Subject: Re: selling SCM solutions
From: "Stephen J. Turnbull" <stephen@xemacs.org>
Date: Sat, 31 Aug 2002 01:09:29 +0900

>>>>> "Tom" == Tom Lord <lord@regexps.com> writes:

    Tom> This is the same respect I've been trying to pay the people
    Tom> who are *not* responding.  Should they respond meaningfully,

Tom, they are responding, and responding meaningfully.  (You're
probably the only person on the list who hasn't learned immense
amounts from the advice you've been given.  Thank you very much for
eliciting it; it has been enormously valuable to me.  And to all those
who have contributed, my earnest thanks.)

They're just not giving you money or referring customers to you.

Nor will I.

I visited regexps.com the other evening looking to see if maybe XEmacs
and arch could make some kind of deal---grumbles about CVS are a
constant background irritation for me.  All I can offer in return is a
testimonial (we have a well-tended website that gets lots of hits,
presumably mostly developers), and links to monkey.el, but what the
hell, you don't have any _paying_ customers yet, do you?

What I found there convinced me that there is no benefit in making a
deal with you; you offer _no_ support capabilities.  arch is not a
product, and on current form won't be for many months, if ever (but
see below).  AFAICT, arch seems to address only one of the main XEmacs
issues, but in current form it would bring uncertain but large
additional costs (with a fat upper tail, too).  The benefits of the
wonderful new architecture to XEmacs are vague at best.

Of BitKeeper, prcs, subversion, RCS, and CVS, sticking with CVS for
the medium term is a satisfactory default.  BitKeeper is technically
and fiscally a serious alternative now (nor does the license bother
everybody at XEmacs, though it would be an obstacle).  In the medium
term, subversion looks to be shaping up very nicely.  Both address
several XEmacs issues with CVS.

All that nets out to a lost sale, this one based not on your "attitude,"
but on real research.  Limited, but sufficient given my available time.

Oh, and don't tell me that "XEmacs is not in my target audience, I
want to get `mind share in new projects.'"  Think indirect.  A sidebar
link to monkey.el would _certainly_ generate some revenue, and an
"arch keeps 200MB of XEmacs repository aloft" (ooh, er!) link'n'logo
in the footer would be good advertising, no?  And .els are a
moderately popular kind of new project.  We add about 1 new 3rd party
library a month to our Packages repository.  You could have had mind
share in _all_ of them, plus the existing ones, for arch.

    Tom> then we could (probably) find the bug and return to business
    Tom> as usual.

Tom, "business as usual" _for arch_ is consumption of capital with no
product sales.  The only person with any interest in that is _you_.
It doesn't even benefit the rest of society, since the "socially
responsible engineering solution" isn't getting installed anywhere.

Of course, there's an obvious way to kickstart arch.  And that is that
somebody (else) with business smarts will grab the sources and start
an arch sales and support firm.  If the ethically-required attempt to
negotiate with you fails, they can go their own way.  They have no
need for you---design is nearly complete (you write "my itches for
arch are mostly scratched"), the implementation is well advanced
(enough for arch to host itself), and the license is, of course, GPL.

It's not like there are less than two dozen people on this list who
could do it.  And you know something?  If arch really is as good as
you say, one of them might do it for the good of the community.


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
 My nostalgia for Icon makes me forget about any of the bad things.  I don't
have much nostalgia for Perl, so its faults I remember.  Scott Gilbert c.l.py