SC issues

[Tom Lord <lord at emf dot net>]

   From: kenner@vlsi1.ultra.nyu.edu (Richard Kenner)

       Back before the GCC steering committee, there was a formally neutral
       arbiter, legally bound to operate in the public interest, who decided
       questions about the evolution of the public GCC sources.  

   No, there wasn't.
   [In theory the FSF had that role, in practice it was Kenner doing
    the usual role of a maintainer.]

A few people who saw drafts of the message you're quoting saw an
earlier version of that paragraph and said "Oh, you mean Kenner".  I
didn't (specifically).  I tried to change it to avoid that confusion.
It didn't work.  My recollection of the times before the SC,
especially long before the SC, has RMS (an officer of the FSF) taking
a more involved role than what you describe.


   I actually have no idea what "legally bound to operate in the public
   interest" might mean, but it's certainly not what *any* FSF maintainer
   does.  

In order to avoid a slew of "I am not a lawyer, however..." posts,
let's just say that the FSF operates under definate constraints
(whether legal, chartered, or ethical) which include protecting and
promoting certain interests of the general public.

I disagree with you when you say that no maintainer is bound by those
constraints.  Many of us have, in the past, and some presumably still
do defer to the FSF's decisions about coding priorities, design
decisions, and engineering process while wearing the maintainer hat.
In some cases that's just voluntary cooperation; in other cases it's
an aspect of employment at the FSF.  Even when the FSF makes no
explicit direction, many of us have tried to make our own decisions
giving highest priority to the FSF's goals.

The situation with the SC is both similar and different from historic
maintainerships.  One the one hand, in theory, if the FSF demanded
some direction and the SC refused, they could lose their status as
"official maintainer".  Unlike most historic maintainers though, the
SC and the companies who have employees that are part of the SC have a
lot of power (by virtue of doing a lot of GCC-related work): it isn't
at all obvious that there is any other practical way to choose a GCC
maintainer than to find a person or organization that all those
parties can agree to.

So in effect, GCC is now a commons maintained by a political process
in which individual user's of GCC and companies not listed in the SC
affiliations have no representation.  Moreover, the SC is relatively
toothless: it's effectively permitted the smallest workable amount of
sovereignty by the companies who are the largest GCC contributors.

Making the SC a more open, accountable, and representative
"government" and granting it greater authority (for example the
authority to define the conditions for corporate citizenship in the
GCC world) go hand in hand.  Why would the currently powerful
companies want to do that?  Altruism?  Just the opposite: to avoid
having to rely on corporate altruism.  Greater SC authority and a more
legitimate political structure would in turn give these companies a
business justification for collectively implementing a better
administration of the commons upon which they depend.  Not altruism,
but improved self-regulation.



   The overriding concern is the interest of the GNU project and the
   free software community, not whatever "the public" might mean.

The GNU project is itself aimed to promote an interest of the general
public.  By transitivity, since GCC is for GNU, GCC is aimed at that
same interest.


   I'm also not sure what "neutral" means here.  Neutral with repect to what
   issue?  The whole point of the SC is that the members are *not* "neutral",
   but each represent a different component of the GCC user community.


Neutral with respect to balancing the long term health and progress of
GCC against the immediate demands on the project made by the largest
contributors, most of whom are businesses, often doing work with a
fairly narrow and short term purpose.

Please don't get me wrong: I do not think that any of these companies
have the goal of sabotaging GCC -- not by a long shot.  Most or all of
them contribute improvements to GCC that no customer has directly paid
for, as far as I know.  But they also don't have a strong business
case for formalizing those contributions, paying for the labor to plan
and coordinate them, etc.  An improved SC structure can give them that
business justification.  (One could argue that the GCC project is
under-resourced: a better SC structure can help to fix that.)

-t