Windows Opening its Source for the Justice Department?
Posted 18 Feb 2000 at 17:57 UTC by dcm
Windows to Open Source
Today the big news hit! Could Microsoft open source code to settle
Justice Department Case? Sun's executives smiled, Linux enthusiasts
stared with an untrusting eye, Microsoft PR executive scrambled.
Immediately Microsoft came out with a press release saying that no such
quote can be attributed to Gates. Greg Shaw, spokesman for Microsoft
said "He did not make any of the comments regarding source code that
were attributed to him".
So why did it come up? Why did it happen? And what does the interview in
question mean for the Open Source community? Well watching the actual
interview will give you a completely different view of the incident.
The interview took place on Bloomberg Television and was intended to be
a Q & A session with Gates about Windows 2000 and what its impact
be on the computing world.
During the interview Gates was asked some interesting questions about
open source. First he was asked if the fact that Sun had opened up their
source meant anything to Gates and his strategy. Luckily the interviewer
caught himself and said that Sun has opened their source to some degree,
but Linux is completely open source.
Gates' response was that they open their source for certain partners,
but more importantly, they provide developers with the tools and
documentation they need to create great applications for their operating
system. He then went on to say that this is much better than getting
code that has been "tinkered with" and could be of questionable quality.
He said that users can be assured that when they purchase Windows 2000,
they are getting the real Windows 20000.
That was the only reference to open source and Windows in the whole
interview. What happened next is what caused the stir.
Once the interview was complete the interviewer was being questioned by
a Bloomberg host about what took place. The press covering the press has
always caused great pride in the industry but what happened during this
"meta-journalism" should go down in the text books as bad journalism.
The interviewer of Gates said that after the cameras were off he
asked Gates if he would be willing to open up Windows' source code to
settle the Justice Department's case. He claims that Gates said yes and
that was the end of the conversation.
Television has always had the advantage of catching someone "on the
record" during an interview, but when there is no camera running,
journalists should pretend they are newspaper journalists and get
another source for such news. In this case, having seen the interview
this author would have to believe Microsoft's claim that no such quote
can be attributed to Gates.
Bloomberg tried their best to not appear like a high dollar Slashdot by
having their Editor in Chief Matthew Winkler say "We stand by our story,
he was asked specifically, and the answer that he gave was yes".
What Does This Mean to Open Source?
So how should the open source community respond to this story? Well
first, the mass media technology press should respond to the scar put on
their profession by such a story, but the open source community should
look at it differently.
If the community ignores the hype and concentrates on what it knows
Gates said, it can see how Microsoft is going to fight the threat of
open source. Gates said that first they provide developers with
"everything" they need to create apps for the operating system. Does
open source do that? No. There are hundreds of examples of great
technologies that have absolutely no API documentation, tutorials, or
even README files. The community has stood by the "read the source"
mantra much too long. If it is to target Microsoft, and that has been
the claim in recent months, it has to provide the tools needed to create
Gates' other attack on open source was his use of the word "tinkering".
Gates was trying hard to plant a seed in the minds of IT professionals
(his target audience) that buying something that, at least, appears to
have no set focus is buying something unstable. While it is known this
is not true how does the community fight such comments? It is true in
some projects that people go in and tinker with code quite a bit. People
are always looking for the latest version of software and when people
take the jump into the unstable branch, even if their knowledge is not
good enough that they should, the community suffers by what news is
spread by that user's experience.
The open source community is a strong one and it has taken the world of
software into new territory. But to survive the community has to listen
to how it is being portrayed to those who have not yet tried it.
Otherwise we can be placed into a position where the new users stop
I agree 100% with the comments and suggestions in the last paragraph.
Some companies and IT professionals are reluctant to using open
source software because they are not sure about what is inside...
especially if someone who is trusted by all big businesses told them to
be very careful about this. But there is even worse: some large
companies are simply banning open source software unless it is
bought from some reputable company and the contract states that the
vendor believes that the software does not infringe any patents.
I think that many companies fear patents even more than viruses or
bad code. Most of us think that software patents are bad, but that does
not change the fact that they exist now and any company can be
sued for infringement. Fighting against a patent is very expensive
(regardless of whether the patent is valid or not) and takes a long
time, so many companies prefer to settle out of court even if they have
to pay license fees.
Let's look at it this way: you are the boss of a large company and
your engineers tell you that they need to add support for the FooBar
protocol in your latest product. They give you the following
- write the code inside your company (estimated cost for the team
working a few months: $80,000),
- buy a closed source library from a reputable company which supports
their product (cost: $10,000),
- download some Open Source library that can be found on the Internet
(download cost: $5)
Let's also consider another thing: you expect to sell thousands of
copies of your product, which should result in significant benefits.
Now, which option would you choose?
If you choose the free library downloaded from the Internet, you can
put a team of testers on it and check that it is stable enough to be
integrated in your product (besides, the Open Source model and the "many
eyeballs" should have killed most bugs already). But are you sure that
there is no patented code inside? Are you sure that there is no IPR
Trojan horse inserted on purpose by the patent holder or accidentally by
some programmer who was not aware of the patent and wrote the code in
good faith? If you use the patented code and integrate it in your
product, you are responsible. If the patent holder sues you and asks
you to stop distributing your product or asks for huge royalties, then
you are mostly on your own. It would be very difficult to counter-sue
the author of the free library (assuming that you can identify who wrote
the infringing code).
If you buy a product from another company, at least they share a
part of the risks (or maybe they are the owner of the patents, so they
are safe). If you write the code yourself and keep it as a trade
secret, you can never be sure that you do not infringe any patents but
on the other hand it will probably be hard for your competitors to
discover that you are using some patented algorithm (some of them are
obvious, some of them not). But if you are using an Open Source
library, it will be trivial for the patent holder to discover and prove
that you are infringing their patent. They do not even have to be
specifically looking for patent infringements in your product: it only
takes one person who knows about some patent, reads the code of the free
library and discovers that it makes use of the patent, then sees that
your product is using that library. From then on, anything can happen
and it is likely that you will loose a significant share of the expected
benefits of your product.
I think that most companies prefer to buy expensive libraries from
reputable sources, not only because they believe (wrongly or not) that
the code is better, but also because they minimize the IPR risks.
I still haven't found any good counter-argument to that (beyond
"software patents are bad"). Suggestions are welcome...
Well, not quite an argument, but a partial solution: Insurance.
There shouldn't be much of a problem purchasing a policy through one of
the insurance markets, e.g. Lloyd's. Whether the transaction cost
is worth it is another matter - it is likely to become so if many
companies start doing it, at least.