🚀 go-pugleaf

Sorry that this post is getting so long, but I thought it wouldn't
make sense if I took out too much of the previous discussion.


In <iRNaeB6w164w@mantis.co.uk> mathew@mantis.co.uk (en font terrible) writes:

>In <1992Jan7.231421.27402@sunb10.cs.uiuc.edu> epstein@sunc2.cs.uiuc.edu (Milt 
>Epstein) writes:
>> In <Pmq7DB12w164w@mantis.co.uk> mathew@mantis.co.uk (en font terrible) writes:
>> >With Microsoft and other commercial licenses, I can use the sample source
>> >code provided in any way I like, and I can do whatever I want with programs
>> >which I have written and linked using their libraries.
>> 
>> Is the first part of this statement really true?  For example, can you
>> take bits of their source code and put them into a program that you
>> are going to sell (without the need to make retribution to them in any
>> way)?  That doesn't sound right to me.
>
>Well, the Quick C for Windows user guide encourages the user to copy 
>one of the sample applications to a new directory and then modify it 
>to build his software on top.  Obviously they expect you to copy their
>sample applications and distribute derivative works based on them.
>
>I've certainly encountered compilers where there was an explicit statement
>saying that you could do what you liked with the sample code.  It's quite
>common to include free stuff.

Someone else also sent me email clarifying the fact that you were
referring to sample code that might come with the compiler.
Unfortunately, when you compare this sample code to GNU code, you are
comparing apples and oranges.  This sample code was explicitly meant
to be used and copied.  However, I bet there isn't the same lack of
restriction on the source code for the compiler itself.  Comparing the
restrictions on the compiler's source to those on GNU source would be
much fairer.


>> >and I can't do whatever I want with programs which I have written and
>> >linked using their libraries.
>> 
>> From a quick read of some files on COPYING GNU stuff I picked up from
>> prep.ai.mit.edu, it looks like any restrictions placed on programs
>> written independently but linked using their libraries have to do with
>> making the libraries available so that the program can be recompiled
>> -- the source for the independently written programs need not be
>> provided. 
>
>If you link your programs with a library under the GLPL, you have to supply
>object code for your programs.

Yes, you have to supply object code, so that customres can recompile
and relink with a perhaps modified library.  But object code is not
source code.


>> Further, they are providing you with the source code and the libraries
>> -- why can't they tell you how you can use code they have written?
>
>Oh, they can.  But I reserve the right to complain when they start telling
>me how I can use code *I* have written.

But basically when you use code that they have written, it as if you
have gone into a joint venture -- you both (you and GNU) have
contributed code, so why can't they specify restrictions on how that
code is used.  You don't have to use their code in the first place,
you know.


>If the GPL said that I had to provide source to any GNU code I used,
>that would be fine.  It's the fact that it says that I have to provide source
>to my code as well that's the problem.

But it doesn't say that -- in the example above, you need to provide
OBJECT code, not source.


>> I don't see how you have answered Jurgen Botz' questions, i.e.
>> basically, how is the GNU license worse than other companies'
>> licenses?
>
>It puts restrictions on me which other companies' licenses don't, and doesn't
>give me much in return (i.e., all it gives me which commercial licenses
>don't is the right to freely copy other people's software).

When you rephrase that as giving everyone the write to freely copy
others' software, it sure sounds like it is giving you (and everyone)
a lot.  It restricts you so that you can't make money and/or not make
source code available when you include some code that someone else
wrote -- but why is that unfair?  That seems like it would benefit
everyone.


>>             If you want/need to use their products so much, then you
>> must put up with their license.  And if you don't like their license,
>> can't you just not use their code?
>
>So why do people complain about AT&T licenses, eh?  Same argument applies,
>does it not?

Sure, and that seems to be a major part of the reason for the GNU
project in the first place.


>> >The FSF tries to get me to overlook these restrictions by bribing me with
>> >the source code for the whole package, and by telling me I can give
>> >away copies of the whole package.
>> 
>> But many people may not consider these to be restrictions.
>
>Can I then claim that many people don't consider "You may not copy this
>software" to be a restriction, and that therefore the AT&T licenses contain
>no restrictions?

You are using some strange logic here -- I did not say that because
many people may not consider the conditions on the GNU license to be
restrictions that therefore we may conclude that there are no
restrictions.  From a practical standpoint, the number (and types) of
people that will be restricted by the conditions on the GNU license
are much fewer than those with a license that has the condition "You
may not copy this software".  With the GNU license, only those who
wish to use GNU code in proprietary (by that I mean restricted source)
code will be affected.  With the other license, any one who wishes to
copy the code (it's not quite clear what you meant by "copy this
software", but it doesn't matter -- the point is clear nonetheless).


>> >Whether you think this is a good deal depends upon how much you value being
>> >allowed to fiddle with other people's source code and copy other people's
>> >software, measured against how much you value being allowed to do what you
>> >want with software you have written (even if you've included bits from
>> >other people's code).
>> 
>> Ahh, but that parenthetical comment makes a big difference -- others    
>> have contributed, so you must accept their conditions for using their
>> contributions.
>
>Right.  That doesn't mean I'm not allowed to disagree with the conditions
>they put on using their contributions, where I perceive the conditions
>to be unreasonable.
>
>> If you want full control over it, shouldn't you have written it yourself?
>
>One of the arguments put forward in favour of GNU is that it will ensure
>that people don't have to reinvent and reimplement things all the time.
>Yet here you are, saying "If you don't like the way I'm telling you to
>distribute your code, go away and reimplement what I've done."

Ahhh, I see now -- you want free access to whatever code you
want/need, but you also want to be able to restrict access to code you
write.  Sounds a little bit hypocritical to me.


>> OK, you've given several examples of types of people that will like
>> the GNU license -- now what types of people will not like it?  One
>> would be someone who wants to use others' work in something that they
>> intend to restrict in some way.
>
>Correct.

So why should they be allowed to control a product that they only
partially wrote?


>> >I'd much rather that the effort being spent on developing FSF software got
>> >spent on developing software that's truly free, so everyone can benefit.
>> 
>> So, who is not benefitting from this setup?
>
>People who need to restrict the redistribution of the software they write
>so that they can recoup the money they had to spend in order to write
>it.  People who want to write software which is free of restrictions.

Well, there's still the alternative of not using their code to begin
with.  And if the only code of theirs you use is the library code when
compiling/linking, you are not required to release your source, so
that doesn't stop you from making your money.

I still don't see that you've given any unbiased reason for finding
the GNU license unreasonable.

-- 
Milt Epstein
Department of Computer Science
University of Illinois
epstein@cs.uiuc.edu