Conference 7.286::digital

Our OS neutrality message is greatly weakened when we produce
platforms which go out of their way to prevent some OS's from
booting (i.e. Alpha XL).  This is getting us a lot of negative
publicity.

However, I believe the recent P6 and UltraSPARC announcements
have the folks in workstation engineering running scared.  So
it wouldn't surprise me if the Alpha XL (or something similar)
will become a Universal platform soon.  Oh, the blessings of
competition!

Ram

    
    	I go with the neutrality issue completely. It is tough enough
    selling against everybody's everything these days.
    
    	Playing Operating System Wars would be a killer. Let the customer
    decide after you present an unbiased case. After all, he/she is the one
    that *has* to live with it...
    
    		the Greyhawk

    Those of us with a few years history in this company will recall
    the OS camps existing even a couple of years ago.  These used to
    spend enormous energy competing against and opposing each other's
    efforts.    The only winners were the Competition.
    
    Now that we can see a way out of these woods it would be 'inadvisable'
    to resurrect those attitudes.  Any OS the customer wants is fine
    by me as long as Digital gets someting out of it. If you can sell
    OpenVMS on Digital hardware then GREAT.  I am currently delivering
    paid SI consultancy to buy in HP kit for a customer!  No problem,
    we still get the SI $$'s and have a chance to quietly influence
    the long term buying pattern.
    
    For every decision, "Now what does Digital get out of this?"
    
    Calvin

    
    re -.1
    
    <<< Neutrality WINS,
    
    Ahh,
    IF you have the applications for the O/S's.
     IF you have the O/S.
      If you have the people set.
    
    neutrality produces generalization in product sets and people.
     generalization produces a weakness unto itself in specific instances.
    neutrality is more costly to maintain, both application wise and
    hardware wise. Everything must work everywhere, everytime.
    So I believe in order to win, you really need to focus on core products
    and services, that while neutral, are the best they can be. We seem to
    be on that road, but we have a long way to go.
    
    -Mike Z.

    Re .3
    
    Most notes with such strong opinions about which technology is
    superior always leave out the IMHO or the like.  History would
    tend to show that none of the above may be the correct answer.
    
    Example:  If you would have told me that the best selling OS
    (by number of licenses sold) in our industry for mid-scale server 
    platforms in the mid-1990's would be UNIX (industry-wide outsold 
    NT by a factor of approx 10 in '94, NT gaining in '95) I would 
    have said *NOT*.  
    
    Of course, we wouldn't have seen Wayne's World, so we wouldn't
    have said *NOT*.  Maybe the future is little Web-server appliances
    with no server platforms, it's just too uncertain.  Gosh, this is
    starting to sound like a Digital commercial I've heard recently...

    I always find it amusing when I read of the impending battles between
    UNIX and Windows NT.  IMHO, the war has been over for quite some time. 
    The only thing holding back the NT onslaught is Microsoft itself, and
    how quickly it can engineer and market NT into commercial/mainstream
    applications.  There are two main reasons for this view:
    
    1. The vast majority of computer users have some familiarity with the
    Windows environment, whether it is 3.1 or 3.11, which allows
    organizations to more freely introduce another Windows operating
    environment.  Now contrast this with the general fear and loathing that
    most computer users have for the UNIX environment (don't get me wrong,
    there are a number of proponents of UNIX, it's just that they pale in
    comparison with Windows lovers.)
    
    2.  The price is right.  We can argue all day that it is the most
    closed and proprietary system in the world, but the relatively low
    price and the name recognition of Windows NT will allow it to capture
    widespread marketshare.  
    
    I work with organizations that operate directly in the DoD Intelligence
    space.  When national security is at stake, the environments chosen
    need to be robust and maintainable.  You would be shocked to see the
    movement to NT, and just how much has been done in the Labs to prep for
    its arrival in the mainstream.  Sorry, UNIX afficienados, you have met
    the enemy, and it is NT.
    
    Chuck
                  

    
    When WNT matures it will be like having VMS availible on all hardware
    platforms..... 
    
    V+1,M+1,S+1 = WNT

    But it will take WNT a long time to even come close to maturing as
    much as VMS.  It still has a LONG way to go - not to say it won't get
    there eventually.  What will make WNT take off is when it can be used
    for timesharing systems.
    
    					Steve

    The real problem with all the OS proponents' claims is that they tend
    to be more statements of personal preferrence or boosterism than the
    sort of measured market forecast on which you'd bet a few billion
    dollars worth of business. 
    
    Any serious study I've seen shows significant market share for both
    UNIX and NT rigth out to the forecast horizon. For that matter
    proprietary operating systems will continue to have a place as well.
    
    I simply can't see betting the company on one particular outcome.
    
    \dave

    
    Re .4 " ... as long as Digital gets something out of it ..."
    
    Sorry about the rathole but I completely disagree with this.
    
    IMO during most SI consultancy projects we are paid to work in the interests
    of the customer and only the interests of the customer. If Digital gets
    something else out of it above its fees, thats fine, but secondary. If the 
    customer needs VM/CMS we should recommend VM/CMS even though we get 
    nothing out of it and it helps the competition. If we are working with the 
    client for the benefit of both then this should be agreed up front and 
    Digital should pay its share.
    
    Also IMO, the attitude that consultancy is a way of getting the customer
    to pay our sales effort is destroying any credibility we have as a serious
    consultancy service provider.
    
    Mike.
    An ex-consultant.
                                   

    Apologies, perhaps I wasn't clear.  We want to stay in business and the
    best way to do that is to keep the customer satisfied.  All I was
    saying was in support of that.  Propose our own technologies if that
    will achieve that satisfaction but if not propose a suitable
    alternative.
    
    What I am against is warring against ourselves internally and
    forgetting that there's a real world out there full of competitors
    who delight (as we do) in eating another's lunch.
    
    I do not believe we have opposing views, simply a different emphasis
    in describing them.
    
    Calvin

    re: .10
    
    This is right on the ball.  Why?  Because the assumption nnn-million
    computers in nnn-thousand corporations will all run a _single_ O/S
    (regardless of its name) in the next decade ignores reality.
    
    There are a variety of BUSINESS REASONS why an O/S is chosen: technical
    merits, standards, available applications, personal biases, strategic
    advantages all come into the mix.  But, to insist that all these
    complex business issues will boil down to single answer (whatever that
    may be) is to insist the world businesses of every type and size will
    all have almost exactly the sames needs, wants, and goals.  That isn't
    going to happen as long as human beings run companies.
    
    And don't forget that when one product becomes dominant and appears to
    be the only game in the future, some irreverant upstart appears and
    moves the nature of computing to some newer, higher plateau.
    
    About the only way I believe you might see a single O/S becoming the ONLY
    player in the market, is to invent some critical networking-type technology
    which only allows that one O/S to play (I'm not holding my breath
    waiting for that).  Outside of that, one could anticipate the birth of
    a dictatorial World Government which will rule information transfer by
    sovereign decree, but that's a separate topic, entirely.  8^)
    
    -- Russ

I'm with Greyhawk on this one: we have finally (mostly) achieved a balanced
message about our 3 operating system strategy, and then the "Dream Machine"
comes along and messes it up.  I agree that NT will eventually dominate the
industry, but this is a minimum of 10 to 15 years away, and we have to make
some money in the meantime.

Microsoft believes that NT will grow up to be "enterprise ready" in about
3 years, and while I think that is a bit aggressive, they just might do it.
But they have completely neglected the idea that no IS Director will bet 
their company (and the IS Directors job) on NT today, or even a few years
from now.  Gartner Group talks about 3 classes of companies: early adopters
who use computing as a competitive advantage and who are willing to take a
serious amount of risk to achieve that end, companies who use technology
to just get the job done but never buy V1 of anything, and clerical type
companies who barely use computers at all.  Digital sells to the first two,
but has very few customers in the last group.  The point here is that the
number of companies in the second group is *much* larger than the number
of companies in the first group.  Microsoft will have some success in 
getting NT into the "bet your business" environments in the early adopters,
but will have *no* success in moving the second group to NT for at least
10 to 15 years.  And whether NT is technically capable of doing the job is
completely irrelevant to whether the IS Director in the second group of
companies will bet their job on Microsoft.

Think about banks doing international funds transfer (miss 1 transaction
and you might have lost $10M -> $100M), or the stock markets (think about
being the IS Director at the NYSE or NASDAQ, and having to explain to the
entire US financial community that trading was shut down because you felt
that they should abandon what has been working for years in order to go to
a brand new operating system), or the military ("Sorry General, we can't
tell you how many enemy missiles are in the air: we need to reboot the 
server") or a production line (which is dealing with things like molten
steel or other liquids which have a tendency to solidify in the valves and
pipes while you are working on the problems with the computers, which means
that you need to replace all the valves and pipes in your factory for a 10
minute computer failure), etc.

These types of customers will not go to NT until it has been out there for
years, and has proven to be solid and reliable in other companies environments.
This is why Microsoft may achieve technical parity with the big boys (MVS,
OpenVMS, UNIX, etc) in a few years, but will not achieve the penetration they
want in the market in that short time.

I have 2 responses when customers ask me which operating system Digital 
recommends for them: one humorous and one serious.

The humorous one is "Well, that is like asking whether Delta Airlines
recommends the aisle or the window (slight smile here for emphasis on the
pun) seat.  They don't care as long as you are on Delta.  And that is the
way we are at Digital..."

The serious one is "There are 3 important factors to consider when buying
real estate: location, location and location.  The 3 important factors to
consider when buying a computer system are applications, applications and
applications.  Now, which applications do you want and where do they run?"

And .-1 has a good point as well: the prediction (which I believe) that NT
will become the dominant O/S.  Dominant .NE. only.  Will NT have the 
majority of desktops and servers by 2010?  I believe so, but that merely
means that it will have 40% of the market by 2000 and 60% of the market
by 2010.  There will still be the 20% or so running UNIX and 20% or so
running everything else.

-- Ken Moreau

    re: .15
    
    While I agree with just about everything Ken said, his off-the-cuff
    analysis of the the year 2010 is offbase, IMHO.
    
    Frankly, I have a hard time believing that some new O/S won't be eating
    NT's lunch by 2010.  The computer business is SO fast moving, SO
    cutthroat, that I can't imagine SOMEONE grabbing something in the
    technology that Microsoft misses and exploiting it into a serious
    challenge to NT.  Heck, I'd be surprised if MS can hold the desktop for
    15 years on the merits of NT alone.  MS is getting larger and slower,
    which means it could become too slow to effectively compete 10
    years from now.
    
    Predicting the face of computing in 15 years is like predicting the
    shape of the next snowflake that falls:  you can analyze all you want,
    but the chances are that you'll be wrong.
    
    Just my take on things.
    
    -- Russ

One big difference between the current situation and previous changes
is that there is now a huge (and still increasing) investment in people's
time for knowing Windows interfaces.  That doesn't necessarily mean
Windows NT, but unless something so obviously better comes along that
it can make a major dent, it's going to be very hard for a competitor
to swim upstream against the MS Windows UI "franchise" (which will have
a strong tendency to carry Windows NT, or some follow-on, with it.)

Inertia goes with mass, and the "mass" is far larger now than it was
10 (or even 5) years ago (that applies both to the UI, and the number
of deployed applications.)

MS/Intel were at the right place at the right time, and that combination
is going to be very hard to beat from an economic standpoint for quite
some time.  (I'd also MUCH rather be in MS's shoes than Intel's -
architectures can be cloned, emulated, etc.  People's time investment
is the stuff long-term dominance is made of.)

As far as MS getting larger.  True.  On the other hand, MS has had
a tendency to build on previous successes, and farm things out once
they're no longer riding the "crest" (i.e. the appear to be more
committed to riding the "technology wave" than a number of other
companies in the past have tended to be, so they may well succeed
in staying in the agressive growth/change stage for a considerably
longer period.)

Microsoft for a second...

A good friend of mine works over in MS ITG, runs the VAX, which incidently
still does the mission critical stuff relative to financials...

We debated over this subject for quite some time last night...

The company line is that NT is going to kill UNIX...at MS, there is no
Netscape, Notes, barely any UNIX, SAP will be implemeted on NT only,
no Java...et. al.  His stance was very 'UNIX is dead, UNIX is too cryptic',
I've been there...I felt the same way while I was there...but then I gave
UNIX a chance, and now I like it just as much as I like NT or VMS for
that matter.

I mean this guy gets visibly upset when I say something like, 'UNIX is
selling, and that means the install base continues to grow', some people
over there will even dispute the numbers...

Is this healthy?  Ignore UNIX and it will go away?  I remember being over there
as a contractor and watching vendor after vendor getting absolutely slaughtered
in presentations over UNIX code base...HP, ATT, CA, etc.  They get mad if you
tell them a solution will be a port from UNIX, they slam NutCracker...they
want everything built from the ground up on NT, and if you even mention that
you like UNIX, you get ostrasized (sp?)...

What I'm saying is that a lot of FUD is used against things that they have
no control over, like Java, like Netscape....and yes, like UNIX.  NT in
it's current incarnation will never be 'better' than our 64 bit UNIX when
it comes to 'full blown, bet your business' applications...and what bothers
me is that this company wants me to ignore that fact, and join some kind
of Operating System Cult...this to me is not very savy, and I really do find
this concept a bit ridiculous...

Also, UNIX is not standing still...CDE, advfs, clusters, Scheduler 3.0, etc.
etc., more and more applications are shipping each day...there is critical mass.

Being as these are our Allies, how do we balance our relationship with them,
and the need to be O/S neutral?

-Stephen

    Don't be surprised if NT doesn't implement 64 bit in the near future!
    Predictions are that it will be 45% NT 45% UNIX by 2000. I agree that
    in 10 years the total picture will be totally differenet but thats
    progress. We need to concentrate on what we can see, which is 2 - 3
    years.
    
    dg
    

	We balance our relationship by agreeing with them. At least those
	folks that have day-to-day contact with MS that is. (I wouldn't
	expect the Digital Unix group to appease MS :))

	I for one, happen to side with the folks at MS a bit. Even tho
	it's a shortsighted view. I'll never be able to profess a love
	for Unix (or OS/2 for that matter). On the other hand, where I
	split off from MS on NT .vs. Unix is that I recognize that Unix
	has its place. It's just not in my office. So when I go to 
	Microsoft, I say I love NT. I don't say I dislike Unix tho. If
	they ask my opinion, I defer comment. 

	ALL of DEC doesn't have to scream "I love NT" from the MS rafters.
	However, those of us who are working on NT probably should push it
	as enthusiasically as they are.

							mike

RE: .19

	There was an article in this weeks InfoWorld on just that.

								mike

    Perhaps the issue will not come down to the technical merits, but to
    the cost.  An ISV increases their cost each time they add support for
    another flavor of UNIX.  This means that the market has a built in cost
    bias towards the support fewer OSs.  This greatly favors NT and will
    help push ISV development dollars towards NT and thus fewer ISV
    development dollars toward UNIX.  I see this happening today and feel
    that it will only accelerate.
    
    A non-cost issue is the Intel/HP deal for the P7 64-bit UNIX OS.  This
    is causing a lot of problems for existing UNIX vendors, who feel they
    are going to be left out.  Also, it is my "personal" opinion that
    customers are not going to look favorably (in the future) on an
    operating system that comes from a hardware vendor(s).  WNT is going to
    end up running on multiple chip architectures.  The Intel/HP UNIX is
    going to run on only one.  I think customers want more flexibility than
    that.
    
    Therefore, I feel that cost and platform choice flexibility could play
    a more important role, overall, than technical merits.  This all favors
    WNT in the long run.  I do not dislike UNIX or VMS, just think the
    market is against them, and it is hard to beat the market.  Try
    shorting the market when it is rising and you will see what I mean.

    re: 13
    
    Point well taken.  Technically superior products often do not 
    enjoy the longer term success of products with less to offer,
    but that catch on in our industry in a big way despite any 
    shortcomings.  I agree with previous entries that have suggested 
    that predicting more than year or so ahead is shaky ground! 

    The OS wars? I personaly don't believe DOS will ever be in a position
    to take over CP/M's installed base. If Digital could just provide a
    machine that runs BOTH. ...Did you ever notice that if you live long
    enough you find yourself in places you have already been, even if you
    don't try to get there??

    As a long term VMS person who has, generally, avoided Unix I have
    recently come across a new devlopemnt which could have an influence on
    the future.  Linux is a FREE Unix for Intel PCs.  Though it is still
    young and probably not ready for main stream applications it does, on
    the surface, appears to be quite good.  There are a full range of FREE
    applications and tools for it (including C, C++, X11R6, etc.).

    With the increasing costs of MS based software they are a growing number
    of people trying this out.  It could be the kick that Unix needs to
    survive, even if it won't help the Unix vendors!

    Gary Harding

If you want to talk about FREE UNIX on the PC, don't forget FREEBSD.  I'm note
entirel sure where this one came from or what the benefits are over lynux, but
my son runs it.  He got the idea from his Internet Service Provider who uses
it in a commercial environment.

-mark

    re .25
    
    Linux is for more than Intel PCs. It also runs on a variety of Alpha
    systems and the Alpha distribution is available through the usual Linux
    distributors. It even has binary compatibility with Digital Unix (but
    take care, I was wearing a Marketing hat when I said that). There's a
    Linux FAQ around, and a selection of newsgroups, but I can't remember
    where without my All-in-1 system, which just went away.
    
    regards
    john

    http://www.azstartnet.com/~axplinux/
    
    is a good place to start on Linux/Alpha
    
    /Charlie

	Nothing is ever truely free. You always get what you paid for.
	This is not a slam on Linux, just on the thought that something
	is for free.

	There's always things like support costs and stuff that are
	"hidden" when you talk about free stuff. Customers pay big
	$$ for things like VMS and NT because there are people who's
	JOB it is to support them. Linux is definately a child of the
	Internet in that the support is done on a "midnight project"
	level. When you are betting your business, you don't want to
	have to depend on someones good graces to solve your problem.

							mike

    How many companies bet their business on BIND, Sendmail, Emacs, TCL/TK?  
    Aren't these all Free software developed on the Internet.  I think if you
    surveyed most internet providers in the US they are running their
    business on a Intel/UNIX, Either BSD or Linux.  As far as no support,
    companies like CyGNUS and BSDI offer support on these UNIX's and free
    software.  NT big disadvantage compared to UNIX is that it is
    Proprietary.  Now NT probably will grow to dominate the market, but
    as other's have stated predicting the computer market more then a
    couple of years out has prooven impossible.
    
    Dan
    

Boy, a lot of activity here...

RE: .16 (Russ Pavlicek saying that NT will be overtaken by something else)

I think that .17 addressed that one: for the very first time in the computer
industry, there is a massive investment (literally millions of people) in a
single system.  Not the fairly minor investment that existed in the 1960's
for IBM mainframes (which seemed big at the time, but is *nothing* compared
to the investment in Windows today), or even the admittedly large current
investment in the many flavors of UNIX systems.  But even that only deals
with computer professionals, who have never really been more than 10K to 20K
people across the world.  (Yes, I know that more people than that *use* UNIX
systems, but most of those could switch to another O/S with only minimal
re-training, because 90% of the time they are inside a GUI application, so
the underlying O/S means nothing to them).  As a proof point, count the 
number of workstations sold per year from *all* vendors, and compare that 
to the number of Windows systems sold.  Intel literally sells more systems 
in a month than the entire workstation industry (Sun, HP, IBM, SGI, Digital, 
etc) together sell in a year.

What that adds up to is a group of people numbering in the 10s of millions
(not the 10s of thousands which has always been the number in the past) who
know and like a particular operating system and style of working.  And the
second part of that sentence (style of working) is more important than the
first part (particular operating system), so we could switch from Windows 3.1
to Windows-NT Workstation 3.51 and 95% of the people would not notice the
difference.

Therefore you now have a set of mid-level to senior level managers all over
the world who bought a system for home, and now want to use the same kind
of system at work.  They understand the applications, they are comfortable
with them, and they refuse to learn grep and awk and vi and roff when they 
could be using Program Manager and Word for Windows.  So there is a built
in bias towards a particular operating environment in business.

Finally, (and this is the key to it all), you have the ISVs.  The see the
market for UNIX as a Tower of Babel, with a maze of twisty little UNIX
versions, all somewhat alike.  There is a fair amount of money to be made
there, but at a fairly high cost because of the differences between the
different implementations (and any vendor who says "My UNIX is better than
your UNIX" simply scares the ISVs, because better = incompatible from a 
developers point of view).

Contrast this with the Windows market.  You code to Win32, and you are done.
Your program now runs on literally 10s of millions (not 10s of thousands 
if you code to SunOS or Solaris, or thousands if you code to anybody else)
of units.  And the market is growing by amazing numbers.  So every ISV I
know of is porting (or re-coding from scratch to take advantage of the
opportunity to go back and fix all the problems they have known about for
years) on their own nickel to Windows.

And of course this completes the circle.  Huge numbers of potential buyers
sucks in the ISVs hoping to cash in, which leads to huge numbers of really
great applications, which of course attracts huge numbers of new buyers...
Round and round we go, building momentum all the while.  And the growing
numbers of developers and support people who know and like Windows simply 
accelerates the process.

So for Russ's theory to be true, this new operating environment which will
wipe out Windows would have to be able to overcome the incredible depth and
breadth of the applications available on Windows, the amount of investment
that each individual and each private business has in their computing
environment (including training and current software, the hardware is almost
inconsequential), and the massive investment that Microsoft is able to make
to keep Windows on the leading edge of functionality, reliability, usability
and all the other factors it will have.  (Yes, I snickered too when I wrote
the words "Windows" and "reliability" in the same sentence.  But look at
Windows-NT and what it has done in just a few years, and project that slope
out a few years with Microsoft spending the kind of money they are spending
toward that end: it *will* happen).

Sorry, Russ, I just don't see it.  Any new "killer app" will be placed on
a Windows box: the investment bankers who are funding the development will
insist on it.  It will also probably be placed on a UNIX box (ex: Hot Java),
but it *will* be placed on a Windows box.

As a case in point, I give you OS/2 Warp.  It is demonstrably superior to
Windows 3.1 and Windows for Workgroups 3.11, and is at least on par with NT 
3.51 and Win-95.  It has perfect emulation of Windows code, so that it will
run all DOS/Win16/Win32 applications seamlessly (better than NT 3.51 will).
I submit it has failed in the marketplace, and will never represent a
serious threat to Windows, because it will never have the massive numbers 
of people who are comfortable with it at home specifying it at work (and
vice versa, the massive number of people who use it at work then buying it
for their own home).  I suggest this is the fate of all competitors to the
Windows dominance.

Now, will NT look the same in 15 years as it does today: certainly not.  
OpenVMS has been changed in major ways from it's original V1 days, but it
is still recognizably VMS.  NT will follow the same path.  It will grow and
mature, but it will never suffer the wild mutant strains of UNIX because 
one company with the marketing and financial power to hold it together owns
it, and will never let it go.  Even if the US Justice Department decides to
try something (and remember that IBM fought them to a draw over many of the
same issues over the same kind of timeframe), you will still have a single
division of Microsoft owning NT.


Finally, a bunch of people brought up Linux.  Cheap, runs on lots of boxes,
lots of software being ported to it, etc etc.  Pardon me, but this is a
mighty tempest in a very small teacup.  For hackers and hobbyists this is a
neat and cool thing, but for production systems this will never happen.  My
points about IS Directors of major corporations not betting their jobs on
NT apply a thousand times over to freeware.  And, with all respect, saying
that because some Web servers run Linux, that it is ready for enterprise wide
implementation on mission critical systems for the Fortune 100, is a joke.
Get a grip, people.  Picture going to an old conservative bank and proposing
that they run their customer accounts on a system for which every hacker in
the world has the sources, which no major company supports, for which they
have no recourse if it fails in any way.  Right.

And the fact that it is free is meaningless.  The cost of the operating
system is practically nonexistent when you compare it to the cost of the
hardware, the computer rooms, the network and communications gear, and most
especially the very expensive people you have to have to make it run.

-- Ken Moreau

I wonder if the user's cost of supporting good, working, free software is
any different from the cost of supporting paid-for software.

___Pete

    re .32;
    
    It really doesn't matter. The very strong perception that freeware is
    	
    	a) less reliable 
    	b) more expensive to maintain 
    
    whether true or false, will keep the big players from ever making the
    experiment.
    
    \dave
    
    \dave

    re: .31  (Ken's well-thought-out reply)
    
    Ken makes good points here, but I believe he has also made a few
    critical assumptions that I am not willing to make:
    
    	1.  MS's huge cash investment in NT will guarantee it will stay on
    		the leading edge.  MS is a company, like any other company.
    		It has done admirably well on maintaining its competitive
    		edge -- but I cannot say that it will stay that way for the
    		next 15 years.  Companies tend to get "stale" at some point
    		in time; the leadership makes some wrong turns and it costs
    		them marketshare (the list of examples here is huge, as we
    		all know).  The MS dynasty is only as solid as their
    		leadership.
    
    	2.  Users are wed to MS-style computing.  I believe users are
    		actually wed to GUI and apps.  Look at NT.  It's not DOS,
    		by any means; it just looks familiar.  Most apps seem to
    		undergo major overhauls every couple of years, so people
    		are forced to learn more.  A supplanting O/S will need a
    		close-enough GUI and ported apps.
    
    	3.  Massive investment will keep it an MS world.  This massive
    		investment gets redone every 7 years, max.  How many users
    		are still using 1988 software and hardware?  Really, not
    		that many.  Many folks needed the extra capabilities that
    		came along and had to buy new hardware and load new
    		software.  Over time, the momentum of #2 is stronger than the
    		momentum of #3, IMHO.
    
    	4.  There will be no technology created which cannot be seemlessly
    		folded into NT.  This is the greatest argument I have
    		against Ken's analysis.  The appropriate technological 
    		breakthrough can change the face of BUSINESS which changes
    		the face of computing.  Look at the FAX machine.  It was an
    		interesting toy a decade ago; you look like a fool if you
    		are in business without one today.  The DOS desktop could
    		absorb this business technology, but will it be able to
    		fully absorb the next technology?  Maybe.  Maybe not.  (See
    		#1)  If not, expect a new player to light the horizon.
    
    Ken, maybe we should agree to teleconference (or whatever is common
    then) on January 1, 2011 and compare notes!  ;^)
    
    -- Russ

    >It really doesn't matter. The very strong perception that freeware is
    >	
    >	a) less reliable 
    >	b) more expensive to maintain 
    >
    >whether true or false, will keep the big players from ever making the
    >experiment.
    
    Maybe, maybe not -- the DoD is currently looking at using Linux
    as the base system software for its new transport plane.

re Note 4299.34 by NEWVAX::PAVLICEK:

>     re: .31  (Ken's well-thought-out reply)
>     
>     Ken makes good points here, but I believe he has also made a few
>     critical assumptions that I am not willing to make:
  ...  
>     	2.  Users are wed to MS-style computing.  I believe users are
>     		actually wed to GUI and apps.  Look at NT.  It's not DOS,
>     		by any means; it just looks familiar.  Most apps seem to
>     		undergo major overhauls every couple of years, so people
>     		are forced to learn more.  A supplanting O/S will need a
>     		close-enough GUI and ported apps.
  
        I think you are both right -- NT gets supplanted when the
        paradigm gets supplanted in a way that needs (or just happens
        to be built upon) a different base, and then NT gets
        supplanted by that base as a consequence.  Perhaps it will be
        voice- and gesture-operated devices worn on the person and
        always able to connect to a wireless network as required. 
        Perhaps something else, or some combination.

        NT gets supplanted when it can no longer do what people want
        to do, or when people prefer something built without it to
        the things built on it.

        Bob

Re 4299.35:

> It really doesn't matter. The very strong perception that freeware is
>	a) less reliable 
>	b) more expensive to maintain 
> whether true or false, will keep the big players from ever making the
> experiment.

Maybe so, but it's a simple factual question.  And I wonder whether anyone
has, in their own context for their own purposes, gotten a factual answer.

___Pete

RE: .34

I agree that we are both making assumptions, and since those assumptions
differ our conclusions will differ.

1) Microsoft will stay leading edge.  Perhaps I stated it poorly (never try to
   construct a well reasoned note after taking your kids to DisneyWorld for
   several days and driving back, then staying up till 11 PM to catch up on
   mail and Notes), but when I said "leading edge" I meant only in terms of
   solving customer problems.  NT does not have to be leading edge in terms of
   technology, but it does have to satisfy the needs of the marketplace better
   than anything else.  And a huge part of that is the investment in training
   and applications which 10s of millions of people have made.  And while it
   is certainly true that at some point Microsoft will fade from the screen,
   I don't believe it will happen in the 10 to 15 year timeframe we are 
   talking about.  And as far as recovering from failure, I give you Chrysler
   and Digital, both of which were down for the count, and both of which have
   come back.

2) I didn't understand your point 2, "users are wed to GUI and apps".  It 
   seemed to me that we are in violent agreement.  You said "a supplanting
   O/S will need a close-enough GUI and ported apps".  Hmm, we already have
   one of those: OS/2 Warp.

3) Investment is recycled every 7 years, and people need extra capabilities
   to keep current.  I don't understand why this works against Microsoft,
   since (it seems to me) they have done very well keeping up with the new
   waves (LANs with Windows for WorkGroups, and now the Internet with Win95).
   Sure people will buy new hardware and software, but do the words "installed
   base" mean anything to you?  Microsoft and Intel have an installed base
   which is larger than any other company in the world, in terms of active
   units, and in terms of old units being upgraded to new units.  The only
   market segment which might come close is automobiles, and there is very
   little brand loyalty among consumers there, where there is a phenomenal
   degree of brand loyalty to Windows.

4) We come closest to our differences with 4, where you said "there will be
   no technology created which cannot be seamlessly folded into NT".  I do
   think that, and you disagree.  It will be fun finding out!

I believe the next level of technology will be a *true* notebook.  It will
be the size and shape of an 8.5"x11" folder, will be less than 2" thick, will
weigh under 2 pounds, will have a high resolution color screen suitable for 
reading under any light conditions, and will come with either a keyboard and 
pointing device (for those who like to type) or a voice control system for 
people who don't want to type.  It will run for >12 hours on a battery.  The 
reason the battery lasts so long is that the unit will not have a hard disk.
It will get all of its storage from service bureaus, just the way we get 
phone service today (you used 10MB for 30 days, plus a peak of 14MB for 2
days, so your charge for storage is $xx.xx).  It will be in constant high speed
communication with its server(s) at all times (in the car, on the train or
plane, at home, **all the time**).  It will have a built in camera and audio.

This unit will replace cellular phones and pagers with video teleconferencing, 
FAX machines with e-mail (why send illegible images when you can send the
document itself), and will go a long way toward reducing the printing costs
of newspapers and magazines with on-line expanded content editions, complete
with up-to-the-second details and full video and sound (think about being on
the train to work reading the WSJ on-line, then flipping over to CNN to see 
the most recent developments on the story you just read, all on one screen).

From Microsoft's point of view, this thing needs very high reliability,
exceptionally resilient networking, and a user interface that literally
everyone can use.  Hey, sounds like Win99 to me!


RE: .37 -< Just a matter of fact >-

>> It really doesn't matter. The very strong perception that freeware is
>>	a) less reliable 
>>	b) more expensive to maintain 
>> whether true or false, will keep the big players from ever making the
>> experiment.
>
>Maybe so, but it's a simple factual question.  And I wonder whether anyone
>has, in their own context for their own purposes, gotten a factual answer.

But it is *not* a simple factual question, it is an *emotional* question.
The IS Directors of major multi-national firms are dealing with so much
technology that they literally cannot keep track of it all.  The operating
system is only one very small part of their problems, which include computers,
data networking, voice networking, databases, storage, applications, etc.
So these people are not up on the latest details of one O/S over another, they 
evaluate it on a much higher level, where cost is only one factor.

What is a factor is their comfort level with it.  If an operating system from
Digital or IBM or HP or Sun or ... breaks, the IS Director calls the CEO of
that firm and raises such a stink that those companies throw their experts 
at the problem and get it fixed.  So from the IS Director's point of view,
(s)he got the right people to solve their problem for the price of a phone
call.  I have personally seen this happen many times at Digital, and IBM made
billions of $ in the 70s because "nobody ever got fired for recommending IBM",
and the reason for that is that everybody knew that IBM would do *WHATEVER*
it took to fix the problem.

If they use Linux, who are they going to call?  I believe this is a much
bigger factor than the technologists realize.  Andrew Kenah's comment does
surprise me, but I will withhold judgement until they actually commit to
an implementation on Linux, as opposed to just "evaluating" it.

-- Ken Moreau

    
    	Now this *is* getting interesting...
    
    	My take, from dealing with both VARs, ISVs, Developers, and Large
    Accounts, is that Intel/Microsoft has already won this round; which
    just might last the length of the VAX (or 12+ years) during its
    ascendency. The question is not who the winner is, but what the
    next one will be?
    
    	Like Ken, I lean toward the notebook as an appliance. What
    powers it, chip-wise, is up for grabs; as for the O/S your guess
    is absolutely as good as mine. But this appliance is going to change
    how our kids work, and play...
    
    	Me, I'll be on the beach - thankful Digital is still cooking along!
    
    
    		the Greyhawk

    I believe that the future IS our kids.  We laughed yesterday when we
    read an Email from a 9th grader that wants to get an Alpha motherboard. 
    We're gonna try to help him find one, who knows where he may end up...
    
    Mark

    Breaking into phone systems and corporate computers?  :-)
    
    				-John

    More like designing a virtual based reality world.
    
    -Mike Z.

    This if off the INFORMATION WEB page this morning:
    
    "Corporations Jump to Microsoft's Other 32-Bit Operating System"
    
    "While Windows 95 rings up consumer sales, the corporate move is on to
    Microsoft's other 32-bit operating system, Windows NT Workstation.  The
    Redmond, Wash, software giant disclosed yesterday that more than 40
    corporations are each deploying at least 10,000 units of NT
    Workstation.  Equally impressive, several consultancies estimate NT
    Server is selling more than 30,000 units per month."
    
    
    

At least the kids know what technology is best...

When I was a kid, I didn't 'lust' after a Honda Accord (NT),
I wanted a BMW 633csi (UNIX).

My take on UNIX versus NT, is that UNIX is a better, faster, more
advanced Operating System, and that NT has borrowed and awful
lot of functionality from it...

Just like features that are expensive and prototyped on high priced
luxury cars, and later find there way into Ford Escorts...remember ABS?

My fundimental problem with Microsoft is this 'fanatical' need to
'kill' UNIX, when all along, UNIX has been NT's mentor in a way,
the Internet, another area that Microsoft wants to 'dominate' was
basically built on UNIX.

I don't see Ford out campaigning to do away with BMW, Mercedes and the rest.

It's irresponsible, short sided, and rather childish IMO, especially
since NT needs to interoperate very well with NT...it's complete silliness
IMO.  

-Stephen

> My fundimental problem with Microsoft is this 'fanatical' need to
> 'kill' UNIX, when all along, UNIX has been NT's mentor in a way,
> the Internet, another area that Microsoft wants to 'dominate' was
> basically built on UNIX.

Let me try and explain:

The PC software market is around $10 Billion.  All other software amounts
to around $50 Billion.  Microsoft already has about 60% of the PC software
market.  In order to continue their record growth and success, they need
to branch beyond the PC market.

Microsoft wants and needs a fair share of the non PC software business.
And as a certain Microsoft executive has said at internal meetings,
he considers 100% a fair share.

- Peter

But I do have a problem trying to manage a heterogenous environment
with an O/S that is built by Engineers whose sole purpose is to
kill one of the environments NT needs to manage...

Take DHCP for example...what a *pain* in the rear.

IMO there is a tremendous market opportunity here for NT based tools
that manage UNIX environments...

And don't tell me it's CA Unicenter...it's a monolithic monster, way
too cumbersome and way to expensive.

Again, this 'NT or the highway' mentality is lame.

-Stephen

    
    > IMO there is a tremendous market opportunity here for NT based tools
    > that manage UNIX environments...
    
    Exactly and that is where Digital is investing a lot of money.  One of
    our first products in this area is POLYCENTER AssetWORKS which provides
    hardware & software inventory of, and software distribution to, Digital
    UNIX, Ultrix, SunOS, Solaris, HP-UX and IBM AIX                

Yeah, I agree...AssetWorks is cool.

    That's as may be (and I have never seen or heard of it and have no
    reason to doubt your assertion) but GOSH, the "--works" monikers are
    so... well...  EIGHTIES.  Ya know??
    
    
    I say this with some realization that it take a lotta dough to make a
    brand or family name...  but this one seems (to these jaded ears) to
    have shot its bolt.
    
    All this imho of course.
    

    P6 has the performance and the marketing muscle of Intel to hurt
    RISC systems sales, including Alpha.  Bravo to those behind the Dream
    Machine.  Lets get market share now.  A server follow-up to the
    Dream Machine is in order.  
    
    Matt Jolma

	As someone who has spent the last year working on Linux
	on Alpha then I couldn't resist replying to this note.

	Microsoft has *almost* rewritten the phrase "nobody
	got fired for buying IBM" (replace IBM with Microsoft).

	The Unix market is huge and we have a great product in
	Digital Unix.  It's 64 bits, well thought out, robust and
	secure.  Linux is vibrant, dynamic and young.  I'd trust
	Linux on the desktop and Digital Unix in the core.
	Windows NT is coming of age.  Microsoft would like to move
	into the server space but right now they're desktop and very
	lucrative it is too.  Unix holds sway in the server space.
	No doubt Microsoft (Windows 95 and NT) will continue to grow
	their market but the Unix market is very large and as a 
	compute engine it is well understood and used.  I'd say that
	as a company we should continue to ride both horses and invest
	our efforts between them.  Customers running hetrogenious
	environments will benefit and we'll make money, which is surely
	the aim - rather than religious war waging?

	Dave

	ps Linux may be the last best hope for Unix.  

    RE: .-1
    
    The IBM analogy for MS is a good one.  One thing I like to point out
    with respect to LINUX is how much it resembles the ISA bus from a 
    S/W perspective.
    
    Think back to 1987 / IBM still owns the PC market, but clone makers
    have begun really encroaching on the PC franchise.  Solution -
    Design a new generation of machine that has so much proprietary content
    that it significantly raises the bar for competitors.  While 
    you're at it, make those competitors show you every add-in board
    they are designing (required for MicroChannel bus registration info),
    and also make them pay a significant royality for both boards and
    systems.
    
    What was the industry's answer?? "We we still have the AT-bus
    architecture (renamed by Compaq to ISA-bus architecture shortly
    thereafter / *Industry*-*Standard* Architecture).  Let's stay with
    that, rather than going the IBM way."  The rest is PC history.
    
    			*	*	*	*
    
    Now I know Microsoft is no IBM (this is positive comment), and their
    franchise on the PS O/S market is still very strong, but consider this:
    
    -	LINUX as an OS is priced like DOS 1.1 - $29.95 (US) with book and disk
    	just about everywhere (and cheaper as well- $0 as download).  The 
    	price of THE UPGRADE of WIN 95 is $80-$100US; the full WIN 95 is
    	more like $200US.  This might be ok in NA and Europe, but with
    	large numbers of PC's being built and sold in the rest of the
    	world, I wonder if all those other PC owners will go for MS?
    
    -	LINUX is essentially like the AT-BUS architecture; available and
        understood everywhere.  No problem with MS attempting to give their
        internal developers a competitive advantage; no hidden API's
        (source code always available); it's based on UNIX, with millions
    	world-wide using and working with this OS.
    
    My point is have we reached a time where the OS will no longer be 
    the property of the most significant application developer, just as
    the PC reached time when the HW architecture could not be controlled
    by the most significant PC manufacturer?
    
    Time, as usual, will tell ... Charles