🚀 go-pugleaf

n-reply-to: Mike Stump's message of Tue, 7 Feb 1995 10:05:51 GMT

In article <D3MK1s.3vw@kithrup.com> Mike Stump <mrs@cygnus.com> writes:
|
|   And what is that?  Maybe there is some amount of conflict of interest.
|   So what?  I don't think it is unresolvable or hurts any.  The usual
|   conflict of interest is, do I work on fixing Cygnus customer bugs, or
|   do I work on net reported bugs?  In some sense, I run into this
|   problem everyday.  It is resolvable, I donate some of my time
|   (Cygnus's time) to the FSF, not 100%, but not 0% either.  I find the
|   amount I can donote to them, and I donate.  I have few complaints from
|   the FSF that my donation is too little.  I think they understand that
|   what I am able to donate is limited.  I wish it weren't.
|
|   Now, as to why this isn't a problem in the grand scheme of things is
|   that Cygnus customers have hit, or will hit the exact same bugs in g++
|   that people on the net will hit, and visa versa.  No _real_ conflict
|   of interest exists in this area.  Just an apparent conflict, or a
|   conflict with respect to a specific bug.
|
|   ...
|   Now as to control over the FSF's software products.  Cygnus has
|   substantial (no, I didn't say absolute) control over some of the FSF
|   software products.  In part, because we have dumped oodles of cash
|   into FSF software products, we were able to attain maintainership
|   status.  We `want' maintainership status, as it does make it easier to
|   install changes.  To us, it is economically more efficient that we are
|   the maintainers.
---

Pre-statement disclaimer: Nothing in the following is intended to
suggest any unethical or disingenuous behavior by Cygnus; everything I
know about the Cygnus folks is positive, technically, ethically,
and in politeness.

That said, the FSF relationship with Cygnus clearly does present the
*potential* for abuse.  Cygnus controls the official version of the
product and Cygnus has paying customers whose interests may be better
served if their code changes or support for their products is available
in the official product before changes or support for other vendors'
products.  That is, if computer vendor X has developed code generation
changes that make GNU C generate higher performance code than the
current, official version, and Cygnus has customers who are compeititors
of vendor X, for whom it has contracted to develop better code
generation for their versions of GNU C, there is a clear conflict of
interest in Cygnus being in control of what code gets integrated into
the official version when.

Again, I have no reason to believe Cygnus has ever said "Well, if you
pay us $100K, your code will get into the official release in a month,
otherwise, it may get in someday."  It is the *potential* for abuse that
worries me.  The conflict of interest clearly does exist, whether it is
causing real problems or not.  I actually believe that Cygnus does look
after the best interests of the official GNU releases of the products it
controls and that Cygnus honestly tries to get changes into the product
as quickly as possible, regardless of their source.  The ethical
dilemma, however, is not one I would want to have to live under.

Note that this isn't really specific to Cygnus - Cygnus is simply more
visible than the individuals who control most of the individual pieces
of the GNU produt list.  This same dilemma affects each of those other
components and, of course, the ones FSF controls, as well.  In each case
decisions have to be made about priorities for developing and for
integrating contributed changes; in each case there is a balancing of
the interests of paying customers and others.  I don't see any practical
way of avoiding that conflict (the only thing that comes to mind is
having the code owned by a foundation that doesn't accept corporate
contributions or individual contributions larger than some relatively
low limit).  I hope the FSF is alert to the possibility, though, since
if such abuse did occur it would inevitably be the FSF's reputation that
was hurt.

Post-statement disclaimer (what I tell you three times must be true,
neh?): my concern is solely that there is the potential for abuse; I do
not believe that Cygnus has done anything even vaguely inappropriate.

scott

--
scott preece
motorola/mcg urbana design center	1101 e. university, urbana, il   61801
phone:	217-384-8589			  fax:	217-384-8550
internet mail:	preece@urbana.mcd.mot.com
--
scott preece
motorola/mcg urbana design center	1101 e. university, urbana, il   61801
phone:	217-384-8589			  fax:	217-384-8550
internet mail:	preece@urbana.mcd.mot.com