To James Gosling @ Sun

Howdy James.

You recently gave an interview to ComputerWorld Australia's Rodney Gedda
(Australian ComputerWorld, February 9th, 2005.)

Part of that interview is located online
(Source:
http://www.computerworld.com.au/index.php/id;562914339;relcomp;1)

I'm curious about some of the statements you've made in that piece.

For brevity & simplicity's sake, I'll iterate through them in point
form.

1) You are quoted as saying that companies which give away software lack
financial viability.

My question to you is:

But isn't that what Sun has elected to do with Open Solaris? Isn't that
what Sun has often done with the various incarnations of Solaris for the
x86? Isn't that what Sun has done with Java for over a decade? How is that
different?

Forgive me for pointing it out, but this tranche of software is in
totality the only well known software that Sun produces in 'separate
product' form. (I'm excluding Solaris for SPARC, which is always bundled
with Sun's SPARC kit, if I'm not mistaken.)

2) You are quoted as saying that 'there are a lot of open source groups
which have to figure out what their economic model is."

My response to you is:

If you'd said this in 1995, it would be an accurate analysis. All
surviving open source groups (i.e businesses) _have_ figured out their
economic model. Which is why they've survived.

3) You are quoted as saying that "everyone who works on these open
source projects [must] pay rent and buy lunch... so where does the money
come from?"

My question to you is:

Surely the fact that these developers _are_ paying their rent and eating
their lunch, must indicate that the exisitng open source business models
_are_ working. Correct?

My firm is one such example, and we're now in our 15th year of profitable
operation. Surely it must not be as perplexing a paradox as it first
seems. If it is, I'm more than happy to explain how it all works.

Gartner has offered up a useful (for once) clue; it goes along these
lines: 95% of the world's software developers do not work for software
publishers, but perform bespoke development. Open source _may_ eat
software publishers' lunch and be a negative causative agent for that
camp. But for the other 95% of the development community, open source is a
potential godsend. For some of their number, as with hardware vendors, it's
in their best interests to propel open source development.

4) You are quoted further: "Gosling also questioned whether it was
fair for academics to engage in sideline open source development while
collecting money for something not integral to the university's
business"

My response to you is:

I find this a particularly quixotic perspective, coming from a senior Sun
employee. I trust you recall that Sun owes its existence and livelihood to
what are two essentially open source developments emanating from those
pesky university academics?

Firstly BSD which was adopted by Sun to form the basis of the Solaris 1.x
operating system (SunOS 4.x.x), came from University of California at
Berkeley. That story is recounted here:
http://www.softpanorama.org/Solaris/solaris_history.shtml

Even the Solaris 2.x lineage, whilst Sys V.4 based, owed plenty to BSD
Unix.

Secondly, the "Berkeley RISC 1 formed the basis for the commercial
Scalable (formerly Sun) Processor Architecture (SPARC) processor"
(Source: http://www.computermuseum.li/Testpage/MicroprocessorHistory.htm)

I would hate to think that Sun feels threatened by other businesses using
the same mechanism it used to bootstrap itself into the path for success.

5) You are quoted further: "These businesses (i.e MySQL, JBOSS and
Red Hat) are more hype than reality. If they don't have a [longer term]
economic model... they are going to have a really hard time."

My question to you is:

Do you agree that these firms are presently profitable? It seems to me
they are. Do you agree that the market for their business is likely to
grow? Once again, it seems that adoption of open source is growing apace
around the world, so these firms are likely to grow commensurately.

If so, where do you perceive the problem for these firms to be?

Additionally, do you agree that even if Red Hat fails, even if MySQL AB
fails and even if JBOSS fails, that (respectively) Linux, MySQL (the
database), and open source J2EE are still very likely to be successful?
Even if other firms take up the cudgels to make them so?

6) You are quoted further: "I actually find it amazing that people
consider MySQL to be an open source project given that there is no one
allowed to do check-ins. When you look at the licence for MySQL you really
do need to feed them money in order to be clean with their licence."

My response to you is:

MySQL AB can claim that MySQL is open source because they've released it
under the primary open source licence, the GPL. Whether or not they accept
outsider contributions is irrelevant. If you or I want, we can grab a copy
of the MySQL source tree and create a CVS repository for the project where
we do allow external contributions. We will have then forked the project,
sure, and we can't call the resulting database MySQL (because MySQL AB
owns that trademark) but it will, in every other repsect, be MySQL. _This_
is what makes it an open source project.

Now, I am almost certain that you know this James, but yet you throw down
obfuscatory remarks to journalists and the public who may not know any
better. That you do this, I can only assume you actually mean to establish
a FUD strategy against this open source project, which is rather
disheartening.

Onto your next point. The MySQL software technology is available under
more than once licence. If the default GPL is not approriate for your
needs, perhaps because you have built a proprietary application on top of
MySQL that you want to re-distribute, then sure, you can pay MySQL to make
the GPL 'share and share alike' clause go away. But for 99.99% of users,
MySQL is GPL, is free as in open code access and is free as in zero cost.
This too I believe you would know, yet you claim otherwise.

7) You are quoted further: "We have been distributing the source to
Java freely since the very first day [and] you can go to our Web site and
download the sources, no problem. In that sense Java has been open source
for a decade."

My response to you is:

James, I know you know the difference between the availability of source
code on one hand and 'open source' on the other. I know you know this
because, like me, you've been playing in the Unix space for over 20 years.
And in that space, source code availability and localhost recompilation,
even of proprietary software, was par for the course. You would know that
just because a vendor gives you source code, it doesn't mean you have
_any_ rights to make use of that source code beyond compile it and run it
as per your licence agreement. As such, it is the furthest thing from open
source. As is Java.

Why you would try and pull the wool over users' eyes with your comment
above I can't quite fathom. Can you elaborate on what you are trying to
achieve here?

8 ) You are quoted further: "Gosling said that people in the open
source world quibble about one clause in the Java licence that
disqualifies it from being open source, but is important for maintaining
compatibility."

You state that open sourcing Java will cause it to bifurcate into
incompatible versions. I keep hearing this from both you and from other
Sun execs. It is provably wrong. I's sure you know this, yet chant this
fallacious mantra every time this issue is raised.

Here's why Java will not likely 'fork' into incompatible versions, and
even if it did, why it matters little to Sun's great game plan:

1) Sun own the Java brand and mark. No one else can push out an un-tested
and incompatible product with that name. This alone means you control what
is and what isn't Java(tm). Java(tm) can therefore never be _incomptible_
with what Sun wants it to be.

2) Every major programming language (except Java) has an open source
implementation, covering compilers, interpreters, libraries and runtime
environment. None of these open source implementations, for C, C++, Perl,
Python, Ruby, PHP and a dozen more, have forked into incompatible
versions. Some of these have been going strong for almost two decades
without that dreaded fork. Why do you believe an open source Java to be
any different?

I'd be keen to understand your viewpoint on these further. If you can
elaborate on your views, or possibly point out any misapprehensions I've
made about your intentions or misinterpretation of your comments, I'm all
ears.

Cheers,

Con Zymaris
CEO
Cybersource
(A 17 year Sun user and customer)

ps: In case you think this is just another whinge against Sun from the
open source fringe, I'd be happy to point out Cybersource were one of the
handful of firms globally which partnered with Sun back in 1994 to push
the Sun's Public Windows Interface specification.

We're with you when you're being honest and making sense. We're perplexed
and hesitant when you aren't.