🚀 go-pugleaf

All standard disclaimers (including that I don't work for the FSF in
any capacity) apply here.

propf@zeus.towson.edu (Gregory Propf) writes:

>I'm the guy that started this thread and I'm still confused.  I am now
>told that libc.a is *not* a GNU library under Linux?!?!  What's all
>this about the 'vendor' libraries.  Are you guys telling that the
>libraries were not written by the GNUs?  If so who does write the
>lib*.a stuff under Linux?  Does Torvalds himself hand code the stuff
>in 32-bit assembler.  I would like to get a real answer to this before
>I start working.  I'd hate to write something for the market and then
>be forced to make it free because I used a few library functions.

I'm not a major Linux user, but as I understand, Linux uses the GNU C
library, perhaps a special branch of it.

The talk about "vendor" libraries appears to be in reference to other
(i.e. proprietary) UNIX-like systems, which usually have libraries
that either come with them by default, or you can buy as part of a
development kit.

>This is one case where I think Microsquash has it all over the FSF.
>When I learned to code in C I went out and bought Quick-C 2.something.
>It cost me about $60 and considering that I only had a 286 it was
>fine.  I wrote thousands of lines of code with that thing.  In the
>docs they basically tell you that you are free to do wahtever you want
>with the libraries.  You can link to them to your heart's content and
>then you sell your code or your binaries or you can post it on the net
>and give it away free.  They don't care what you do with them.

I think you'll find this is incorrect.  If you look at the licenses
included with the software, what you pay for is the privelege of using
their software (this is a lot how their licenses read), you do not own
it, so I understand from the licenses on their packaging.  You simply
have a right to use it on one machine, and maintain enough copies to
ensure you won't lose it easily.  Microsoft and other vendors are
pushing the copyright thing even further...  there is lobbying from
"intellectual property" groups going on right now to move this definition
to count loading it into the machine at run-time as a "copy" of
the software under legal terms, and yet more rumbling in the wings.

My point simply being that they very much care about what you do with
the software.  You are simply allowed to use it in an "appropriate"
manner.

This kind of thing is a point of contention between the
BSD-like license folks and the GNU-like license folks, i.e. totally
free to do what you want with the code vs. enforced freedom of access
to and distribution of source code to all who use the binaries.

A lot of people say that you must include the source code directly with
the binaries if you use the GPL, but I believe it simply says it must
be available for some nominal distribution fee, and you can't restrict
arbitrary redistribution on the binaries or sources...  I'm going to look
it up.

>The assumption is that people buy compilers to write code and then they
>will likely sell it.  They would not sell you a compiler that you
>could only use if you paid them royalties.  And this is Microsquash
>we're talking about here.  Not exactly the most humanitarian bunch of
>people.  The FSF should put their libraries out under a similar
>licence (or anti-licence if you like)

I thought about this long and hard, since I for business reasons I needed
to join their "Developer's Network", as they call it.  They make it pretty
clear that even the use of their stuff is restricted to never looking
too deeply into it to figure out how it works (even for personal interest,
say), using it only in certain conditions (like you are told it is
your responsibility that if the software is installed on a network,
to ensure it is not used simultaneously by more people than you have
licenses, to the point of requiring you to buy a license manager
software package), etc.

The FSF comes from a different point of view.  You can do whatever
you want with the code except restrict someone else's similar rights
once you give it to them.  Somewhere in the explanation about how the
FSF got started with Richard Stallman it points out his intense dislike
of not enabling other programmers to have access to source code to
change as they need or like.

>One of the posters said something really intelligent.  'The LGPL is
>one of the big obstacles to business use of Linux'  or something like it.

How is the LGPL and obstacle?

The LGPL simply states that the source code for the libraries themselves
that are under the license must be available (and likewise freely
re-distributable), *not* the main body of the code.

If you enhance the libraries themselves, then you must make the source
code to those enhancements available as well.

Thus, if you wrote a commercial program using the GNU C library (which
is now under the LGPL), you can license the main body of the program
as you please, but you must make available the libraries and any
enhancements to the libraries you make.  Frankly, I don't see what
all the fuss is about.

>I agree.  Businesses will generate tons of code for an OS if they
>think they can get some of their investment back by selling it.  If
>you can't be sure that you can write a program with your compiler and
>then market it your incentive to write code is stifled.  This is a
>shame because Linux has it all over Win or WinNT in terms of general
>robustness.  I mean I've got 7 apps up under X right now.  Windoze
>crashes after #3 or so and then still can't run all at once... - GP

First of all, neither the GPL nor the LGPL stop you from selling
the code, they just ensure that redistribution of any binaries and
source code covered by them is free.

The main grumble seems to be "why can't we license the code so we can
make money on it?".

The main answer is "that's the whole point of the GPL in particular,
and the LGPL is a weaker statement of the same thing.  If you use the
code as a starting point for work, you must give the same rights you
had to those you give it to.".

I think there are other arguments than the "intellectual property"
ones, myself.  I'm a programmer, and I've been burned so many times
by proprietary code it isn't even funny.  What about protection for
the users, by allowing them to fix problems that you can't or won't
fix?  Such a set of problems is costing my company probably over
$1 million for a new product line that might otherwise not be
necessary.  That isn't even including the upgrade costs of the users
in the field!

In the end I think it is a tough issue, but if you really have problems
with it, there exists the BSD-like copyrights that don't restrict
people taking the code and licensing it as they please.  It's not
fundamentally a problem with free software, mostly a difference of
opinion in what "free" means.

Comments, etc. are welcome.

Erich Boleyn


-- 
Erich Stefan Boleyn                \___  E-mail (preferred):  <erich@uruk.org>
Mad Genius wanna-be, CyberMuffin       \____    home #:  +1 (503) 226-0741
<A HREF=http://www.uruk.org/">WWW Site</A>  |  USnail: 924 S.W. 16th Ave, #202
Motto: "I'll live forever or die trying"     \        Portland, OR, USA  97205