[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Bundling and operating systems



On Sat, 18 Dec 1999, you wrote:
> > 1) Every modern operating system currently being shipped comes with a web
> > browser, generally integrated into the OS in some way. Even hoary old SCO
> > Unix, last of the "old line" Unix systems, now comes with a bundled web browser.
> > Even for people who do not use the Internet capabilities of the browser, it is
> > still being used for viewing help files, filesystem browsing, etc. 

> Is this logic supposed to offset the harm cause to consumers?

As mentioned elsewhere: if 51% of customers want a capability, then they are
getting the product cheaper if it's bundled. First, the cost is being spread
across the whole customer class. Secondly, the company saves money on
packaging, stocking charges, store shelf space purchases, marketing costs,
accounting, etc., and can pass that savings on to the customer. 

In short, if 51% of customers want a capability, then the majority of customers
benefit from the bundling.  But you do raise other issues. 

> 1) those who do not want a browser

A minority in today's Internet age. Most iMacs, for example, are purchased in
order to browse the Web (Apple marketing data confirms this, according to at
least one Apple advocate that I've seen posting). 

> 2) those who might want one but already have one (see the Bill Gates
> piece)
> 3) those who might want one but prefer their own choice

I believe that #3 is the main problem. Microsoft's current OEM contracts do not
allow OEM's to bundle whatever browser they want with the operating system.
Compaq, for example, might want to bundle Netscape as the default browser --
that is, have Netscape as the browser icon on the main screen, Netscape as the
default browser evoked whenever an HTML file is encountered, etc.  Microsoft's
current OEM contracts do not allow OEM's to bundle whichever browser their
customers are demanding (as is demonstrated by the now-famous contract
cancellation letter that Microsoft sent Compaq when Compaq dared put Netscape's
icon out there on the desktop). 

I believe that OEM's should be allowed to bundle whatever programs with the OS
that their customers desire, in the manner that customers desire. As a vendor
my goal is to please my customers, and Microsoft's policies would prohibit that
(if I were a Microsoft vendor, which, thankfully, I am not). 

> Bundling always causes two affects:
> 
> 1) the price in increased to cover the cost

The price of a specific instance of the product is increased to cover the cost.
But the price, ON AVERAGE, of the solution that consumers buy may actually
decrease, due to the decrease in packaging, manufacturing, marketing, and
distribution costs.  It depends upon how many consumers want the functionality
that is proposed to be bundled. 

> 2) choice is removed from the consumer in both the short and long run

Only if consumers are not given the choice to use alternatives. Microsoft's OEM
contracts clearly deprive consumers of that choice, and thus are
anti-competitive. 

> Therefore, I repeat, anyone who argues for bunding is only attempting to
> harm consumers.

Beware of blanket statements. They are usually used by people who are more
interested in bluster than facts. That includes this blanket statement (grin). 

I do not believe that Microsoft set out to provide value to the consumer. I
believe that Microsoft set out to commit anti-competitive acts.
Their actions regarding OEM licensing and allowing OEM's to
include Netscape as the default browser should make that clear
enough. Nevertheless, I do believe that this is one instance  where bundling
decreases the average cost of a solution to consumers. My evaluation is that,
by bundling the components that make up Internet Explorer into the Windows
operating system, Microsoft both a) provided more value to the consumer (by
allowing better integration of applications with the Internet), and b) provided
a more cost-effective solution for the majority of consumers (due to the
reduced packaging and distribution costs involved in a bundled vs. standalone
solution).  I base this evaluation upon two facts: a) Internet penetration is
now past 50% of personal computer owners (in fact, "to browse the Internet" is
the most-cited reason that consumers buy a computer today), and b) as a user
of the "K" Desktop Environment under Linux/FreeBSD/Unix, I have seen how
integrating the browser into the user interface has made the environment easier
to write, more powerful, and easier to use -- i.e., provided more value.  The
value doesn't change just because the perpetrator is Microsoft rather than
a loose coalition of German freeware authors. 

Of course, the big difference is that with KDE I have a choice -- I can use it,
or not. And Microsoft does not provide a choice. But the bundling itself is not
the cause of the lack of choice. Rather,  Microsoft's onerous licensing terms
are a cause of the lack of choice.

Finally, regarding my credentials: Lewis has implied, in private EMAIL and in
another EMAIL on this list, that I might be an employee of Microsoft
or a holder of Microsoft stock. Both assertations are laughable on the face of
it. My Linux and Open Source credentials are longstanding. At Executive
Consultants in 1995 I was project lead for the first non-Internet-related
commercial use of Linux in the state of Louisiana, porting a school
administrative system from Unix to Linux. Linux servers now run half the school
districts in the state of Louisiana due to my work. From there I went to
becoming chief technical officer of Linux Hardware Solutions, a vendor of
pre-configured Linux systems. When LHS was bought out by VA Linux Systems, I
then went to work for Enhanced Software Technologies, the makers of BRU, the
leading tape backup solution for Linux and Unix, and oversaw the migration of
their network from SCO Unix to Linux (see my home page at
http://members.tripod.com/e_l_green/editorial.html for a description of the
network architecture that I created for us -- Linux is both our server OS
and our desktop OS). I am currently the lead designer for a next-generation
software project underway at EST, though I can't talk about details on this
list due to non-disclosure reasons. One thing I will and can disclose is that
Linux is platform #1 on its list of server and client platforms supported (we
do all of our development under Linux, then port to other Unix-like systems,
and only then consider non-Unix-like systems), and Windows NT and Windows 2000
will only be supported as a client platform (and not under the initial
release)-- Microsoft products will NEVER be supported as a server platform by
any EST product as long as I have any say over the matter. 

'Nuff said. I hope this satisfies those who were wondering whether I was a
Microsoft employee or Microsoft stock holder.. I most definitely am not.

-- Eric Lee Green    mailto:e_l_green@hotmail.com     
       Software Engineer, Enhanced Software Technologies Inc.
       Former CTO, Linux Hardware Solutions
       http://members.tripod.com/e_l_green