Subject: Re: update server alternatives
From: Tom Lord <lord@emf.net>
Date: Tue, 29 Oct 2002 00:35:19 -0800 (PST)



       > Note that Red Hat mostly does _not_ produce its product, 

Yeah.

I'm a (laughably minor) shareholder and have various personal
interests in their success.  I want them to do a lot more growing and
producing.  I think it's their "out" from the currently losing
competition with IBM and HP/C.  It'll also make them a better
employer.


	  [Aside: I really don't understand why you think Red Hat is
	  missing the point of open source.  As far as I can tell,
	  most of their "Linux" development work is aimed at quality
	  control of the RH distribution,

Which is fine until you get to the details.  I think there is a
disconnect there, because instead making that QA their contribution to
the community (by striving to push it upstream and make it intrinsic
in how free software is developed) they try to use it as
differentiation.  They are trying to turn an FSB into a de facto
proprietary company -- to exploit the weaknesses of the community to
their advantage rather than trying to fix those weaknesses.  Look, the
business advantages of free software have to do with cost sharing,
opportunistic development, customer autonomy, user-driven development
-- in general, community based engineering.  RH's current most
lucrative business models are mostly about vendor lock-in combined
with the parasitic exploitation of the community: the negation of
those FS advantages.


	  and as such the most important communication is among maintainers of
	  individual packages.  Most of the rest is done by assigning Red Hat
	  employees to work as developers in the publicly developed projects
	  such as the kernel.  

In my opinion, these guys aren't behaving in the manner they would if
their jobs didn't depend on their RH-biased participation: there's a
discordance of values at work, and it puts those particants in a
schizophrenic [*] position.  I'm frequently embarrassed for them.  My
diagnosis is that they aren't unambiguously resourced and incented in
the direction of advancing those projects in the best directions and
that this is reflected in the public projects as a harmful hegemony of
(false) "practicality".  (Some people accuse me of being really
confrontational but, truth be told, I hold back on these guys a lot
out of sympathy and my own apparently pathetic attempt at
realpolitik.)

In addition to unionization, we need a consciousness-raising program
plus a much stronger CPSR-like movement.  "We are, we are, we are the
engineers, we can drink all of fourty beers...." (google for the
lyrics variations, disregard the mysogynist verses, and attend to the
"our job is to save your asses so piss off and let us do our thing"
verses to get my point).



	Their eternal mistake of "not funding arch" is a pretty
	small one, I think....]

I almost agree (surprise!).  

Their eternal mistake of not funding _me_ in particular is a fairly
narrow concern that is more mine than most people's.  I can flame them
on social justice grounds and on the basis of my particular history
with Cygnus (and it's connection to the SV/stanford-pansy mixer-party
mafia), but then there are plenty of social justice causes to go
around so, realistically, I'm just a drop in that particular bucket.

`arch', on the other hand, has a subtle but signficant role to play in
FSB product design.  Their mistake in not funding `arch' is of the
"shooting themselves in the foot (or the gut)" variety.  Here's how I
put it in another thread, from another list:

    I've talked [...] with some senior engineers at some vendors
    -- people overseeing O(terabyte) of source.  There is most definately
    sympathy there for the parts of our views we hold in common.
    Unfortunately, there's also a lot of stagnation and beaurocratic
    grid-lock.  It doesn't help that the free software community hasn't
    been able to align on a source mgt. technology direction.

In other words, there's a business leadership failure here, and I'd
like to think that part of RH's role in the world is to be leaders.

Failing them, I hope MSFT moves towards a walled-garden variety of
"shared source" -- for which `arch' would be invaluable.  Why?  Not
because I like either MSFT or walled-garden licenses and not because I
expect to get any money from MSFT -- just for the shadenfreude value
of it.


     > The USPS has historically contracted the bulk of its transport
     > to the common carriers, and both transshipment and channel
     > reliability as such have made the quality of postal service it
     > a joke at least since the demise of the Pony Express.

Hmm.  In my experience, the USPS premium services are an excellent
value, extraordinarilly reliable, and significantly more flexible than
fed-ex.  If what you say about outsourcing is true, that just deepens
the mystery of how they pull it off (is USPS still running a
deficit?).  But this is FSB, not a list about package delivery, and
your analogy is still (to me at least) completely vague.


     > I think that Red Hat is going to want to control the
     > distribution channel in the same way for the same reason of
     > quality control as Federal Express.


"_The_ delivery channel".  Belief in the essentialness of such a thing
is the hegemonic bug in your world-view.  Personally, I see no need
for any such thing.  It's a concept borrowed from proprietary
software.


-t



[*]: "it puts those particants in a schizophrenic position"

   You can ask our resident philologist gumby-cum-david-wallace about
   the etymology of schizophrenic, apparently.