Apple Reference & Presen…rary 8 (Internal Edition)/Apple R&P Lib Internal v8.0.iso/3-Presentations/Apple Computer Inc./Industry Competition/ROMs/Software/IBM Microsoft Realign OS⁄2

Text File | 1990-06-24 | 13.3 KB | 269 lines | [TEXT/GEOL]

Read-Only Memo
David Garr & Michael Mace, Apple Competitive Analysis
Need to Know Confidential


IBM, MICROSOFT REALIGN OS/2 DEVELOPMENT


_______
SUMMARY

On September 17 IBM and Microsoft once again realigned their development plans
for OS/2, saying that OS/2 development for Intel machines will be done by IBM.
Microsoft will make only "development contributions."

Some of the early press claimed that this signals a new age of harmony between
the two companies, but we think it means just the opposite:  OS/2 and Windows
(and therefore IBM and Microsoft) are now competing with one-another much more
directly than ever before.  This has some important implications:

• Microsoft is now free to market and develop Windows more aggressively.  This
should make Windows a bigger competitive threat to Apple.
• Windows' increasing success should delay OS/2’s adoption even further.

Opportunities for Apple:

• Microsoft and IBM will create even more customer confusion as they jockey for
position.
• Developers will feel even more abused than before, and therefore will be more
amenable to Apple evangelism.

We’ll summarize the IBM/Microsoft announcement below, then discuss what we
think it means.  If you saw our earlier link regarding the announcement, skip
down to the section labeled "WHAT DOES IT MEAN?"


______________
WHAT HAPPENED?

For the second time in a year, IBM and Microsoft have clarified their
relationship and the relative positioning of DOS, Windows, and OS/2.  In a
press release issued by IBM (and endorsed by Microsoft), they also spelled out
the division of labor between the two companies:

•  Microsoft has primary control over development of DOS and Windows.
•  IBM has primary control over development of OS/2 for Intel processors.
•  Microsoft has primary control over development of a portable version of OS/2
which we believe is targeted primarily for RISC processors.

In essence, IBM controls OS/2 for the next couple of years.  Everything will be
cross-licensed between IBM and Microsoft, so IBM won't be a sole source, but it
is in position to drive the development of OS/2.

Here's the new positioning:

• DOS is for "entry level systems with less than 1 MB of memory and will be
enhanced over time."
• Windows "provides a graphical solution for users in a DOS environment."  It's
for personal productivity on systems with 1 MB or more of memory. "Typical" RAM
is 2 MB to 4 MB.  [The 2 MB recommendation is a subtle slap from IBM against
Windows.  Apple should play it back to customers when Microsoft claims that a
typical Windows machine has 1 MB.]
• OS/2 is an "advanced function operating platform," offers "full
multitasking," and is "ideal for larger, more complex applications, such as
mission-critical and line-of-business applications and is recommended for
servers."  Minimum RAM is 2 MB, and a "typical" configuration is 3-6 MB.

Those are pretty weak positioning statements, but the one that really counts is
the OS/2 identity, which has IBM's MIS customers written all over it.

IBM also said a "lite" version of OS/2, using 2 MB of RAM to boot, is on the
way.  OS/2 for the 386 will be shipped to selected customers later in 1990, and
generally available in 1991 (we think mid-1991).  This version will have faster
performance, run multiple DOS applications, and have better device drivers.
Note:  IBM is already using its control to seed pre-release software to
preferred customers.


__________________
WHAT DOES IT MEAN?

Here’s our current thinking about what this announcement means to IBM,
Microsoft, software and hardware developers, and Apple:

IMPORTANCE TO IBM:

IBM WANTS TO COMPLETE OS/2.  Like Apple, IBM now controls its own hardware and
operating system (PS/2s running OS/2).  IBM’s first OS/2 goal is to just finish
the thing.  The second goal may be to add hooks for proprietary IBM products
like OfficeVision.  After that, we wouldn’t be surprised to see features that
took advantage of IBM’s special hardware--Micro Channel, the multimedia boards,
and so on.

THE FORCE DRIVING IBM IS THE DESIRE FOR ACCOUNT CONTROL.  Many of the software
tools, such as OfficeVision, that IBM needs in order to hold its traditional
accounts (and make them buy more mainframes) are built on OS/2.  The success of
these products has been held up by Microsoft's slow OS/2 development.
Therefore, since IBM desparately wants OS/2 to succeed, it has taken control
over OS/2 and therefore has more control over its own destiny.

IBM WILL PROBABLY NOT PUSH WINDOWS.  Some analysts say that because IBM has
cross-licensed Windows, that IBM has now "endorsed" Windows and will help to
market it.  We doubt this will happen, primarily because of IBM’s desire for
OS/2 to succeed quickly.  IBM already resells Windows in some situations, but
only when customers or market situations demand it.  We don’t expect that to
change.

WINDOWS COMPATIBILITY IN OS/2 MAY BE DELAYED.  Microsoft had indicated that the
next release of OS/2 would run Windows applications directly, without emulating
DOS, but it's not clear if that's high on IBM’s priority list.

INTERFACES IN THE PC WORLD MAY DIVERGE.  The Windows and OS/2 interfaces might
diverge.  When Microsoft was codeveloping OS/2, it claimed that OS/2's
Presentation Manager would look and feel like Windows 3.0.  However, now that
IBM has control over OS/2, and Microsoft appears to have more freedom with
Windows 3.0, the two interfaces might further diverge rather than become more
similar.

The features and interfaces for OS/2 might also diverge.  It is unclear how the
development of the various OS/2 versions will be coordinated between IBM and
Microsoft.  Presumably there will be some effort to ensure that IBM’s
OS/2-Intel has the same features as Microsoft’s portable OS/2.  However, IBM
has an incentive to try to add new features to OS/2-Intel, in order to beat
Windows.  This may lead IBM to break with Microsoft over some features.  Adobe
Type Manager (vs. TrueType) could be the first example.

Even if the basic features of OS/2 don’t diverge, we think details of its look
and feel probably will.  Just as Windows does not look exactly like OS/2 today,
Microsoft’s OS/2 probably won’t look exactly like IBM’s.  This is a potential
marketing point for Apple, but only if it can convince customers to care about
interface details.  Only Macintosh users do today.

PATRIOT PARTNERS MAY PLAY AN IMPORTANT ROLE.  IBM and Metaphor have just formed
a new company called Patriot Partners to create a combination graphical user
interface and an object-oriented development environment.  In retrospect,
several recent deals by IBM now make a lot more sense.  The license from Go
Corp., the creation of Patriot Partners, and this announcement regarding OS/2
all have one thing in common:  They allow IBM to move further away from
Microsoft.  This seems to have finally become a major priority at IBM, which in
the past tended to view Microsoft as an irritant but not a threat.

IBM could theoretically use its relationship with Patriot, Go, and even NeXT to
add new features to its version of OS/2.  IBM has a strong incentive to do
this, and the world is certainly full of companies that would be glad to help
IBM take a chunk out of Microsoft’s hide.


IMPORTANCE TO MICROSOFT:

MICROSOFT HAS GIVEN RESPONSIBILITY FOR OS/2 TO IBM.  OS/2 has been a huge
marketing failure for Microsoft.  Bill Gates has now given that failure to IBM,
convinced IBM that it should be happy to get it, and positioned IBM to take the
blame from developers and customers if OS/2 continues to falter.

WINDOWS IS NOW FREE TO GROW.  We think Microsoft is now free of its earlier
commitment not to add some of OS/2’s features to Windows.  So you should expect
future versions of Windows to add long file names, more robust multitasking,
multimedia capabilities, and so on.  We also expect Windows to be put in ROMs.
This will allow it to run in cheaper machines and portables.

Microsoft had also been limited in how strongly it could market Windows.
Microsoft said things like "OS/2 is the platform of the 90s."  We suspect that
Microsoft's message will change.  It will be happy if Windows dominates the
marketplace for years, and it will probably promote it in that way.

The force driving Microsoft is the desire to control the PC industry.
Microsoft already has a stranglehold on the operating system, but it wants to
control applications.  The longer OS/2 is delayed and the stronger Windows
becomes, competitors late to market with Windows products, such as Lotus and
WordPerfect, will become weaker.

MICROSOFT WILL BE READY FOR RISC.  Microsoft is betting that OS/2 will never be
wildly popular on Intel-based machines; by the time customers are ready for
OS/2, RISC-based systems will have arrived. Therefore, it believes that by
giving up some control over Intel-based OS/2, it loses nothing.

We don’t know, but we think Microsoft would not have made this agreement if the
project were not making good progress.  We think OS/2-RISC will be in beta by
late 1991, and released before the end of 1992.  This increases the liklihood
that companies like Compaq would introduce RISC machines within two years.

In the long run, the migration to RISC could be very important for the PC
industry, and Microsoft is positioning itself to control that transition.  We
think that portable OS/2 will be ported to the Intel i860/960, IBM's POWER
processor (used in the RS/6000), SPARC, and MIPS.

MICROSOFT AND IBM ARE COMPETING.  Think of the situation as a gigantic bet:
Microsoft bets that IBM can't make OS/2 succeed in the next couple of years,
while IBM bets that it could lock up the corporate market if it didn't have to
work through Microsoft.

We think Microsoft can improve Windows faster than IBM can improve OS/2, so we
give Microsoft the edge.  But it's important for Apple to understand what's
going on:  Windows and OS/2 are now on parallel tracks, competing with
one-another.  This will push both Microsoft and IBM to move faster.


IMPORTANCE TO SOFTWARE DEVELOPERS:

DEVELOPERS FEEL EVEN MORE ABUSED.  OS/2 developers must feel like they've been
dumped out of the frying pan and into the fire.  On the one hand, they didn’t
trust Microsoft to finish OS/2 on time.  On the other hand, even though IBM
will be pushing hard to make OS/2 a success, IBM’s record for PC software has
been poor.  And if Microsoft begins improving Windows more aggressively, that
will push back the acceptance of OS/2.  Companies like Lotus probably feel
generally abused right now.  This should make them more receptive to evangelism
from Apple, especially once they understand the seriousness of Apple’s drive to
increase market share.

Currently Windows is outselling OS/2 by a margin of more than ten to one.  And
we suspect that developers believe that Microsoft will market and develop
Windows better than IBM can OS/2.  Therefore, we expect software developers to
increase their commitment to Windows 3 and decrease their commitment to OS/2.


IMPORTANCE TO PC COMPATIBLE MANUFACTURERS:

IBM MIGHT MANIPULATE OS/2 TO ITS ADVANTAGE.  In the past, Microsoft and IBM
codeveloped OS/2.  When a version was finished, IBM could immediately sell it
to its customers, while Microsoft would begin to tailor it for PC compatible
vendors' hardware.  Most PC compatible vendors have not been shipping OS/2 for
their machines until six months or more after IBM's version has been available
for its machines.  Now that IBM has solidified its control over OS/2, it could
lengthen that delay.  This means other PC companies may have trouble getting
new versions of OS/2 promptly. Furthermore, the announcement that IBM is
seeding OS/2 for the 386 to favored IBM customers must be chilling to IBM's
competitors.

Therefore, we think it is unlikely that the PC compatible vendors will heavily
promote OS/2.  Instead, they will probably increase their commitment to
Windows.

Compaq could be an exception.  We suspect it may be formulating long-term plans
with Microsoft for OS/2 RISC right now.


IMPORTANCE TO APPLE:

MORE CONFUSION IN THE PC MARKET.  As usual, the impact on the rest of the
industry will be more confusion.  In spite of the new positioning statements,
we think that it will become increasingly clear that Windows is now competing
against OS/2 more than ever.  This will disturb developers, worry customers,
and give the industry analysts fresh material.  Apple's message to the outside
world should be:  The growth path in the PC world is now even more confused and
difficult.  The fact that the Macintosh operating system runs on the complete
Macintosh product line and that Macintosh provides an excellent software
migration path could be used to sway these frustrated customers and developers
into looking at Macintosh systems.  More than ever, Macintosh is the rational
alternative.

FASTER INNOVATION FROM THE COMPETITION.  With both Microsoft and IBM competing
to try to develop the operating system of the 90s, we expect that there will be
speedier innovation.  Apple will be challenged to innovate faster than it has
in the past.

APPLETALK SUPPORT COULD SUFFER.  Previously, Microsoft and IBM had promised to
merge Microsoft's LAN Manager with IBM's LAN Server.  However, this
announcement states that IBM is now driving that process.  IBM’s server product
may not support everything in LAN Manager.  Support for the AppleTalk protocol
could be one casualty.


____________________
We welcome your comments and questions.  Please link us at COMPETITION.