🚀 go-pugleaf

In article <1991Jun19.141134.19083@dg-rtp.dg.com>, wood@gen-rtx.rtp.dg.com (Tom Wood) writes:
> I'm trying to get a handle on the GLGPL.  In particular, would
> building and using a shared version of libg++ be a simple way to meet
> the terms of the GLGPL.  As I see it, an application vendor could, by
> using a shared libg++, not have to do *anything* different in the way
> he does business.  The users can get the sources to the library,
> modifiy it, and not have to deal with the executables that use it in
> any fasion.

Unfortunately, the use of shared libraries makes it virtually impossible
to enforce per-cpu or per-network copying restrictions commonly used
by software hoarders. I have brought this up with RMS in the past,
but he is more interested in persuing the goals of freeing software
directly than in making what he perceives as a concession to
convince computer manufacturers to incorporate GLGPL code.

I mean no criticism of RMS by this: no one is forced to use GLGPL
code unless they want to, and the software hoarders have to reason
to expect the support of RMS and FSF in developing their code.
But in the current less-than-ideal world, a system vendor must
act to make the system attractive to both software sharers and
the software hoarders who sell code that customers demand.

I respect RMS's stand on the principles involved here, but my own
personal opinion is that he could reach his goals more quickly
by permitting software hoarders to produce "works that call
the library" to be distributed without the restrictions in the GLGPL:
this would make it easy for system vendors to choose GNU libraries
as standard equipment on the basis of their quality.
I believe that GNU will succeed by demonstrating that free
software, by better incorporating the best efforts of many
people, is of superior value than hoarded software.

Craig Hansen
craig@netcom.com