Roy Schestowitz wrote:
> Will Microsoft ever take a mainstream product open source?
Microsoft has taken some "products" open source, mostly libraries and
utilities intended to help developers create Microsoft-only
applications using Visual C++, Visual Studio, and the related languages
and libraries.
There is a catch. Microsoft still owns the copyrights on the core
libraries, the runtime libraries. Furthermore, if you sell a product
using calls to these run-time libraries, Micrososft expects you to give
them a cut.
Thus, Microsoft can "cash in" on code that it has offered in "Open
Source", but it's more like a BSD style license.
It gives Microsoft the ability to create their own proprietary version
of your code if you seem to lose interest.
> ,----[ Quote ]
> | I have always thought that Microsoft anti-GPL stance was strange given
> | the control it gives developers about what their competitors can do
> | with the code. In certain circumstances the GPL would appear to make
> | much more sense for Microsoft than a more permissive license that
> | allows others to benefit commercially from its work.
Microsoft's first compiler was BASCOM, it was a BASIC compiler for
CP/M.
For the first two releases, your paid about $400 per developer for the
language,
but the run-time code was free. As a result, many who would have gone
for
a compiler like C or Forth opted to go with BASIC.
But with Version 3, they decided that the developer revenue wasn't
enough.
They decided that end-users should pay $400/user for the run-time.
The company I was working for at the time (1983) had about 200 users
using an accounting package based on this language. Since Microsoft
wasn't
offering bug fixes or other support, they had to lay off 1/2 of their
staff to cover
the royalty payments. I was the last to go.
I didn't even blame Microsoft. In fact, I applied for a job. They
wanted samples of my
code. My employer was even willing to let me give them samples of the
work
I had done for them, as long as I put copyright notices on every page,
sent it via
certified/registered mail, and had the registration number stamped on
every page.
I sent them copies of user interface code, operating system changes I
had made to CP/M,
(to support networking) and a 20 page response to interview questions.
> | Anyway, Microsoft and the GPL? It would seem unlikely. It's worth
> | noting that when Steve Ballmer referred to Linux as "a cancer"
> | the issue he had was with the license, not the code.
> `----
Yes. Ballmer knows that Microsoft used a bunch of BSD code in Windows
NT, especially Windows 2000.
The problem is that Linux also used the GPL version of the BSD code.
But the Linux version has been getting enhanced and refined and has new
features - while the BSD version hasn't been as well supported. This
means that if Microsoft wants a patch, feature, or enhancement, they
have to go to all of the original contributors and negotiate payments
and/or royalties. It's much easier to negotiate with someone who has
just written a little ditty and has no idea whether people will like it
or not, than with someone whose little ditty has become a coveted
> http://www.businessreviewonline.com/os/archives/2007/01/will_microsoft.html
> It's already ironic how they messed about with the GPL just before striking a
> deal with Novell. Never say never.
Remember, the deal with Novell does not give them rights to any GPL
code not originally written by Novell.
What it does give them is the hooks into XGL and XEN. This makes it
possible to create a Linux implementation that can be comparable to the
MAC OS/X with the virtual Windows machine.
In exchange, Novell gets $350 million, and the right to have users call
the Windows libraries from WINE.
Either way, Microsoft gets a "revenue hook" into Linux without having
to commit everything to GPL.
Remember, Microsoft did the same thing with the Mac in 1984. By making
Word, Excel, and Powerpoint run on the Mac, Microsoft got almost as
much in software royalties as Apple's Software division, maybe even
more.
If Microsoft knew that they could get $30-40 per Linux machine just for
offering the "windows compatibility libraries" for Linux, compared to
the $30-$60 per Windows machine that it already gets from the OEM when
they have to do all of the kernel and driver work, it's a Win-Win for
Microsoft. Let's say that the OEM pays Novell $40 per machine for
Win/Lin. Novell pays Microsoft $30 per machine. Novell then has the
support of the OSS community as well as binary drivers, and the
hardware vendors can focus on the Linux drivers.
The Win/Lin interface means that Microsoft can offer Word, Excel,
PowerPoint, Access, Project, and Visio for Linux, without having to
reengineer the entire system for Linux.
Net result, Microsoft doesn't have to depend on someone like Bill Gates
or Nathan Myrvold to be a kernel guru, but can focus on high-value
applications which can bring the higher profit margins. Project and
Visio would be good examples of high value applications that Microsoft
could continue to sell at high margin values. Office prices are
already eroding, with most people being able to purchase it with a PC
for $150.
The thing is that if Microsoft DIDN'T play nice with Linux, and Linux
ever made it to the retail shelves, which it inevitably will, probably
in the next 3-6 months, Linux would be compared favorably with Mac OS/X
while Windows Vista would still be a 32 bit system running 32 bit
applications, and running a very small suite of applications.
It would be a bit like comparing the corner 7-11 (Windows), to the big
indoor shopping mall with 300 stores under one roof, including
franchises (Linux). The point is that OSS software is becoming widely
recognized and is well recieved by those who use it. It might not be
"as good" as Office, but it's good enough to topple the barriers to
adopting Linux.
> They also mocked OLPC and then asked to participate in the project
> (because it brings Linux to many millions of children).
Keep in mind that when Apple put it's little Apple ][ machines in
schools, it sold millions of Apple ][s at a time when personal
computers were such a new concept that most businesses couldn't see a
practical need for them. Before long, the kids grew up, and they were
bringing their Apples to work with them. Sure, some people switched to
CP/M, but even that was a byproduct of Apple's efforts. By the time
IBM entered the market, they had to displace almost 2 million
personally purchased Apple, Commodore, and CP/M machines. When IBM
started with only 1/4 million, Compaq licensed MS-DOS and came in with
almost twice as many.
Suddenly the OEMs have a new choice. They can choose Linux/Unix(OSX).
But the customers aren't quite willing to drop Windows completely. At
the moment, the choice seems to be OSX. But OEMs aren't going to be
willing to sit back and sell "32 bit Windows-only machines" while Apple
sells OSX/Windows machines and grows to 15% of the marketplace.
The OEMs aren't stupid. Over 1/2 the machines sold in the last 12
months were 64 bit machines, including AMD-64 and Intel Duo machines.
These machines provide no significant gain over 32 bit machines when
running Windows XP, but when running Linux, they run at almost 3x the
speed of comparable Windows machines (Linux is less limited by disk
access).
> Ethics? what's that gonna cost me?
Not ethics or lack of ethics. Bill is retiring. Bill was selling over
1 million shares a day back in October, and may be ready to dump a
bunch more. Steve is isn't a techno-geek. He needs to look at where
he has the best chance of staying in the market, where he can get the
best profit, and how he can keep the company solvent in the wake of any
substantial transition to Linux or OSS.
More importantly, he needs to keep an eye on the future. Perhaps the
reason Bill is retiring is because he can see into the future. If you
could see into the future and knew that you would be spending the next
10 years of your life in a wheel-chair, you might not like the vision,
but you could make appropriate plans. You could move into an elevator
building, you could make sure that you can work from home, you can
learn to telecommute, and you can make sure you have access to public
transit nearby.
If you knew that Windows would become a "second string" operating
system, would you wait until all of your applications were orphaned, or
would you make arrangements to make sure that your available products
could run on the hottest new contenders?
Vista isn't even officially out yet, and most of the industry has been
saying:
1. Windows Vista is to little - not enough real features that
actually improve productivity.
Eye candy is nice, but it doesn't provide ROI.
2. Windows Vista is too late - the 32 bit version seems obsolete
when the processors are 64 bit.
the 64 bit version won't run 32 bit applications. Linux
would let users run 64 bit Linux applications while
letting them also run 32 bit Windows applications.
3. Windows Vista is too much - It eats too much memory, hogs too
many resources, soaks up hard drive.
yet with all of these additional resources, there is almost
no real gain in productivity problems.
4. Windows Vista didn't solve any problems - Vista didn't solve the
security problems, it didn't solve the
information life-cycle problems, and it still takes hours to
find a file for a simple project 20 staff-year project.
5. Windows Vista creates new problems - The EULA is almost
offensive. It pretty much says "yes you paid for
the license, and purchase the PC because the license was
included, but we won't let you use it, unless
you keep asking nicely. We can disable it any time we want.
6. The law of diminishing returns - People who already have Windows
don't need Vista. People who don't have
Windows now probably never will. There is no compelling
reason to spend twice the price for the same
performance, fewer features, and more restrictions, just to
get a cute little 3D GUI full of "eye candy".
Sure, Microsoft might be able to pull another "force feed" maneuver,
like they did with XP back in 2000, but the last time, this stunt
triggered a huge interest in Linux and OSS. This time, the CIOs have
made sure that every machine they have purchased for the last 3 years
is capable of running Linux. They have made sure that every document
can be read by an application that runs on Linux. They have made sure
that at least 80% of the machines used by their employees could be
switched to Linux in less than 48 hours. CIOs, CTOs, and CEOs who have
not made such contingincy plans, will likely be looking for new jobs if
they commit to a force-fed VML agreement that involves higher costs of
any kind.
Keep in mind that if the CIO wants to try and upgrade existing
machines, instead of just replacing them, it will be much easier to
upgrade these machines to Linux, which is a known quantity, than to try
and determine which machines can be converted to Vista, and which will
require hardware replacement.
Remember too, that SANs, External USB and Firewire drives, and Flash
keys have made the transition to Linux much easier. If I can spend a
less than $200 and have a full back-up of my personal data (which I
will need to do anyway), and can install Linux along with XPG and XEN,
in less than at hour, and I can reinstall Windows XP as a Xen client,
then I have even less need for Vista, which won't even allow me to run
the basic system as a XEN client. I have to buy a bunch of features I
don't need or use, just to be able to run Vista as a virtual machine,
and the cost is as high as $299 (MSRP), but I can get Win4Lin Pro for
only $69, and Crossover for only $69.
http://www.win4lin.com/
http://www.codeweavers.com/products/cxoffice/
So why would I want to spend $299 for Vista again?
> ,----[ Quote ]
> | It seems to me that IT fraud comes in two main forms: those perpetrated
> | for money or power, and involving failures to act ethically.
> `----
>
> http://blogs.zdnet.com/Murphy/?p=776
|
|
