include the location of that documentation
in the diagnostic output of latex, and requiring distributors of
modified versions of separately-distributed style/class files to do
the same, with a waiver of the documentation requirement if the
file/suite is renamed (thereby not misrepresenting the modified
version as any longer being a substitute for the original).  This
certainly would pass the DFSG and would clearly inform users of what
sort of LaTeX they're getting.

For example, if I modify article.dtx, I must either rename it (thereby
no longer calling it article.dtx) or include diagnostic output showing
where on the local filesystem a file is that clearly documents the
differences between article.dtx as distributed by the LaTeX3 project
and my article.dtx (for example, "corrected typo in line 300 that
stopped articles with fewer than 83 pages from having more than 17
footnotes".)

Then again, maybe I'm missing the point :-)


Chris
-- 
Chris Lawrence <[EMAIL PROTECTED]> - http://www.lordsutch.com/chris/


pgphu9zxJdYdC.pgp
Description: PGP signature

Re: User's thoughts about LPPL

[Richard Braakman]

On Tue, Jul 16, 2002 at 07:52:09PM -0400, Boris Veytsman wrote:
>B. The *name* TeX is reserved for Knuth's program. If you program
>is called TeX, it must satisfy triptest. You can NOT correct bugs
>in this program, you cannot do Debian QA for it -- you either take
>it as is or rename it.

That might be why ours is called teTeX.

> 2. Debian people seem not to realize that LaTeX (and TeX) is BOTH a
>program and a language -- and a language requires
>standardization. The nightmare of incompatible HTML dialects proves
>this point well. Yes, standards limit freedom in some way. However,
>do you really want your grocer to have a freedom to call 800g a
>kilogram? 

Being BOTH a program and a language is not unusual.  The same goes for
perl, python, ruby... even gcc.  Other things also require standardization,
such as Internet protocols (telnet, ftp, ssh, the TCP/IP stack in the
kernel) and document formatting languages such as docbook.  Yet all
of these are under free licenses, and there does not seem to be rampant
malignant mutation going on.

>A. Standardization. I want a LaTeX document to be compiled and
>printed exactly in the same way at my desk, at my publisher's desk,
>at my student's computers etc UNLESS I or students or publishers
>want otherwise.

The "UNLESS" part is precisely the freedom that is being discussed here.
Do you want the freedom to change the way a LaTeX document is compiled
and printed, or not?

>Suppose a user is near blind and wants all documents to be printed
>in a big fontsize. He can create a program (in latexese called
>style) bigsize.sty and add to all his documents a line
>\usepackage{bigsize}.

What if the user wants this to happen automatically, without having to
change every document he wants to print or view?

What if the user wants to create documents that can be easily exchanged
with others?  He would have to add or remove the usepackage line at
every exchange, and will probably forget sometimes.  It would be even
more difficult to use a shared version control system to collaborate
on documents.

> By doing this he makes a decision about
>document formatting. He is free to do this under LPPL. On the other
>hand, the authors of the documents know that the formatting of
>their works is exactly same UNLESS a user made an explicit decision
>to change.

It should be possible to make such a decision without modifying the
documents.

> To summarize: I think LPPL strikes a necessary balance between
> standardization and flexibility. This balance was tested by 20+ years
> of TeX, which is licensed under exactly same conditions.

No, TeX is in the public domain.  Only the name is restricted, and
that's not a filename restriction.  The distinction is important if
filenames are part of the technical interface.

Richard Braakman


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Walter Landry]

Boris Veytsman <[EMAIL PROTECTED]> wrote:
> I think Debian team overlooks a couple of points.
> 
> 1. Debian already uses software other than LaTeX under the "no changes
>unless the files are renamed" clause. This is Don Knuth's TeX and
>MF suite *and* the relevant fonts. Let me remind you that the
>licensing of TeX is rather peculiar:
>
>A. The program itself is in the public domain -- you can do
>whatever you want with the code or its parts
> 
>B. The *name* TeX is reserved for Knuth's program. If you program
>is called TeX, it must satisfy triptest. You can NOT correct bugs
>in this program, you cannot do Debian QA for it -- you either take
>it as is or rename it.
> 
>The same is going for Knuth's Computer Modern fonts. You can do
>whatever you want with the lettershapes -- as long as you do not
>call your product CM.

These simple conditions on the overall program name sound like they
fall within the scope of DFSG #4.  Restrictions on individual file
names do not.

> 2. Debian people seem not to realize that LaTeX (and TeX) is BOTH a
>program and a language -- and a language requires
>standardization. The nightmare of incompatible HTML dialects proves
>this point well. Yes, standards limit freedom in some way. However,
>do you really want your grocer to have a freedom to call 800g a
>kilogram? 

Debian has several C and Java compilers.  They support different parts
of the languages and are certainly incompatible with each other in
subtle and mysterious ways.  They are all free software and yet,
somehow, the sky has not fallen.

>As a LaTeX user I have two requirements:
> 
>A. Standardization. I want a LaTeX document to be compiled and
>printed exactly in the same way at my desk, at my publisher's desk,
>at my student's computers etc UNLESS I or students or publishers
>want otherwise.
> 
>B. Flexibility. I want a possibility to completely change
>appearance of any document I received -- IF I WANT IT.

These are nice goals, but they do not make free software.  Debian's
definition of free software means that it satisfies the DFSG.
Whatever your motives may be, if your program doesn't satisfy the
DFSG, then it doesn't go in main.

Regards,
Walter Landry
[EMAIL PROTECTED]


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[M. Drew Streib]

On Tue, Jul 16, 2002 at 05:47:20PM -0700, Walter Landry wrote:
> These are nice goals, but they do not make free software.  Debian's
> definition of free software means that it satisfies the DFSG.
> Whatever your motives may be, if your program doesn't satisfy the
> DFSG, then it doesn't go in main.

And that is what it all comes down to.

I also have sympathy for some of the goals of non-free licenses, but
they are just that... non-free. 

In all of the work people have done on the LSB, we made decisions
early on to enforce the trademark of the name "LSB" against a written
specification, rather than to make portions of software non-free. 
(Given that most of the software was already written and free, it also
wasn't much of a choice.)

-drew

-- 
M. Drew Streib <[EMAIL PROTECTED]>
Independent Rambler, Software/Standards/Freedom/Law -- http://dtype.org/


pgpKtm7tAhLfG.pgp
Description: PGP signature

Re: User's thoughts about LPPL

[Boris Veytsman]

> Date: Tue, 16 Jul 2002 19:18:02 -0500
> From: Chris Lawrence <[EMAIL PROTECTED]>
> 
> I don't think anyone here has a problem with a license that says "If
> your LaTeX doesn't pass such and such a validation suite, you can't
> call it LaTeX, but you can do whatever else you want to do with it."
> 

This requirement is possible for TeX -- but it reqiured a lot of Don
Knuth's time and effort. A similar requirement for LaTeX is impossible
because of the sheer bulk of the code.

Probably the story about Computer Modern Fonts would be
instructive. These fonts are distributed under the license described
before: you can change everything, BUT you must rename. One of TeX
distributions (NTex) used slightly modified CM fonts without changing
file name. This provoked a very sharp responcse form
Knuth. Unfortunately the modified fonts got their way in a couple of
Linux distributions (Slackware and Redhat, as far as I remember).

Nevertheless it turned out that CM fonts are NOT the best for
non-American typesetting. Since CM fonts are not going to change, a
new set called EC was created, based on Knuth's work. This was done
according to the license: the fonts are free, the names are
copyrighted. The current distributions include EC set, which is the
default for many installations.

-- 
Good luck

-Boris

Furious activity is no substitute for understanding.
-- H.H. Williams


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Boris Veytsman]

> Date: Wed, 17 Jul 2002 03:41:42 +0300
> From: Richard Braakman <[EMAIL PROTECTED]>

> 
> On Tue, Jul 16, 2002 at 07:52:09PM -0400, Boris Veytsman wrote:
> >B. The *name* TeX is reserved for Knuth's program. If you program
> >is called TeX, it must satisfy triptest. You can NOT correct bugs
> >in this program, you cannot do Debian QA for it -- you either take
> >it as is or rename it.
> 
> That might be why ours is called teTeX.
> 

The reason is different. I'd suggest reading something first before
making statements.

> 
> >A. Standardization. I want a LaTeX document to be compiled and
> >printed exactly in the same way at my desk, at my publisher's desk,
> >at my student's computers etc UNLESS I or students or publishers
> >want otherwise.
> 
> The "UNLESS" part is precisely the freedom that is being discussed here.
> Do you want the freedom to change the way a LaTeX document is compiled
> and printed, or not?
> 

I'm afraid you mix two different ponits here. You mix the freedom to
change the document look and the freedom to change the typesetting
engine. 

I have a freedom to change the look of any LaTeX document if its
license allows me to do this (the document license has nothing to do
with the LaTeX license). I do not want my system admin to have a
freedom to change the look of all my documents without my explicit
wish.

> >Suppose a user is near blind and wants all documents to be printed
> >in a big fontsize. He can create a program (in latexese called
> >style) bigsize.sty and add to all his documents a line
> >\usepackage{bigsize}.
> 
> What if the user wants this to happen automatically, without having to
> change every document he wants to print or view?

Then he can put the changes into his own format (say, newlatex.fmt),
and instead of doing 
$ latex myfile.tex
do
$ newlatex myfile.tex

This is really easy to do (and does not require superuser privileges,
btw).



> 
> What if the user wants to create documents that can be easily exchanged
> with others?  He would have to add or remove the usepackage line at
> every exchange, and will probably forget sometimes.  It would be even
> more difficult to use a shared version control system to collaborate
> on documents.


Again, for permanent change the user can create a new format and
symlink tex (the execeutable) to a new name. This can be done
completely transparent.




> 
> > By doing this he makes a decision about
> >document formatting. He is free to do this under LPPL. On the other
> >hand, the authors of the documents know that the formatting of
> >their works is exactly same UNLESS a user made an explicit decision
> >to change.
> 
> It should be possible to make such a decision without modifying the
> documents.

It *is* possible -- there are about 1001 way to change the way TeX and
LaTeX works without breaking lppl. Precisely because of this LaTeX3
team insists on a simple island of predictablity in this sea of
changes. 

-- 
Good luck

-Boris

Most people are unable to write because they are unable to think, and
they are unable to think because they congenitally lack the equipment
to do so, just as they congenitally lack the equipment to fly over the moon.
-- H.L. Mencken


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Branden Robinson]

On Tue, Jul 16, 2002 at 09:43:33PM -0400, Boris Veytsman wrote:
> I'm afraid you mix two different ponits here. You mix the freedom to
> change the document look and the freedom to change the typesetting
> engine. 

From your perspective, that may be a meaningful distinction; from
Debian's, it is not.

Clause 3 of the Debian Free Software Guidelines states:

Derived Works

The license must allow modifications and derived works, and must
allow them to be distributed under the same terms as the license
of the original software.

> I have a freedom to change the look of any LaTeX document if its
> license allows me to do this (the document license has nothing to do
> with the LaTeX license). I do not want my system admin to have a
> freedom to change the look of all my documents without my explicit
> wish.

In that case, you should specify in your documents' licenses which
version of LaTeX you require your documents to be viewed with.  Of
course, your documents would not licensed in a DFSG-free way if you did
so, but if you're seeking their distribution by Debian that isn't very
important.

> Then he can put the changes into his own format (say, newlatex.fmt),
> and instead of doing 
> $ latex myfile.tex
> do
> $ newlatex myfile.tex
> 
> This is really easy to do (and does not require superuser privileges,
> btw).

The ease of alternatives to modifying source code is not important.  The
right of the user to create modifications and derived works is.

Interestingly, Frank Mittelbach is asserting that it is not the intent
of the LPPL to forbid mere modification of LaTeX source code:

Message-ID: <[EMAIL PROTECTED]>

"it is not the intention to disallow modification without
redistribution.  the license has a paragraph labeled recommendation that
talks about that but that is already by its title a suggestion only"

Perhaps the LaTeX community should appoint a spokesman to the Debian
Project so that we do not get contradictory statements about what is
acceptable?

> Most people are unable to write because they are unable to think, and
> they are unable to think because they congenitally lack the equipment
> to do so, just as they congenitally lack the equipment to fly over the moon.
>   -- H.L. Mencken

-- 
G. Branden Robinson|Good judgement comes from
Debian GNU/Linux   |experience; experience comes from
[EMAIL PROTECTED] |bad judgement.
http://people.debian.org/~branden/ | -- Fred Brooks


pgpTMKUpzxaui.pgp
Description: PGP signature

Re: User's thoughts about LPPL

[Branden Robinson]

On Tue, Jul 16, 2002 at 10:55:25PM -0500, Branden Robinson wrote:
> In that case, you should specify in your documents' licenses which
> version of LaTeX you require your documents to be viewed with.  Of
> course, your documents would not licensed in a DFSG-free way if you did
> so, but if you're seeking their distribution by Debian that isn't very

s/if/unless/

> important.

Sorry about the typo.

-- 
G. Branden Robinson|
Debian GNU/Linux   |   Extra territorium jus dicenti
[EMAIL PROTECTED] |   impune non paretur.
http://people.debian.org/~branden/ |


pgpU84CXHhJkg.pgp
Description: PGP signature

Re: User's thoughts about LPPL

[David Carlisle]

  I think Frank et al's concerns could be addressed fairly easily by
  requiring distributors of modified versions of the entire LaTeX suite
  to document the changes and include the location of that documentation
  in the diagnostic output of latex, and requiring distributors of
  modified versions of separately-distributed style/class files to do
  the same, with a waiver of the documentation requirement if the
  file/suite is renamed (thereby not misrepresenting the modified
  version as any longer being a substitute for the original).  This
  certainly would pass the DFSG and would clearly inform users of what
  sort of LaTeX they're getting.

  Then again, maybe I'm missing the point :-)

Yes I think so.

Notifying users is part of it but not the main part.

LaTeX is a document markup language the primary aim is to have
portable documents. Thus anything that claims to be latex (or tex, or
the computer modern fonts) should produce the same output.

LaTeX has the extra constraint that unlike a compiled program
the full source of latex is visible to every latex document.

Some people have suggested that latex should allow arbitrary changes but
only allow the name "latex" to be used if the resulting program meets
some published interface. That is fine for a compiled program which can
implement a published interface via an implementation that isn't seen by
the application. However it is a technical non starter for a macro
language. If you change latex in _any_ way, adding \relax to any
definition then the observable behaviour of the program will alter.

So for a macro language saying that it meets some published interface is
equivalent to saying that no changes have been made at all.

C the language is specified by international standard. It makes sense to
talk of different free or not free implementations of that language.

You simply can not do that for LaTeX. A different implementation will
implement a different language. That is just the way TeX works.

So LaTeX has the perfectly reasonable (and apparently DFSG compliant)
restriction that if you change the code so that it no longer implements
the same language, you call it something different.


David


_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[David Carlisle]

> These simple conditions on the overall program name sound like they
> fall within the scope of DFSG #4.  Restrictions on individual file
>  names do not.

Those conditions [on cm fonts] _are_ at the level of individual file
names.

You can not call your font metric file  cmr10.tfm unless it is byte for
byte the same as Knuth's cmr10.

Unlike the LaTeX version of this rule which has, as far as I know, never
been "used in anger" Knuth has in the not so distant past used this rule
to explictly force the withdrawl of some modified font metrics that
were being distributed with some linux distributions.

The modified metrics "improved" the interletter spacing in some cases,
and at the same time destroyed the use of TeX as a portable document
system.

David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Javier Bezos]

> I think Frank et al's concerns could be addressed fairly easily by
> requiring distributors of modified versions of the entire LaTeX suite
> to document the changes and include the location of that documentation
> in the diagnostic output of latex, and requiring distributors of
> modified versions of separately-distributed style/class files to do
> the same, with a waiver of the documentation requirement if the
> file/suite is renamed (thereby not misrepresenting the modified
> version as any longer being a substitute for the original).  This
> certainly would pass the DFSG and would clearly inform users of what
> sort of LaTeX they're getting.
> 
> Then again, maybe I'm missing the point :-)

David Carlisle said:

> Yes I think so.

I think so, too. The problem is not the distribution itself
but the documents created using that distribution. If I get a
document by "John Smith" (somehow), how can I see if _his_
system had a modified latex?

> LaTeX has the extra constraint that unlike a compiled program
> the full source of latex is visible to every latex document.

And modifiable from every latex document. I would like to
note that the behaviour of latex can be changed **freely**
from within a document without even touching the latex
files, by means of an extension mechanism named "packages"
(amongst others). All macros defined in latex (ie, the
whole latex program) are public and can be redefined
without any restrictions by these packages. To put it
in other words, with lppl you can rewrite latex in full
(even for a single document and from within the document!).

Walter Landry said:

> Here is a hypothetical.  Let's say that someone wants to add support for
> Klingon into Latex.  So they hack something together which, by necessity,
> changes a few standard files, and it works for them without breaking anything
> else.  You reject the patch because it isn't really a good i18n solution, it
> only works for Klingon.  You also think that Klingon is a silly thing to add
> support for, so you'll probably never add it in.  However, for the people
> interested in writing Klingon (e.g. Hollywood screen writers and trek fan
> fiction writers), this is a good solution.  In this case, you are preventing
> people from having seamless support for Klingon.

This is a really good argument *in favour* of LPPL! If someone
adds support for Klingon by modifying the LaTeX kernel, the
resulting documents will have a restricted distribution
because they won't compile correctly in other systems. This
is an _actual_ restriction. But if instead a package -- extending
and modifying latex as desired -- is created and called from
the document, it will complain about a required but missing
package and you will be able to locate and get the package,
and then typeset the document. Otherwise, you will be frustrated
because you could have a 'correct' document displaying nothing
without any explanation.

Javier
[A latex user who don't like some things of latex and who
changes it *freely* by means of packages.]


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Brian Sniffen]

David Carlisle <[EMAIL PROTECTED]> writes:

>> These simple conditions on the overall program name sound like they
>> fall within the scope of DFSG #4.  Restrictions on individual file
>>  names do not.
>
> Those conditions [on cm fonts] _are_ at the level of individual file
> names.
>
> You can not call your font metric file  cmr10.tfm unless it is byte for
> byte the same as Knuth's cmr10.

That's not true.  He hasn't registered a trademark on the name, I
believe.  So the most the license can do is prohibit you from calling
a *derivative work* of Knuth's cmr10 file cmr10.tfm.

If I write my own font -- or take a free font from elsewhere -- I may
name it cmr10.tfm and distribute it without violating Knuth's
license.  I may even do so in parallel with a TeX distribution.

-Brian

-- 
Brian Sniffen
[EMAIL PROTECTED]


pgpVWgOUVkX8N.pgp
Description: PGP signature

Re: User's thoughts about LPPL

[David Carlisle]

> That's not true.  He hasn't registered a trademark on the name, I
> believe.  So the most the license can do is prohibit you from calling
> a *derivative work* of Knuth's cmr10 file cmr10.tfm.

er yes that's the usual rider. Although it's not clear to me that you
can do that and still use it with TeX (unless you rename TeX) as that
would be a change to tex-the system. (But Knuth's licence comments are
not especially clear on edge cases).

Which doesn't alter my main point that the TeX and font licences are
essentially the same as LaTeX's so any move to move latex to Debian's
non free area should also take TeX and the computer modern fonts at the
same time.

David


_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Jeff Licquia]

On Wed, 2002-07-17 at 05:29, David Carlisle wrote:
> Those conditions [on cm fonts] _are_ at the level of individual file
> names.

As a data point, there is some disagreement within Debian as to whether
non-functional components, such as documentation, should be held to the
same standard as functional ones.  Additionally, there is the question
of defining "non-functional" data; some kinds of data, such as fonts,
have functional impact, and it's not clear where to draw the line (if
indeed there is a line to be drawn).

I would encourage the curious to look in the debian-legal archives
(http://lists.debian.org/debian-legal/) for more information.  In
particular, there are some good references in: 

http://lists.debian.org/debian-legal/2002/debian-legal-200204/msg00028.html

So as not to delve into that argument again, it might be more productive
to shelve the question of the CM fonts and pay closer attention to files
(like tex.web) that are more clearly functional.  It's my understanding
that we have yet to take action against any packages currently in Debian
main pending a more clear resolution; if the time comes to act, the CM
fonts can be moved to non-free then.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Walter Landry]

Javier Bezos <[EMAIL PROTECTED]> wrote:
> Walter Landry said:
> 
> > Here is a hypothetical.  Let's say that someone wants to add support for
> > Klingon into Latex.  So they hack something together which, by necessity,
> > changes a few standard files, and it works for them without breaking 
> > anything
> > else.  You reject the patch because it isn't really a good i18n solution, it
> > only works for Klingon.  You also think that Klingon is a silly thing to add
> > support for, so you'll probably never add it in.  However, for the people
> > interested in writing Klingon (e.g. Hollywood screen writers and trek fan
> > fiction writers), this is a good solution.  In this case, you are preventing
> > people from having seamless support for Klingon.
> 
> This is a really good argument *in favour* of LPPL! If someone
> adds support for Klingon by modifying the LaTeX kernel, the
> resulting documents will have a restricted distribution
> because they won't compile correctly in other systems. This
> is an _actual_ restriction. But if instead a package -- extending
> and modifying latex as desired -- is created and called from
> the document, it will complain about a required but missing
> package and you will be able to locate and get the package,
> and then typeset the document. Otherwise, you will be frustrated
> because you could have a 'correct' document displaying nothing
> without any explanation.

You missed the point, which is that the changes in the Latex kernel
had to be made in order for Klingon (or whatever language) to work.
Don't tell me that there will never be a need to change the internals
in order to make something work.  You can't anticipate everything that
will happen in technology for the 100 years.

Regards,
Walter Landry
[EMAIL PROTECTED]


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Nick Phillips]

On Wed, Jul 17, 2002 at 11:23:17AM +0100, David Carlisle wrote:

> Some people have suggested that latex should allow arbitrary changes but
> only allow the name "latex" to be used if the resulting program meets
> some published interface. That is fine for a compiled program which can
> implement a published interface via an implementation that isn't seen by
> the application. However it is a technical non starter for a macro
> language. If you change latex in _any_ way, adding \relax to any
> definition then the observable behaviour of the program will alter.
> 
> So for a macro language saying that it meets some published interface is
> equivalent to saying that no changes have been made at all.

I think you are mistaken. You are assuming that the engine used to process
those macros will also not be changed; it would be quite possible to change
LaTeX in such a way that it produced identical output from all valid LaTeX
input whilst adding other functionality, if you modified it to use
 "under the hood".

Why should this not be allowed?


> You simply can not do that for LaTeX. A different implementation will
> implement a different language. That is just the way TeX works.

Not true. Maybe not *usefully* false, but the original creator of a work is
often the last person who should be judging in what ways it may be useful.

If your intention is to ensure that anything called LaTeX produces identical
output to the then-current "official" implementation, then apply Occam's
Razor and say that.

If you are happy for people to take the code and do anything they like with
it (bar distributing a functionally incompatible version described as
"LaTeX"), then why not keep things simple and say so -- allow all kinds of
modification and distribution in the license, but control a "LaTeX" trademark
in such a way that no-one can distribute incompatible versions.

This would probably be more appropriately performed via an entity set up
to manage the development and maintenance of LaTeX - a "LaTeX Foundation"
or similar. The constitution etc. of that organisation would decide how
responsibilities for maintaining the "official" version were delegated,
removing all the messy bits about maintainers etc. from the LaTeX license.

And everybody would live happily ever after. We hope.

I'd have thought that the solution above would be a more appropriate way
to handle the question of "so, what happens if the maintainer goes AWOL?"
in any case -- there would be an entity dedicated to looking after LaTeX,
no matter what misfortunes befell individuals working on the project.


If setting up a whole new body to perform this task is more than you are
willing or able to undertake, then you might consider asking one of the
existing similar bodies, such as the ASF or SPI to take on this role.




Cheers,


Nick
-- 
Nick Phillips -- [EMAIL PROTECTED]
Today is the first day of the rest of your life.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Javier Bezos]

>> This is a really good argument *in favour* of LPPL! If someone
>> adds support for Klingon by modifying the LaTeX kernel, the
>> resulting documents will have a restricted distribution
>> because they won't compile correctly in other systems. This
>> is an _actual_ restriction. But if instead a package -- extending
>> and modifying latex as desired -- is created and called from
>> the document, it will complain about a required but missing
>> package and you will be able to locate and get the package,
>> and then typeset the document. Otherwise, you will be frustrated
>> because you could have a 'correct' document displaying nothing
>> without any explanation.
> 
> You missed the point, which is that the changes in the Latex kernel
> had to be made in order for Klingon (or whatever language) to work.

As I said, that's not exact.

> Don't tell me that there will never be a need to change the internals
> in order to make something work.

Of course, I won't tell you that. I repeat that the internals *can*
be changed without touching a single file from the LaTeX kernel,
and currenty there are several packages doing that. An example:
the hyperref package patches some internals of the LateX kernel
to provide hyperlinks and more in pdf documents, and no file
from LaTeX has been modified.

> You can't anticipate everything that
> will happen in technology for the 100 years.

Definitely right, and as you can see LaTeX already provides
a mechanism for that (which could explain why LaTeX is still
alive after 20 years). The lppl ensures that this mechanism is
used, or otherwise the free distribution of LaTeX documents
won't be possible and it will suffer severe restrictions.

Regards
Javier




-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Glenn Maynard]

On Thu, Jul 18, 2002 at 09:28:22AM +0200, Javier Bezos wrote:
> Of course, I won't tell you that. I repeat that the internals *can*
> be changed without touching a single file from the LaTeX kernel,
> and currenty there are several packages doing that. An example:
> the hyperref package patches some internals of the LateX kernel
> to provide hyperlinks and more in pdf documents, and no file
> from LaTeX has been modified.

If the core can be changed in any way without changing it directly,
then you can break output exactly as well by this mechanism as you
could by editing it directly.

If so, then there's no point in forcing people to use it; they can break
stuff anyway.

If not, then there are apparently things you can't do with this mechanism.

> Definitely right, and as you can see LaTeX already provides
> a mechanism for that (which could explain why LaTeX is still
> alive after 20 years). The lppl ensures that this mechanism is

I can modify gcc, remove short-circuit evaluation, and distribute it,
breaking almost every C program in use.

Yet, C has been around for 30 years, without any of these restrictions,
without falling apart.

> used, or otherwise the free distribution of LaTeX documents
> won't be possible and it will suffer severe restrictions.

You're saying that if I edit the core directly instead of using the
mechanisms you mention, free distribution of documents is no longer
possible, and Latex will suddenly have severe restrictions?

Huh?

Sorry, but there's no connection here.  If I edit the core directly and
break everything, I can still freely distribute my Latex documents and
Latex will not suffer any restrictions.

Of course, if I distribute my broken copy of Latex and call it Latex, it
might cause some grief, since people who think they're getting Latex are
getting something else.  That's why some people use licenses that say
that if you distribute changed versions, you need to call the program
something else.  No matter how much (or how subtly) I break Latex, and
no matter how many people I distribute it to, it's not going to cause
preventable problems if I call it something else.

Of course, that means there's another tex parser out in the wild that
may not be completely compatible with Latex, but that can't be stopped.

-- 
Glenn Maynard


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[David Carlisle]

> Additionally, there is the question
> of defining "non-functional" data; some kinds of data, such as fonts,
> have functional impact

for a system like latex the fonts (or at least their metrics) have as
much impact as the rest of the system. Modifying the font metrics is
even more likely to change the final document layout than changing the
latex macros (most of which are not used in any given document run).

>  if the time comes to act, CM fonts can be moved to non-free then.

In that case probably it's best if we just all come back then.
It will be a lot of work finalising the details of a rewrite of LPPL
and if the only benefit of that is that you declare LaTeX suitable for
the free part of Debian, that effort will be completely wasted if TeX
and the fonts are not in the free part.

I'd like to see LaTeX classed as Free by Debian (because it is Free)
but distributing LaTeX separately from TeX would be non sensical
and lead to massive user confusion. So if TeX and the CM fonts were in
non-free I'd suggest you distribute latex from there as well, even if
latex had a licence that you would be happy to classify as free.

So I don't think we can do anything about the latex licence until 
then. (This is a personal response to your comments, Frank may have
different ideas, especially as he's spent a lot of time redrafting
LPPL this last month and is (I thought) almost there as regards
addressing any concerns raised by Debian with the old version.)

David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[David Carlisle]

> you missed the point, which is that the changes in the Latex kernel
> had to be made in order for Klingon (or whatever language) to work.
> Don't tell me that there will never be a need to change the internals
> in order to make something work.  You can't anticipate everything that
> will happen in technology for the 100 years.

If the user wants klingon they go

\usepackage{klingon}

klingon.sty can redefine _absolutely every single line of latex code in
the latex distribution_ and still be compliant.

What more do you want?

If you want more and want to change the kernel itself rather than
redefining it on the fly (perhaps just for optimisation reasons) you can
do that as well as long as you don't call it latex, see the quote from
modguide.tex I posted a few minutes ago.

David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Javier Bezos]

Glenn Maynard said:

> If the core can be changed in any way without changing it directly,
> then you can break output exactly as well by this mechanism as you
> could by editing it directly.

So what...?
 
> If so, then there's no point in forcing people to use it; they can break
> stuff anyway.

Yes of course. If someone uses \usepackage{nonsense} in the
document where nonsense is a package breaking latex, the document
won't work (never). I don't see your point.

> If not, then there are apparently things you can't do with this mechanism.

Thanks for saying "apparently" :-). We are repeating and repeating
again than you can rewrite latex in full, if you want.

>> Definitely right, and as you can see LaTeX already provides
>> a mechanism for that (which could explain why LaTeX is still
>> alive after 20 years). The lppl ensures that this mechanism is
> 
> I can modify gcc, remove short-circuit evaluation, and distribute it,
> breaking almost every C program in use.

And you can write a package doing that. Except for the fact
that you cannot change the behaviour of "your" C compiler
from within a C "document" (ie, a C program), while you may
fix the wrong behavior of LaTeX with your own package or
even from within the latex document.

> Yet, C has been around for 30 years, without any of these restrictions,
> without falling apart.

So what...? Did I say a single word saying that LPPL is
good for C and, say, GPL is wrong? What I'm saying is that LaTeX
is not C and therefore the licence could have some differences.
If C and LaTeX are so long-lived depends on lots of factors,
but the right license (or standardization) might be one of them.
(BTW, except for TeX itself, the tetex code is distributed under
GPL and i think that's fine.)

> You're saying that if I edit the core directly instead of using the
> mechanisms you mention, free distribution of documents is no longer
> possible, and Latex will suddenly have severe restrictions?

Nope. The _distribution_ of LaTeX documents and the LaTeX
documents, _not_ LaTeX itself.

> Of course, if I distribute my broken copy of Latex and call it Latex, it
> might cause some grief, since people who think they're getting Latex are
> getting something else.

I repeat what I said a few messages ago -- the problem is not
redistributing LaTeX but distributing documents created using a
modified latex. I'm going to remember it:

>> but the documents created using that distribution. If I get a
>> document by "John Smith" (somehow), how can I see if _his_
>> system had a modified latex?

Packages and lppl provide freedom for both modifying latex and
distributing documents that work from the USA to India or
Iran or Australia or Russia or... Otherwise, documents will be
condemned to personal use only.

Regards
Javier


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Brian Sniffen]

> In article <[EMAIL PROTECTED]>, David Carlisle <[EMAIL PROTECTED]> writes:

> If you want more and want to change the kernel itself rather than
> redefining it on the fly (perhaps just for optimisation reasons) you can
> do that as well as long as you don't call it latex, see the quote from
> modguide.tex I posted a few minutes ago.

But that isn't all: if the restriction were just that I couldn't
distribute my changed (perhaps just very heavily commented) LaTeX
kernel under the name LaTeX, that part of the license would be
unambiguously DFSG-free.  It's the requirement that I change every
file name as well, coupled with TeX's heavy dependence on file names,
that is the problem here

> What more do you want?

Would a statement in the license that *either* of the following must
happen be acceptable to the LaTeX project?

* The modified copy of the Program is distributed under a name which
  clearly distinguishes it from Standard LaTeX, the unmodified copy.

* Any files which share names between the unmodified copy and the
  modified copy must be identical in content.  You may modify files
  only if you change their names.

-Brian


pgpLCIgDcb82z.pgp
Description: PGP signature

Re: User's thoughts about LPPL

[David Carlisle]

> Would a statement in the license that *either* of the following must
> happen be acceptable to the LaTeX project?
> 
> * The modified copy of the Program is distributed under a name which
>   clearly distinguishes it from Standard LaTeX, the unmodified copy.
> 
> * Any files which share names between the unmodified copy and the
>   modified copy must be identical in content.  You may modify files
>   only if you change their names.

such a statement appears already in the modguide.tex file (which is
referenced from the licence) as I posted this morning.
there is the additional requirement that you have to make sure your
modified files are not picked up by latex (if you also distribute that).

that is 
you can have (if you want, although no one does) 
debianlatex
that has its own article.cls that does whatever you want.

But you can't do that if that modified file is on latex's default path 
so that
latex 
uses the same file.


point 2 in modguide.tex which I post again:


\begin{enumerate}
\item
  Give your system a distinguished name, such as \nstex, which clearly
  distinguishes it from \LaTeX{}.

\item
  Ensure that it contains no file with a name the same as that of
  a file in the standard distribution but with different contents.
  (If this is not possible then you must: 
  \begin{itemize}
  \item
ensure that files from the non-\LaTeX{} system cannot be
accidentally accessed whilst using a standard \LaTeX{};
  \item ensure that each file from the non-\LaTeX{} system clearly
identifies itself as a non-\LaTeX{} file on the terminal and in the
log file.)
  \end{itemize}

\item
  Ensure that the method used to run your system is clearly
\label{mcon:command}
  distinct from that used to run Standard \LaTeX; e.g.~by using a
  command name or menu entry that is clearly not \texttt{latex}
  (or \texttt{LaTeX} etc).

\item
  Ensure that, when a file is being processed by your system, the
  use of non-standard \LaTeX{} is clearly proclaimed to the user by
  whatever means is appropriate.

\item Ensure that what is written at the beginning of the log file
  clearly shows that your system has been used, and that it is 
  not Standard \LaTeX{}.
  See the file \texttt{cfgguide.tex} for how to achieve this.

\item
 Clearly explain to users that bug reports concerning your 
 system should not be sent to the maintainers of Standard
 \LaTeX{}. 
\end{enumerate}

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Boris Veytsman]

> Date: Thu, 18 Jul 2002 04:15:20 -0400
> From: Glenn Maynard <[EMAIL PROTECTED]>

> 
> If the core can be changed in any way without changing it directly,
> then you can break output exactly as well by this mechanism as you
> could by editing it directly.
> 

No, because to change the core you need to make an explicit decision,
like using a command

newlatex file.tex

instead of

latex file.tex

LaTeX team does not mind you using or distributing a changed
LaTeX. It does not want you to use it AND call it LaTeX.

> 
> Of course, if I distribute my broken copy of Latex and call it Latex, it
> might cause some grief, since people who think they're getting Latex are
> getting something else.  That's why some people use licenses that say
> that if you distribute changed versions, you need to call the program
> something else.  No matter how much (or how subtly) I break Latex, and
> no matter how many people I distribute it to, it's not going to cause
> preventable problems if I call it something else.
> 


That is exactly the intention of LPPL as far as I understand it.

-- 
Good luck

-Boris

Kime's Law for the Reward of Meekness:
Turning the other cheek merely ensures two bruised cheeks.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Steve Langasek]

On Thu, Jul 18, 2002 at 09:28:58AM +0100, David Carlisle wrote:

> In that case probably it's best if we just all come back then.
> It will be a lot of work finalising the details of a rewrite of LPPL
> and if the only benefit of that is that you declare LaTeX suitable for
> the free part of Debian, that effort will be completely wasted if TeX
> and the fonts are not in the free part.

> I'd like to see LaTeX classed as Free by Debian (because it is Free)
> but distributing LaTeX separately from TeX would be non sensical
> and lead to massive user confusion. So if TeX and the CM fonts were in
> non-free I'd suggest you distribute latex from there as well, even if
> latex had a licence that you would be happy to classify as free.

FWIW, you wouldn't even need to ask. :)  Debian policy requires our
'main' archive to be self-contained; even if a piece of software meets
the DFSG, if it depends on other software that does not, it's placed in
a separate archive section called 'contrib'.

And I definitely think moving LaTeX to contrib would be a loss for all
Debian users.

Cheers,
Steve Langasek
postmodern programmer


pgpfzcxDKsaiu.pgp
Description: PGP signature

Re: User's thoughts about LPPL

[Jeff Licquia]

On Thu, 2002-07-18 at 03:28, David Carlisle wrote:
> 
> > Additionally, there is the question
> > of defining "non-functional" data; some kinds of data, such as fonts,
> > have functional impact
> 
> for a system like latex the fonts (or at least their metrics) have as
> much impact as the rest of the system. Modifying the font metrics is
> even more likely to change the final document layout than changing the
> latex macros (most of which are not used in any given document run).

Well, yes.  OTOH, substituting pictures can also change layout, and
pictures are clearly non-functional data.

You may be right; fonts may be too "functional" to waive the DFSG for
them.  But that's a bridge we'll have to come to at some point, and it
affects far more things than just TeX.

> >  if the time comes to act, CM fonts can be moved to non-free then.
> 
> In that case probably it's best if we just all come back then.
> It will be a lot of work finalising the details of a rewrite of LPPL
> and if the only benefit of that is that you declare LaTeX suitable for
> the free part of Debian, that effort will be completely wasted if TeX
> and the fonts are not in the free part.

I hope not.  Hopefully, the license you craft with our input will be a
stronger license, and will more clearly reflect your priorities.  I
think there have been several cases where we've identified
characteristics of your license that do not reflect your stated goals.

> I'd like to see LaTeX classed as Free by Debian (because it is Free)
> but distributing LaTeX separately from TeX would be non sensical
> and lead to massive user confusion. So if TeX and the CM fonts were in
> non-free I'd suggest you distribute latex from there as well, even if
> latex had a licence that you would be happy to classify as free.

Yes.  We cover this problem with a section called "contrib", which
contains DFSG-free software that depends on non-free software.  A lot of
Java software falls into this section, for example.

So if the LPPL ends up being DFSG-free but TeX is not, we won't take
that away from you.

> So I don't think we can do anything about the latex licence until 
> then. (This is a personal response to your comments, Frank may have
> different ideas, especially as he's spent a lot of time redrafting
> LPPL this last month and is (I thought) almost there as regards
> addressing any concerns raised by Debian with the old version.)

I haven't seen his response yet, and am looking forward to it.  I urge
him (and you) to stay engaged if you would, as I think our discussion
has been profitable in many ways.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[David Carlisle]

> Well, yes.  OTOH, substituting pictures can also change layout, and
> pictures are clearly non-functional data.

document formatting systems have fonts they use in all documents.
They don't usually have pictures they use in all documents.

Changing the font metrics isn't like changing a picture file
it's like changing your png renderer so that the same file produces a
different image. ie the result document changes with no apparent change
to that source document.

> I hope not.
well that was just an initial reaction. I suspect having got this far we
will do a LPPL 1.3 version:-)


David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Glenn Maynard]

On Thu, Jul 18, 2002 at 12:55:43PM +0200, Javier Bezos wrote:
> >> but the documents created using that distribution. If I get a
> >> document by "John Smith" (somehow), how can I see if _his_
> >> system had a modified latex?

Others (eg Boris) seem to be saying that the Latex developers don't
mind if you rename Latex to something else and modify it such that
\usepackage{article} in a document does something different than it
does in the real latex.  You're contradicting this, so I'll discontinue
this thread unless Frank or David speaks up to clarify this.

-- 
Glenn Maynard


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Mark Rafn]

On Thu, 18 Jul 2002, David Carlisle wrote:

> Changing the font metrics isn't like changing a picture file
> it's like changing your png renderer so that the same file produces a
> different image. ie the result document changes with no apparent change
> to that source document.

I agree with David - changing font metrics is similar to changing the 
engine.  If I can't change the engine (even if the result is incompatible 
with the original), I don't have free software.

A license that prevents me from making such a change to a png renderer and 
distributing the result is not free either.
--
Mark Rafn[EMAIL PROTECTED]  


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Sam Hartman]

> "Javier" == Javier Bezos <[EMAIL PROTECTED]> writes:

Javier> Thanks for saying "apparently" :-). We are repeating and
Javier> repeating again than you can rewrite latex in full, if you
Javier> want.

MMM, someone on the Debian side here should write up an instructive
rant on what source code means focusing on the preferred form for
editing definition.  I disagree that the source code (as we think of
it) is available to all documents.  The source to LaTeX isn't even on
my system.  The source goes through a bunch of transformations from
the .dtx files to produce the .cls and other files.

The LaTeX community's point remains--you can change any aspect of the
behavior of LaTeX from within a document.  But you do not get the
editable form of LaTeX source from within a document.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Jeff Licquia]

On Thu, 2002-07-18 at 13:22, Mark Rafn wrote:
> On Thu, 18 Jul 2002, David Carlisle wrote:
> 
> > Changing the font metrics isn't like changing a picture file
> > it's like changing your png renderer so that the same file produces a
> > different image. ie the result document changes with no apparent change
> > to that source document.
> 
> I agree with David - changing font metrics is similar to changing the 
> engine.  If I can't change the engine (even if the result is incompatible 
> with the original), I don't have free software.
> 
> A license that prevents me from making such a change to a png renderer and 
> distributing the result is not free either.

OK, let me clarify here.

I am trying to avoid rehashing an old argument that isn't getting
resolved now, and is really orthogonal to the question of whether LaTeX
is DFSG-free or not.

The question of "data freedom" or "font freedom" or whatever affects
many more packages than TeX, and doesn't affect TeX in any way that it
doesn't affect these other packages.  In addition, we have concerns
about real executable code here that no one disputes is functional, so
it's not like the fonts are the only problem here.

Again, if you're interested, read the thread I posted links to earlier. 
In the meantime, please, let's keep focus.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Boris Veytsman]

> Date: Thu, 18 Jul 2002 14:12:44 -0400
> From: Glenn Maynard <[EMAIL PROTECTED]>

> 
> On Thu, Jul 18, 2002 at 12:55:43PM +0200, Javier Bezos wrote:
> > >> but the documents created using that distribution. If I get a
> > >> document by "John Smith" (somehow), how can I see if _his_
> > >> system had a modified latex?
> 
> Others (eg Boris) seem to be saying that the Latex developers don't
> mind if you rename Latex to something else and modify it such that
> \usepackage{article} in a document does something different than it
> does in the real latex.  You're contradicting this, so I'll discontinue
> this thread unless Frank or David speaks up to clarify this.
> 


To say the truth, I do not see a contradiction here.

Anyway, here is a quotation from modguide, which is referred to by the
LPPL (the version currently in force):

It is possible that you need to produce a document processing
system based on standard \LaTeX{} but with functionality that
cannot be implemented by using the approved configuration files
and complying with the restriction on the code that is allowed in
them.  In other words, you may need a system which is sufficiently
distinct from Standard \LaTeX{} that it is not feasible to do this
simply by using the configuration options we provide or by
producing new classes and packages.

If you do produce such a system then, for the reasons described
above, you should ensure that your system is clearly distinguished
from Standard \LaTeX{} in every possible way, including the
following.

Then the guide lists the ways you should take to make your system
clearly distinguishable from the standard, including name change.

I read this quote as clear permission to perform any modification as
long as you do not claim your system to be LaTeX and take care it
cannot be taken for LaTeX accidentally.

-- 
Good luck

-Boris

Operator, please trace this call and tell me where I am.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Javier Bezos]

> On Thu, Jul 18, 2002 at 12:55:43PM +0200, Javier Bezos wrote:
 but the documents created using that distribution. If I get a
 document by "John Smith" (somehow), how can I see if _his_
 system had a modified latex?
> 
> Others (eg Boris) seem to be saying that the Latex developers don't
> mind if you rename Latex to something else and modify it such that
> \usepackage{article} in a document does something different than it
> does in the real latex.  You're contradicting this, so I'll discontinue
> this thread unless Frank or David speaks up to clarify this.

There is no "real" latex as opposed to "unreal/modified" latex.
What David is saying is that you can create a format based on latex
which is no longer latex but balhblahlatex -- a consciuos act
is necessary to typeset that document using something different
to "latex" and while it's true that the document (the tex file)
could look identical, it won't be latex anymore, and if it doesn't
work in your system that will be the fault of the author of the
document which is distributing a document not being latex, but
which looks like latex, without any notice. If latex can be
modified, that document could be both created and distributed
as a "latex" document, and that is what I'm saying.

Javier



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[David Carlisle]

> I think you are mistaken. You are assuming that the engine used to process
> those macros will also not be changed; it would be quite possible to change
> LaTeX in such a way that it produced identical output from all valid LaTeX
> input whilst adding other functionality, if you modified it to use
>  "under the hood".
> 
> Why should this not be allowed?

it is allowed.
pdftex for example produces different output from the same input.
you could use the command "latex" for that as it doesn't involve any
changes in LPPL'ed code, although tetex calls the command pdflatex
as a user convenience so they can easily get access to both forms.

omega on the other hand (unicode TeX) did make a few changes to latex
(I think in only a couple of lines) to use omega features. The omega
version of latex is called "lambda" rather than latex. Is that really
too much to ask? So the omega developers take latex and just change it
as they see fit, they are not restricted in any way the kind of chages
they make. The end user sees "lambda" which is almost exactly latex but
changed a bit and using a different engine underneath. I can't see how
anyone would benefit if this version was distributed as latex.


> If you are happy for people to take the code and do anything they like with
> it (bar distributing a functionally incompatible version described as
> "LaTeX"), then why not keep things simple and say so -- allow all kinds of
> modification and distribution in the license, but control a "LaTeX" trademark
> in such a way that no-one can distribute incompatible versions.


It is just about conceivable that the latex project could consider
registering "LaTeX" (although certain rubber based products might make
that difficult) but LPPL isn't just used for the core LaTeX: it has
turned out to be remarkably popular (and successful in its stated aims
of keeping LaTeX both Free and stable). 100's  of contributed
packages are now distributed under LPPL.

It is not reasonable that the author of a package such as
"indentfirst.sty"
for example (which consists of exactly 4 TeX tokens) should be expected
to go to the trouble of trying to legally register that name.
the LPPL as it is currently drafted gives other users some assurance
that if their documents have
\usepackage{indentfirst}
then their documents will behave as they expect (indenting the first
paragraph of sections) and not invoke some other completely different
code that someone thought was an "improvement".

> 
> 
> If setting up a whole new body to perform this task is more than you are
> willing or able to undertake, then you might consider asking one of the
> existing similar bodies, such as the ASF or SPI to take on this role.
> 

The latex3 project does exist as some kind of entity (has web sites and
bank accounts etc) but as I say that isn't really any help in the
general case. Most GPL'ed code is not from FSF and similarly most
LPPL'ed code is not from the LaTeX3 project.

David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Nick Phillips]

On Fri, Jul 19, 2002 at 04:04:33PM +0100, David Carlisle wrote:

> it is allowed.
> pdftex for example produces different output from the same input.
> you could use the command "latex" for that as it doesn't involve any
> changes in LPPL'ed code, although tetex calls the command pdflatex
> as a user convenience so they can easily get access to both forms.

I would have expected this to require modifications to LaTeX in order to
call a different underlying engine -- but it's been a while since I used
LaTeX heavily, so...


> It is not reasonable that the author of a package such as
> "indentfirst.sty"
> for example (which consists of exactly 4 TeX tokens) should be expected
> to go to the trouble of trying to legally register that name.

No, agreed, but...

> the LPPL as it is currently drafted gives other users some assurance
> that if their documents have
> \usepackage{indentfirst}
> then their documents will behave as they expect (indenting the first
> paragraph of sections) and not invoke some other completely different
> code that someone thought was an "improvement".

This falls down because the "indentfirst" package is likely to be a relatively
small work, and until and unless it is incorporated into the LaTeX
distribution, there is absolutely no assurance that someone else somewhere
else won't come up with something with the same name, similar intent, but
critically-different results, without having heard of the first package.

In short, the protection of which you speak is only really present when the
package is included in a distribution large and widespread enough to justify
a trademark in any case.

Presumably CTAN will only provide one package with any given name, and this
fact alone is enough to provide the degree of protection required for such
a package, is it not?


> The latex3 project does exist as some kind of entity (has web sites and
> bank accounts etc) but as I say that isn't really any help in the
> general case. Most GPL'ed code is not from FSF and similarly most
> LPPL'ed code is not from the LaTeX3 project.

If there are a large number of third parties using the LPPL, all the more
reason to keep it simple, IMHO. I would at least move the sections regarding
maintainership into a separate document, and vigourously attack the rest
of the draft to remove unnecessary complication.


While I'm at it, thanks for spending the time both on LaTeX, and such
potentially head-banging-against-brick-wall discussions as this.

We appreciate it.



Cheers,


Nick
-- 
Nick Phillips -- [EMAIL PROTECTED]
Seriousness is in the eye of the beholder.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Frank Mittelbach]

Nick Phillips writes:
 > On Fri, Jul 19, 2002 at 04:04:33PM +0100, David Carlisle wrote:
 > > It is not reasonable that the author of a package such as
 > > "indentfirst.sty"
 > > for example (which consists of exactly 4 TeX tokens) should be expected
 > > to go to the trouble of trying to legally register that name.
 > 
 > No, agreed, but...

I think what needs to be added here is that indentfirst.sty is one of the
"extreme" cases. At the other end of the spectrum (also extreme) are very huge
packages. But both extremes are extremes. the majority of packages are in the
middle and it seems hopeless to think that for the

[EMAIL PROTECTED]:~ > locate /usr/TeX/texmf/tex/*.sty | wc -l
   1375

it seems hopeless to think that registered names would be possible.  (.sty =
 latex packages to be loaded by \usepackage)



 > > the LPPL as it is currently drafted gives other users some assurance
 > > that if their documents have
 > > \usepackage{indentfirst}
 > > then their documents will behave as they expect (indenting the first
 > > paragraph of sections) and not invoke some other completely different
 > > code that someone thought was an "improvement".
 > 
 > This falls down because the "indentfirst" package is likely to be a 
 > relatively
 > small work, and until and unless it is incorporated into the LaTeX
 > distribution, there is absolutely no assurance that someone else somewhere
 > else won't come up with something with the same name, similar intent, but
 > critically-different results, without having heard of the first package.

actually it is part of the tools-distribution of LaTeX but i don't see that
this is any guarantee either

 > In short, the protection of which you speak is only really present when the
 > package is included in a distribution large and widespread enough to justify
 > a trademark in any case.
 > 
 > Presumably CTAN will only provide one package with any given name, and this
 > fact alone is enough to provide the degree of protection required for such
 > a package, is it not?

It is not. The combination of three things make it work

 a) LPPL
 b) literate programming conventions within the LaTeX community
 c) CTAN

why is that so?

 1) if the person knows about indentfirst he kows that he is not allowed to
   modify it without giving it a new name. he is of course allowed to have
   some completely other work given that name, but he will know that CTAN is
   going to suggest to give it a different name (so that latex user can use
   both packages) --- he is of course free to distribute it from some other
   archive.

 2) if he doesn't know about it, he might start in good faith with his
   "indentfirst". if he uses dtx as documentation system (as most people do)
   then when turning .dtx into .sty latex will warn him that it already knows
   some indentfirst.sty. This will bring us back to 1) early in the game. of
   course there is no guarantee for this whatsoever so he might find out only
   later when submitting to CTAN.


The practice has shown that this combination works. 

CTAN conventions alone are not enough. They would keep CTAN sane but they
would not ensure that major or smaller distributions of LaTeX start to add
their own changes without renaming the involved files. They are entitled to
make such changes and in fact encouraged to do so, but not in a
way that the changed packages appear from within a LaTeX system ie when
running a document requesting the original package as being that original.

The problem is that in 99% of the cases the wish to make a modification is not
due to a clearcut bug fix but involves an aesthetic component. Note that we do
not argue the validity of a wanted change. we argue that it is best if that
change is offered identically to everybody and by requiring a name change this
is automatically happening as then if that changed package is used anywhere in
a document by somebody  the situation is either

 - it compiles at a certain site because the changed package is installed
   there, or 

 - one gets the information that package X doesn't seem to be installed

first is perfect, second is good and at least much better than haveing the
document behave differently in subtle or less subtle ways.

Note that there is no question of quality involved. The individual quality of
varioref versus varioref2 might be better/worse/uncomparable. The importance
is that it is either identically available at all sites when running LaTeX or
absent, but not slightly different.


 > If there are a large number of third parties using the LPPL, all the more

of the  1375 packages used as an example above 416 directly refer to LPPL and
at at least 400 hundred more are style files belonging to bigger packages
where the whole package is put under LPPL. Then there are a number of styles
which have (if you like a simplified LPPL) lines like

%%%   copyright = "Copyright 1995, 2000 American Mathematical Society,
%%%all rights reserved.  Copying of this file 

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

Boris Veytsman <[EMAIL PROTECTED]> writes:

>B. The *name* TeX is reserved for Knuth's program. If you program
>is called TeX, it must satisfy triptest. You can NOT correct bugs
>in this program, you cannot do Debian QA for it -- you either take
>it as is or rename it.

No.  You are quite wrong.  Provided it still passes triptest, you can
call it TeX.  You certainly can correct bugs or do Debian QA, provided
the changes still pass triptest.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

David Carlisle <[EMAIL PROTECTED]> writes:

> LaTeX is a document markup language the primary aim is to have
> portable documents. Thus anything that claims to be latex (or tex, or
> the computer modern fonts) should produce the same output.

But you have *no* way to assure this, short of trademarking the name
"latex". 

I can write something which is radically (or minorly) deviant from
latex, but not a derivative work, and I can totally fudge up your
goal.

Indeed, I can do two things:

Make a derivate work of latex, which is variant, and called
"special-non-latex".  

Make a package with no derivatives of latex at all, which contains a
single symlink: 'latex -> special-non-latex'.

Happy with that?


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

Boris Veytsman <[EMAIL PROTECTED]> writes:

> > Date: Thu, 18 Jul 2002 04:15:20 -0400
> > From: Glenn Maynard <[EMAIL PROTECTED]>
> 
> > 
> > If the core can be changed in any way without changing it directly,
> > then you can break output exactly as well by this mechanism as you
> > could by editing it directly.
> > 
> 
> No, because to change the core you need to make an explicit decision,
> like using a command
> 
> newlatex file.tex
> 
> instead of
> 
> latex file.tex
> 
> LaTeX team does not mind you using or distributing a changed
> LaTeX. It does not want you to use it AND call it LaTeX.

Ah, but it can't do anything about me making a totally separate
package, calling it LaTeX, and having it invoke my
changed-LaTeX-with-a-different-name.

Thomas


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Frank Mittelbach]

Thomas Bushnell, BSG writes:
 > David Carlisle <[EMAIL PROTECTED]> writes:
 > 
 > > LaTeX is a document markup language the primary aim is to have
 > > portable documents. Thus anything that claims to be latex (or tex, or
 > > the computer modern fonts) should produce the same output.
 > 
 > But you have *no* way to assure this, short of trademarking the name
 > "latex". 
 > 
 > I can write something which is radically (or minorly) deviant from
 > latex, but not a derivative work, and I can totally fudge up your
 > goal.
 > 
 > Indeed, I can do two things:
 > 
 > Make a derivate work of latex, which is variant, and called
 > "special-non-latex".  
 > 
 > Make a package with no derivatives of latex at all, which contains a
 > single symlink: 'latex -> special-non-latex'.
 > 
 > Happy with that?

yes.

for the kernel it is a bit tricky, but for packages under LPPL (and the
majority of software which was put by their authors under LPPL) it is not a
problem.

the moment somebody has a document that loads your fudged package into LaTeX ,
LaTeX will detect that you are trying to sail under a stolen flag and that is
the whole purpose. 

Note that there is no intention to discriminate against producing a better or
even only different version of a package.  the intention is to ensure the
users expectation  that if he/she puts a document through two LaTeX systems it
will  either

 - produce the same results
 - or stop and tell that some component (for example your new package derived
   from some other package) is not available at one site

frank


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Frank Mittelbach]

Thomas Bushnell, BSG writes:
 > Boris Veytsman <[EMAIL PROTECTED]> writes:
 > 
 > >B. The *name* TeX is reserved for Knuth's program. If you program
 > >is called TeX, it must satisfy triptest. You can NOT correct bugs
 > >in this program, you cannot do Debian QA for it -- you either take
 > >it as is or rename it.
 > 
 > No.  You are quite wrong.  Provided it still passes triptest, you can
 > call it TeX.  You certainly can correct bugs or do Debian QA, provided
 > the changes still pass triptest.

sorry but I fear it's you that is quite wrong. The triptest is only there to
help you determine that your implementation is okay. you are neither allowed
to fix bugs or add extra features (new commands, or whatever).

theroretically (now Goedel turns up again:-) a program is only allowed to
call itself TeX if it produces for all inputs exactly the same output compared
to the master copy in stanford (there are technically a bunch of exceptions
related to floating point stuff in dvi production, but that isn't related to
the argument.

have a look at Don's home page. there you find that upon his death TeX version
number goes up to \pi and from there on all bugs are by definition features.

of course you are allowedto rename the source files and produce whatever you
wish from them, but you are not allowed to call the resulting thing "TeX"
again not even if you have the most valid bug fix upon your sleave.

frank


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

Frank Mittelbach <[EMAIL PROTECTED]> writes:

>  > No.  You are quite wrong.  Provided it still passes triptest, you can
>  > call it TeX.  You certainly can correct bugs or do Debian QA, provided
>  > the changes still pass triptest.
> 
> sorry but I fear it's you that is quite wrong. The triptest is only there to
> help you determine that your implementation is okay. you are neither allowed
> to fix bugs or add extra features (new commands, or whatever).

>From tripman.tex:

  If somebody claims to have a correct implementation of \TeX, I will not
  believe it until I see that \.{TRIP.TEX} is translated properly.
  I propose, in fact, that a program must meet two criteria before it
  can justifiably be called \TeX: (1)~The person who wrote it must be
  happy with the way it works at his or her installation; and (2)~the
  program must produce the correct results from \.{TRIP.TEX}.

  \TeX\ is in the public domain, and its algorithms are published;
  I've done this since I do not want to discourage its use by placing
  proprietary restrictions on the software. However, I don't want
  faulty imitations to masquerade as \TeX\ processors, since users
  want \TeX\ to produce identical results on different machines.
  Hence I am planning to do whatever I can to suppress any systems that
  call themselves \TeX\ without meeting conditions (1) and~(2).
  I have copyrighted the programs so that I have some chance to forbid
  unauthorized copies; I explicitly authorize copying of correct
  \TeX\ implementations, and not of incorrect ones!

Sure sounds to me that if numbers (1) and (2) have been met, it can be
called TeX.

The issue is not about bug fixes or extensions, but about whether
something is "faulty", and the test--as here carefully specified by
Knuth--is whether you are happy with how it works for you, and it must
produce the canonical output from the trip test.

Is there something that contradicts that?

Thomas


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Alexander Cherepanov]

21-Jul-02 01:29 Frank Mittelbach wrote:
> Thomas Bushnell, BSG writes:
 >> Indeed, I can do two things:
 >>
 >> Make a derivate work of latex, which is variant, and called
 >> "special-non-latex".
 >>
 >> Make a package with no derivatives of latex at all, which contains a
 >> single symlink: 'latex -> special-non-latex'.
 >>
 >> Happy with that?

> yes.

> for the kernel it is a bit tricky, but for packages under LPPL (and the
> majority of software which was put by their authors under LPPL) it is not
> a problem.

> the moment somebody has a document that loads your fudged package into
> LaTeX ,
> LaTeX will detect that you are trying to sail under a stolen flag and that
> is the whole purpose.

As I understand it's allowed under LPPL to rename, for example,
article.cls to article.fcl and then hack it. And article.fcl
will naturally contain

  \ProvidesClass{article}

(and not

  \ProvidesClass{article-hacked}

). Now let's create a new article.cls (from scratch) which only
contains

   \input{article.fcl}

(there is no need for symlinks really:-)
After that there is no way for LaTeX to detect this hacked class.

This works of course for any package and for the kernel. But for the
kernel the situation is worse. To hack the kernel one can create a new
file which starts with \input latex.ltx [1] and then contains all
necessary redefinitions. The point is that LPPL does not require to
change any identification string or typeout banner in this case
because no LPPLed file is changed.
And this created file can be named latex.tex for example, or .fmt
file after dumping format can be renamed to latex.fmt, or something
like that, so that starting latex as usual will load a hacked format.

[1] In fact it's necessary to redefine \dump temporary because
latex.ltx ends with \dump.

Did I misunderstand LPPL?

Sasha




-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Steve Langasek]

On Sun, Jul 21, 2002 at 01:29:36AM +0200, Frank Mittelbach wrote:
> Thomas Bushnell, BSG writes:

>  > Indeed, I can do two things:

>  > Make a derivate work of latex, which is variant, and called
>  > "special-non-latex".  

>  > Make a package with no derivatives of latex at all, which contains a
>  > single symlink: 'latex -> special-non-latex'.

>  > Happy with that?

> yes.

> for the kernel it is a bit tricky, but for packages under LPPL (and the
> majority of software which was put by their authors under LPPL) it is not a
> problem.

> the moment somebody has a document that loads your fudged package into LaTeX ,
> LaTeX will detect that you are trying to sail under a stolen flag and that is
> the whole purpose. 

Are you using the word 'package' here in the same sense as Thomas?
AIUI, Thomas is referring to creating a Debian package -- not a TeX
package -- that is called 'latex' and which provides a mechanism (a
symlink or an execve hack) for directly invoking his
modified-and-renamed version of LaTeX by the original name.  Would LaTeX
really be able to detect this subterfuge?

Steve Langasek
postmodern programmer


pgpyI1dx0zPbV.pgp
Description: PGP signature

Re: User's thoughts about LPPL

[Walter Landry]

Frank Mittelbach <[EMAIL PROTECTED]> wrote:
> Thomas Bushnell, BSG writes:
>  > Indeed, I can do two things:
>  > 
>  > Make a derivate work of latex, which is variant, and called
>  > "special-non-latex".  
>  > 
>  > Make a package with no derivatives of latex at all, which contains a
>  > single symlink: 'latex -> special-non-latex'.
>  > 
>  > Happy with that?
> 
> yes.
> 
> for the kernel it is a bit tricky, but for packages under LPPL (and the
> majority of software which was put by their authors under LPPL) it is not a
> problem.
> 
> the moment somebody has a document that loads your fudged package
> into LaTeX , LaTeX will detect that you are trying to sail under a
> stolen flag and that is the whole purpose.

I must have missed something.  Why will the special-non-latex detect
that it is not the real latex?  I didn't think that it had internal
checks (which I could take out anyway).  Why would the end user know
that anything different has happened?  Unless they read the output,
which I thought the LPPL people thought didn't happen enough.

Regards,
Walter Landry
[EMAIL PROTECTED]

p.s. Thanks a lot, Frank, for keeping up the discussion.  I'm amazed
at your stamina.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Frank Mittelbach]

Thomas Bushnell, BSG writes:
 > >From tripman.tex:
 > 
 >   If somebody claims to have a correct implementation of \TeX, I will not
 >   believe it until I see that \.{TRIP.TEX} is translated properly.
 >   I propose, in fact, that a program must meet two criteria before it
 >   can justifiably be called \TeX: (1)~The person who wrote it must be
 >   happy with the way it works at his or her installation; and (2)~the
 >   program must produce the correct results from \.{TRIP.TEX}.
 > 
 >   \TeX\ is in the public domain, and its algorithms are published;
 >   I've done this since I do not want to discourage its use by placing
 >   proprietary restrictions on the software. However, I don't want
 >   faulty imitations to masquerade as \TeX\ processors, since users
 >   want \TeX\ to produce identical results on different machines.
 >   Hence I am planning to do whatever I can to suppress any systems that
 >   call themselves \TeX\ without meeting conditions (1) and~(2).
 >   I have copyrighted the programs so that I have some chance to forbid
 >   unauthorized copies; I explicitly authorize copying of correct
 >   \TeX\ implementations, and not of incorrect ones!
 > 
 > Sure sounds to me that if numbers (1) and (2) have been met, it can be
 > called TeX.

it does so.  i once made the remark that the way tripman alone is phrased I
can write a program that

 - reads in files (and ignores their content)
 - writes out two or three files by dumping the results expected by TRIP.TEX

then i only have to feel happy about it to be able to call it TeX. :-) in
other words you can always trip wordings (as several try to prove to me too)

Don has clarified the definition of what is TeX and what not on several
occasions. You may be right that it is not codified in a license (though the
text in the tripman isn't a license either)


 > The issue is not about bug fixes or extensions, but about whether
 > something is "faulty", and the test--as here carefully specified by
 > Knuth--is whether you are happy with how it works for you, and it must
 > produce the canonical output from the trip test.
 > 
 > Is there something that contradicts that?

i think so yes, for example, Don's home page 
other may be able to refer you to more explicit quotes.

frank


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Frank Mittelbach]

Steve Langasek writes:
 > On Sun, Jul 21, 2002 at 01:29:36AM +0200, Frank Mittelbach wrote:
 > > Thomas Bushnell, BSG writes:
 > 
 > >  > Indeed, I can do two things:
 > 
 > >  > Make a derivate work of latex, which is variant, and called
 > >  > "special-non-latex".  
 > 
 > >  > Make a package with no derivatives of latex at all, which contains a
 > >  > single symlink: 'latex -> special-non-latex'.
 > 
 > >  > Happy with that?
 > 
 > > yes.
 > 
 > > for the kernel it is a bit tricky, but for packages under LPPL (and the
 > > majority of software which was put by their authors under LPPL) it is not a
 > > problem.
 > 
 > > the moment somebody has a document that loads your fudged package into 
 > > LaTeX ,
 > > LaTeX will detect that you are trying to sail under a stolen flag and that 
 > > is
 > > the whole purpose. 
 > 
 > Are you using the word 'package' here in the same sense as Thomas?

clash of meanings for names :-) I was talking of "packages in the LaTeX sense"
not of packaged works by Debian.

 > AIUI, Thomas is referring to creating a Debian package -- not a TeX
 > package -- that is called 'latex' and which provides a mechanism (a
 > symlink or an execve hack) for directly invoking his
 > modified-and-renamed version of LaTeX by the original name.  Would LaTeX
 > really be able to detect this subterfuge?

no LaTeX would not, but "LaTeX" wouldn't be involved in this example.

by producing "special-non-latex" you are required to change its identifaction
strings which means that this program will identify itself to the user as
"not-latex" no matter what it is called as a debian package. Of course by
packaging it with a different name and/or on operating level symlinking it to
some other name you can clearly try to (and probably succeed in several case)
confuse your users (receivers of your Debian packages) that they run "latex"
as it would run on other machines.

but if the user calls 

 latex foo.tex

he would then see "This is special-non-latex version xyz" on the terminal.
Not perfect but there you are. And if this Debian package also provides a
special terminal driver that whenever it sees the string "special-non-latex"
replaces it with LaTeX2e then the fudging will be even more difficult to
detect ... yes i agree there are always ways to fudge

but back to what i was referring to as a package: a LaTeX work like
"geometry.sty". such a work is usable within LaTeX by loading it. There
symlinking isn't working, though Alexander Cherepanov correctly pointed out
that with enough energy you can built wrappers within LaTeX to make it go
undetected by the standard mechanisms and becomes visible only by explicitly
looking through LaTeX log files.

Walter, i think, asked if one can't remove that checking code through another
(independent) modification. The answer is yes, easily, but only by either

 - forking the latex kernel, ie running on a non-latex in which this whole
   discussion is irrelevent because users starting with such a kernel know
   that they can't expect their documents to behave same as if run under the
   latex kernel, or by

 -  loading a package into the document that does this change to the kernel in
which case they know it as well.

btw, would it be acceptable to you if LPPL would say,

 in case of modification you either

  - do what LPPL asks for now (i.e. rename ...), or

  - you keep the LaTeX package file name but replace

\ProvidesPackage{varioref}
[2000/08/22 v1.2g package for extended references (FMi)]

   by

\ProvidesModifiedPackage{varioref}
[2000/08/22 ]
[2000/08/22 v1.2g package for extended references (FMi)]   % that's the  
derived info

frank


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

Frank Mittelbach <[EMAIL PROTECTED]> writes:

> by producing "special-non-latex" you are required to change its identifaction
> strings which means that this program will identify itself to the user as
> "not-latex" no matter what it is called as a debian package. Of course by
> packaging it with a different name and/or on operating level symlinking it to
> some other name you can clearly try to (and probably succeed in several case)
> confuse your users (receivers of your Debian packages) that they run "latex"
> as it would run on other machines.

Heck no!  special-non-latex might have a different identification
string, but one of the jobs of the little "latex" package could be to
mutate the identification string also as it gets printed out.  

Now nobody wants to confuse anybody.  And the kinds of notice
requirements that you might actually want are actually perfectly
reasonable: requiring, for example, modified versions not to claim in
their ID string to be the original; requiring changes to be clearly
marked, and so forth.

> but if the user calls 
> 
>  latex foo.tex
> 
> he would then see "This is special-non-latex version xyz" on the terminal.

Naw, not if this weird "latex" thing also mutated the output of
special-non-latex on the way.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

Frank Mittelbach <[EMAIL PROTECTED]> writes:

>  - reads in files (and ignores their content)
>  - writes out two or three files by dumping the results expected by TRIP.TEX
> 
> then i only have to feel happy about it to be able to call it TeX. :-) in
> other words you can always trip wordings (as several try to prove to me too)

I think "feel happy about" is going to be violated here.  That first
clause might look squishy, but it can't be squished indefinitely.

> Don has clarified the definition of what is TeX and what not on several
> occasions. You may be right that it is not codified in a license (though the
> text in the tripman isn't a license either)

What Don Knuth says is really quite irrelevant.  If he's given
permission to use the name (which tripman sure looks like), then it's
that permission that matters, even if he later regrets his
carelessness.

Thomas


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

Frank Mittelbach <[EMAIL PROTECTED]> writes:

> i think so yes, for example, Don's home page 
> other may be able to refer you to more explicit quotes.

Knuth's home page is large.  Do you have a specific reference?


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Jeff Licquia]

On Sun, 2002-07-21 at 04:20, Frank Mittelbach wrote:
> btw, would it be acceptable to you if LPPL would say,
> 
>  in case of modification you either
> 
>   - do what LPPL asks for now (i.e. rename ...), or
> 
>   - you keep the LaTeX package file name but replace
> 
> \ProvidesPackage{varioref}
> [2000/08/22 v1.2g package for extended references (FMi)]
> 
>by
> 
> \ProvidesModifiedPackage{varioref}
> [2000/08/22 ]
> [2000/08/22 v1.2g package for extended references (FMi)]   % that's the  
> derived info

I don't see any problem with that.  I assume LaTeX would provide
ProvidesModifiedPackage and do something to make it clear that the
package is not "official LaTeX", but would otherwise be equivalent to
ProvidesPackage.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Walter Landry]

Frank Mittelbach <[EMAIL PROTECTED]> wrote:
> Walter, i think, asked if one can't remove that checking code through another
> (independent) modification. The answer is yes, easily, but only by either
> 
>  - forking the latex kernel, ie running on a non-latex in which this whole
>discussion is irrelevent because users starting with such a kernel know
>that they can't expect their documents to behave same as if run under the
>latex kernel, or by

I'm getting really, really confused.  In this case, the only
indication that a person running the modified program will get is
modified id strings and output banners.  Yet earlier you (or someone
defending the LPPL) claimed that that is insufficient to reduce the
confusion of users.  Is it enough, or isn't it?

Regards,
Walter Landry
[EMAIL PROTECTED]


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[David Carlisle]

> But you have *no* way to assure this, short of trademarking the name
> "latex". 

That is a very tired argument.
Of course it is true as written, but it ignores the fact that LPPL has
been remarkably successful in its stated aims.

Prior to the latex2e licence (which from which LPPL was derived)
"latex" could be (and often was) locally modified and re-distributed.
It got so bad by around 1990 that passing a latex document from one site
to another was largely a matter of luck.


I had some of my latex packages modified on ftp sites in the US and I
only discovered this when I started getting bug reports, and finally when
I got the user to send me a full tracing log it turned out the code
had been changed and uploaded on to the ftp site without changing the
version number or any indication as to who'd changed it.

the latex2e licence (later generalised to become the LPPL by replacing
"latex" by "the program", so that other people could easily use it on
contributed packages) stabilised things greatly and now there is again a
real sense in which latex is a stable document format that you can use
to pass focuments round the world in a reliable manner.

Yes if someone wants to deliberately spoof latex using all new code that
mirrors the latex distribution they can do that. But that was never the
danger: the danger was people making well intentioned local
"improvements".

Indeed, I can do two things:


> Make a derivate work of latex, which is variant, and called
> "special-non-latex".  
> 
> Make a package with no derivatives of latex at all, which contains a
> single symlink: 'latex -> special-non-latex'.
> 
> Happy with that?

Happy with the first but not the second which, taken with the first,
would be producing a derived work and re-using the latex name.

David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Glenn Maynard]

On Mon, Jul 22, 2002 at 09:38:47AM +0100, David Carlisle wrote:
> > But you have *no* way to assure this, short of trademarking the name
> > "latex". 
> 
> That is a very tired argument.
> Of course it is true as written, but it ignores the fact that LPPL has
> been remarkably successful in its stated aims.

Why do you believe that moving these things to a separate document,
asking people to do this in good faith, would be any less successful
than trying to force them to in a license?

If they know the license is unenforcable and/or avoidable, then they're
already acting in good faith.

-- 
Glenn Maynard


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

David Carlisle <[EMAIL PROTECTED]> writes:

> Prior to the latex2e licence (which from which LPPL was derived)
> "latex" could be (and often was) locally modified and re-distributed.
> It got so bad by around 1990 that passing a latex document from one site
> to another was largely a matter of luck.

Perhaps latex is a miserably poor interchange format.  Or perhaps
the language needed a clear standard and clear documentation.  After
all, the way the world of C programmers solved this problem was by
careful standardization, not by insisting that there should be Only
One C Compiler.

Perhaps the best thing to do is to have a notice requirement so that
users know what they are getting.

But geez, if the only thing that matters is guaranteeing
interchangeability, no matter what, then just make it propietary and
be done with it.  Then you can pretty much do whatever you please.  Of
course, it's not free.

Freedom includes the right to do things that you (and even I) think
are stupid.  Debian stands for freedom.

> > Make a derivate work of latex, which is variant, and called
> > "special-non-latex".  
> > 
> > Make a package with no derivatives of latex at all, which contains a
> > single symlink: 'latex -> special-non-latex'.
> > 
> > Happy with that?
> 
> Happy with the first but not the second which, taken with the first,
> would be producing a derived work and re-using the latex name.

Naw, they'd be done by totally different people without even any
collusion. 


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[David Carlisle]

> Perhaps latex is a miserably poor interchange format.  Or perhaps
> the language needed a clear standard and clear documentation.  After
> all, the way the world of C programmers solved this problem was by
> careful standardization, not by insisting that there should be Only
> One C Compiler.

It may not be ideal that the language is standardised via its canonical
implementation, and that implementation is in a macro expansion language
that makes every line of the implemenattion visible to the document,
that is, means that any change in implementation will cause some
documents to behave differently. However despite that, latex is a
remarkably good and stable document format, and one can port documents
across the world and across nearly two decades of time with remarkably
few problems. The licencing arrangement is one reason for that success.

> Freedom includes the right to do things that you (and even I) think
> are stupid.  Debian stands for freedom.

So do I.  LaTeX is free software. I want to keep it that way.


> Naw, they'd be done by totally different people without even any
> collusion. 

It is the responsibility of the person distributing the renamed work to
ensure that it is distributed under a licence that avoids it be renamed
back to latex. I would say that the different person who symbolic linked
it back to latex would be breaking that licence (which need not be
LPPL). But I'm not a lawyer, if yow lawyer convinces you that is
allowed, so be it. Any licence can only protect against certain.
Debian did a good job with DFSG in acknowledging that GPL was not
suitable in all circumstances. If you have two licences it's clear that
some things will be allowed in one and not the other, even if both are
Free as defined.

LaTeX is distributed with a free Licence that most independent people
have taken as meeting the DFSG. The Licence has proved very successful
in keeping LaTeX stable while allowing arbitrary modified versions.
It has also proved very popular in the latex community with many third
party packages being distributed under that licence.

In amongst all the mis information and scaremongering, these threads
have shown a few places where the text of the licence could be
improved/clarified, and Frank's long message asks for a few more places
where clarification would be helpful. We can, and no doubt will, re-draft
the licence to address these valid concerns but I don't see how your
message helps in any way to get towards clarification.

Yes the LPPL does not stop anyone writing something from scratch and
calling it latex. The chance that Debian or anyone else would distribute
such a thing is rather small.

As Frank said (about the TeX and Font licences rather than LPPL)

> Now i'm not saything this is legally inforcable the way he said it (i have no
> idea), for TeX there is a trademark, though for Computer Modern there is
> probably none (definitely not for the 72 individual font names. Nevertheless
> Debian wouldn't get a good press if it would generate modified versions of
> such programs and fonts and distributed them under the original names.


David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Javier Bezos]

> On Mon, Jul 22, 2002 at 09:38:47AM +0100, David Carlisle wrote:
>>> But you have *no* way to assure this, short of trademarking the name
>>> "latex". 
>> 
>> That is a very tired argument.
>> Of course it is true as written, but it ignores the fact that LPPL has
>> been remarkably successful in its stated aims.
> 
> Why do you believe that moving these things to a separate document,
> asking people to do this in good faith, would be any less successful
> than trying to force them to in a license?

Experience. The question «how to modify book.sty?» is a FAQ, and
the problem happened before the lppl was introduced (and it was
one of the reasons it was adopted, I think).

Javier


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Javier Bezos]

> Freedom includes the right to do things that you (and even I) think
> are stupid.  Debian stands for freedom.

And lppl is intended to give you the right to do stupid things (yes
you can do them), but without perjudicing the right of all latex
users to have a latex working correctly and with documents which
can be distributed freely.

Maybe I'm wrong, but the rest of the message sounds like
if you wanted to begin a flame war. Do you really thing the
LaTeX maintainers are a mafia? They are trying to improve the
lppl, and I thinks that's nice.

Javier


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Boris Veytsman]

> From: [EMAIL PROTECTED] (Thomas Bushnell, BSG)
> Date: 22 Jul 2002 02:27:04 -0700

> 
> David Carlisle <[EMAIL PROTECTED]> writes:
> 
> > Prior to the latex2e licence (which from which LPPL was derived)
> > "latex" could be (and often was) locally modified and re-distributed.
> > It got so bad by around 1990 that passing a latex document from one site
> > to another was largely a matter of luck.
> 
> Perhaps latex is a miserably poor interchange format.  Or perhaps
> the language needed a clear standard and clear documentation.  After
> all, the way the world of C programmers solved this problem was by
> careful standardization, not by insisting that there should be Only
> One C Compiler.
> 


I am afraid you do not know about the recent history of gcc. 

When cygwin people split their branch from RMS, in couple years the
situation became amusing. At some time I had to have three gcc
compilers on my Linux box. One could compile the kernel and pretty
much nothing else. Another compiled most of other programs but some
fraction that was compiled by the third one. Of course even some code
allowed compilation by all three, the executable would be different --
but who cares?

The situation was resolved when RMS overcame his obstination and
delegated the full responsibility to gcc team. Note that neither C
standard nor any particular properties of the language helped. There
is no guarantee this will not happen again -- actually there is almost
a guarantee it will.

This situation was merely annoying. With LaTeX it would be
disastrous. It would contradict our professed goals.

Again, our goal is exemplified in the way TeX makes calculations. It
shuns floating point numbers precisely because this might produce
slightly different output on different machines. Instead all
calculations are integer with the base 5.36e-7cm. We want our output
to be this exact. 

Again, we consider our goals to be compatible with the free software
goals. If you do not agree with the way we are doing things, this is
fine. But I think you owe us to understand what we are doing and why
we are doing this before making a judgement.

-- 
Good luck

-Boris

The little pieces of my life I give to you, with love, to make a quilt
to keep away the cold.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Frank Mittelbach]

Javier Bezos writes:
 > 
 > > Freedom includes the right to do things that you (and even I) think
 > > are stupid.  Debian stands for freedom.
 > 
 > And lppl is intended to give you the right to do stupid things (yes
 > you can do them), but without perjudicing the right of all latex
 > users to have a latex working correctly and with documents which
 > can be distributed freely.

may I suggest a change of one word in Javier's reply:

 not "working correctly"

 but "working indentically"

as long as the starting point is the LaTeX format. If a fork (which is allowed
and simple without any file cascading problems) happens on the that starting
point then the above isn't any longer true, but then a user who runs with
"pslatex" or "pdflatex" or "debianlatex" knows that he doesn't get that.

it is not that the LPPL tries to enforce "the best way" on the level of an
individual package and i think at least to some here this is the suspicion
still. please for get that. lppl simply ensures that a latex package licensed
under it has a fair chance to be used in "latex" documents all over the world
with identical output (no assumption on how good that package is or anything).


frank


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

Javier Bezos <[EMAIL PROTECTED]> writes:

> And lppl is intended to give you the right to do stupid things (yes
> you can do them), but without perjudicing the right of all latex
> users to have a latex working correctly and with documents which
> can be distributed freely.

Huh?  Even if it were in the public domain, it couldn't impact the
right of latex users to have a working latex, unless you mean that
there is some fundamental human right to have Canonical Latex.  And
surely you don't mean *that*.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Walter Landry]

David Carlisle <[EMAIL PROTECTED]> wrote:
> LaTeX is distributed with a free Licence that most independent people
> have taken as meeting the DFSG.

Where did you get this from?  I have doubts as to whether any
independent people (i.e. not affiliated with Debian or the LaTeX
project) have really considered the issue.

Regards,
Walter Landry
[EMAIL PROTECTED]


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Jeff Licquia]

On Mon, 2002-07-22 at 03:38, David Carlisle wrote:
> 
> > But you have *no* way to assure this, short of trademarking the name
> > "latex". 
> 
> That is a very tired argument.

And this is not?

> Of course it is true as written, but it ignores the fact that LPPL has
> been remarkably successful in its stated aims.

"Tired", I observe, is entirely in the eye of the beholder.

Say, rather:

"I am unwilling to compromise on this matter, or consider other avenues
of meeting our mutual goals."

> Prior to the latex2e licence (which from which LPPL was derived)
> "latex" could be (and often was) locally modified and re-distributed.
> It got so bad by around 1990 that passing a latex document from one site
> to another was largely a matter of luck.

Registering "LaTeX" as a trademark would have given you much more power
(i.e. real power) to discourage such things without requiring such high
standards for others wanting to play around with the code.

> > Make a derivate work of latex, which is variant, and called
> > "special-non-latex".  
> > 
> > Make a package with no derivatives of latex at all, which contains a
> > single symlink: 'latex -> special-non-latex'.
> > 
> > Happy with that?
> 
> Happy with the first but not the second which, taken with the first,
> would be producing a derived work and re-using the latex name.

The second is outside of your power to prevent, as it (or its
equivalent) can be done without reference to your work, and therefore
does not fall under your legal purvey.

For example, here is a Python wrapper that makes no reference to LaTeX:

-
#!/usr/bin/python

import sys
import os

program = sys.argv[0] + "-fake"

os.execv(program, sys.argv)
-

Now, if my hacked LaTeX is named "latex-fake", I can symlink my program
above to "latex", and "latex foo.tex" will invoke my fake LaTeX.  You
have no recourse because you retain neither any legal right to my
program above nor any trademark rights to the name "latex".

Don't reinvent the wheel.  If you want the legal assurance of a
trademark, just go and get one.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Glenn Maynard]

On Mon, Jul 22, 2002 at 01:22:50PM -0500, Jeff Licquia wrote:
> Don't reinvent the wheel.  If you want the legal assurance of a
> trademark, just go and get one.

It seems that people who havn't been willing to act in good faith (eg.
people who wouldn't follow guidelines for this if they didn't appear
to have the force of the copyright) simply havn't been aware that the
intent of the copyright is easily sidestepped, and the Latex people
are willing to depend on this continuing to be true.

-- 
Glenn Maynard


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Jeff Licquia]

On Mon, 2002-07-22 at 11:05, Boris Veytsman wrote:
> > From: [EMAIL PROTECTED] (Thomas Bushnell, BSG)
> > Date: 22 Jul 2002 02:27:04 -0700
> > 
> > Perhaps latex is a miserably poor interchange format.  Or perhaps
> > the language needed a clear standard and clear documentation.  After
> > all, the way the world of C programmers solved this problem was by
> > careful standardization, not by insisting that there should be Only
> > One C Compiler.
> 
> I am afraid you do not know about the recent history of gcc. 
> 
> When cygwin people split their branch from RMS, in couple years the
> situation became amusing. At some time I had to have three gcc
> compilers on my Linux box. One could compile the kernel and pretty
> much nothing else. Another compiled most of other programs but some
> fraction that was compiled by the third one. Of course even some code
> allowed compilation by all three, the executable would be different --
> but who cares?

We, as a project, understand this perhaps better than you do.  We
currently ship three different C compilers for woody: 2.95 in most
cases, 2.96 for certain architectures, and 3.0 for one architecture
(hppa).

But the gcc people have different goals than the LaTeX Project, and such
a situation is acceptable for the short term.

> This situation was merely annoying. With LaTeX it would be
> disastrous. It would contradict our professed goals.

Would it really contradict your professed goals to have three
LaTeX-alike systems floating around, one named LaTeX, one named FooTeX,
and one named BarTeX?

Again: there is no dispute that you have a valid concern in the name.

> Again, we consider our goals to be compatible with the free software
> goals. If you do not agree with the way we are doing things, this is
> fine. But I think you owe us to understand what we are doing and why
> we are doing this before making a judgement.

If I were to accuse LaTeX of anything, it would be a sloppy attitude
towards legality in general.  This isn't new; see the KDE mess of days
past for an even worse example.

We've gone over part of this, in that your license (as recently posted)
doesn't accurately reflect your real intentions, and it's my
understanding that this is being resolved.  But this constant
backchannel about filenames and trademarks and so on is another.

It gets thrown around a lot in passing as a disclaimer, but there's
really good reason for the LaTeX Project to consult a lawyer about some
of this stuff.  I understand that this costs money; OTOH, it seems to
some of us that you aren't clear on the legal ramifications of some of
the stuff you're doing, and a real legal opinion would do wonders to
clear a lot of that up.

(Not that Debian is trying to welsh out of helping you, of course.  I'm
not, anyway.)


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Boris Veytsman]

> From: Jeff Licquia <[EMAIL PROTECTED]>
> Date: 22 Jul 2002 15:02:28 -0500
> 

> 
> Would it really contradict your professed goals to have three
> LaTeX-alike systems floating around, one named LaTeX, one named FooTeX,
> and one named BarTeX?
> 

Of course not. Actually there are several systems floating around
besides LaTeX: AMSTeX, plain TeX, ConTeXt to name a few. There is a
healthy competition between them; eg it seems that (AMS)LaTeX
practically killed AMSTeX. However, a document in AMSTeX will always
remain a document in AMSTeX, which is important for us. 

To say the truth, it is ok to have kgcc, egcs, gcc on the same
system. The problem is, you need to decide what is *the* $CC for each
program.

> 
> It gets thrown around a lot in passing as a disclaimer, but there's
> really good reason for the LaTeX Project to consult a lawyer about some
> of this stuff.  I understand that this costs money; OTOH, it seems to
> some of us that you aren't clear on the legal ramifications of some of
> the stuff you're doing, and a real legal opinion would do wonders to
> clear a lot of that up.
> 


I agree with you. Unfortunately I am not a member of the LaTeX3 team,
so I cannot say what is the financial situation of the project (AFAIK,
not very good) and whether they can afford a consultation. However, it
might be a good idea to finance a lawyer consultation from TUG
funds. Or, if it is impossible, we can probably start a collection. I
pledge to contribute to such collection if it starts.

-- 
Good luck

-Boris

Let us treat men and women well;
Treat them as if they were real;
Perhaps they are.
-- Ralph Waldo Emerson


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Frank Mittelbach]

Jeff,

 > > I am afraid you do not know about the recent history of gcc. 
 > > 
 > > [...]
 > 
 > We, as a project, understand this perhaps better than you do.  We
 > currently ship three different C compilers for woody: 2.95 in most
 > cases, 2.96 for certain architectures, and 3.0 for one architecture
 > (hppa).
 > 
 > But the gcc people have different goals than the LaTeX Project, and such
 > a situation is acceptable for the short term.
 > 
 > > This situation was merely annoying. With LaTeX it would be
 > > disastrous. It would contradict our professed goals.
 > 
 > Would it really contradict your professed goals to have three
 > LaTeX-alike systems floating around, one named LaTeX, one named FooTeX,
 > and one named BarTeX?

the answer from me is no, that is already the case, eg LaTeX, pdflatex,
Omega's lamda, pslatex, and perhaps one or two I missed.

The point is not the fork on that level it is the fork on the package
level. LaTeX users, just as pdflatex users, etc. expect their documents if
processed at one site with LaTeX (or with pdflatex, etc) to come out the same
if processed with LaTeX (or with pdflatex, etc) at another. That is true as
long as the fork of any latex packages is using a new name.

It is also clear to the LaTeX users that if they use a fork which internally
supports arbitrary replacements of such packages (ie one that employs the
general filename remapping features then this feature is gone.

This is also why the trademark argument doesn't really help as a
suggestion. you can trademark latex but that doesn't keep its several hundred
(and growing) subcomponents conformant unless you start trying to trademark
such names as well which isn't really an option. I think this is what David
meant when he said "tired argument" as it was the third or fourth time the the
same chain of argument started to come up (including the same replies).

the trademark LaTeX would be helpful to avoid your python example, ie somebody
deliberately trying to pass a user a forked kernel as the original, but that's
all (might be enough to actually check for applying for some, but it doesn'
make the real problem (as we see it) go away.

 > If I were to accuse LaTeX of anything, it would be a sloppy attitude
 > towards legality in general.  This isn't new; see the KDE mess of days
 > past for an even worse example.

this is probably correct, and may have to be rectified and I think you for
pointing out some of those things or giving me ideas certain stuff.
 > 
 > We've gone over part of this, in that your license (as recently posted)
 > doesn't accurately reflect your real intentions, and it's my
 > understanding that this is being resolved.  But this constant
 > backchannel about filenames and trademarks and so on is another.

yes, i would be glad if that could stop, I tried with 

  http://lists.debian.org/debian-legal/2002/debian-legal-200207/msg00422.html

perhaps some answers to that could bring us finally again to discussing a new
license text.

 > It gets thrown around a lot in passing as a disclaimer, but there's
 > really good reason for the LaTeX Project to consult a lawyer about some
 > of this stuff.  I understand that this costs money; OTOH, it seems to
 > some of us that you aren't clear on the legal ramifications of some of
 > the stuff you're doing, and a real legal opinion would do wonders to
 > clear a lot of that up.

yes, we have to think about that perhaps (and check the financial etc
implications)

 > (Not that Debian is trying to welsh out of helping you, of course.  I'm
 > not, anyway.)

for that i would like to thank you

frank


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

Boris Veytsman <[EMAIL PROTECTED]> writes:

> To say the truth, it is ok to have kgcc, egcs, gcc on the same
> system. The problem is, you need to decide what is *the* $CC for each
> program.

And if it's posix, there's c89, which is guaranteed on Posix systems
to be the ANSI C compiler.  

But there is no reason to prevent a system administrator, site, user,
or group of people, from being non-Posix compliant, and having a
program called c89 which is deviant from ANSI C in some respects.
The system would not be Posix-compliant any more, but so what?  Isn't
it for the users and administrators of the system to decide if they
want a Posix compliant system?

The one thing that would be a disaster is if GCC tried to enforce
Posix naming conventions through license terms!  But this is exactly
what latex is doing...


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

Frank Mittelbach <[EMAIL PROTECTED]> writes:

> The point is not the fork on that level it is the fork on the package
> level. LaTeX users, just as pdflatex users, etc. expect their documents if
> processed at one site with LaTeX (or with pdflatex, etc) to come out the same
> if processed with LaTeX (or with pdflatex, etc) at another. That is true as
> long as the fork of any latex packages is using a new name.

If this were true of all LaTeX users, then nobody would ever have a
need to create a variant of latex under that name, and there would be
no reason to have this clause in the license.

So how about this: I'll stipulate that you actually know what all
LaTeX users want in this regard, and you don't need the license
provision.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[David Carlisle]

> Registering "LaTeX" as a trademark would have given you much more power
> (i.e. real power) to discourage such things without requiring such high
> standards for others wanting to play around with the code.

It wouldn't have given any protection at all to users of the package
longtable (which wasn't at that point part of the core distribution)
who were unfortunate enough to have downloaded the package from 
the main US TeX archive where the file had been modified by someone (I
still don't know who) uploading an "improved" (broken) version.

> The second is outside of your power to prevent, as it (or its
> equivalent) can be done without reference to your work, and therefore
> does not fall under your legal purvey.

As I say I'm not a lawyer, but it seems to me that the wrapper
is done without reference to the original latex but is done (as
originally described) with reference to the derived work.
Your vesion that just aliases every program xxx to xxx-fake
could no doubt claim to to be independent of everything. So be it.
If someone wants to deliberately avoid the spirit whilst following the
letter of any agreement so as to deliberately confuse the end user, then
there is not much that can be done about that.

David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Anthony Towns]

On Mon, Jul 22, 2002 at 09:38:47AM +0100, David Carlisle wrote:
> > But you have *no* way to assure this, short of trademarking the name
> > "latex". 
> Of course it is true as written, but it ignores the fact that LPPL has
> been remarkably successful in its stated aims.
> Prior to the latex2e licence (which from which LPPL was derived)
> "latex" could be (and often was) locally modified and re-distributed.
> It got so bad by around 1990 that passing a latex document from one site
> to another was largely a matter of luck.

Err, are you sure this is largely due to the license change, and not to
other changes in the Unix world? Since 1990 we've had the rise-and-rise
of Linux and BSD systems which tend to include tetex as part of the core
distribution, and don't need any modification. Likewise, sunfreeware.com
seems to have been founded in 1993 or so, with possibly a similar effect
on the Solaris types.

Cheers,
aj, who was still in primary school in 1990...

-- 
Anthony Towns <[EMAIL PROTECTED]> 
I don't speak for anyone save myself. GPG signed mail preferred.

 ``If you don't do it now, you'll be one year older when you do.''


pgp29ulSQR0qP.pgp
Description: PGP signature

Re: User's thoughts about LPPL

[David Carlisle]

> Err, are you sure this is largely due to the license change, and not to
> other changes in the Unix world? 

I don't want to disapoint you but it's most likely true that most tex
use doesn't happen in the unix world:-) (although as it happens a good
part of latex was written on a Debian system...)

It's hard to guess exactly why things happen, but conversely that
stability has been very important and I wouldn't want to risk it
lightly by a careless change to the licence.

It seems to me that we can address all the  technical issues that have
been raised with LPPL.

I think the one area where there will probably be disagreement is over
the "renaming rule" however that eventually gets worded.

However the disagreements there from the Debian side seem to be
characterisable as "it can't work" or "I'd have no respect for someone
who uses such a licence". It seems to me that's fine, we're not asking
people to think it's necessarily the best licence in the world and use
it on all their software. Just to accept that it is a free licence in
the sense of the DFSG. That is the good thing about the guidelines;
they acknowledge that one licence can not be suitable for all programs
and that a range of licences needs to be supported. It was always the
intention that LPPL met the guidelines and so it seems unlikely that
given a relatively small amount of rewording we can not get something
that everyone does agree meets the guidelines.

As someone commented earlier just meeting the guidelines isn't
necessarily enough for Debian to distribute the program in their free
tree. Which is also true, It's up to Debian to decide how to lay out
their distribution and can take whatever decision they want on a case by
case basis.

David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Thomas Bushnell, BSG]

David Carlisle <[EMAIL PROTECTED]> writes:

> I think the one area where there will probably be disagreement is over
> the "renaming rule" however that eventually gets worded.
> 
> However the disagreements there from the Debian side seem to be
> characterisable as "it can't work" or "I'd have no respect for someone
> who uses such a licence". 

Um, no.  The real objection is: it's not DFSG free.

The other comments are attempts to convince you to switch, by pointing
out that the requirement doesn't actually achieve your goals, and that
there is a better way to achieve them.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[David Carlisle]

> Um, no.  The real objection is: it's not DFSG free.

Last time I asked for an objective list of places where people thought
LPPL didn't meet the DFSG, someone posted such a list and Frank I think
addressed all the raised points in his last draft, didn't he?

> The other comments are attempts to convince you to switch, by pointing
> out that the requirement doesn't actually achieve your goals, and that
> there is a better way to achieve them.

But the only alternative I've seen offered is registering the name
"latex" which might help for latex itself but isn't really a viable
alternative for all the other LPPL'd packages, many of which are small
(some very small) contributions that people have given in their spare
time, it's not reasonable to ask people to register those names.

David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Colin Watson]

On Tue, Jul 16, 2002 at 07:52:09PM -0400, Boris Veytsman wrote:
>If Debian wants to declare (and presumably delete from the main
>distribution) the software under this license, it would be
>hypocrisy to keep TeX and fonts. I wonder whether people realize
>that this means a complete disaster for the GNU info system? GNU
>info is prepared with a program called texinfo, which is basically
>a special TeX format. 

You only need TeX if you're processing texinfo documents into DVI files,
not for the simple generation of info, HTML, and so on. Thus, although I
would much prefer to see TeX stay in main if possible because of the
knock-on effect it would have on the rest of the distribution, it would
certainly not be any kind of disaster for the GNU info system if we
ended up deciding it was non-free.

-- 
Colin Watson  [EMAIL PROTECTED]


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[Mark Rafn]

> David Carlisle <[EMAIL PROTECTED]> writes:
> > However the disagreements there from the Debian side seem to be
> > characterisable as "it can't work" or "I'd have no respect for someone
> > who uses such a licence". 

I regret making that comment, and I apologize for it.   I intended to say that I have less respect for someone who demands
control over his work than someone who is willing to let it be free.  
This is very different from "I have no respect.".

In fact, I have a great deal of respect for both the latex developers for 
making an extremely useful and stable piece of software, and for the latex 
license people who've been willing to wade into this discussion.  It could 
increase further by coming up with less restrictive license terms, but in 
no case will it become zero. ;)

On 23 Jul 2002, Thomas Bushnell, BSG wrote:
> Um, no.  The real objection is: it's not DFSG free.
> The other comments are attempts to convince you to switch, by pointing
> out that the requirement doesn't actually achieve your goals, and that
> there is a better way to achieve them.

Exactly.  Some of this may have been unclear, as there were parts of my 
messages that were intended to explore whether your intent was actually 
free and parts that hoped to convince you to relax your requirements, and 
I didn't delineate them well.

Opinion of intent:
Understanding your goal a bit , I think I can state that it is not
possible to release software that is both free and prevents users from
being given a modified copy.  The question remains whether an
implementation of these goals can be found that is weak enough to still be
free and strong enough to discourage unnecessary forks.

Recommendation:
Better, IMO, would be for your requirement to be less general than "users 
must expect that documents work the same on all derived works of latex".  
We could easily find a way to accomodate a requirement "users must be able 
to find out that they have a modified version and must be able to get a 
pristine copy of the sources".

I believe that the second goal is all you're going to get anyway (from any
semi-free license), so I urge you to adopt that and make a license that
does it well rather than one that does the first badly.
--
Mark Rafn[EMAIL PROTECTED]  


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: User's thoughts about LPPL

[David Carlisle]

> Understanding your goal a bit , I think I can state that it is not
> possible to release software that is both free and prevents users from
> being given a modified copy.

I agree with that as you write it, but I don't believe that saying you
must call the modified copy something else is the same as saying you
can't modify. And I can't see any reading of the DFSG that would imply
that either.

David

_
This message has been checked for all known viruses by Star Internet
delivered through the MessageLabs Virus Scanning Service. For further
information visit http://www.star.net.uk/stats.asp or alternatively call
Star Internet for details on the Virus Scanning Service.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

spokesman (was Re: User's thoughts about LPPL)

[Frank Mittelbach]

Branden Robinson writes:

 > The ease of alternatives to modifying source code is not important.  The
 > right of the user to create modifications and derived works is.
 > 
 > Interestingly, Frank Mittelbach is asserting that it is not the intent
 > of the LPPL to forbid mere modification of LaTeX source code:
 > 
 > Message-ID: <[EMAIL PROTECTED]>
 > 
 > "it is not the intention to disallow modification without
 > redistribution.  the license has a paragraph labeled recommendation that
 > talks about that but that is already by its title a suggestion only"
 > 
 > Perhaps the LaTeX community should appoint a spokesman to the Debian
 > Project so that we do not get contradictory statements about what is
 > acceptable?

Branden, pardon me, but i think this is funny. seems that you think the LaTeX
community needs as spokesman which is the very thing that I think debian
needs.

as far as this discussion goes a lot of people (some of which with
@debian.org) have contributed in one way or another, are they all spokesmans
for debian? if so I probably give up that this discussion as it feels
pointless and impossible to reach any conclusion. if not then please tell us
who is spokesman for debian (eg you and or Jeff)?

as for spokesmen of the LaTeX community i can't give you any as the community
has members and they have opinions and express them for which I am grateful
even if i don't agree with them always (just as i grateful if members of
debian contribute)

as for spokesmen for the LaTeX Project Team as the copyright owner of LPPL
this is me and David Carlisle.

does that help?

frank


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: spokesman (was Re: User's thoughts about LPPL)

[Jeff Licquia]

On Wed, 2002-07-17 at 01:31, Frank Mittelbach wrote:
> Branden Robinson writes:
>  > Perhaps the LaTeX community should appoint a spokesman to the Debian
>  > Project so that we do not get contradictory statements about what is
>  > acceptable?
> 
> Branden, pardon me, but i think this is funny. seems that you think the LaTeX
> community needs as spokesman which is the very thing that I think debian
> needs.

I'm curious.  From my perspective, the Debian people seem not to have
been contradicting themselves so far; I certainly cannot point out
anything that I would disagree with.  Do you perceive differently?

I'm not trying to make a point, either; if you see us as contradicting
each other, it's likely that we're not communicating our points
effectively enough.

> as far as this discussion goes a lot of people (some of which with
> @debian.org) have contributed in one way or another, are they all spokesmans
> for debian? if so I probably give up that this discussion as it feels
> pointless and impossible to reach any conclusion. if not then please tell us
> who is spokesman for debian (eg you and or Jeff)?

Speaking for myself, I would gladly defer to Branden if it were deemed
necessary for further progress.

> as for spokesmen of the LaTeX community i can't give you any as the community
> has members and they have opinions and express them for which I am grateful
> even if i don't agree with them always (just as i grateful if members of
> debian contribute)

The same is true, I think, of Debian.

One possible important difference: there is, I would imagine, a much
higher degree of consensus about the Debian Social Contract and DFSG
within Debian than I expect there is in the LaTeX user community over
licensing principles.  We debate these issues on a regular basis - very
loudly, in some cases.  I think it's safe to say that the LaTeX Project
probably doesn't have quite as much of an investment into legal
inquiries. :-)


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: spokesman (was Re: User's thoughts about LPPL)

[Boris Veytsman]

> From: Jeff Licquia <[EMAIL PROTECTED]>
> Date: 17 Jul 2002 02:02:25 -0500

> 
> One possible important difference: there is, I would imagine, a much
> higher degree of consensus about the Debian Social Contract and DFSG
> within Debian than I expect there is in the LaTeX user community over
> licensing principles.  We debate these issues on a regular basis - very
> loudly, in some cases.  I think it's safe to say that the LaTeX Project
> probably doesn't have quite as much of an investment into legal
> inquiries. :-)
> 


Note that Debian here consists mostly of developers. I am a *user*,
not a primarily developer of LaTeX

-- 
Good luck

-Boris

 *  This is complicated.  Has to do with interrupts.  Thus, I am
 *  scared witless.  Therefore I refuse to write this function. :-P
-- From the maclinux patch


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: spokesman (was Re: User's thoughts about LPPL)

[Sam Hartman]

> "Jeff" == Jeff Licquia <[EMAIL PROTECTED]> writes:

Jeff> On Wed, 2002-07-17 at 01:31, Frank Mittelbach wrote:
>> Branden Robinson writes: > Perhaps the LaTeX community should
>> appoint a spokesman to the Debian > Project so that we do not
>> get contradictory statements about what is > acceptable?
>> 
>> Branden, pardon me, but i think this is funny. seems that you
>> think the LaTeX community needs as spokesman which is the very
>> thing that I think debian needs.

Jeff> I'm curious.  From my perspective, the Debian people seem
Jeff> not to have been contradicting themselves so far; I

But we do seem to be contradicting our actions and unclear on the
implications of DFSG 4.

I think DFSG 4 means that you can require renaming or patch files of
the sources.  It also seems that you can  require renaming of the distributions.

What Debian finds unacceptable is the assertion that we must break the
TeX or LaTeX API (hey you said it was a language) in order to make
some changes.

I.E.  if we find a bug in article.cls, it's not OK to rename that file
because then \documentclass{article} will either fail or get the old
file rather than our changed version.  The argument is that in
practice we cannot follow DFSG 3 because we cannot change the software
and maintain API compatibility with existing documents if we fix bugs.

The TeX license is OK because it mandates what we call the program,
but does not say anything about the API.  Even if the binary is called
uglytex, it's still easy for me to run it over .tex files.  If those
files use macros defined in plain.tex, those macrso can (at our
option) continue to work in uglytex.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: spokesman (was Re: User's thoughts about LPPL)

[Boris Veytsman]

> From: Sam Hartman <[EMAIL PROTECTED]>
> Date: Thu, 18 Jul 2002 06:20:32 -0400

> 
> The TeX license is OK because it mandates what we call the program,
> but does not say anything about the API.  Even if the binary is called
> uglytex, it's still easy for me to run it over .tex files.  If those
> files use macros defined in plain.tex, those macrso can (at our
> option) continue to work in uglytex.
> 

This is exactly the same with LaTeX. If you create a new format
newlatex.fmt and symlink /usr/bin/tex to /usr/bin/newlatex (this is
the UNIX TeX way to use formats), then you have a complete freedom to
load newarticle.cls whenever your document calls article.cls.

In macro languages you can redefine everything -- this is the beauty
of macro languages.

-- 
Good luck

-Boris

Thank goodness modern convenience is a thing of the remote future.
-- Pogo, by Walt Kelly


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: spokesman (was Re: User's thoughts about LPPL)

[Sam Hartman]

> "Boris" == Boris Veytsman <[EMAIL PROTECTED]> writes:
Boris> This is exactly the same with LaTeX. If you create a new
Boris> format newlatex.fmt and symlink /usr/bin/tex to
Boris> /usr/bin/newlatex (this is the UNIX TeX way to use
Boris> formats), then you have a complete freedom to load
Boris> newarticle.cls whenever your document calls article.cls.

Hmm. An interesting point.  From a purely pragmatic standpoint I
suspect Debian would simply choose to keep its current LaTeX packages
rather than write infrastructure for this sort of support.  It would
certainly be enough to convince me to stop using LaTeX.

It would be ironic if Debian ended up forking LaTeX because of a
license designed to promote standardization.


But you're right, if the LaTeX license allows this  it may be DFSG free.

There's another issue though.  When I discussed TeX, I carefully
avoided discussing whether the TeX license allows us to have a
modified version of TeX (which does not pass trip) invoked by
/usr/bin/tex.  We cannot call that program TeX.  However it is not
clear to me that having an executable called TeX to preserve makefiles
that depend on being able to shell out to /usr/bin/tex counts as
calling a program TeX.

The Debian community may decide that restricting API constants like
filenames or command names violates DFSG 3, by precluding large
classes of derived works where as restricting what a program can be
called in some vague sense referring more to documentation, claimed
feature set, and program output is OK.

We had a bit of a discussion of this issue on IRC and some people
pointed out to me that even if requiring renaming of files is OK, LPPL
may still violate DFSG 3 on some technicalities.  IF we resolve all
the big issues we can come back to this point.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: spokesman (was Re: User's thoughts about LPPL)

[Thomas Bushnell, BSG]

Sam Hartman <[EMAIL PROTECTED]> writes:

> But you're right, if the LaTeX license allows this  it may be DFSG free.

And, we can create another package that has a symlink from latex to
patched-not-really-latex.  That other package would not be, in any
way, a derivate of latex, and thus it isn't subject to the naming
requirements on latex derivatives.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

what is allowed with TeX and CM fonts (was Re: User's thoughts about LPPL)

[Frank Mittelbach]

Thomas Bushnell, BSG writes:
 > Frank Mittelbach <[EMAIL PROTECTED]> writes:
 > 
 > > i think so yes, for example, Don's home page 
 > > other may be able to refer you to more explicit quotes.
 > 
 > Knuth's home page is large.  Do you have a specific reference?

sorry, seems i have thrown you a red herring myself. His intentions seem to be
only briefly touched at on his home page (the reference to TeX gets absolutely
frozen upon his death is there but not much more)

i took the effort (though i think this is a different discussion therefore
finally changed the subject) to try and pin down some references. best i was
able to do is


@Article{Knuth:TB11-4-489,
  author =   "Donald E. Knuth",
  title ="{The future of {\TeX} and {\MF}}",
  journal =  j-TUGboat,
  volume =   "11",
  number =   "4",
  pages ="489--489",
  month =nov,
  year = "1990",
  ISSN = "0896-3207",
  bibdate =  "Wed Jul 18 18:33:43 MDT 2001",
  bibsource ="ftp://ftp.math.utah.edu/pub/tex/bib/tugboat.bib;
 
http://www.math.utah.edu/pub/tex/bib/index-table-t.html#tugboat";,
  acknowledgement = ack-bnb # " and " # ack-nhfb,
}

in there he say (beside other things):

As stated on the copyright pages of Volumes B, D, and E, anybody can make use
of my programs in whatever way they wish as  long as they do not use the names
TeX, METAFONT, or Computer Modern. 

The copyright page of volume B then says:

 The progam for TeX is in the public domain and readers may freely incorporate
 the algorithms of this book into their own programs. However, the use of the
 name TeX is restricted to software systems that agree exactly with the program
 presented here.

that seems to me more than trip test complience.

while I'm at it, the copyright page for volume E is similar and states that
any font named cmr10 (and so on) has to be fully compatible with the one
produced from the programs in the book.

Now i'm not saything this is legally inforcable the way he said it (i have no
idea), for TeX there is a trademark, though for Computer Modern there is
probably none (definitely not for the 72 individual font names. Nevertheless
Debian wouldn't get a good press if it would generate modified versions of
such programs and fonts and distributed them under the original names.


frank


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: what is allowed with TeX and CM fonts (was Re: User's thoughts about LPPL)

[Henning Makholm]

Scripsit Frank Mittelbach <[EMAIL PROTECTED]>

> probably none (definitely not for the 72 individual font names. Nevertheless
> Debian wouldn't get a good press if it would generate modified versions of
> such programs and fonts and distributed them under the original names.

Please avoid the fallacy of assuming that every right required by the
DFSG is something that some in the Debian project itself intends to
actually do. The requirements are there much more to make sure our
users have them than to make sure *we* have them.

[This being said without implying anything about the freedom or lack
thereof of the Computer Modern fonts].

-- 
Henning Makholm  "Det er jo svært at vide noget når man ikke ved det, ikke?"


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: what is allowed with TeX and CM fonts (was Re: User's thoughts about LPPL)

[Frank Mittelbach]

Henning Makholm writes:
 > Scripsit Frank Mittelbach <[EMAIL PROTECTED]>
 > 
 > > probably none (definitely not for the 72 individual font names. 
 > > Nevertheless
 > > Debian wouldn't get a good press if it would generate modified versions of
 > > such programs and fonts and distributed them under the original names.
 > 
 > Please avoid the fallacy of assuming that every right required by the
 > DFSG is something that some in the Debian project itself intends to
 > actually do. The requirements are there much more to make sure our
 > users have them than to make sure *we* have them.

I wasn't intending to imply that. but i was implying that (at least by moral
rights) you don't have them in this particular case, or in other words that
you are faced with a similar situation compared to LPPL and that is not
unimportant.

As it was pointed out several times, we think it is pointless for us to
continue any discussion if at some point in the future TeX and CM fonts are
moved to non-free. Then we would like to see LaTeX there as well.

So it is important that Debian comes internally to a conclusion what their
position to DEK's licenses or interpretation of his licenses is.

 > [This being said without implying anything about the freedom or lack
 > thereof of the Computer Modern fonts].

i was looking for the reference exactly because i would like to see that being
settled as well, as i think it is to some extend a prerequisite --- after all
LPPL was modeled after (in our opinion successful model of the TeX license or,
say, DEK's idea on a license).

cheers
frank

ps answers to your other posts coming up in a different window, stay tuned


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]