OT : Making money from Java

It's Judson's thread about Micro Focus compiler fees/runtime which
triggered this one in my mind.

If Java is FREE can somebody explain how Sun Microsystems makes its
money from the language ?

Jimmy

James J. Gavan wrote:
> It's Judson's thread about Micro Focus compiler fees/runtime which
> triggered this one in my mind.
>
> If Java is FREE can somebody explain how Sun Microsystems makes its
> money from the language ?
>
> Jimmy

They sue people.

Donald

Donald Tees wrote:
> James J. Gavan wrote:
> 
>
>
> They sue people.

Well let's say that's a less informative answer than that you use
Clarion. And the price for Clarion, or have you discovered a freebie ?

Jimmy

> James J. Gavan wrote: 

There's a pretty good discussion about this on a blog:

[url]http://discuss.fogcr.com/joelonsoftware5/default.asp?cmd=show&ixPost=158765[/ur
l]

The conclusion is basically that they don't.

- Oliver

Also - companies like IBM licence Java from Sun

Which then leads to reciprocal deals.


IBM sells Websphere products
Sun says "Hey, make Websphere for Solaris"

Now Sun has their own products and other products pushing their own other
products....

For a better understanding read what Sun's COO has to say.

http://blogs.sun.com/roller/page/jonathan/20040712

DU

"Oliver Wong" <owong@castortech.com> wrote in message
news:jxtef.113259$S4.69557@edtnps84... 
>
>    There's a pretty good discussion about this on a blog:
>
> [url]http://discuss.fogcr.com/joelonsoftware5/default.asp?cmd=show&ixPost=158765[/
url]
>
>    The conclusion is basically that they don't.
>
>    - Oliver
>

James J. Gavan wrote:
> Donald Tees wrote:
> 
>
>
> Well let's say that's a less informative answer than that you use
> Clarion. And the price for Clarion, or have you discovered a freebie ?
>
> Jimmy

I work over the net, using PC-Anywhere on a high speed line. It is a
straight contract job, hourly rated. The customer owns all the software,
and it runs only on his machines. I do not have copies of any of the
developmental software, so the problem does not arise. PC-Anywhere cost
me $79, US. Clarion is a fairly expensive program, but I do not know the
price.

On my own machines, I have been trying out all sorts of stuff, all of it
either what I consider reasonable prices, or that I own valid copies of
from over the years. I spend about $1500 per year on new software of
various sorts, and get 10 or 15 products for that. I still do most of my
own work on my version 5.0 Fujitsu Cobol. I went to Fujitsu when MF went
to user fees, and I although I kept them current from Version 3 through
5, I stopped updating when they also got prohibitively expensive. The
upgrades after that were mainly for the net, and since that did not
apply, it struck me as a waste of money.

I still make money on that software, but it is reaching the point that I
have to decide whether to re-write or abandon. I wrote the skelton of
that system in the late 70's, to run on an Altos(using MF Cobol), so I
cannot complain, it's had a good run, but it's also had one too many
conversions. I maintain it using Fujitsu 5.

Which brings us to your answer.

Most of the companies that write "free" software make their money
writing software for a price. They really sell hours of programming time
directly to the person that wants the software written. That begs the
question "AND THE CUSTOMER ALLOWS THEM TO PUBLISH IT?" The answer is
yes.  It is the only sensible way.

You know the value of a software library. A source code software
library. Would you rather write software using a source code library
that includes everything written by several million programmers, or
would you rather start from scratch, yourself?

If you put what you write back into the library, you earn the right to
use everything in the library in your own code. What is the most
sensible development route?

Last question.  You want something written.  The best thing in the open
software market is product x, and you have a copy that "sorta does what
you want".  It comes 75% of the way, but does not do the job. Product X
is written by Company Y, who give it away free over the net. They are
are a contract programming firm, run by five programmers, all of whom
contract out at an hourly rate.  Who do you hire to get it to where you
can use it?

They make their money the same way we do, by hiring out at hourly rates.

Donald

P.S. "they sue people" was a joke, and should have had a smiley. If you
do a net search on "Sun Lawsuit" I expect you will see it.

Donald

defaultuser wrote:
> Also - companies like IBM licence Java from Sun

Thanks defazultuser.

Jimmy

James J. Gavan wrote:
> It's Judson's thread about Micro Focus compiler fees/runtime which
> triggered this one in my mind.
>
> If Java is FREE can somebody explain how Sun Microsystems makes its
> money from the language ?

Consulting, certifications, training, and web traffic.  :)


--
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~
~   /   \  /         ~        Live from Montgomery, AL!       ~
~  /     \/       o  ~                                        ~
~ /      /\   -   |  ~          daniel@thebelowdomain         ~
~ _____ /  \      |  ~      http://www.djs-consulting.com     ~
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
~ GEEKCODE 3.12 GCS/IT d s-:+ a C++ L++ E--- W++ N++ o? K- w$ ~
~ !O M-- V PS+ PE++ Y? !PGP t+ 5? X+ R* tv b+ DI++ D+ G- e    ~
~ h---- r+++ z++++                                            ~
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~

James J. Gavan <jgavandeletethis@shaw.ca> wrote:
> It's Judson's thread about Micro Focus compiler fees/runtime which
> triggered this one in my mind.
>
> If Java is FREE can somebody explain how Sun Microsystems makes its
> money from the language ?
>
> Jimmy

You should note that the word "free" has two meaning here:

1) can you get Java implementation free of charge?

2) can you do what you want with the implementation?

Sun Java is distributed free of charge, but is not free in the second
sense. In fact, the main motivation for Sun to distribute Java free of
charge is to have a platform which they (at least partially) control.

Namely, there is Java the language and Java Virtual Machine (JVM). JVM
is quite similar to an operationg system: it offers numer of services
to the running program. The program can not run if JVM services are
not present. Sun designed JVM to be "portable": JVM can run on multiple
processors and on top of various operationg systems. In effect, JVM
can serve as a portability layer, making underlying operationg system
irrelevant.

Why this matters for Sun: ATM about 90% desktop machines run operating
systems form Microsoft. Historically, Microsoft used its position as
operating system vendor to promote its own application (see for example
Netscape accusations). So a vendor either has to be independent of
Microsoft OS or risk unfair competition. Like all major software vendors
Sun can not just ignore 90% of desktop market. But JVM allows to be
independent of Microsoft and still deliver on Microsoft platforms. The
same problem affects other companes, so they teamed with Sun to promote
Java.

Now, for this tactic to be effective JVM has to be popular, so Sun is
willing to distribute copies free of charge. They (and other companies
too) distribute may development tools for Java free of charge, again
this is part of their tactic to make JVM popular.

Note that Sun retains much control over Java (they had exclusive control,
but gave up part of to to other companies). One example may be "native"
Java compilers: it makes perfect sense to compile Java program into
native executables. One can gain some speed in this way, but more important,
the program is then independent of JVM. In particular, native compiler
would allow small "standalone" Java executables (important if you want
to minimize dowload size and suspect that the recipient has no (or incorrect
)
JVM). But you will not get such a beast from Sun: this goes against their
tactic.

To put thing in more general perspective: compilers are relatively small
programs, which require moderate amount of work to create. Namely, a toy
compiler can be built in a few days, small usable one in few months.
Mature compiler accumulates features (and code) over longer period so
it is more likely to represent few man years and industrial leader
may represent hundreds of man years. But the the law of diminishing
returns works strongly here: extra features take most of the effort.

OTOH compilers play pretty crucial way in software developement, so
there is motivation to create new ones.

Coming back to question of "free" software: the one who controls
software controls the society. Namely, quite a lot of things in
our life is controlled by software. For example, if a mailing program
does not allow a user to send e-mail to some address, it means that
the user can not send e-mail to this address. Of course, the user
may s workarounds (for example use another program), but is
software in controlled by a single entity which deliberatly put some
limitation, then it may happen that there is no workaround and even
if there is one the control still may be quite effective. Also,
if software restricts information presented to the user, the
user may even not notice that the control is present.

Some people belive that ability to modify software is very important
for our freedom. One of the persons who quite early formulated
such view was Richard Stallman:

http://www.gnu.org/philosophy

You may find ideas above controversial: some people strongly object the
very idea, other accept the idea but disagree with Stallman on many
details. But there is a number of people who volunteered their time to
write software motivated by the idea.

Note, that making _all_ software free would work against current
software houses. But part of Stallman's idea was peacefull coexistence
with closed-source software. So, free software can be used by
commercial companies as long as they do not try to turn it into non-free
software. In particular GNU C compiler was originally written by Stallman,
but now it contains many contribution from firms like IBM and Apple.

--
Waldek Hebisch
hebisch@math.uni.wroc.pl