[PLUG] SCO Clarification

[Jeme A Brelin]

[I started writing this message this morning and got pulled away from the
computer until just now... I'm going to finish it and send it before
reading the rest of the list.  Enjoy.]

On Sat, 1 May 2004, John Meissen wrote:
> Jeme A Brelin <jeme at brelin.net> said:
> > UML seems to have filled that niche pretty nicely, I'd say.  Other
> > systems will surely be developing compatible user mode equivalencies
> > and VMware will go out of business.  Hope you don't depend on the
> > software for anything critical else you might get stuck with a piece
> > of software that is unmaintained and unmaintainable.
>
> It's a nice fantasy, but not particularly realistic.

If by "realistic" you mean "the way things are today", you're right... but
that's a very narrow view.

There is an old guard standing watch over the systems of the past that
have led to the creation and maintenance of the "powers that be".  Some
are direct benefactors and wielders of that power and others are merely
"conservative" and don't know how to look to something better.  Through
methodic subversion and the occassional outright rebellion, we can push
those powers aside and develop new systems that are as nice as our
fantasies.

> While it may or may not apply specifically to VMware, as long as
> software is patentable there will be proprietary software for which
> there will not be a quality OSS alternative. And before you go quoting
> jpeg vs png and ogg vorbis vs mpeg, note that I did not say ALL patented
> software.

There are surely problems with patent law!  However, when administered at
all (actual review of patent applications for non-obviousness and prior
art), it's difficult to see how a patent would cover code that couldn't be
simulated without violating the patent.

Your examples are quite fine.  It's an interesting tactic to give the
counter-argument and then dismiss it out of hand.  Is that something like
a pre-emptive strike?

> And even when there aren't patent issues Open Source Software has to
> compete on quality, where it often fails.

There are some applications for which there is no clear Free equivalent,
sure, but there are capabilities in Free Software that are just not part
of the design considerations in the proprietary software world.

While one could say that, say, sodipodi doesn't do half of what
Illustrator does, you could also argue that Illustrator isn't nearly as
scriptable nor are its data files nearly as configurable or flexible in
their interface with other programs.

The GNU project got the ball rolling with writing Free replacement
software for existing proprietary applications and utilities.  Free
Software has progressed to include more than just duplicating the
proprietary software functionality, but evolving new kinds of applications
and systems that do things "the unix way".

So there are certain applications that can be compared
feature-for-feature, but so long as you use the proprietary work as the
standard, the Free Software will fall short.  A comparison in the other
direction will have the opposite result.  They are different beasts.

> OSS development only works if there are interested developers. A lot of
> OSS projects reach a stage of minimal functionality and then development
> stops, mostly because the few developers involved lost interest or don't
> have the time to make the project a priority.

This is true of all kinds of human endeavor... not just "OSS projects".

> Unless there's a business model based on the application there's no
> guarantee of any support or specific development plans.

Just because there's a business model based on the application doesn't
mean there's a guarantee of any support or specific development plans.

> If I were a business that depended on the continued development and
> support of a specific project I don't know that I'd want to bet my
> future on that development model.

The smart money, of course, doesn't bet the future on a business that
depends on the continued development and support of a specific project.
Flexibility is key.

Don't look for the solution that depends on the piece of software that is
most likely to be around in ten years.  Look for the solution that doesn't
depend on any particular piece of software and allows for drop-in
replacements or easy re-engineering.

> When an OSS application becomes unmaintained it's true that it's still
> maintainable since the source is available. But the cost of paying for
> even one developer to do that is significantly more than the cost of the
> license I'd have to pay for a commercial/proprietary product.

Well, this would be true if you had to pay developers for every single
software product you run, but that's just not the case.  For most
computer-dependent mid-to-large-sized businesses, the cost of software
licensing and support is more than the cost of a developer or two.

When I worked at ELI, members of my team were specifically told to spend
time enhancing a few of the key Free Software and merely open source tools
that we used and contribute back to the community.  Other tools, we used
in their mainstream development branches and didn't modify at all.  The
network of companies hiring developers to enhance different products made
for a large group of programmers working on the complete set of tools for
less money than licensing a proprietary replacement set for each company.

> And finally, there's marketing. I learned a long time ago that a bad
> product with good marketing will be more successful than a good product
> with bad marketing, even when that good product is close to free.

Yeah, what we call "marketting" today is what Adam Smith called a "market
flaw" in 1790.  There are lots of ways to make it difficult for people to
make informed decisions.  So long as there is power to be gained from
exploiting other people, a few will continue to attempt exploitation.
That's just one more damned good reason to get rid of the exploitation and
the power by eliminating proprietary software once and for all time.

> So yes, there will always be proprietary software.

You're coming from a sharply narrowed viewpoint, I think.

> > Rich Shepard wrote:
> > > Also, there's nothing in the GPL that prevents an application being
> > > sold (e.g., Red Hat's distribution, Slackware's distribution, SuSE's
> > > distribution) with source and the GPL.
>
> > Well, no, but how many copies are you going to sell when the folks who
> > buy it can and will simply put it on their website/DAV/P2P system and
> > share it freely?
>
> So you're saying there's really no point in any of these distributions
> existing?

I have no idea how you read that into what I wrote.  I just wrote that, as
the pipes get bigger and the connectivity gets wider, there will be less
and less interest in plopping down cash for mere distribution of
information.  The vendors MUST include some added value that is
demonstrably worth the price alone.  Hence, the software that they produce
will merely be a tool for ensnaring more customers to the services that
have real value.  In other words, they won't be selling software anymore.

> Interesting, since without them Linux wouldn't be enjoying it's current
> increase in popularity.

Again with the narrow view.  This is a transitional phase and it is surely
going to give rise to transitional (hybrid) organizations.  But we are
already seeing the distribution business disintegrate and the services
business come to the fore in these organizations.  RHI's recent change in
development and distribution model is a good example of the transition in
action (though it's pretty clearly a dead-end model that is only
attracting some of the former Red Hat and rpm-based distribution customers
looking to keep up to date and nobody else -- Gentoo and Debian have the
fastest growing share and I don't think it's coincidence that they are not
associated with proprietary development and for-trade distribution).

J.
-- 
   -----------------
     Jeme A Brelin
    jeme at brelin.net
   -----------------
 [cc] counter-copyright
 http://www.openlaw.org