From: Nathan Mhyrvold
Sent: Tuesday, April 16,1995 4:57 PM
To: Bill Gates, Russ Seigelman; Craig Mundie (raigmu); Dan Rosen
(drosen); Pat Ferrel (patfer); Paul MaAritz (paulma); Peter Neupert (petern)
Subject: Internet strategy
There has been a flurry of email about Netscape and our general Internet
development strategy. This email is my contribution to this topic.
My assumptions going into this, as discussed in my previous Internet
memo, are that:
- "Internet standards" in the sense of the current public domain
committee driven standards are a red herring and are not a competitive
threat. One reason is that most of the key people behind public domain
projects are busy writing business plans so as they can cash in. This is
true of the Netscape team, but also the many other rapidly
commercializing services. The big issue to be concerned about is the
same issue that we have faced in the past - propretary standards coming
from competing software companies. Netscape is certainly on of the many
companies who will try to promote their proprietary extensions (and
entirely new protocols) on the world.
Given this, our natural strategy is to try achieve a number of goals.
- Superset Internet protocols and standards with our own value added
extensions. As platform specific work on the internet, we want it to be
done on our platform. As proprietary technology and protocols are used,
we want them to be ours - in as many broad mainstream areas as is
reasonably possible. We don't need to own every protocol in every area,
but we want to be an important player. There certainly is a danger that
Netscape, or another company, could establish enough APIs and
proprietary protocol extensions that they would wind up owing the
"Windows internet platform".
PaulMa is quite correct in comparing this situation to Novell, which
successfully established a "sub-platform" (for a set of network
services) within the context of our client operating system. This does
not mean that Netscape needs to be a direct competitor - I am optimistic
that we can have a positive relationship with them - but out of the many
possible future directions for them and us, if we are not carefull they
will evolve towards being a direct competitor in this manner.
- The natural way for us to do this supersetting is using our current
technology agenda in PC computing. This means using monickers, OLE
objects, Forms3 forms and every other Windows technology that is
applicable as part of our extensions to the currrent Internet world. In
addition we also have to look at developing some new things that have no
equivalents in the PC world, such as security and billing, so it is not
all about reworking existing stuff.
- We will move to using TCP/IP, and thus will benefit from the ever
cheaper connectivity which is a central part of the Internet.
- We will allow access to any Internet service.
.. I have seen conflicting mail on this topic - some suggests that they
are willing to do things like license us technology and have us license
them some. Even if we ultimately are competitors. I see some value in us
doing this and trying to cultivate them as quasi-partners ..