This is the mail archive of the mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

why is the SC relevant (was Re: SC issues revised)

There have been two sorts of replies to "SC issues revised": some on
the technical matters, some on the political matter of the SC.   This
message is about the political matters, though it is for a technically
informed audience.

       	The fallacy in your reasoning is assuming that these [long
	term, cross-cutting technical issue] discussions do not occur
	already -- that electing members of the GCC SC through a
	popularity contest and/or soliciting formal corporate
	sponsorship is going to change or improve this situation.

What you're calling my fallacy isn't quite what I'm advocating.

	All of the companies that use and benefit from the GNU
	toolchain (AMD, Apple, ARM, CodeSourcery, Compaq, HP, IBM,
	Intel, Motorola, Red Hat, SGI, Sun, SuSE, and Swox, just to
	name a few) contribute to the toolchain because of their own
	business self-interest which can be justified to their
	respective owners or shareholders.  All of these companies
	have contributed to the GNU toolchain, continue to contribute
	to the GNU toolchain, and have more contributions coming down
	the pike.

More or less agreed.  However, I believe that the business interest of
some of those companies is larger than is reflected by their
contributions.  Evidence in support of this belief can be found in the
long standing problems that aren't addressed with dispatch: the
testing infrastructure, for example; the generally low priority given
to deep architectural improvements and code clean-ups.  If GCC were an
important proprietary project run entirely in-house at one shop, there
would likely be a general budget for such activities and on-going
aggressive planning for prioritizing of those tasks within the budget.

We're at a point in business history where the business roles of Free
Software and Open Source projects is changing.  For example, the
GNU/Linux platform and various key applications are being increasingly
viewed by more and more companies and customers as:

	- a de facto industry standard

	- a cost effective development model, leveraging the
	  efficiencies of inter-organizational cooperation

	- a platform whose licensing structure benefits customers

	- a platform and set of applications that can compete
	  effectively against purely proprietary systems such 
	  as Microsoft's

GCC and the toolchain are cornerstone elements.  They are as important
to many of these companies and customers as any proprietary compiler
technology.  They merit similar levels of investment and pro-active,
coordinated development.

	I am fairly certain that none of the companies want to create
	a new, independent bureaucracy to administer funding of GNU
	toolchain development.

And I'd guess you're right.  That's not what I'm suggesting.
Currently, funding is mostly administered through in-house spending
and through custom development contracts for which competing providers
can bid (I'm leaving out academic and hobbiest contributions).  That
shouldn't change.

What should change is:

	1) Formalizing of the detailed agenda for longer term, general
           interest improvements.  "Discussions on a mailing list" is
           about the least-effective non-empty infrastructure for
           formulating long-term plans.

	2) Given a formalized agenda, lower the barrier to multi-party
           custom development contracts: where several customers who
           compete in other domains have a clear path to pooling
           resources for their common interests.  Being able to point
	   to an SC RFI document in a contract is an example of a 
	   lower barrier in this area.

	3) Given the idea of multi-party custom development contracts,
	   lower the transaction costs still further by having 
	   open ended "contribute to GCC for our shared needs"
	   contracts: long term funding for people to do what they do
	   best -- figure out how to make the best use of their time
	   improving GCC in general ways without having to separately
	   negotiate a development contract every step of the way.

The expanded role for the SC that I see in this is a planning role and
evaluation role:  giving us an objective and detailed perspective on
what can be considered "general improvements" and evaluating
contributions and proposed contributions against that plan.  For that
role, the SC would need to be more politically legitimate and would
need an independent budget (which can be small compared to the size of
the total spending on GCC).

Yes, I did rantingly ask the question "When do I get to vote for SC
members?" but I didn't intend that to be a serious proposal for a
general, Internet-wide election of the SC.  The entire series of
questions in that rant was only aiming to point out that the SC's
legitimacy and neutrality aren't particularly reinforced by its
existing political structure.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]