RMS Talk at University of Toronto Mississauga

[Christopher Browne]

On 7/6/07, tleslie <tleslie-RBVUpeUoHUc at public.gmane.org> wrote:
> On Fri, 2007-07-06 at 14:48 -0400, phiscock-g851W1bGYuGnS0EtXVNi6w at public.gmane.org wrote:
> > > Part of the problem lies in the ambiguity that lies in "What does
> > > Linux mean?  Kernel?  Or distribution?"
>
> And if you deem it a distro, and use desktop,
> then by lines of code it should be referred to as
>
> X11/Linux/Gnu ? or just X11 and drop the linux and gnu part as they
> would be so minor :)

Or should it be called X11/KDE/Perl/Linux? The "if it makes a
significant code contribution" argument heads down a pretty slippery
slope, and frankly, one that is quite distinct from the "GNU"
argument.

Part of RMS' argument is that without GCC and GNU fileutils, there
would be no Linux, and nothing else would have gotten connected to it.

While it's true that Linux uses GCC and a pretty full set of GNU
binary and file utilities, I actually *STRONGLY* disagree with part of
the premise.  I would agree that if GCC and other GNU stuff had not
been there, we'd probably not be using today something called "Linux."

However, RMS seems to take the stance that since virtually everyone
uses GCC, there wouldn't be any "free Unix" without it, and I think
that particular position is nonsense.

There were other C compilers at the time that could have grown into
suitable tools, including the following options:

http://en.wikipedia.org/wiki/Portable_C_Compiler
http://en.wikipedia.org/wiki/TenDRA_Compiler
http://www.atari-forum.com/wiki/index.php/Sozobon_C

None were quite as good as GCC has become, but there most certainly
was enough demand for a "free Unix" that if the FSF guys had not
grabbed the Pastel compiler and retargeted it to compile C rather than
Pascal (therefore note: the FSF did not start the project!), I believe
that some other C compiler (or set thereof) would have emerged
instead.  There was enough desire for "free Unix" for this to be a
"historical inevitability."

I believe that the predominance of GCC has actually had some very
*negative* effects; it IS NOT a fast compiler, and it has a complex
enough architecture that for a long time, development on it largely
stalled, and for the last 10 years, it hasn't been the FSF that has
managed GCC.  That RMS "rests on the GCC laurels" as his justification
for demanding credit for the outgrowth of Linux seems mighty weak to
me.  GCC isn't RMS' project any more.  It hasn't been an FSF project
for a long time, either.

For a long time, the effects of GCC pretty much scorched the earth for
any other compiler projects.  People have taken up efforts on TenDRA
again, and there are quite a number of "small C" compiler projects out
there.  GCC was, in the mid '90s, pretty much scorching its own earth,
as there were big fights amongst the developers between complaints by
Ada folk of poor quality work everywhere else, combined with people
doing various pretty incompatible efforts on Pentium-specific
enhancements.  It wasn't until EGCS had some success at drawing
together efforts outside the FSF "fold" that development could really
continue usefully...

> i guess RMS would probably say if you want to short form it
> when you speak, then say  GNU  and don't even say Linux ..
> I wonder if anyone ever asked him that?
>
> Of course where do you draw the line too,
> i mean the C programming language has a pretty big
> impact on the GLX to.

Yeah, and C and Unix came out of Kernighan, Ritchie, Pike, and Bell
Labs.  If inspiration is supposed to be rewarded with getting your
name in the OS, then it seems way MORE dishonest to leave them out
than to leave GNU out.

RMS has some valid points, but he seems to be willing to not notice
some of the bits of history that are inconvenient to his positions...
-- 
http://linuxfinances.info/info/linuxdistributions.html
"...  memory leaks  are  quite acceptable  in  many applications  ..."
(Bjarne Stroustrup, The Design and Evolution of C++, page 220)
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists