Subject: Re: ok, FSBs created schwag; now bet the farm on something better
From: Tom Lord <lord@regexps.com>
Date: Mon, 17 Sep 2001 01:51:44 -0700 (PDT)


Ian,

	You asked:

	Where do you expect the money to come from?  When Microsoft
	started doing serious research, they had a secure revenue
	stream.  Most FSBs are living essentially hand to mouth, 
	[....]

I'm only able to judge the situation on the basis of SEC filings and
news stories.  I think there's enough cash available to make some
progress without breaking the banks of the funders.  The costs of Free
Software R&D needn't be astronomic.  We don't need to pay taxes on 700
acres in San Jose, for example.  The dominant expense is several
smallish teams of FTEs, secure enough in their employment to undertake
serious projects with seriousness of purpose.  You can augment those
FTEs with appropriately incented engineering bandwidth from the host
corporations, whenever that happens to be available and amenable.

There's also plenty of ways to at least try to implement cost recovery
-- though certainly there's risk involved.  I'll leave specifics to
conversations in other, site-specific channels (sorry).  If nothing
else, research is *interesting* and sometimes research produces
valuable intermediate results.

There's a cost hump to get over: it can easily take a bit of time for
programs of practical research to positively impact the bottom line of
their host corporations but once there's momentum, there should be a
continuous positive impact.  Vital tricks include: not quashing
projects prematurely; not executing projects with a thrashing agenda;
and creating an interface between research and the hosts that balances
entrepreneurship and transfer, within the contexts of the hosts,
against sheltering researchers from distraction and interference.

The danger of postponing getting over that hump until you are really
rich is that you may be waiting until its too late.  Practical
research, especially for FSBs at this point in history, is part of a 
reasonable strategy for getting (and then staying) really rich.
I don't think the existing open source basis set is strong enough
to withstand what I see coming from various proprietary software 
organizations.



	The standards for the research tax breaks are not that high
	when you're building software.

I do read the IRS pubs differently than you, but I am most definitely
not a tax attorney -- so I concede that point.




       Why shouldn't the proper place for FSB research be where it has
       historically been: in the academic community?

To a first approximation, academic researchers do fundamental
research, aimed at figuring out what software *can* do.  I'm
advocating investment in practical research, aimed at implementing
basis sets which enable tactical plays: what software *does* do.

Though academics sometimes do practical research (such as when they
aim to create spin-off corporations or otherwise leave academia), it
doesn't fit the publish-or-perish culture or the
advance-human-knowledge charter.  Most practical research
traditionally takes place in corporate labs.

Moreover, practical research should not take place in isolation from
other business activities.  Continuous technology transfer and
entrepreneurial experimentation are part of the process.  You can't
expect those things of universities unless you *over* fund those
activities.  So really, corporate R&D labs can be the less expensive
approach.

Regards,
Tom Lord