Conference 7.286::digital

hmmmm... does this mean we've sold OpenVMS to Microsoft ?

  In a sense, we did that years ago when old Mr. "Punch-a-wall-
  as-a-management-technique" moved into Gates's stable.

                                   Atlant

    
    Atlant -
    
    	So you remember those days, eh?
    
    	Now Mr. Marine Major Sir is a media star. Makes you wonder, don't
    it?
    
    		the Greyhawk

    re: .0
    
    What concerns me is that if we are diligent, we will make opportunities
    for sales of OpenVMS to WNT shops, but if we are lax, we give Microsoft
    the tools necessary to eat our customer base.
    
    I'm not saying we shouldn't have done it, but I think we had better
    stay sharp and not let our new opportunity turn into our new problem.
    
    If we aren't prepared to push this possibility for the next few
    years, we will find ourselves at a serious risk.
    
    -- Russ

    Well, me beeing a Unix person I have some mixed feelings about this
    announcement. Digital UNIX wasn't even mentioned. I'm only waiting for
    the next weeks edition of "Computer Woche" (the USA Today type of German
    computer paper). I can see it coming:
    
    	Digitals new OS strategy OpenVMS/WindowsNT - Is UNIX dead?
    
    Just my 2 Pfennige. Cheers from Munich
    
    	Martin

    
    
    
    on the good side: VMS 'll get a GUI ... right? :-)
    on the bad side:  i see this as possibly an easier way for
                      customers to ditch an expensive OS

  Gives a whole new meaning of VMS being "Window" dressing :-)

  so, should I start looking for a VGA for the Turbo Laser ?

Re: .4

>>    What concerns me is that if we are diligent, we will make opportunities
>>    for sales of OpenVMS to WNT shops, but if we are lax, we give Microsoft
>>    the tools necessary to eat our customer base.

But in either case we sell Alphas, right?  Isn't that the point?

Re: .5 and similar

Just an observation, not a direct commentary on your note: people in the
various OS camps frequently seem paranoid.  Don't mention Unix ==> Unix is
dead; don't mention VMS ==> VMS is dead.  Maybe this is less true for NT, I
don't know.  With a three operating system strategy, you can't mention all
three operating systems in every announcement.

						Brian

     My cut at this is that it's an attempt to preserve, not expand,
     the huge but wanning VMS installed base. I would suspect that many
     of our customers have investments in (VMS) applications that make it
     cost prohibitive to migrate to a new (NT) platform in the near
     term but not necessarily in the future. The popularity of Windows NT
     is exploding and (VMS) customers are taking it seriously.

     This strategy may provide VMS customers with the security to
     maintain their current investments (VAX-ALPHA/VMS) and provide
     the flexibility to migrate/integrate 3rd party (Windows based) 
     applications in the future. Thus, no reason to ditch their
     current (hardware) investments.

     The annoucement suggests that VMS is targeted for the mission
     cirtical - backroom computing environment (no real surprise)
     to take advantage of clustering and such with the client
     piece as Windows NT driven.

     That's my interpretation - FWIW
    
     Paul


     

|	:
|	:
|
|   	to allow seamless interoperability of OpenVMS with Windows 
|   NT.
|
|	:
|	:
|
|   			Over the next year, Digital will add new 
|   functionality to OpenVMS in the areas of 64-bit addressing

	Presumably this means we're going to see 64-bit NT in a not
	too distant future. Hope Intel will not steal (too much of)
	the show.

	Also interesting to note that Microsoft recently announced
	a convergence of Windows and Windows NT (towards the latter).
	I guess this means we're going to see EDT, PATCH and BACKUP
	on millions of desktops ;-)

	>Per

    I just finshed reading a blurb in the EE Times magazine and it
    basically said that we will be porting the Win32, MFC and OLE
    subsystems to OpenVMS.  In return, we will provide clustering
    technology to Microsoft.
    
    From my perspective, this gives VMS a new lease on life by allowing
    NT and Windows95 applications to run it without having to do hardware
    emulation.  However, this changes the pricing game.  I figure that the
    price for OpenVMS will have to come down.  Intrinsically, I don't see
    why OpenVMS has to be expensive.  The development work has already been
    paid back.   The costs that remain are customer support and
    maintenance.
    
    I wonder if there are some more skunk work mechanations in progress.
    .start speculation
    Suppose that the "OpenVMS on Pentium" April Fools gag was really
    true.  (That would explain why it was yanked so quickly).  Then,
    we would be in position to offer an alternative to WindowsNT.  That
    is, Win32 on OpenVMS on Pentium.  Now that would be neat.
    .end speculation
    
    Well anyway, long term, we are in for an interesting ride with
    the new interfaces available on OpenVMS.
    
    -David

>    Suppose that the "OpenVMS on Pentium" April Fools gag was really
>    true.  (That would explain why it was yanked so quickly).

One way to find out: post an "OpenVMS on MIPS" gag.

    Speculation seems to be about all there is on this right now, which is
    a shame because several of my customers are real interested. Hard info
    is seemingly non-existent. Nice to hear EE Times picked up on it
    though.

It appears not all the press is receiving the message clearly.  Here's the
initial customer response from DECUServe, from a customer who read
ComputerWorld's version of the story.  PC Week's version is reportedly similar.

         <<< EISNER::$2$DIA7:[NOTES$HIVOL]BUSINESS_PRACTICES.NOTE;1 >>>
                            -< Business Practices >-
================================================================================
Note 382.92      Greater Boston LUG - Why is Digital Killing VMS        92 of 95
EISNER::MAZZONI "Michael E Mazzoni"                  21 lines   8-MAY-1995 17:02
             -< 'Official' migration path from OpenVMS announced >-
--------------------------------------------------------------------------------
The May 8, 1995 issue of COMPUTERWORLD has a front-page article titled "NT,
OpenVMS to Blend".  The article claims that today, DEC is to lay out a
migration path designed to slowly pass the torch from OpenVMS to NT.  Wesley P.
Melling, VP of Digital's OpenVMS systems business group in quoted as saying
that the OpenVMS/NT/Unix strategy isn't working, and that ISVs were flocking to
NT and leaving OpenVMS in the dust. DEC is teaming up with Microsoft, and in
the future NT applications will run on OpenVMS (but not vice versa). 

I sincerely hope that this article is an out-and-out misunderstanding of the
facts on the part of COMPUTERWORLD.  For if it's not, we have the DEC VP in
chargte of VMS telling the world that VMS is dead, and NT is your future. 
O.k., so us cynics can say that we knew for a long time that VMS was dead - all
you had to do was read the handwriting on the wall with respect to the
abandonment of VMS by ISVs and the selling off of DEC software applications. 
Still, I got the impression that DEC had seen the error of its ways and was
trying to shore up VMS and reverse the trend.

Now if this article is true, we've got yet another case of DEC lying to its
customers, and lying big time.  How can DEC expect customer loyalty when it
keeps misleading them and lying to them?  If DEC says one thing then does
another, how can anyone trust them?  This really sucks.

    Is there a clear concise explanation anywhere which outlines exactly
    what is our strategy, what is being integrated and how, and how should
    sales be positioning this with our customers ?
    For many large accounts who still have considerable investment in VMS,
    but intend to migrate, is it better (more profitable for DEC and better
    strategy for the customer) to push in the direction of Unix or
    Windows/NT ?  
    Can we get some clear messages and a direction here ?  Help !

    Contact Ken Swanton - he's in charge of OpenVMS Marketng.  I believe
    there are presentations available that describe the strategy.
    
    					andrew

    What I heard at DECUS was different and generally positive.
    
    * We have customers who *love* VMS, and who *love* VMS on Alpha.
    * We have customers who are watching ISV's drop support for VMS.
    * We have customers who are feeling pulled away from VMS by the ISVs,
      not pushed away by Digital.
    
    Many customers where feeling forced to choose *when* they were going
    to migrate from VMS.  Now they feel they have an additional choice,
    the choice of staying with VMS indefinately.
    
    
    BTW, as long as ISVs continue to choose Unix and Windows (Win32),
    then the Unix customers are happy.  There may be a day (maybe not too
    too far away) when ISVs choose just Windows.  Then a similar solution
    will probably be appropriate for the Unix market as well.
    
    Finally, the irony is that the original vision of "DECwindows" is coming
    together nicely, and it only took about a decade.
    
    								-mr. bill

    What I heard was:
    
    	- The "presentation" API is not being ported to VMS
    	- Don Harbert said that they'd do for UNIX what seemed
    	  best
    
    My take on it was that the idea was for VMS systems to keep the
    data and NT systems still be on the desktop.  It was suggested to
    me that if we tried to make it possible to use VMS "instead of"
    NT at all levels that Microsoft would be unlikely to help us.
    
    I agree that the announcement was confusing and easily misinterpreted.
    
    					Steve (at DECUS)

Re .11:

>    I just finshed reading a blurb in the EE Times magazine and it
>    basically said that we will be porting the Win32, MFC and OLE
>    subsystems to OpenVMS.  In return, we will provide clustering
>    technology to Microsoft.

David, this tells me more than the whole 12 screen Live Wire press release.
				/AHM

    re: .8
    
    No, we don't necessarily sell Alphas!  Customers can also migrate to
    Pentiums (and beyond).  This becomes even more the case when the media
    begins to say how we're ready to dump OpenVMS (as cited earlier). 
    
    Many customers won't buy a "hot box" if they think the vendor is
    unstable.  If customers think we've given up on OpenVMS, many may flee
    to NT, but if they think we're closing shop and going home, they could
    easily go with multiple Intel platforms to back-end new client-server
    efforts.
    
    Customers going to NT isn't necessarily bad, but it certainly isn't
    necessarily good either.
    
    We MUST remain alert and make sure our customers understand a single
    coherent message regarding this announcement and the products which
    (presumably) will follow!
    
    -- Russ

    I, too, attended the Decus presentation and actually misunderstood
    what Wes Melling stated (his fault or mine?) I really thought
    that OpenVMS would be extended to support all NT applications.
    
    Others thought this too as was clear in the ensuing Q&A.
    It only then became clear that only NT applications that 
    make no use of presentation layer
    functionality will be able to run on the API which I gather will
    be an extension/upgrad of the Bristol kit.
    
    re roelof (at DECUS)
    
    Hey Steve, where are you :-)

>What level of compatibility / integration are we talking about?
>
>      1.  Will VMS be extended to run Windows NT / Windows 3.1 / DOS binaries
>          off-the-shelf?
>
>      2.  Will VMS be extended to be compatible with Windows NT at the source
>          code API level, including the graphical user interface, so that you
>          can design a program for NT, get it working there, and port it more
>          or less by recompiling?
>
>      3.  Or are we building APIs to ensure that back-end server applications
>          using new APIs can be ported between VMS and NT?
>
    In the session Wes Melling excluded 1 ("simple recompile") and in the
    Q&A excluded 2.
    
    A large part of the presentation was devoted to promoting a three
    tier model (presentation, application services, data services)
    and that Digital and MS were supporting this 3 tier model so that
    - in time - any well designed (NT) application would not directly
    call presentation layer functionality but would do that through an
    API layer. This is - in retrospect - where the confusion comes
    from I think. The - in time - bit didn't come through clearly in
    the presentation...
    
    re roelof

> It only then became clear that only NT applications that 
> make no use of presentation layer functionality will be
> able to run on the API...

  Which two Win/NT applications are these (that don't require
  the presentation layer)?

                                   Atlant

There seems to be a conflict between the two statements
in this string that (a) we're not porting the presentation
layer (b) we are porting MFC.

I know MFC isn't a presentation layer, but the majority of
its code seems to be about building GUI software (the
rest is fairly obvious stuff, like CString, CFile, ...).

I could imagine a situation in which (a) and (b) could
both be true -- if, for example, we invented a remoteable
implementation of GDI32 so that an MFC app running on
the VMS system would actually be putting bits out 
on an NT desktop.  But I don't think that is intended.

    The idea as I understood it was that applications coded against a
    "middleware" provider library would be able to use OpenVMS for the
    data repository function.  Wes named a handful of middleware products
    which he implied were in wide use, but I am not familiar with that
    segment of the market and can't comment.
    
    What this clearly does NOT mean is that you can take shrinkwrapped
    NT applications and run them on VMS.
    
    				Steve

    re .24:
    
  �Which two Win/NT applications are these (that don't require
  �the presentation layer)?
    
    We have written quite some stuff for Win32 which doesn't have any
    presentationm layer at all. (But it already runs on VMS anyway...).
    

    
    	Comment from a sales guy -
    
    	If you cannot take shrinkwrapped Windows NT applications and run
    them under OpenVMS, we are wasting both our time and money in the
    marketplace.
    
    	Anything else will be perceived as "snake oil" and customers will
    continue to desert VMS in droves.
    
    	Maybe Wes ought to talk to *real* users in both MIS departments
    and VARs before he performs marketing hari-kari :-)
    
    			the Greyhawk

    re -.1
    
    I agree. Don't do things in a half-assed manner. Got us into enough
    trouble already.

    My initial reaction to this announcement is this is another stake in
    VMS's heart.  Why would any customer wait 12-18 months for this effort
    to be completed when they can spend 5K on an NT workstation and start
    porting right now?  Another botch with VMS.  And for those who think
    you can't make money with proprietary OS's, please explain why IBM
    is raking in tons of money with the AS/400?  Quite a contrast on how
    two companies with 'legacy' products are handling the doom-and-gloom
    'mainframes are dead' mentality.  One company is basically walking
    away from customers, and one plows forward very profitably.
    
    What a waste of a great product!
    
    Brad.

    Before anybody seriously considers dropping VMS and moving to Windows
    NT, I'd recommend they try actually using Windows NT as a production
    operating system for a while. I've tried it. My experience says that NT
    backups, for example, are not of the quality that VMS and Unix users
    are accustomed to (though there are probably 3rd party solutions
    around). Windows NT is not VMS, and vice versa. 
    
    But - we (and our customers) need some _facts_ on this announcement,
    not press-style speculation. Please...
    
    regards
    john

The way I heard it Melling was an industry analyst
(with Aberdeen Group??) when he came up with this idea,
then Digital hired him to implement it.  I talked to
the VP of IT for a large bank recently.  He had just
talked to Melling and loves the idea.  I gather
that this is good stuff for large IT departments that
use things like RTR, banks for example.  Imagine the
political heat that VMS defenders (ACMS and Rdb
customers for example) are taking these days in large 
corporations.  This announcement gives our loyal 
customers a reasonable plan that they can defend and
follow.  I hope that it (like the Oracle announcement)
will help us sell lots of high-margin servers.
My impression is that the potential buyers of such 
systems are the target of this announcement.

    ISVs are, in general, writing apps for NT and UNIX and dropping 
    support for VMS. I think this is an attempt to bridge the gap
    between ISVs and VMS. Digital will rely on some quasi-pseudo
    NT port to run NT apps on VMS. Maybe this is an adequate stop-gap
    message to appease the VMS installed base. I hear the annoucement
    went over very well at DECUS. Sure would be nice to get more
    insight though.


    

 This is what I understand of this, after attending 2 days
of training a month ago:

- there are (in the very near future) 2 *major* operating
paltforms, competing for the ISV's: Unix and WNT.

- Digital has 3; Unix is "safe" but OpenVMS needs some plan
to align to this 2-OS-world.

- WNT (currently and in Digital's view) will not be robust
enough for several years.

- OpenVMS IS robust enough for mission critical, AND has many
similarities to WNT

conclusion:

- continue to enhance OpenVMS in the areas of availability, 
integrity, performance, security and more

- support all necessary WIN32 APIs in OpenVMS  (there are 
apparently 3000 APIs, not all of them will be available)

result:

- ISV's can now develop application for WNT and DEPLOY THEM ON
OpenVMS!

- OpenVMS customer will continue to benefit from OpenVMS AND see
enhancements (like the new file system, enhanced clustering etc.)

I fail to see how this could be a threat to Unix. It's a "mid-life"
(or maybe late-life ;-) kicker for OpenVMS.

And for more info, I'd also suggest to contact Mike Cuccia (in Ken 
Swanton's group.)

    
    	My customers got the same message... heard the same rumor. VMS is
    dead..... or will be dead. I told them that rumors can be started by
    anybody and that our committment to VMS is "stronger" (I hope I'm
    right .... at least that's what I understand). We don't need to be 
    giving out confusing statements right now about VMS, I sure hope that
    the source of this rumor wasn't generated by something we did/said.
    
    Rick
    
    
    

Damage report Mr. Scott!


  DEC PLANS OPENVMS TRANSITION TO NT
  May 7, 1995, PC Week
  
  * Digital will start a gradual phaseout of its venerable OpenVMS 
    operating system by coupling it with Microsoft Corp.'s Windows NT to 
    support large-scale client/server applications that use key features 
    from both environments.
  * DEC also plans to port its All-In-1 application suite and the 
    OpenVMS Systems Management tools to NT. DEC officials say the moves 
    are part of a short-term strategy to keep users from abandoning 
    OpenVMS en masse as ISVs their applications directly to NT and Unix 
    platforms. Some OpenVMS users "were succumbing to the pressure and 
    going to Unix," Melling said.
  * DEC's long-term strategy, he said, is for NT to gradually replace 
    OpenVMS on its Alpha workstations and servers.  


  WINDOWS NT, OPENVMS TO BLEND
  May 6, 1995, Computerworld
  
  * Conceding that its three-headed operating system strategy left users 
    with a splitting headache, Digital lays out a migration path 
    designed from OpenVMS to Microsoft Corp.'s Windows NT
  * Digital has recognized that offering competing operating systems  
    OpenVMS, NT and Unix  and asking users to pick one just did not work 
    well, said Wesley P. Melling, VP of Digital's OpenVMS systems 
    business group Digital is teaming with Microsoft to ensure that for 
    the first time, an application written for NT will also run on 
    OpenVMS.
 

  DIGITAL TO ANNOUNCE PLANS FOR INTEGRATING OPENVMS WITH NT - USERS
  WILL GET ACCESS TO NEW NT APPS
  May 7, 1995, InfoWorld
  
  * Digital announces this week at DECUS plans to
    link its OpenVMS operating system for Alpha and VAX systems to 
    Windows NT
  * Aims to provide its OpenVMS and Unix customers with access to
    the new generation of Windows NT applications
  * Ultimately IS managers will be able to manage from a Windows desktop 
    the information architecture of an enterprise that runs on OpenVMS 
    and Unix

    
    	Co-worker just had a very interesting suggestion...
    
    	Let Bill Gates do the announcing, at least the press will pay
    "proper" attention to "exactly" what *he* says.
    
    	And out of the mouths of....
    
    
    		the Greyhawk

>Melling was an industry analyst (with Aberdeen Group??) 
    
    Wes worked for many years for the Gartner Group.

    First, Wes Melling came from The Gartner Group, which is well regarded
    for its market analysis of the computer industry.  I might add that a
    lot of customers follow what Gartner has to say.
    
    Second, anyone who under estimates the power of
    	1. Microsoft, and
    	2. The strong desire by users for standardization in software, and
    	3. That one OS, Windows, dominates the desktop and has not hurt
           sales of PCs one bit, but in effect has helped grow the business
    is a fool.
    
    Perhaps Melling, who analyzed the market for years from a reasonably
    independent position at Gartner, and others within Digital, have come
    to the conclusion that if we embrace Microsoft we will prosper
    enormously like so many of those companies that jumped on the Microsoft
    bandwagon years ago.
    
    Yes, there are risks.  But let me respond to a previous note about IBM
    making all that money on AS/400s.  On the front page of the WSJ a few
    days ago was a very negative article about IBM and there "current"
    profits vs. "what lies down the road".  They continue to fight
    MicroSoft with OS/2 and pour money into this rat hole that is going to
    lose.  Mark my word, IBM will pull the plug on OS/2.  W/NT is going to
    dominate and kick OS/2's butt.  PowerPC is not taking the market by
    storm.  Alpha runs circles around it and the Intel P6 and P7 will
    certainly out perform it.  AS/400 sales are on the decline, mainframe
    sales were up because of some delayed (pent-up-demand) sales, and AIX get
    negative revues as an enterprise-wide server platform.  They also got
    bad revues on the performance of the CMOS chips used in the mainframes. 
    The bottom line is that the good times (high profit) will not last. 
    They are not positioned that well for the future.
    
    Perhaps by embracing the future (W/NT) and quite possibly the
    inevitable (single OS dominates server market), Digital will in the
    long run, come out on top.  I said long run.
    
    If Digital can set the stage for its OpenVMS base to move to W/NT, and
    the absolute best server platform on which to run W/NT is Alpha, and
    Digital is way ahead of the competition in embracing W/NT and is viewed
    as a company that saw the future, just like it did with 64-bit chips,
    and embraced it, not fight it or run from it, then we may come out
    smelling like a rose while our current OpenVMS customers stand next to
    us.
    
    "The only thing we have to fear is fear itself"

As usual, the Greyhawk has put his finger on both the problem and the solution:

RE: .28

>    	If you cannot take shrinkwrapped Windows NT applications and run
>    them under OpenVMS, we are wasting both our time and money in the
>    marketplace.
    
And that is exactly the intent of this announcement: to get shrink-wrapped NT
applications which can run under OpenVMS (both VAX and Alpha).

Does this mean that we will take {Intel|MIPS|NEC|PowerPC} NT binaries and run
them under OpenVMS?  No.

Does this mean that we will make NT run under OpenVMS?  No.

Does this mean that we are planning to kill OpenVMS and switch over to NT?  No.


What it does mean is that we are going to make it easy for ISVs to develop
their applications under NT, and then choose to deploy them on OpenVMS.  We 
are doing this by taking Win32 API and the Microsoft Foundation Class (MFC)
library and putting them on OpenVMS, so that applications which are developed
using (for example) Visual C or Visual C++ using the Win32 API and the MFC 
can be re-compiled and linked under OpenVMS, and then shrink-wrapped as the
OpenVMS version of that product.

ISVs are like anyone else: they want to the most $$$ for the least amount of 
work.  By making it easy to take the source code of their applications and
deploy it on OpenVMS, they get an expanded market for very little investment.
And this market is one that is attractive to many ISVs: the high-end server
market, which demands absolute availability, reliability, security, fail-over,
scalability, and all of the other things which OpenVMS clusters has that 
absolutely *NO ONE ELSE* in the industry has right now.

Yes, NT clusters are in development, but the engineers who are working on it
would be the first to admit that they won't match the functionality of OpenVMS 
clusters for years to come.  And by the time they do match today's cluster
functionality, OpenVMS will have developed some extra things which our high-end
customers will be demanding.

I just got back from DECUS, and I can tell you that every customer and ISV I
talked to was totally positive on this.

-- Ken Moreau

    re .39
    "PowerPC is not taking the market by storm.  Alpha runs circles
    around it..."
    
    reality check, please: although only on the market for *half* the
    time Alpha has been on the market, PowerPC systems have outsold Alpha
    systems more than *12 to 1*. 

re                      <<< Note 3860.41 by PCBUOA::KRATZ >>>

>    re .39
>    "PowerPC is not taking the market by storm.  Alpha runs circles
>    around it..."
>    
>    reality check, please: although only on the market for *half* the
>    time Alpha has been on the market, PowerPC systems have outsold Alpha
>    systems more than *12 to 1*. 

Sorry, but you can't link the 2 statements from .39 together, they are in 
different contexts.  PowerPC *HAS NOT* taken THE market by storm no matter 
how many have been reported as sold/made/delivered.  And in a totaly 
different context, Alpha indeed runs rings around PowerPC, just look at 
the performance numbers.

Jilly

    
    re .last
    >PowerPC *HAS NOT* taken THE market by storm no matter how many have
    >been reported as sold/made/delivered.
    
    I don't agree with that, sorry.  Apple can't make PowerMacs fast enough
    to meet demand.  If PowerPC hasn't taken the market by storm, then
    what adjective(s) would you use to describe Alpha sales (which finally
    reached 100k after 2 1/2 years, vs. PowerPC's million systems in the
    first 10 months on the market)?  Kratz
    

> If only we had shown a little more foresight when Apple came calling,

What should we have done differently?  I believe we did try to get Apple
to use Alpha, but were less convincing than IBM, partly because Alpha
existed only on paper, while IBM could point to the multi-chip RS6000
systems and say "PowerPC will be just like that, only cheaper."  I don't
know at what point Motorola got involved -- but that surely made a difference
to Apple -- they already used 68xxx's, and wouldn't like to deal solely
with a competitor.

    
    So what is Digital's internal systems strategy...?
    
    Well, according to the DECUS 'The Digital Systems Infrastructure
    Project' seminar by George Champine, our path is DCE and CORBA on
    Digital UNIX, with the possiblity of moving to W/NT down the road.
    
    Of course the question was asked, what about OpenVMS, to which my
    recollection of the response was 'well this is just one companys
    implementation, others may/will choose OpenVMS."  Note: these seminars
    were taped, perhaps someone might verify?  (AD008)
    
    While attendance at this hour seminar was sparse (20 people maybe),
    it did leave a few customers puzzled.
    
    Mike M.

RE: .44

> But Apple's
> Power Macintoshes are taking their market by storm ... and that niche is big
> enough to make Apple one of the top two or three PC vendors, despite all the
> Intel clone competition.

How do you reconcile "taking the market by storm" with a report in the WSJ
a number of weeks ago that Apple's world-wide market share reportedly
*slipped* slightly during the past year (in spite of the arguably "pent up"
demand for Power Macs among existing Mac users.)  There is no indication
at all that I'm aware of that most people that buy Windows PCs today would
even consider switching to a Mac.  (I'm probably not a good statistical
sample, but 4-5 years ago I might have considered a Mac - I wouldn't today -
it simply has too small an installed base with all the ramifications that
has around software.)

> Unfortunately, I think we learned the wrong lesson, e.g.
>
>    "VAX/VMS is the wrong religion.  If we convert to the church of Intel/
>     Microsoft, our reward will be to have Alpha inherit Intel's 90% of
>     the PC market, regardless of issues like compatibility or pricing."

    1) Your note sounds an awful lot like "religion" to me (arguing
       that the 10% and shrinking market share is better than the 90%
       and increasing market share).

    2) RE: "regardless of issues like compatibility or pricing".  Anyone
       who thinks compatibility and price are irrelevant are going
       to lose REGARDLESS of what market they're trying to break into.
       (And yes, I continue to think our pricing strategy for Alpha
       is broken.  You don't get market share for charging a higher
       premium.  You're lucky to get it if you charge the same price
       and have a better product.  Significant inroads generall requires
       both a price advantage and a superior product.)

Anyway, the past is the past.  Given the situation NOW, what do you suggest?
That we become the 2nd-string player in the Mac market?  (Personally,
if we're going to re-visit history, I think we should have had small memory
accesses in the architecture, and pulled all the stops to get Intel/HP
to adopt it rather than persuing their own path.  BUT, that's also
the past.  The question we have to ask is GIVEN OUR CURRENT POSITION,
where do we go from here to be successful?  The past my be interesting,
and analyzing it for learning is useful - but you can't change it.)

	There's alot of crying over spilled milk here. Mac isn't on
	Alpha. Get over it. It's over, done, kaput. Lets move on to
	what we have. Alpha IS starting to become synonymous with
	performance when it comes to Windows NT. We need more of
	a thrust with less expensive (and faster) systems and better
	marketing and more incentives to port to Windows NT Alpha. Sitting
	around crying in our beer about Mac doesn't pay the bills. Right
	now, ramping up in a big way our Windows NT effort could pay the
	bills in the not too distant future.

	(This is all NOT at the exclusion of other operating systems BTW)

							mike

I just returned from DECUS and was asked to take a look at his string of 
notes.

We have been briefing the op customers (selected by the ABU & SBU) with 
excellent results.  The strategy is being received very well at these 
briefings.  The wider press returns (beyond CW & PCWeek) have been quite 
good with good analyst support, as well.

Attached to this reply you will find a copy of the OpenVMS/Windows NT 
Executive Summary which has been distributed to customers..  This should 
help to clear up some of the questions and concerns you have expressed 
with your replies to this note.  

On Monday I will post answers to some of the questions in this note and 
provide a pointer to more information on the strategy.

Regards,

Mike







	 
		Integrating OpenVMS with Windows NT

Overview: Enterprise client/server computing with the world's
          largest applications' library


Integrating OpenVMS with Windows NT

Millions of Digital customers rely on the OpenVMS operating system for 
its rich functionality, price-performance leadership, scalability, high 
availability and very high reliability. Many customers, however, face 
pressure to move to industry-standard operating systems, and they require 
the long-term availability of third-party applications. 

First and foremost, Digital Equipment Corporation is committed to 
addressing the concerns of its OpenVMS customers and increasing the value 
of their new and existing investments. 

Digital is teaming with Microsoft Corporation to develop an integrated 
systems environment that creates synergy between the unequaled, 
bet-your-business capabilities of OpenVMS and the state-of-the-art 
functionality and growing applications' library of Windows NT.

This new direction for OpenVMS allows customers and business partners 
to capitalize on the enterprise strengths of OpenVMS while dramatically 
increasing the number of available applications. 

The OpenVMS and Windows NT operating systems have a natural affinity for 
each other. In many respects, they share a common heritage expressed 
through highly similar system design and distributed system 
capabilities.

Digital believes that combining the best features of Windows NT with the 
best features of OpenVMS will establish an environment that provides the 
richest set of applications, tools, and overall functionality. 
 
The Future Is Now
 
Many customers and software developers have already discovered the 
compatibility between OpenVMS and Windows NT. They are writing 
mission-critical applications today that draw upon the strengths of each 
operating system.
 
Digital's significant engineering investment will make the integration of 
OpenVMS and Windows NT even easier and more transparent, creating a seamless 
environment in which

	- OpenVMS functions as the "unlimited" high-end server 
	  for Windows NT with new capabilities in 64-bit 
	  addressing, log-structured file system and OpenVMS 
	  Cluster enhancements.
 
  	- Enterprise applications, functionality and data are kept 
	  secure on scalable, high-availability servers, but are 
	  easily integrated with Windows applications.
 
  	- Application code developed on Windows NT is easily 
	  deployed on OpenVMS and Windows NT servers with 
	  Digital and partner tools and utilities.
 
  	- OpenVMS users have seamless access to Windows 
	  presentation services and to Windows NT application 
	  logic.

	- Applications and application components are developed 
	  using industry-standard middleware.

  	- Existing applications and new applications reside on the 
	  same server, and are easily integrated into the new 
	  environment.

 
  
Developers produce code that can be deployed across heterogeneous servers, 
including non-Digital systems.

Ultimately, the entire information architecture of the enterprise can 
appear as a seamless extension of the Windows desktop.

Starting in July, Digital will offer training courses to help application 
developers and architects begin writing for the new environment. They will 
immediately be able to apply what they learn.
 
In addition, Digital will offer a full set of OpenVMS and Windows NT 
implementation services to assist customers.

Three-tier client/server computing
 
The full potential of the OpenVMS and Windows NT synergy is realized 
through a three-tier client/server model of computing. The three software 
tiers are characterized as follows:
 
  
	- The third tier is where data, mission-critical applications 
	  and services are managed. 

  
	- The application or functional logic tier governs the rules 
	  by which the business applications operate. 
 
  
	- The presentation tier, or desktop, is dominated by 
	  personal computers.

Each software tier is architecturally separate within the overall 
application. The three tiers are integrated using market-accepted 
middleware, such as ObjectBroker, COM/OLE, and DECmessageQ.

For customers who have standardized on Windows at the desktop, the 
suggested approach is to write presentation logic for the desktop and 
to write application logic in portable code for Windows NT and 
OpenVMS servers. 

OpenVMS systems should serve as the overall data repository for the 
Windows and Windows NT systems.   OpenVMS systems also will continue 
to house existing applications.
 
As a result, the new desktop and LAN-based applications can access the 
same data as existing OpenVMS applications. And business-critical enterprise 
data is secure on the third tier.
 
This three-tiered approach provides the ultimate flexibility. Applications 
can be deployed across multiple vendor environments and operating systems. 
Customers maintain the flexibility to move easily from one vendor 
to another, should that become necessary. 

Another major benefit is the availability of the largest application base 
in the world, Windows, within an environment that uses the latest technology. 
 
Most importantly, business-critical enterprise data applications are 
available on reliable OpenVMS servers operating in 24 x 365 mode.

 
Immediate Benefits
 
High availability: OpenVMS servers and clusters provide around-the-clock 
24 x 365 services and data access for Windows based application and file 
servers.
 
Data integrity: Data integrity requirements are uniquely supported by the 
proven data-integrity features of OpenVMS. The new log-structured file 
system, which continuously backs up on-line data, further extends 
OpenVMS data-integrity features.
 
System and network management: Digital's POLYCENTER, NetView products 
manage large networks of systems, including OpenVMS, Digital UNIX, and 
Windows NT.
 
High performance: Industry-leading performance is provided through the 
Alpha 64-bit architecture and the new log-structured file system.
 
UNIX interoperability: The OpenVMS and Windows NT environment interoperates 
with UNIX through a wide array of industry standards, such as 
POSIX, XPG and CORBA.
 
Flexibility: Applications can be developed once, then deployed across 
multiple platforms and re-deployed as requirements change or a change 
in vendor is desired.
 
Cost-effectiveness: Low-cost, industry-standard PC tools can develop 
applications for the OpenVMS and Windows NT environment.

Business advantages: OpenVMS users will have access to the fast-growing 
Windows NT server application portfolio, running on Windows NT and OpenVMS. 

In Summary
 
The new OpenVMS direction builds on the past successes and brings forward 
a new integrated approach to enterprise-scale computing. It promises a 
unified environment for software development and deployment that leverages 
the mission-critical attributes of OpenVMS and the volume-market attributes 
of Windows NT.

OpenVMS is positioned for the future without sacrificing existing investments 
and advantages. The new direction ensures the availability of leading-edge 
server applications through business partners by reducing their software 
development costs and dramatically increasing their markets. 

Current OpenVMS customers will immediately see the value of their 
investments grow and will continue to gain the advantages of mission-critical 
OpenVMS computing. Both current and new customers will reap the benefits 
of an integrated enterprise environment that is application-rich, highly 
cost-effective and unquestionably dependable and secure.

There is every reason to begin harnessing the combined 
power of OpenVMS and Windows NT today.


Digital believes the information in this publication is accurate as of its
publication date; such information is subject to change without notice.
Digital is not responsible for any inadvertent errors. Digital conducts its business 
in a manner that conserves the environment and protects the safety and health of 
its employees, customers, and the community.

Digital, the DIGITAL logo, DECmessageQ, ObjectBroker, OpenVMS, and 
POLYCENTER are trademarks of Digital Equipment Corporation.

NetView is a registered trademark of International Business Machines 
Corporation.

UNIX is a registered trademark in the United States and other countries, licensed 
exclusively through X/Open Company, Ltd.
Windows and Windows NT are trademarks of Microsoft Corporation.

Copyright 1995 Digital Equipment Corporation. All Rights Reserved.

I too was at DECUS and heard Wes talk 3 times. The following is my 
cut at re-stating what I heard... don't take it as Digital's official 
statements around this announcement.

One thing Wes made VERY clear in his presentations was the marketshare of 
the various operating systems. His prediction is that by the year 2000, the 
OVERALL (ie: across all vendors) split of SERVER OS's will be:

  40% Unix
  40% WNT
  20% All the rest

Of that 20%, approx 5-6% is OpenVMS, the rest is MVS, Tandem/Guardian, 
OS400, etc.

Doesn't look good for OpenVMS, right ??  WRONG !!!  Assuming this is a $100B
market (VERY conservative number), OpenVMS (which is 100% Digital revenue!) 
would be $5 - $6B..... NO large company sneezes at or walks away from that
type of revenue.

We have to fight for our piece of the Unix pie along with dozens of other
Unix vendors, the same goes for WNT. If we could get OUR SHARE of the Unix
and WNT market up to where the OpenVMS market is, we'd no longer be a $14B
company, we'd be a $17-$18B company. OpenVMS, while not a "market leader"
when compared to the overall Unix or WNT market, is a market leader compared
with any one specific implementation of any other OS. 

Application vendors (ISVs) have been walking away from OpenVMS.... walking 
towards Unix for a long time. However, a large number have just recently 
starting walking (running ?) towards WNT. They're not stupid, they see that 
by offering their applications on both Unix and WNT, they cover 80% of the 
server marketplace. However, one major concern for any server application is 
the reliability of the underlying hardware/OS architecture. We're making 
great strides with our Unix clusters, and we're working on WNT clusters, but 
neither is up to what we offer for 24x365 computing on OpenVMS (yet). And, 
even once we do have 24x365 clusters for WNT and Unix, there will not be the
years of track record we have with OpenVMS.. it'll be "bleeding edge"
technology. Some large production shops are not comfortable with unproven 
environments.

A major thrust of this announcement is that we keep our 5-6% market share 
for OpenVMS by providing a method for ISVs to easily offer their NT 
applications on OpenVMS without being at all concerned with "proprietary" 
system calls, interfaces, etc. They design the SERVER application once, and 
can offer it to customers on a non-production-proven WNT server (until Digital 
brings clusters to WNT), or on the benchmark OS for 24x365 operations, 
OpenVMS. The intent is NOT to provide support for shrink-wrap "desktop" 
applications like Excel, Word, etc. We'll get our piece of that thru the WNT 
market on both Alpha and Intel platforms (our piece of the 40% above). The
intent is NOT to compete with Unix applications like Data Warehousing -
we'll get our piece of that thru the Unix market (our piece of the 40%
above). The intent is to offer a highly reliable server platform for the
gaggle of WNT client/server applications which are currently under
development on an OS which has proven itself to be the most reliable server
OS in the marketplace. 

Will this attract thousands of new customers and users?? Unknown. But take SAP
for example. They have decommitted from OpenVMS. They run on Unix, so we get
a piece of the Unix/SAP market, but that's all we get. However, SAP has
committed to a WNT client/server port. Assuming all this OpenVMS/WNT stuff
happens and works, we would be able to offer SAP on OpenVMS again using the
WNT Server code from SAP, while the front-end runs on WNT clients. This is
not intended to detract from our Unix sales of SAP, but to provide an
additional revenue source from customers who want the NT C/S version of SAP
with a highly available and reliable server environment. 

Wes and others are working to correct the mis-interpretations by some of the 
trade rags and provide additional internal info. However, the feedback from 
"most" customers at DECUS was positive. 

Just my interpretations.

Arlan

I'd like to make some comments on the comments in the preceding entries.

My comments below indicated by >>>


re: .4

    							.... we had better
    stay sharp and not let our new opportunity turn into our new problem.


>>> Absolutely right, and everyone involved in this understands that.


re: .5


				....     I can see it coming:
    
    	Digitals new OS strategy OpenVMS/WindowsNT - Is UNIX dead?
     
>>> NO ONE put this spin on the strategy announcement. We worked hard to 
>>> insure that it did not get interpreted that way. Digital's UNIX
>>> people were involved in the developing the right spin to specifically 
>>> avoid this problem. It worked very well (probably because the truth is
>>> we were NOT attacking UNIX).



re: .10

|
|   	to allow seamless interoperability of OpenVMS with Windows 
|   NT.
|
|	:
|	:
|
|   			Over the next year, Digital will add new 
|   functionality to OpenVMS in the areas of 64-bit addressing

	Presumably this means we're going to see 64-bit NT in a not
	too distant future. Hope Intel will not steal (too much of)
	the show.

>>> No. This announcement has nothing to do with 64-bit NT.
>>> 
>>> The reference to "seamless interoperability" and OpenVMS "64-bit 
>>> addressing" are two different points (probably why they were stated 4 
>>> paragraphs apart). As you'll see from reading the Executive Summary.
>>> there are 2 major thrusts to the strategy -
>>> 	
>>>	1) Integrate the OpenVMS environment with Windows NT environment, 
>>>	   so that customers can develop & deploy applications, manage, etc. 
>>>	   as though they are one environment. That's the goal we are reaching 
>>>	   for in the integration area. 
>>>
>>>	2) Drive ahead with OpenVMS as the unlimited high end to the 
>>>	   windows environment, by expanding it's advantages in enterprise
>>>	   scale dependability, security, data integrity and scale. 64-bit 
>>>	   OpenVMS, log-structured file system, wide area clusters, etc are
>>>	   examples of new OpenVMS features that fit into this area. These 
>>>	   are the reasons that people will continue to buy OpenVMS in the 
>>>	   hybrid OpenVMS with Windows environment.
>>>
>>> We are investing in both of these areas. 


re: .11

 .start speculation
    Suppose that the "OpenVMS on Pentium" April Fools gag was really
    true.  (That would explain why it was yanked so quickly).  Then,
    we would be in position to offer an alternative to WindowsNT.  That
    is, Win32 on OpenVMS on Pentium.  Now that would be neat.
    .end speculation
    

>>> Definitely NOT under consideration. Our strategy leverages NT affinity
>>> for an integrated environment of OpenVMS WITH Windows NT. 


re: .14


It appears not all the press is receiving the message clearly.  Here's the
initial customer response from DECUServe, from a customer who read
ComputerWorld's version of the story.  PC Week's version is reportedly 
similar.


>>> CW and PC Week did indeed get the story wrong.
>>>
>>> However, EE Times, Communications, Week, INFORMATIONWEEK, INFO WORLD,
>>> NetworkWorld, and DN&R got it right and largely positive. It's hard to
>>> fathom the mind of a journalist, so I won't speculate on why most of 
>>> the press got it right and CW and PC Week got it wrong. We will of
>>> course be pursing them to turn them around as we roll the strategy out.


re: .21 - .25
    
    Others thought this too as was clear in the ensuing Q&A.
    It only then became clear that only NT applications that 
    make no use of presentation layer
    functionality will be able to run on the API which I gather will
    be an extension/upgrade of the Bristol kit.

>>> I'll post a separate reply on the question of Win32 API's and what 
>>> applications are targeted. 


re: .28

    	If you cannot take shrink wrapped Windows NT applications and run
    them under OpenVMS, we are wasting both our time and money in the
    marketplace.

>>> Not true in the least. OpenVMS is targeted at the enterprise level 
>>> applications and customers who use more than "shrink wrap" functionality.
>>> We are targeting ISVs who are selling and implementing complex 
>>> solutions. When they do buy out of the box apps they modify them.
>>> The strategy is aimed at making sure that the complex applications and 
>>> enabelers are available for OpenVMS, not at Word or Excel. That's not
>>> OpenVMS' strength.
     
    	Anything else will be perceived as "snake oil" and customers will
    continue to desert VMS in droves.

>>> Not true. We have spoken to over 200 accounts (ABU, ISV and 
>>> Distributors).  We specifically targeted the OpenVMS base. The response 
>>> has been overwhelmingly positive from all. They see it as a commitment 
>>> to OpenVMS and addresses a need they have. End users see it as a way
>>> to ensure that the applications they worry about are available. The ISVs 
>>> see this as way to reduce their cost and expand their market opportunity.

    	Maybe Wes ought to talk to *real* users in both MIS departments
    and VARs before he performs marketing hari-kari :-)

>>> Cheap shot and wrong. Wes has spent the past 8 years talking to and
>>> advising top IS people in large accounts and ISVs worldwide. They paid
>>> big money to get his advice and acted on it. He ran the IS operation for 
>>> a billion dollar healthcare provider. His credentials in the area of
>>> knowing the "real" world stack up well against anyone's in the industry.


re: .34

>>> This is a good short hand description of the logic of the strategy.



  

    re: .53
    
> re: .4
    							.... we had better
>    stay sharp and not let our new opportunity turn into our new problem.

> >>> Absolutely right, and everyone involved in this understands that.

    I'm glad we agree on this point, but I believe you've missed the scope 
    of my comment.
    
    The "we" here is not Marketing, or Product Management, or BP et al --
    it's _ALL_ of Digital.
    
    The Field has to be armed to the teeth to handle this announcement
    properly.  You can lay odds that our competitors will do everything in
    their power to beat the "OpenVMS is dead" drum louder than ever now.  If
    the Field is not given adequate support to fight back and win business
    _OVER THE NEXT 3+ YEARS_, we'll find ourselves waving goodbye to our
    customer base as they migrate to WNT (not necessarily on Alpha) and to
    other platforms like HP-UX as the other vendors sing the "if-you-have-
    to-migrate-go-with-the-best" song.
    
    We can turn this announcement into major profit _IFF_ we _ALL_ stay in
    the fight and _IFF_ we continue to support sales efforts for the long
    haul.  If we think the job is done in 12 months and say, "Well, the
    Field can just handle it from here; we've published our statements and
    customers have no reason to abandon us," then we might well find
    ourselves in a hole the depth of which this company has _NEVER_ seen!
    
    Like I said, it isn't a bad move -- it can be a _REALLY_ big win.  But
    you can't win a war by declaring victory early and going home.  And 
    that's one mistake I've seen repeated in Digital with frightening 
    regularity over the years.
    
    This is a _VERY_ high stakes move.  We're betting our OpenVMS installed
    base against our future.  We had _all_ better be ready to play out this
    hand.
    
    -- Russ

Re 3860.5:

>	I can see it coming:
>     	Digitals new OS strategy OpenVMS/WindowsNT - Is UNIX dead?

Some feedback from a (large European) customer, a large UNIX user: "I got a
sneak preview on the new OS strategy: VMS and NT... pfew!"  The same
customer asks why, if it's a good idea to broaden the VMS market by making
it more accessible to and from WNT, the same doesn't apply to UNIX.

___Pete

    It would be VERY desireable to get some clear, concise, TECHNICAL
    information to the field people who are face-to-face with the customers
    on a daily basis. Like Digital Consulting. It is a continuing
    embarassment to be at the customer site as a supposed Digital expert
    and not have any more information about a major issue than the
    customer.
    
    It doesn't inspire confidence in Digital's ability to get its act
    together. Not does is suggest that Digital can support the New World
    Order we've just announced.
    
    \dave

    .58
    
    spot on.

    Like I said (.13), where's the beef, marketing-folks ?
    
    Till the concise technical summary of what will actually be delivered
    (not when, just what) arrives, confusion will continue inside and
    outside Digital. Conusion => lost opportunities => bad news.

    Re the last few,
    
    I agree, as one who has to try to convert this announcement from
    marketing hype to technical implementation detail, where's the
    background material ??.
    
    Steve.

    see vtx ir doc no. mr02oz for dataquests play on this ---pretty
    positive

Subject: I: IDC'S OPINION ON DIGITAL & MICROSOFT PLAN TO INTEGRATE OPENVMS &
	    W/NT

From:	MSBCS::RDVAX::MOORE "MARTHA MOORE DIGITAL LIBRARY NETWORK 226-2918 LJO2
12-May-1995 1721" 12-MAY-1995 17:22:26.08
To:	@OPSYS.DIS
CC:	
Subj:	fyi

Unix and Advanced Operating Environments
Filing Information:
	Date:	April, 1995
	IDC #:	????
	Volume:	1. Software
	Tab:	3. Vendor 
Strategies

Digital and Microsoft Plan to Integrate OpenVMS and Windows NT
Copyright:  International Data Corporation, reproduction of this 
document is prohibited.
Analyst:  Dan Kusnetzky, Philip Johnson, and Jay Bretzmann
IDC Opinion
Digital recently announced a new strategy for its OpenVMS operating 
system. It will position OpenVMS as the robust, enterprise-class 
platform supporting OpenVMS and Microsoft NT clients in a three-tier 
distributed processing model. IDC believes Digital has a high 
probability of success with the venture with those users who are loyal 
to OpenVMS based upon the following.
 	Microsoft needs more time to develop robust communications, 
management, and TP software  for NT
 	Digital is currently providing a significant portion of  the 
necessary development tools, connectivity software, and system 
integration services today.
 	OpenVMS and Windows NT are architecturally similar.
For the next several years, this appears to be a win-win relationship 
for both vendors. Digital�s primary challenge is to maintain its 
present network management and system software differentiation while 
Microsoft duplicates these capabilities within NT. IDC believes the 
decline of the OpenVMS platform will slow dramatically this year. We 
see this alliance as more of a stabilization move rather than a strong 
growth opportunity.
Announcement Highlights
On May 8th, Digital Equipment Corporation announced a new strategy for 
OpenVMS in conjunction with Microsoft, a long time Digital partner, 
under which Digital will enhance OpenVMS by adding several new 
features and the two companies will jointly build an integrated 
systems environment supporting three-tier distributed processing.
Digital will enhance OpenVMS by adding the following features
 	64-bit addressing
 	Log-structured file system
 	Kernel threads
 	Enhancements to OpenVMS and Windows NT cluster technology
Digital, Microsoft,  and other software partners will jointly 
integrate the following key development tools, interfaces, and 
middleware with OpenVMS
 	Digital�s ObjectBroker, a CORBA-compliant object broker which is 
compatible with Microsoft�s Object Linking and Embedding (OLE) 
technology
 	Digital�s Reliable Transaction Route (RT.) for Windows NT providing 
highly reliable store-and-forward messaging.
 	Application development tools initally including Visual BASIC, C++, 
Natural, Forte, and Entera.
 	Win32 application programming interfaces (APIs) will be integrated 
into OpenVMS in three phases. The first phase will include the 
integration of Win32s, Microsoft Foundation Class (MFC) libraries, 
and OLE2. Phase 2 will include server focused Win32 APIs. The final 
phase will include the integration of relevant APIs from 
Microsoft�s BackOffice.
Digital and Microsoft will jointly develop training and service 
products.
Snapshot Analysis
This announcement consists of three different ideas, each will be 
addressed separately in his flash. They are 1) Microsoft and Digital 
jointly attacking enterprise-class opportunities, 2) Digital 
integrating Windows NT interfaces and middleware into OpenVMS, and 3) 
Digital�s promotion of three-tier distributed application 
architectures.
Digital and Microsoft are attempting to building upon one another�s 
strengths with respect to system software. Digital has long been 
recognized as a premier provider of robust, distributed systems 
solutions of all sizes; Microsoft has long held the majority of the 
desktop market, but has not enjoyed the same degree of success for 
server solutions. IDC�s report, 1994 Data and Near-Term Outlook for 
AOEs, (#9628, January 1995) points out that Windows NT server is the 
fastest growing (175.9%) of the Advanced Operating Environment 
Server/Host Systems; however it�s the smallest in term of overall 
license shipments (100,000). Together, both companies will reap the 
rewards of NT mindshare and market presence backed up by a robust 
enterprise computing platform.
The announcement of Digital�s new OpenVMS strategy comes on the heels 
of another joint venture Microsoft entered into with Computer 
Associates (See Computer Associates Throws its Weight Behind Windows 
NT, #9945, April 1995). Here again, the focus was on providing near-
term solutions that will accelerate the acceptance of NT as a key 
component technology in enterprise-class environments. CA�s Unicenter 
is a network and system management product with a proven track record 
of success on mainframe and Unix systems. It�s availablility on 
Windows NT lends comfort to IT managers who are reluctant to adopt 
untried enterprise hardware and software offerings. Now Digital lends 
its credibility as a proven operating system and scalable hardware 
supplier to the Windows NT cause, hoping to tap into its application 
development momentum.
Digital has recently lost many of its application partners to Unix and 
Windows NT platforms. Even though OpenVMS is a world-class operating 
system environment, new application developers are not attracted to 
the platform because they perceived it to be a declining technology in 
terms of price/performance and they had residual concerns about the 
true openness of OpenVMS. Most of the best and brightest developers 
are now honing their skills on Unix and Windows NT where they will be 
able to sell these talents for many years to come.
As a consequence, Digital�s OpenVMS application portfolio is aging. It 
hopes the addition of Win32 interfaces, MFC, and OLE2 will help it 
provide new applications for OpenVMS with the use of software 
development tools like CORBA, Forte, Etc. It will also give loyal VMS 
developers a chance to build new skills for a platform (NT) that�s 
architecturally similar to VMS, and unlike a transition to Unix, NT 
represents a basic level-set for both VMS and Unix developers.
An open issue is whether or not ISVs will port their Windows and 
Windows NT software to OpenVMS. IDC observes that this is the second 
time Digital has tried to use this strategy to reinforce OpenVMS. The 
first is when Digital added POSIX interfaces to VMS and renamed it 
OpenVMS. This action wasn�t sufficient to attract Unix ISVs. What�s 
more, rewritten applications were poor performers. It took Digital 
several releases to improve the performance to acceptable levels.
Unlike SunSoft�s Wabi, this is a source level strategy rather than a 
binary or executable strategy - an important distinction. While source 
level approaches are easier and more flexible it puts the development 
burden on developers rather than on Digital. IDC believes Digital 
chose this approach partly because of the difficulties SunSoft faced 
during its attempt to support a fully binary based implementation. 
Furthermore, it isn�t clear that ISVs will accept this additional 
development burden of an additional port, QC release, and ongoing 
support NT code on OpenVMS. IDC believes Digital must now persuade 
developers that this is a worthwhile utilization of their limited 
engineering resources and funding. If Digital and Microsoft are able 
to convince developers to pick up the gauntlet that they have thrown 
down, higher performance and more reliable software is a likely 
outcome. 
Digital�s previous assertion that OpenVMS was the correct choice "from 
the desktop to the data center" was never accepted by the majority of 
Digital�s installed base. OpenVMS was viewed as too large and complex 
as a desktop operating system. OpenVMS personal productivity 
applications - word processors, spreadsheet and database software - 
were priced at a premium when compared to those available on 
Microsoft�s Windows. Furthermore, 
Both Digital and Microsoft have been promoting multi-tier distributed 
processing architectures for quite some time. Multi-tier distributed 
processing archtiectures (see Figure 1.) offer a number of advantages 
including those in the following list.
 	Lowest cost hardware configuration for certain business problems
 	Efficient use of technology and computing power
 	High performance distributed solutions
 	Easy expansion and growth
 	Investment in personal computers, workstations and general purpose 
systems made for other applications can be reused as part of 
another solution
Figure 1.	Multi-tier distributed 
processing architectures
 
Overwhelmingly users have chosen Windows as their desktop environment. 
Unix, NetWare, and to a much smaller extent, Windows NT Server were 
their selection for workgroup environments. Even those OpenVMS users 
who have decided to adopt Unix as their enterprise server do not 
typically shut off their systems and send them to the shipping dock. 
These users simply put new or updated applications on Unix. Typically 
they chose HP, IBM, or Sun over Digital�s own Unix platforms because 
these vendors have put strong migration and coexistence programs into 
place. This announcement will do little to persuade them to stay on 
OpenVMS. Loyal OpenVMS users, on the other hand, are likely to see 
this announcement as a wonderful opportunity to gain Windows 
development experience and will stay with OpenVMS for the foreseeable 
future.
IDC is not convinced this announcement will stop the move many users 
have been making to Unix as their enterprise server. These users 
believe Unix is "open" and OpenVMS is "closed" even through OpenVMS 
was branded as XPG3 and later XPG4 compliant by X/Open long before 
several major Unix competitors were similarly banded. IDC is aware of 
situations in which end-user company management has overruled their 
own technical staffs� recommendation to maintain a strong focus on 
OpenVMS and required a move to Unix because they�ve read media 
accounts of the imminent demise of OpenVMS.
Market Impact
IDC believes this announcement moves Microsoft one step closer towards 
being considered a viable partner for enterprise-class opportunities. 
If developers are willing to port their software to OpenVMS, Digital 
may have a better distributed solution strategy. This may potentially 
slow the defection of the loyal OpenVMS base, but certainly it is not 
a strong enough story to reverse the trend completely. This move 
supports IDC�s view that Windows NT will see a significant growth in 
both revenues and shipments in 1995.
It is possible that this strategy will actually help growth of the 
OpenVMS market. This perspective is predicated upon Microsoft�s 
failure to develop equivalent Windows NT capabilities to eventually 
offer a lower price alternative to OpenVMS.
International Data Corporation		Page 3

Unix and Advanced Operating Environments
Filing Information:
       Date:   April, 1995
       IDC #:  ????
       Volume: 1. Software
       Tab:    3. Vendor

  Strategies

  Digital and Microsoft Plan to Integrate OpenVMS and Windows NT
  Copyright: International Data Corporation, reproduction of this
  document is prohibited.

  Analyst: Dan Kusnetzky, Philip Johnson, and Jay Bretzmann

  IDC Opinion
  -----------

  Digital recently announced a new strategy for its OpenVMS operating
  system. It will position OpenVMS as the robust, enterprise-class
  platform supporting OpenVMS and Microsoft NT clients in a three-tier
  distributed processing model. IDC believes Digital has a high
  probability of success with the venture with those users who are
  loyal to OpenVMS based upon the following.

  Microsoft needs more time to develop robust communications,
  management, and TP software for NT

  Digital is currently providing a significant portion of the
  necessary development tools, connectivity software, and system
  integration services today.

  OpenVMS and Windows NT are architecturally similar. For the next
  several years, this appears to be a win-win relationship for both
  vendors. Digital's primary challenge is to maintain its present
  network management and system software differentiation while
  Microsoft duplicates these capabilities within NT. IDC believes
  the decline of the OpenVMS platform will slow dramatically this
  year. We see this alliance as more of a stabilization move rather
  than a strong growth opportunity.


  Announcement Highlights
  -----------------------

  On May 8th, Digital Equipment Corporation announced a new strategy
  for OpenVMS in conjunction with Microsoft, a long time Digital part-
  ner, under which Digital will enhance OpenVMS by adding several new
  features and the two companies will jointly build an integrated
  systems environment supporting three-tier distributed processing.
  Digital will enhance OpenVMS by adding the following features

    o 64-bit addressing
    o Log-structured file system
    o Kernel threads
    o Enhancements to OpenVMS and Windows NT cluster technology


  Digital, Microsoft, and other software partners will jointly
  integrate the following key development tools, interfaces, and
  middleware with OpenVMS


  Digital's ObjectBroker, a CORBA-compliant object broker which is
  compatible with Microsoft's Object Linking and Embedding (OLE)
  technology

  Digital's Reliable Transaction Route (RT.) for Windows NT providing
  highly reliable store-and-forward messaging.

  Application development tools initally including Visual BASIC, C++,
  Natural, Forte, and Entera.

  Win32 application programming interfaces (APIs) will be integrated
  into OpenVMS in three phases. The first phase will include the
  integration of Win32s, Microsoft Foundation Class (MFC) libraries,
  and OLE2. Phase 2 will include server focused Win32 APIs. The final
  phase will include the integration of relevant APIs from
  Microsoft's BackOffice.

  Digital and Microsoft will jointly develop training and service
  products.


  Snapshot Analysis
  -----------------

  This announcement consists of three different ideas, each will
  be addressed separately in this flash. They are 1) Microsoft and
  Digital jointly attacking enterprise-class opportunities, 2) Digital
  integrating Windows NT interfaces and middleware into OpenVMS, and
  3) Digital's promotion of three-tier distributed application archi-
  tectures.

  Digital and Microsoft are attempting to building upon one another's
  strengths with respect to system software. Digital has long been
  recognized as a premier provider of robust, distributed systems
  solutions of all sizes; Microsoft has long held the majority of the
  desktop market, but has not enjoyed the same degree of success for
  server solutions. IDC's report, 1994 Data and Near-Term Outlook for
  AOEs, (#9628, January 1995) points out that Windows NT server is the
  fastest growing (175.9%) of the Advanced Operating Environment
  Server/Host Systems; however it's the smallest in term of overall
  license shipments (100,000). Together, both companies will reap the
  rewards of NT mindshare and market presence backed up by a robust
  enterprise computing platform.

  The announcement of Digital's new OpenVMS strategy comes on the
  heels of another joint venture Microsoft entered into with Computer
  Associates (See Computer Associates Throws its Weight Behind Windows
  NT, #9945, April 1995). Here again, the focus was on providing near-
  term solutions that will accelerate the acceptance of NT as a key
  component technology in enterprise-class environments. CA's Unicenter
  is a network and system management product with a proven track record
  of success on mainframe and Unix systems. Its availablility on
  Windows NT lends comfort to IT managers who are reluctant to adopt
  untried enterprise hardware and software offerings. Now Digital
  lends its credibility as a proven operating system and scalable
  hardware supplier to the Windows NT cause, hoping to tap into its
  application development momentum.

  Digital has recently lost many of its application partners to Unix
  and Windows NT platforms. Even though OpenVMS is a world-class op-
  erating system environment, new application developers are not at-
  tracted to the platform because they perceived it to be a declining
  technology in terms of price/performance and they had residual
  concerns about the true openness of OpenVMS. Most of the best and
  brightest developers are now honing their skills on Unix and Windows
  NT where they will be able to sell these talents for many years to come.

  As a consequence, Digital's OpenVMS application portfolio is aging.
  It hopes the addition of Win32 interfaces, MFC, and OLE2 will help
  it provide new applications for OpenVMS with the use of software
  development tools like CORBA, Forte, Etc. It will also give loyal VMS
  developers a chance to build new skills for a platform (NT) that's
  architecturally similar to VMS, and unlike a transition to Unix, NT
  represents a basic level-set for both VMS and Unix developers.
  An open issue is whether or not ISVs will port their Windows and
  Windows NT software to OpenVMS. IDC observes that this is the second
  time Digital has tried to use this strategy to reinforce OpenVMS. The
  first is when Digital added POSIX interfaces to VMS and renamed it
  OpenVMS. This action wasn't sufficient to attract Unix ISVs. What's
  more, rewritten applications were poor performers. It took Digital
  several releases to improve the performance to acceptable levels.
  Unlike SunSoft's Wabi, this is a source level strategy rather than a
  binary or executable strategy - an important distinction. While source
  level approaches are easier and more flexible it puts the development
  burden on developers rather than on Digital. IDC believes Digital
  chose this approach partly because of the difficulties SunSoft faced
  during its attempt to support a fully binary based implementation.
  Furthermore, it isn't clear that ISVs will accept this additional
  development burden of an additional port, QC release, and ongoing
  support NT code on OpenVMS. IDC believes Digital must now persuade
  developers that this is a worthwhile utilization of their limited
  engineering resources and funding. If Digital and Microsoft are able
  to convince developers to pick up the gauntlet that they have thrown
  down, higher performance and more reliable software is a likely
  outcome.

  Digital's previous assertion that OpenVMS was the correct choice
  "from the desktop to the data center" was never accepted by the
  majority of Digital's installed base. OpenVMS was viewed as too
  large and complex as a desktop operating system. OpenVMS personal
  productivity applications - word processors, spreadsheet and data-
  base software - were priced at a premium when compared to those
  available on Microsoft's Windows. Furthermore, Both Digital and
  Microsoft have been promoting multi-tier distributed processing
  architectures for quite some time. Multi-tier distributed proces-
  sing archtiectures (see Figure 1.) offer a number of advantages
  including those in the following list.

    o Lowest cost hardware configuration for certain business problems
    o Efficient use of technology and computing power
    o High performance distributed solutions
    o Easy expansion and growth
    o Investment in personal computers, workstations and general
      purpose systems made for other applications can be reused
      as part of another solution

   - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

                             Figure 1.
       Multi-tier distributed processing architectures

   - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



  Overwhelmingly users have chosen Windows as their desktop environment.
  Unix, NetWare, and to a much smaller extent, Windows NT Server were
  their selection for workgroup environments. Even those OpenVMS users
  who have decided to adopt Unix as their enterprise server do not
  typically shut off their systems and send them to the shipping dock.
  These users simply put new or updated applications on Unix. Typically
  they chose HP, IBM, or Sun over Digital's own Unix platforms because
  these vendors have put strong migration and coexistence programs into
  place. This announcement will do little to persuade them to stay on
  OpenVMS. Loyal OpenVMS users, on the other hand, are likely to see
  this announcement as a wonderful opportunity to gain Windows devel-
  opment experience and will stay with OpenVMS for the foreseeable
  future.

  IDC is not convinced this announcement will stop the move many users
  have been making to Unix as their enterprise server. These users
  believe Unix is "open" and OpenVMS is "closed" even through OpenVMS
  was branded as XPG3 and later XPG4 compliant by X/Open long before
  several major Unix competitors were similarly banded. IDC is aware of
  situations in which end-user company management has overruled their
  own technical staffs' recommendation to maintain a strong focus on
  OpenVMS and required a move to Unix because they've read media
  accounts of the imminent demise of OpenVMS.


  Market Impact
  -------------

  IDC believes this announcement moves Microsoft one step closer towards
  being considered a viable partner for enterprise-class opportunities.
  If developers are willing to port their software to OpenVMS, Digital
  may have a better distributed solution strategy. This may potentially
  slow the defection of the loyal OpenVMS base, but certainly it is not
  a strong enough story to reverse the trend completely. This move
  supports IDC's view that Windows NT will see a significant growth
  in both revenues and shipments in 1995.

  It is possible that this strategy will actually help growth of the
  OpenVMS market. This perspective is predicated upon Microsoft's
  failure to develop equivalent Windows NT capabilities to eventually
  offer a lower price alternative to OpenVMS.

  International Data Corporation

    
    Doesn't a server need to run a Network Operating System ? Currently
    this is PATHWORKS for VMS, so does the new announcement mean that the 
    LAN Manager v3 functionality will be built into VMS as part of the
    Win32 API implementation, and no longer offered as a separate product
    (i.e is this the end of PATHWORKS ?)
    
    Alternatively is PATHWORKS going to be brought up to WNT (LAN Manager v3) 
    functionality ?
    
    Or have I got it all totally wrong ? :-)
    
    Just a thought,
    Pete

    
    Some comments from someone working in the SAP R/3 space:
    
    - SAP & OpenVMS
      -------------
      Digital has paid dearly (IMHO) for an initial R/3 strategy that
      stressed VMS' benefits over UNIX.  While VMS was a very good
      strategy for production sites running SAP (which intended for 
      replacing MVS and VAX enterprise applications with client/server
      "canned" modules that are fully integrated), SAP decided to only
      support UNIX and NT.  This left us with many clients that were
      convinced to go with VMS over UNIX because it was "better", and
      then faced with a migration to UNIX anyway.  In one case I know,
      they bought VMS because it would support SMP while OSF/1 would
      not; then they found themslves the owner of a 2-cpu machine for
      more than a year that would not support SMP (SMP support is only
      now arriving for SAP on UNIX).  I am not being critical here
      either of the SAP or OSF/1 eng support groups; both have been 
      doing a tremendous job of bringing Digital customers an SAP 
      solution that will meet their needs.  I would just like everyone  
      to see how difficult it can be to swim against the ISV tide.
       
    - SAP & Windows/NT
      ----------------
      SAP does provide support for both UNIX and NT-on-Intel.  Current
      indication is that the Intel-based NT version does not scale well 
      for larger clients, but that RISC-based implementations will
      not be supported until the NT market "matures" more.  I have
      taken this to mean SAP recognizes the inability of NT to handle
      the kind of complex 3-tier server configuration that can be 
      done with UNIX (ex. - 2x8400's being used as a High-availability
      DBMS back end, with multiple SMP 2100's being used as the application
      servers).  I also believe that SAP will remained focused on UNIX
      as a primary platform for the big customers (including US) in order
      to maintain S/W margins until NT becomes a larger threat.
    
    
    What's the point?  IF WE HAD BEEN ABLE TO OFFER A "HYBRID" OF NT ON THE
    APPLICATION SERVERS AND VMS ON THE DATA SERVERS, SAP might still be 
    supporting  VMS as a target platform, because it would allow
    NT-favorable customers to effectively scale an NT solution up to full
    size.  Crazy?  Currently SAP has authorized DG to do a similar deal
    with a NT/UNIX Hybrid (UNIX as the DB server).
    
    While SAP may be a dead issue, perhaps there are other solutions that
    need the strength of VMS at the core with NT in the middle tier.  WE
    ARE THE ONLY COMPANY CURRENTLY ABLE TO OFFER A SINGLE H/W BASE THAT 
    WILL SUPPORT THIS KIND OF STRATEGY!  Perhaps customers will know that
    we will eventually cede the DB core to NT, but only when NT is ready 
    for this function (est as 2-4 years for this to be a reality).  In the 
    meantime, we have the best possible alternative to having to go UNIX,
    especially if the management tools on the NT platform will permit 
    management of both NT and VMS systems transparently.  Isn't this worth
    pursuing?  Having met and listened to Wes Melling while he was still
    with Gartner, I am hoping he has some of this in mind.
    
    Charles  
    
                                    

    >Alternatively is PATHWORKS going to be brought up to WNT (LAN Manager v3) 
    >functionality ?
    
    As far as I know, yes (and that's regardless of the announcement
    discussed in this topic - has been in the plans/works anyhow -, 
    but probably helps there too). Talk to the PW product manager, 
    Beth (NAC::) Joseph, if you need details.
    
    ...petri

Re .55:

>Some feedback from a (large European) customer, a large UNIX user: "I got a
>sneak preview on the new OS strategy: VMS and NT... pfew!"  The same
>customer asks why, if it's a good idea to broaden the VMS market by making
>it more accessible to and from WNT, the same doesn't apply to UNIX.

That's pretty strange given that Bristol's press release in .67 makes it clear
that they plan to sell Wind/U V2.1 on Alpha/OSF half a year Digital gives it
away on VMS.
				/AHM

There's a lot more to this announcement than the Bristol piece.  For example,
providing access to the "Dollar" file system from NT and enhancing NT with
OpenVMS cluster technology.

					Steve

	Not to mention being able to run NT services on an OpenVMS
	cluster. THAT'LL be interesting!! I'd like to see something
	like Exchange server running on OpenVMS :) :) :) If they
	can pull it off, it'll be a neat trick.

							mike

>> Some feedback from a (large European) customer, a large UNIX user: "I got a
>> sneak preview on the new OS strategy: VMS and NT... pfew!"  The same
>> customer asks why, if it's a good idea to broaden the VMS market by making
>> it more accessible to and from WNT, the same doesn't apply to UNIX.
> That's pretty strange given that Bristol's press release in .67 makes it clear
> that they plan to sell Wind/U V2.1 on Alpha/OSF half a year Digital gives it
> away on VMS.

*We* are doing it and will stand behind it (presumably) for VMS.  We are
*not* doing it and will *not* stand behind it on UNIX.  There's a
difference, and customers are not blind to that.

Got a note yesterday from a former Digital colleague now at Sun, who says
"My first response to this announcement was `They must be crazy.'  I guess
the Digital frame of mind has evaporated faster than I expected."

You can expect much of the UNIX market to react the same way.

> There's a lot more to this announcement than the Bristol piece.  For example,
> providing access to the "Dollar" file system from NT and enhancing NT with
> OpenVMS cluster technology.

And why not provide access to AdvFS?  Etc., etc., etc.  If the strategy is
valid for VMS it's valid for UNIX.  Why is it necessary in mid-1995 even to
have to point this out?

___Pete

    Does anyone have any official corporate info as to our statement
    regarding this announcement and continued investment in UNIX etc.
    
    I just received a fax from a major customer who is very concerned that
    this may mean we are de-investing in UNIX, and have no long term UNIX
    strategy.
    
    Greg 

	You know, it never ceases to amaze me that our customers can't
	understand that we can devote resources to both Unix AND VMS.

	Pay Unix some attention=we're dropping VMS

	Pay VMS some attention=we're "de-investing in Unix"

	Folks, there is no one operating system that does everything
	well. (OS/2 lovers: Please, don't jump forward at this time)
	Unix does some things very well, VMS does 24x7 better than
	anyone, Windows NT, especially on Alpha, does some things
	very well.

	You don't use a monkey wrench to bang in nails, do you?

							mike

re.73:

Major customers were supposed to be briefed before the announcement. 
Someone in your organisation should have been doing this, I suggest you 
call Mike Powell's office to find out who, when, and where.

Dave.

    re:       <<< Note 3860.74 by AXEL::FOLEY "Rebel without a Clue" >>>



<	You know, it never ceases to amaze me that our customers can't
<	understand that we can devote resources to both Unix AND VMS.

<	Pay Unix some attention=we're dropping VMS

<	Pay VMS some attention=we're "de-investing in Unix"

    
    	Reminds me of my ex wife!  Just replace the words "customers" with
    "kids", UNIX with DAD and VMS with MOM
    
    tony
    (who's only sorta kidding!)

Re 3860.74:

> You know, it never ceases to amaze me that our customers can't
> understand that we can devote resources to both Unix AND VMS.

Yeah, ain't they amazing, them customers, what they can't understand?  Too
bad they're the ones with the bucks, because if they were really smart
they'd sure be buying up a storm with us.  I guess they just never heard of
our great strategic thinking and dynamite marketing strategies!  :-)

Now back to business: if it's a good strategy for VMS it's a good strategy
for UNIX, but we're not doing it on UNIX.  Is it stupid of (those darned)
customers to notice that?

Or are we back to the days of telling them they really don't want UNIX?

___Pete

>Now back to business: if it's a good strategy for VMS it's a good strategy
>for UNIX, but we're not doing it on UNIX.  Is it stupid of (those darned)
>customers to notice that?

  I thought one of the fundamental reasons for this strategy was
  due to ISVs dropping support for VMS. (NOT Unix) Based on
  previously replies, ISVs are porting to Windows / Unix platforms.
  This leaves the VMS installed based in a very precarious position
  and thus the annoucement. The Unix installed based it not in the
  same position. Digital Unix is alive and well, receiving rave
  reviews and has the comfort of continued ISV investment. Does this 
  have any significance from the customers point of view ?

>>Now back to business: if it's a good strategy for VMS it's a good strategy
>>for UNIX, but we're not doing it on UNIX.  Is it stupid of (those darned)
>>customers to notice that?
> I thought one of the fundamental reasons for this strategy was
> due to ISVs dropping support for VMS.

That's our *internal* reason, and it relates only to what's politely called
"preserving the installed base", unless you think that somehow we may be
able to grow VMS sales again in competition with WNT and UNIX.  The real
basic reason for everything we do, in the capitalist sense, should be "make
the most possible money by selling stuff to customers."

Potential customers don't care an iota about our internal motivations.

> The Unix installed base is not in the
> same position.  Digital Unix is alive and well, receiving rave
> reviews and has the comfort of continued ISV investment.  Does this
> have any significance from the customers point of view?

You bet it does!  Virtually every (current and potential) large UNIX user I
talk to also wants to interoperate as closely as possible with WNT -- some
even have plans for how they'll evolve from UNIX to WNT over the next
several years -- but they will never adopt VMS in order to do this.  If we
can do it with UNIX better than our competitors, we will get that business,
but if all we can do is to point to the same third party, that's no
advantage and we're reduced to selling our hardware in competition with
other very reputable hardware.

Which position do you prefer to be in?

___Pete

Pete,
	if what you say is generally true (i.e. Unix customers
want to migrate to WNT) than your concern is very valid, and you
should probably make those concerns known to Lipcon and whoever
is in charge of Unix strategy.

	The VMS - WNT initiative is Wes Melling's, and it is, as
far as I can tell, based 1) on his charter (i.e. OpenVMS business
growth) and 2) the belief that WNT and Unix are the two mainstream
OS's and they will both have their own faithful followers.  Also,
my feeling is that this move (VMS/WNT) is primarily aimed at the
ISV community, where this Unix OR WNT might be more prevalant (?)

but then again, what do I know ....?

cheers,

- Lars

    and lets face it, VMS and NT have LOT in common......
    
    Alan
    

    Why can't we all just be friends and learn to live togeather?
    
    Listen.  I'm tired of the UNIX folks complaining about OpenVMS.
    
       	     I'm tired of the OpenVMS folks complaining about Unix.
    
    	     I'm tired of the WNT folks acting smug like a Prom Queen
              on Prom Night with 50 Dates...
    
    The OpenVMS Partners were told the OpenVMS/Digital-Unix/WNT strategy
    after DECUS in DC and for the first time during all our company's 
    religious WARs someone made good sense. (Wes M.)
    
    If you can't position all three OS's for Digital, If you can't talk 
    nicely about all three OS's, and if you don't know how to sell these
    OS's I suggest some new sales training starting with:
    
    
    WNT/Windows 95     The Desktop is owned (65%Windows/10%Xwindows/7%Apple) 
                       and controlled by this OS. Apple is weakening and 
                       losing market share Unix is a small player on the 
    		       desktop.  
                      
                       Digital sells Intel and Alpha in this market and 
    		       a great number of Laptops, Desktop and First Tier
                       servers on these platforms.  Our products are among
                       the best and the fastest.
    
    Unix Applications  To some Users/Buyers Unix is an application and 
                       they won't be swayed differently.  This isn't bad,
                       Sell them some of our Unix...
                       
                       We have 64 bit Butt-kicking Digital-Unix, Linux on Alpha
                       and have a real credible story about Standards based
                       computing that's the best in the Unix Marketplace
                       USE IT.
    
    OpenVMS            At 6% of the Computer market place, OpenVMS is the only
    		       Cost Effective 24x7x365 continuous processing system
                       in the Industry.  We're not talking tier 1 servers
                       supporting a couple hundred PCs but clusters that 
                       continually serve 100, 200, 500, and over a 1000 
                       Gigabytes of Data and Databases with the best 
                       performance availablity and reliablity in the industry.
    
                       We have to start building on what we have and the 
                       NT/OpenVMS alliance makes us part of the growing
                       Desktop - Desktop/Server market. 
    
    
    
    I will personally come across the wires and trottle the next person
    who says that Digital Unix (with Polycenter, Clusters or anything 
    else loaded in) competes TODAY with high end (over 300 gigabytes of 
    Storage) Servers (OpenVMS,MVS, even AS400) as a 24x7x365 solutions. 
    
       (I will gladly prostrate myself, acknowledge my heresy, and
         confessing my sin to the Unix person who can show me a Unix system 
          that's servicing more than 300 Gigabytes, with continual 24x7 service 
           and availablity after 9 months or more, with more than 1000 users. 
            -- I can name three sites in Dallas of this caliber running 
             OpenVMS off the top of my head) 
          
    
    I will personally "BOOT" to the Head any people still "EDUCATING" 
    technology recommenders how much better OpenVMS is than UNIX or NT 
    or force-fitting it into the wrong markets.
    
    Sell these products where appropriate and with the least amount of 
    customer resistance. Unless the customer is paying us to educate 
    them on the nuances between OSs or is paying for our informed opinion.
    
    Sell OpenVMS/WNT/Unix Clusters into Every Market for high availablity,
    Scalablity, and reliablity and remember we've had over 12 years
    of Experiance with OpenVMS Clusters to build Unix and NT Clusters
    with.
    
    And remind people that we Service Everthing that We sell and many 
    items we don't.. Unlike most other vendors in the marketplace...
    
    --
    
    
    Unix sales... Take your unfair share of the 40% Unix Market at the 
    expense of SUN, HP, and IBM 
    
    (After Digital-Unix captures the Unix Market then we'll help capture 
     the Desktop/WNT Market for you. Never fight a war on two fronts...)
    
    NT sales..   Take your unfair share of the 40% WNT market at the 
    expense of SG, Unisys, NEC, and the other NT vendors.
    
    OpenVMS sales..  Take your unfair share of the other 20% of the Market
    at the Expense of IBM (MVS and AS400s), HP 3000s and 9000s, Tandum, 
    Stratus, Unisys and the need for high end NT systems.
    
    
    Unix, OpenVMS and NT represent the broadest range of computing
    solutions available from Digital and available to the marketplace
    ever... 
    
    
    Never be ashamed of our heritage and legacy... We are Digital and have
    the best engineers on Earth...
    
    Go forth and sell Digital for we are growing to be giants in the market
    place. (and don't make me come across the wires and smack you around;-)
    
                       
    John W.
    http://www.montagar.com/~johnw/		      
    

re .82

Nicely spoken.

I do have a question though.  Where are these desktop figures from?
    
>    WNT/Windows 95     The Desktop is owned (65%Windows/10%Xwindows/7%Apple) 
>                       and controlled by this OS. Apple is weakening and 
>                       losing market share Unix is a small player on the 
>     		        desktop.

Apple is selling over 1,000,000 Macs a quarter.  The last
numbers I heard for Xterminals and Workstations were
200,000 and 500,000 annually.

Do you have numbers claiming there are more than 18 million
(the Mac installed base) X terminals and workstations?

X windows is the GUI for unix.  How can unix be a small player
on the desktop if Xwindows controls 10%?

Also, according to the Mac trade press, a study of what people actually
use, not everything purchased in the last 10 years, shows Macs are
gaining share on the desktop.

- Peter

>    I will personally come across the wires and trottle the next person   
>    who says that Digital Unix (with Polycenter, Clusters or anything    
>    else loaded in) competes TODAY with high end (over 300 gigabytes of    
>    Storage) Servers (OpenVMS,MVS, even AS400) as a 24x7x365 solutions. 
    
    
Come trottle me then (What is trottle?  Will I like it?).
    
    My customer has recently migrated some critical parts of his
    application to UNIX.   It isn;t 24*7 and it isn;t 300GB, but his
    business brings in US$9B during the 500 hours a year the system needs to
    be up.  During that time it runs 1000 tps at times.
    
    I think UNIX is just over the cusp of being able to support
    critical applications.  A few years ago it would have been far too
    risky, today it is manageable, and in three years it will be
    commonplace.
    
    SInce it takes quite a while to bring up a mission critical application,
    now is the time to start. 
    
    I am curious how many applications need to have a central 300GB data
    base.  My experience is that many of them can (and should) be
    distributed.  I admit it is more expensive in real costs to distribute,
    but there are other advantages.
    
                                        

> Furthermore, the different flavors of DOS / Windows
> operating systems and Intel / Alpha PCs do not add up to one ("this") market.

That's one of the fundamental mistakes Digital mistakes - it views the world
from the perspective of the base system manufacturer, rather than ISVs
and end users.  I'll grant you it's true that they aren't exactly the same
market.  However, to take an anology from another market - all cars that
run on gasoline share the same infrastructure.  Cars that don't run on gas
are at a disadvantage, and non-cars (e.g. horse/buggy) are at extreme
disadvantages.  (Even "simple" underlying differences like a rotary
engine can put you at a disadvantage, other things being equal, and
even if those underlying things have some advantages.)

To the degree that applications and end users do not see significant
difference, the DOS/Windows family ARE in fact the same market from
the perspective of infrastructure including both commodity hardware
and, more importantly, application availability and consistent user
interface.  (For example, this is being written on an Intel/NT home
system that runs standard windows and DOS apps - the ISVs don't care
that it happens to be Windows NT rather than "Old Windows" or DOS -
the apps "just run", which is all that matters to me, the end user.
No "other" interfaces.  No "different versions" of applications.
Quicken, MS Office, a few games, win16 apps off the net, etc., all run,
and I get the benefit of no one-app-crashes-another that I get with
"Old Windows", which more than makes up for the loss of a couple
of DOS apps I very rarely used, and really don't need any more given
better replacements.  But, I don't have to deal with a different UI
or a different hardware platform, it was just a software upgrade,
no strings attached.  I could even trivially back off in the presence
of any problems to my same set of files running under DOS/Old Windows.)

I do agree that unless Alpha/NT or other RISC processors can run 386 apps,
they are at an *extremely* serious disadvantage that effectively cuts
them out of portions of the market (in addition to the portion excluded
by NT being different than "standard Windows & DOS".)

>    (c) the RISC-processor desktop OS business - against each other,
>        MacOS, and any other serious players which emerge.  Here the
>        DOS/Windows customer must migrate to a new OS.  If the cost/
>        benefit ratio is too high, Intel PCs and Power Macs continue
>        to jointly control the desktop and notebook markets.

There is a LOT of cost to migrating between OSes in terms of user training.
What makes you think that:

    1) Apple and Apple ISVS would embrace Alpha in the Mac market enough
       to ensure that enough native apps are availableto make the platform
       attractive (and that Alpha would be anything but an also-ran even
       in the limited Mac market)?

    2) That Digital could sustain an investment for a 4th major
       operating system?  (Or 5 if you ask for this and NextStep.)

IMHO part of Digital's problem at the moment is that the focus is already
too diluted.  When we go to ISVs for a port, we're already talking 3 OSes.

> Here's a possible formula for success.

    1.  License the NeXTstep operating system, to get its graphical user
        interface, ease of use, Display Postscript support, and object-
        oriented development tools.

    2.  Combine it with Digital Unix.  (Both systems are based on a Mach
        kernel, which could make this step easier.)

    3.  Rigorously evaluate the usability of the system and fix problems.

    4.  Add emulators or binary translators for Microsoft Windows and/or
        Macintosh software.

    5.  Throw in some "neat stuff" like World-Wide-Web browsers, support
        for voice recognition, etc.  NeXT may be able to help with ideas.

    6.  Sell the resulting operating system as shrinkwrap software which
        can run on any Alpha PC meeting certain hardware specifications.
        Price it at $100 - $150 a copy retail, as opposed to the $800 or
        so that NeXT used to charge for a user copy.

>        This facilitates the development of a healthy Alpha PC industry.

What makes you think that Digital is going to be able to get a return
on this available only on Digital hardware?  The reason why MS can afford
to sell OSes relatively cheap is volume.  Digital hasn't shown any inclination
toward volume even for Alpha NT, let alone a brand new OS that wouldn't
be able to run many of the existing DOS/Windows apps.  (Translators aren't
a highly desirable option if you're going after volume, because it's basically
cost-shifting from the vendor to the end-user - a strategy that rarely works.)

>    7.  Build an Alpha PC with a fast Alpha chip, CD-ROM, 1GB hard disk,
>        32MB RAM etc. which is priced the same as a 90 MHz Pentium.  Let
>        buyers choose whether they want Windows NT, Digital NeXTstep, or
>        both (extra $).

This would be good.  Note, however, that your new OS isn't likely to go
anywhere if it only runs on high-end desktops (the same criticism you
just levelled against the current Alpha WNT offerings, which is valid.)
Also, last I checked, OSF was much less memory-friendly than Alpha NT
(let alone Intel/Windows.)

>    8.  Extensively advertise the new OS and the Alpha PC.

Again, a hardware vendor launching a new OS (or at least a miniscule
minority GUI) that runs only on it's own hardware, and low-volume hardware,
is likely to be DOA as far as ISVs are concerned (and would likely be,
at best, 4th on the porting list after Windows, Mac, and X.)

> Items 6, 7, and 8 are the hard parts.  Digital and NeXT have killer products
> (Alpha PC, NeXTstep) which languish in small niches because of their pricing
> and marketing.  Yet both would need to cooperate in lowering their prices to
> have a chance of going after the desktop OS market.

My personal opinion is that you'd have a real difficult time trying to crack
the desktop/portable market with a Next GUI that ran in 8Mb, ran on x86s, ran
almost all existing Win16 apps, and sold for less than real Windows.  (No,
I'm not joking.  I think an awful lot of technical people grossly underestimate
how likely it is that people that view computers as a means to solving their
problems are like to learn a new GUI, for example, plus potentially a completely
new set of apps.  For example, my father recently got a Mac portable after
having a Mac previously.  He was very unlikely to even consider switching,
and that would have been the case as well if he had previously had a Windows
machine.  He wanted "what he had before with slightly different attributes",
not something radically different.  If you had *given* him a Next portable
he *might* have taken it.  It's almost as likely, if not more likely, that
he would have said "I can't afford the time to learn a different system".)

re .82,

I managed a Unix system which ran for 6 months with no operator
assistance, admittedly with only 60-100 users at any given time.
The only reason that the system was shutdown was for a memory
upgrade!

From what I gather, I seem to be one of the few people who enjoys
using both Unix and VMS from both a programmer and user viewpoint.
I guess that must make me something of a weirdo!

Chris.

RE: .88

I have more I'll write later, but I thought this tidbit was particularly
interesting in terms of a hardware vendor trying to launch an OS (and
particularly given that OS/2 was held up as an example).  This wasn't
even a case where the system vendor was trying to use it to leverage
their own hardware architecture...

...
  OPERATING SYSTEMS:
  
  GARTNER NOTE ON SERVER OPERATING SYSTEMS (NETWARE AND OS/2)
  Gartner Group, April 26, 1995
...
  *  OS/2 program at IBM responsible for billions of dollars of losses;
     no signs of turning profitable; Gartner recommending that clients
     make no new commitments to OS/2
  (This bleak forecast for OS/2 echoes what Dataquest recently predicted:
   that IBM will be forced to sell or spin off OS/2 to stop the bleeding)
...

> From what I gather, I seem to be one of the few people who enjoys
> using both Unix and VMS from both a programmer and user viewpoint.
> I guess that must make me something of a weirdo!

Meet another weirdo.

But in my earlier notes here I began with what a *customer* said about the
VMS/WNT thing, continued with what an *ex-employee competitor* said about
it, and then said some things about the *market*.  I'm not talking about my
own personal tastes, or technical merit, or any of that old-fashined stuff.

I'm talking about what people want to buy, and how they perceive whether
we'll provide it to them.  I repeat: if tighter integration with WNT is
good for VMS, it's also good for UNIX.  So why don't we do it?

___Pete

re: .91

>I repeat: if tighter integration with WNT is good for VMS, it's also good
>or UNIX.  So why don't we do it? 

It's really very simple:

1.  The Digital UNIX people do what they think is best for Digital UNIX.

2.  The OpenVMS people do what they think is best for OpenVMS.

If you really want to know the answer to your question, go ask the Digital
UNIX people.  There's certainly nothing stopping them from implementing a
similar strategy if they thought it was in their best interests to do so.
 
Ray

>> I repeat: if tighter integration with WNT is good for VMS, it's also good
>> for UNIX.  So why don't we do it?

> It's really very simple:
> 1.  The Digital UNIX people do what they think is best for Digital UNIX.
> 2.  The OpenVMS people do what they think is best for OpenVMS.
> If you really want to know the answer to your question, go ask the Digital
> UNIX people.  There's certainly nothing stopping them from implementing a
> similar strategy if they thought it was in their best interests to do so.

Perhaps they weren't informed ahead of time?  (Speculation.)  Those silly
customers seem to think "Digital" is one company, but you're right, we know
better.  There are people doing what they think best for [___].

Who's doing what's best for Digital?

___Pete

> Perhaps they weren't informed ahead of time?  (Speculation.)  Those silly
> customers seem to think "Digital" is one company, but you're right, we know
> better.  There are people doing what they think best for [___].
> 
> Who's doing what's best for Digital?

Here's another point of view -- these people are doing what's
best for Digital precisely by doing what's best for the stuff
they work on.  In The Bad Old Days, we'd have spent 5 years coming
up with the Grand Unified Strategy for Every Piece of Software on 
Every Operating System.

(I'm miffed because no-one mentioned what we're doing to help
 RSTS/E customers migrate to Windows NT).

re: .93 -< Those silly customers ... >-

(discussion about if it is good for OpenVMS, why isn't Digital UNIX doing it)

My understanding, in talking with both OpenVMS and Digital UNIX Product 
Management, is that both groups went to their customers and asked if the
customers saw any benefit to this, for the respective O/Ss.  OpenVMS customers
loved it, and UNIX customers yawned: they didn't care.

Bristol Technologies has an off-the-shelf implementation of this for various
UNIX systems, and I have to believe that it came up during the discussions
between our two companies (I wasn't there, but I can't imagine no one noticed
it).  Bottom line is that the technology is there for both O/Ss, but only 
one of the O/S customers cares, so Digital is not funding product development
for which there is no perceived market.

What a revolutionary concept! :^)

Now if you believe that there is a market for this on Digital UNIX, bring your
numbers to Digital UNIX Product Management!  Hey, if there is a market, that
means more revenue for Digital...

-- Ken Moreau

    I think that .95 really drives to the heart of the issue - I talk Alpha
    and O/S stuff with customers all the time, and if anything a stronger
    position on NT- UNIX would hurt our UNIX message.  This thing is
    cultural and emotional, but the fact is that a lot of UNIX folks
    wouldn't use anything from Gates and Co. if it came with a $500 bill in
    every box.  These are the same people who scrutinize at every new NT
    related announcement from Digital to see whether it implies that we are 
    really just using Digital UNIX as a way to trick them into NT land.
    
    Similarly, there are a lot of OpenVMS IS professionals who have spent a
    sizable portion of their careers justifying why they use OpenVMS rather
    than (anyone's) UNIX.  For them, NT offers a palatable, politically
    correct, no humble pie needed, POTENTIAL future migration path.  By
    offering them NT apps on OpenVMS and clusters, we legitimize their
    position, ease the pressure, and give them the tools they need to
    remain Digital customers.
    
    In short, I'm pleased with the announcement, feel that it's been given
    exactly the right emphasis from both the OpenVMS and UNIX camps.  My
    only regret was that we were unable to manage the negative spin that
    some elements of the press took on.
    
    bonj

>I do have a question though.  Where are these desktop figures from?
    
>>    WNT/Windows 95     The Desktop is owned (65%Windows/10%Xwindows/7%Apple) 
>>                       and controlled by this OS. Apple is weakening and 
>>                       losing market share Unix is a small player on the 
>>     		        desktop.

    IDC chart from last year... Window's GUI was 65% Xwindows 10%, apple
    was 7% and the rest were terminals, cash registers and various
    specialty devices with a GUI.
    
    
    
    
>Apple is selling over 1,000,000 Macs a quarter.  The last
>numbers I heard for Xterminals and Workstations were
>200,000 and 500,000 annually.

    4million plus 700,000 other desktops  
    
    Last year over six million PCs were sold by the top three PC vendors alone
    
    The Desktop war is over... Window and Microsoft has won except for
    niche markets...
    
    Last year PCs outsold character cell terminals for the first time in 
    history...
    
    
>Do you have numbers claiming there are more than 18 million
>(the Mac installed base) X terminals and workstations?

    Estimates are over 100 million Windows systems world wide...
    
>X windows is the GUI for unix.  How can unix be a small player
>on the desktop if Xwindows controls 10%?

    Unix is only a small part of the market.. Many Unix applications (as 
    do OpenVMS applciations still use character cell terminals)
    
>Also, according to the Mac trade press, a study of what people actually
>use, not everything purchased in the last 10 years, shows Macs are
>gaining share on the desktop.

    
>- Peter

    I'm not here to bash Apple but they are a small part of the Market and 
    appear to be running on some hard times.
    
    Windows is the only clear winner and leader in the marketplace.
    
    And the market place is where this type of Defacto standard setting
    is done...
    
    John Wisniewski
    
    

>             <<< Note 3860.85 by SCHOOL::NEWTON "Thomas Newton" >>>

>Re: .82

>>>  WNT/Windows 95     The Desktop is owned (65%Windows/10%Xwindows/7%Apple) 
>>>                     and controlled by this OS.
>>>
>>>                     Digital sells Intel and Alpha in this market and 
>>>    		       a great number of Laptops, Desktop and First Tier
>>>                     servers on these platforms.  Our products are among
>>>                     the best and the fastest.

>The desktop is owned and controlled by DOS, Windows 3.x, and MacOS.  Windows
>95 is not shipping now.  Furthermore, the different flavors of DOS / Windows
>operating systems and Intel / Alpha PCs do not add up to one ("this") market.

    I listed this market as WNT/Window 95 for the detailed impaired among
    us.  
    
    Windows (any OS from Microsoft) appears like a single market to most
    users.. And Microsoft is keen to keep the Single Market, Single OS, 
    Single Interface message alive and well...
    
    

>In the first category, I expect that Windows NT will be a threat to VMS in a
>few years, for exactly the same reasons that personal computers of all types
>took market share away from minicomputers.

>To sell Alpha desktops, it would help us to have several serious bets in the
>third category.  The three most important standards are probably

    You won't sell alpha Desktops into the mass market.. Too expensive
    That's why Digital and other Computer vendors have an Intel Line...

>We have Windows NT, which runs some off-the-shelf Windows software.  We blew
>our chance to have a native Macintosh OS as well.  That leaves Unix.
    
>The attraction of Unix today is that it is standards-based, and suitable for
>server use, and more stable than Windows.  Its drawbacks are usability and a
>lack of off-the-shelf software.
    
    Don't discount the Synergy that WNT/OpenVMS have with the Anti-Unix
    Customers.
    
    Why do we have to "FIX" customers emotional problems... Let them 
    fight their own dragons I'm going to sell them something...

>    1.  License the NeXTstep operating system, to get its graphical user
>        interface, ease of use, Display Postscript support, and object-
>        oriented development tools.
    
          There is no Software for NeXTstep.  We're playing in the Mass
    market here... 
    
>    2.  Combine it with Digital Unix.  (Both systems are based on a Mach
>        kernel, which could make this step easier.)
       
         So Like this is an improvement over DCE, CORBA, and the other 
         tools that will have software written to them.     
    
    
>    3.  Rigorously evaluate the usability of the system and fix problems.

         Goes without reservation for all our products...
    
>    4.  Add emulators or binary translators for Microsoft Windows and/or
>        Macintosh software.
        
         You folks are Hung up on this Macintosh software... The Company's 
         I do business with use very little Mac and are moving twords total
         Windows Desktops.
            
    
>    5.  Throw in some "neat stuff" like World-Wide-Web browsers, support
>        for voice recognition, etc.  NeXT may be able to help with ideas.

         Have you seen the Highnote Ultras?  Voice recognition, Netscape,
         Trumpet and Pathworks support... We don't need a Dead OS's two 
         year old "State of the Art" ideas...
    
    
>    6.  Sell the resulting operating system as shrinkwrap software which
>        can run on any Alpha PC meeting certain hardware specifications.
>        Price it at $100 - $150 a copy retail, as opposed to the $800 or
>        so that NeXT used to charge for a user copy.

         Why don't we sell a Linux Shrink Wrap for Alpha PCs and be done
    	 with it. (for like $25-$50)  Most support for PC tools are things
         are already in there...
    
>        This facilitates the development of a healthy Alpha PC industry.

         Linux and WNT will do that without NeXTstep...
    
>    7.  Build an Alpha PC with a fast Alpha chip, CD-ROM, 1GB hard disk,
>        32MB RAM etc. which is priced the same as a 90 MHz Pentium.  Let
>        buyers choose whether they want Windows NT, Digital NeXTstep, or
>        both (extra $).

         Why? The Power PC is already giving the MAC power users the
    additional power they want.  Why would we spend the engineering 
    resources to add Alpha into a shrinking market...
    
>    8.  Extensively advertise the new OS and the Alpha PC.

        *Sigh*
    
>Items 6, 7, and 8 are the hard parts.  Digital and NeXT have killer products
>(Alpha PC, NeXTstep) which languish in small niches because of their pricing
>and marketing.  Yet both would need to cooperate in lowering their prices to
>have a chance of going after the desktop OS market.

    The war is over.... check the market leaders and come in for the 
    big win... PCs are like TVs today and getting more so every month...
    A simple GUI appliance that's a given for any application deployment.
    
    Let's worry about two/three tier Client Server and getting it right 
    instead of tilting at windmills...
    
    John W.
    

	I gonna try to state this "safely" so bear with me...

	- Bristol's stuff is/will be available on all sorts of UNIXi
	  (including, I'm told, ours)...
	- for many reasons (most of which are stated here) that may not
	  be as "exciting" as the OpenVMS version...
	- this is *not* aimed at the shrinkwrap market...the technology
	  allows development on Visual C++ on Windows NT with subsequent
	  compile/link/execute on the target platform...the target market
	  *is* in-house developers who may want the advantages of both
	  an "industry standard" development platform *and* a choice of
	  deployment platforms with their strategic advantages
	  (i.e. OpenVMS=scalable, highly available, etc)...
	- despite this, shrinkwrap vendors could come along for the ride
	  if they so choose simply by compile/link on the target platform...
	  its simply that many won't perceive much of a market to justify
	  the effort...
	- Bristol is working on the GUI piece of it as well...thus
	  a real WNT program really would execute on OpenVMS (provided it was
	  compiled and linked there)......

	this is my understanding...please correct me if I've screwed
	anything up.....

	_kelley

    >
>                    <<< Note 3860.86 by HGOVC::JOELBERMAN >>>
>                              -< Come trottle me >-
>
>>    I will personally come across the wires and trottle the next person   
>>    who says that Digital Unix (with Polycenter, Clusters or anything    
>>    else loaded in) competes TODAY with high end (over 300 gigabytes of    
>>    Storage) Servers (OpenVMS,MVS, even AS400) as a 24x7x365 solutions. 
>    
>    
>Come trottle me then (What is trottle?  Will I like it?).
>   
    Excuse me... I ment Throttle;-)
     
>    My customer has recently migrated some critical parts of his
>    application to UNIX.   It isn;t 24*7 and it isn;t 300GB, but his
>    business brings in US$9B during the 500 hours a year the system needs to
>    be up.  During that time it runs 1000 tps at times.
>    
    
    You are not listening... 24x7x365 isn't 500 hours or 800 hours it's
    everyday, everyway, no downtime.  Period. No One Does it better than
    OpenVMS....NOT UNIX NOT MVS, NOT HPUX, NOT DIGITAL UNIX, NOT WNT
    
    When downtime happens the Plant Blows up
    
    When Downtime happens people Die
    
    When Downtime happens the customer looses MILLIONS OF DOLLARS PER
    MINUTE.
    
    It's not just some trivial Niche business... it represents 8 Billion
    Dollars of OpenVMS sales in marketplace...
    
>    I think UNIX is just over the cusp of being able to support
>    critical applications.  A few years ago it would have been far too
>    risky, today it is manageable, and in three years it will be
>    commonplace.
    
    
    And for a customer that demands UNIX this is an excellent solution.
    
    But when a customer comes to me and asks how they can achieve 24x7x365
    they don't care about being on the cusp, or the latest polycenter 
    announcements, they want the track record of a system that has done
    this before... In most of these situation Unix doesn't measure up.
>    
>    SInce it takes quite a while to bring up a mission critical application,
>    now is the time to start. 
    
    With Untried technology? With Green Clustering... Not with my
    customers.... You have to have a track record to win these types
    of sales... for many customers that demand a proven track record
    and have no preference in the religious wars OpenVMS is the 
    best way to go (and the most profitable for Digital Equipment
    Corporation) 
>    
>    I am curious how many applications need to have a central 300GB data
>    base.  My experience is that many of them can (and should) be
>    distributed.  I admit it is more expensive in real costs to distribute,
>    but there are other advantages.
>    
    I can name 5 DEC customers in Dallas right now, more if you target the 
    companies we should be going after... Your experiance is limited to the 
    Markets you see... I and many of the folks in OpenVMS see a much
    broader view of the industry...  And the large, critical database
    business is outthere... and not all of it can be easily distributed
    today.
    
    

>    When downtime happens the Plant Blows up
>    

Really?  That's a pretty poor design.  It's an inescapeable
fact that no matter how hard you try and prevent it, machines 
**will** stop running.  Usually when you don't want 'em to.


Actually, it used to be in DEC's corporate philosophy (or something
like that) that we didn't sell computers into situations where
system failure would out lives at risk.  Is that still true?
If not, when was it changed?

    
    We sell computers to chemical plants and semiconductor manufacturers
    that do have backup and human intervention.. but after several years
    of continuous availablity -- They are still relying on OpenVMS as a
    first line of defense and very very few ever go to DEFcon II;-)
    
    No matter how hard you try to destroy a highly available, reliable and
    properly configured OpenVMS cluster it doesn't go down.
    
    Parts of the system may fail, but the Cluster keeps on going.
    That's why customers buy them...That's why we sell them..
    That's why we're providing Cluster technology on WNT and Unix too...
    
    And I've got the stories of Customer Neglect and OpenVMS's resiliance 
    to prove it;-)
    
    John W.

RE: .101

	I think John meant Planet Hollywood Dave. :) :) :) :)

						mike

>    No matter how hard you try to destroy a highly available, reliable and
>    properly configured OpenVMS cluster it doesn't go down.
 
    No wonder the CIA wouldn't let these things be sold to the Russians
    during the cold war. Any hope of nuking the USSR out of existance
    would have failed as the rest of the country might well have evaporated
    but those OpenVMS clusters would have ran and ran and ran... :-)

    I think we may be on to something here, lets get the Energizer Bunny
    to do our CLUSTER adds, they keep running and running and running...
    :-).
    

    
    	Now that really is a GREAT idea; besides most CIOs I know can
    relate to the Energizer Bunny, even if they can't remember its
    corporate sponsor.
    
    
    		the Greyhawk

    .100
    
   > When downtime happens the customer looses MILLIONS OF DOLLARS PER
   > MINUTE
    
    
	so 2M per minute = 120M per hour = 2,880,000,000 per day
        times 365 days = $1,051,000,000,000 each year
    
    I apologize, I didn't realize there were companies that dealt with that
    kind of money.  
    
    
    

RE: .107

	Word for MacOS running emulated on a VMS system running on a 
	PowerPC? Seriously Thomas, it's time to lay off the coffee or
	switch to decaf. The last week of your notes/proposals are
	"challenging" to say the least. The fact that the would cost a 
	FORTUNE to implement and would probably as fragile as china
	never seemed to show up in any of your postings. Really, they
	would definately ping the "gee whiz!" meter, but this is a 
	business. If someone wants to do all that, it's best done as
	an exercise in a grad school.

	To use your previously posted automobile analogy against you,
	you don't use a Caddy to go 4 wheeling just like you don't use
	a Hyundai to race at NASCAR. There is a place for all the systems
	you mentioned, but not all of them have to run the same software.

						mike

re:: .108 -< I am impressed >-

>	so 2M per minute = 120M per hour = 2,880,000,000 per day
>       times 365 days = $1,051,000,000,000 each year

Actually, for places like big international banks, those numbers are low.  
$3B is not that big a deal when you are dealing with international monetary
funds transfers and currency trading.

Big New York banks, and their counterparts across the world, do *transactions*
of much larger than that 24 hours/day.  They don't count this as *revenue*, but
the fact remains that this much money passes through their systems 24 hours per
day 365 days a year, and if the system is down, those transactions don't occur
and the bank loses money.

-- Ken Moreau

    .111
    
    It was stated that the company would lose millions per minute.  I was
    suprised by that.  I do believe that if a service company is down for a
    bit more than once it could lose its customers and go out of business. 
    But even that should not be a trillion dollar loss.
    
    

>    It was stated that the company would lose millions per minute.  I was
>    suprised by that.  


John does like to exaggerate just a little, but there are MANY companies who 
do lose bunches of money whenever their production systems are down. I 
support Intel.. they run their FABs on OpenVMS because when they're not 
making chips, they're losing money ($thousands/hour). Many of my 
"production" customers run OpenVMS because downtime costs real money. For 
customers who don't mind a few minutes of downtime here and there, the 
high-availability of OpenVMS is not as big of a selling feature as the
perceived openness of Unix or WNT. If they're down for 1-2 minutes (or more 
depending on the application) while DECsafe runs through all the restart
scripts, they may have some frustrated users, but they don't lose real
money, air traffic controllers don't lose sight of planes, factories don't 
halt production, etc. OpenVMS does not have to "restart" the application on 
another node if the primary node fails, it's already up and running on one 
or more nodes in the cluster accessing the same shared files. Unix is headed 
this way, but it ain't there yet.

Bottom line, OpenVMS clusters *can* (with proper configuration, management, 
and software) be up 24 hours a day, 365.25 days per year... as good as a 
fault tolerant system. I agree that Digital Unix has made some major strides 
in the high availability area, but even with DECsafe, when a system failover 
occurs, there may be a minute or more of time when the application is NOT 
available... for most customers that's acceptable: for some, it's a major 
problem - it's these customers who still want and need OpenVMS, and will 
welcome the WNT server applications on a highly reliable server platform.

Arlan

>    The only fix is to go to open commodity markets, which promise (a) a
>    very definite attraction to all customers (regardless of whether your
>    prices are "too low"), and (b) the chance that once customers see the
>    utility of the commodity base product, they will willingly fork over
>    large sums of money for high-value plug-compatible accessories (like
>    the OpenVMS $Dollar filesystem).

One thing that is a tad ironic, is that very frequently the things that
wind up being commodities don't start out that way.  (i.e. survival
of the fittest tends to chose what actually becomes commodities over
the long term.)  Trying to "set standards" that are open from scratch
very frequently fails (partly because the people pushing them don't
have complete vested interest in general, or when they do, it's frequently
an already-to-late reaction to something that has already shown its
worth.)

This calculation:

>	so 2M per minute = 120M per hour = 2,880,000,000 per day
>       times 365 days = $1,051,000,000,000 each year

makes a leap that isn't justified, as I realized from the post about supporting
Intel's fabs.  The leap is, "if losing a computer for a minute costs millions,
then you are earning millions every minute."

In fact, any time you have a process where everything has to go right, breaking
any single step can cause loss of the whole batch of product.

Think about a semiconductor fab line, where there's $millions worth of chips
in the oven at a time -- but each of those chips has already had several
hours of time invested in it.  If something goes wrong with the oven, those
chips are now jewelry items...  but the second, third, etc. minute of downtime
are "free".

    re: 24X7X365=OpenVMS
    
    How many nuke plants have you been in lately? I know several that are
    seriously handicapped if the cluster is down, and actually ere shut
    down because they hadn't clustered yet.
    
    re: megabucks/minute...
    One of the largest banks in N.Y has it's backup datacenter in Syracuse,
    mirroring the VAXcluster in Buffalo, trading $$$ all over the planet.
    Big Blue couldn't handle the peaks, never mind the power failures. Yet
    the VAXcluster just reboots/reforms and goes on about it's business.
    The operators scramble like mad IPL'ing the mainframes, and never even
    bother to check the VAX'n. They "know" it'll come right back up.
    
    I havn't yet seen any unix shop be able to "depend" on their computing
    machinery like you can with VMScluster...
    
    .mike.

> Keeping Alpha in the closed market invites a repeat of the unexpected,
> painful, VAX/VMS disaster a few years back.  Placing it in the open market
> will ensure its chances to succeed beyond our wildest dreams.

No disagreement at all there.  The question is whether a "very cooperative"
agreement is the right way to achieve the end, or whether there is really
a "primary driver".

Basically:

    - Noone else will believe that Alpha can be a commodity unless Digital
      does and acts that way (a necessary, but nowhere near sufficient,
      condition.)

    - Digital must continue to drive Alpha, and can't afford to significantly
      direct efforts at a lot of other things (which includes relying
      on significantly minority APIs and ABIs.)  Digital has a very
      limited set of resources, and we have to focus on things where
      we can gain and sustain competitive advantage.  Remember we're
      already doing 2 architectures and 4 ABIs relative to Compaq
      focusing on 1 each (counting the [DOS] Win16/x86 / Win32/x86
      family of ABIs as one.)  One can easily argue that Digital is
      still way too lacking in focus (relative to significant competitors
      that are gaining ground) to be more than marginally successful.

    >               <<< Note 3860.108 by HGOVC::JOELBERMAN >>>
    >                          -< I am impressed >-

    >.100
    
   >> When downtime happens the customer looses MILLIONS OF DOLLARS PER
   >> MINUTE
    
    
>	so 2M per minute = 120M per hour = 2,880,000,000 per day
>        times 365 days = $1,051,000,000,000 each year
    
>    I apologize, I didn't realize there were companies that dealt with that
>    kind of money.  
    
 
    Good Math but companies tend to loose only as much as they're worth
    then they go out of business... Most company's never do the risk 
    analysis of how much down time would really cost them.  
    
    One of the most compelling reasons for designing a bullet proof
    business computer system is that company's that experiance a Disaster
    and loose their ability to transact business have an extremely high 
    failure rate -- in addition to  losing millions per minute (although some
    financial companies have that potential of loss)...
    
    Consider that the failure rate for a company's disaster generated
    down time is (from a gartner report) 
    
     1-12 hours  30% of companies would fail
    12-24 hours  60% of companies would fail 
    24-48 hours 100% of companies would fail
    
    (No Fortune 500 (that gartner was aware of) has ever had their business 
    interrupted for 48 hours and survived to be in the next issue...)
    
    It's interesting to note that Failure means going out of business or a 
    loss of all current corporate assets... Million per min... Absolutely,
    up to the worth of a company...
    
    Even with business continuation insurance, customers tend to be very 
    fickle in who they buy from if you can't service them in an emergency,
    they'll find someone else... and usually continue to do business with 
    that someone else...  You can't replace existing customers...
    
     
    John W..

    
    

>    This agreement will let us ship OpenVMS and Digital/NeXT Unix as THE
>    APIs/ABIs for the whole workstation and server industry ... e.g., not
>    just our share today, but including IBM, HP, SGI, and Sun's share for
>    basically no immediate work.  Lots of leverage there.  That gives us
>    time and money with which to plan later stages of integration.

Unfortunately, the lion's share of growth in information services isn't
in the workstation/server "industry".  Railroad manufacturers (providing
"server class systems") still exist, but they were dwarfed by the automotive
industry (and the associated trucking industry, which is based largely
on the automotive infrastructure) long, long ago.

But, as I pointed out before, shrink-wrapped software is what matters
in a number of cases.  Today, there is DOS/Win16=>Win32 and Apple 68K=>
PowerPC.  Right now, basically everything else is nowhere.

> Which dealer do you think will prosper in the PC-priced transportation market,
> the one with the Intel car only, or the one that has every car on the list
> including the Intel car?

That depends, is the marginal profit for the non-Intel cars below that
for Intel cars?  It certainly looks like that's a possibility (if lower
volume things come with higher costs), in which case, the Intel-only ones
will in fact prosper more...  (Note that a lot of stores that were VHS+Beta
dropped Beta to increase margins/decrease costs, in spite of the fact they
were targetting a slight smaller market - a situation not uncommon today
between Windows and Mac.)

More isn't necessarily any better, more may just be more (and have the effect
of reducing margins by increasing costs.)

(You're also making the implicit presumption that PowerPC will in fact
be price and price-performance competitive with x86 running predominantly
x86 apps, or that there will be equivalent native apps at an equivalent
price with at least as good performance, which is an assertion that is
at best on a weak foundation.)

    If you reas my first note you will see that I say my customer does a
    US$9B business over a period of 500 hours per year.  That is 300K per
    minute, but in fact is is skewed.  Some minutes may only be 50 or 100K,
    while other minutes can be to 7M.   The fact that only 500 hours is
    critical is not helpful because those 500 hours are fixed and the
    system must run during those 500 minutes.  That is why they have
    invested in two sites separated by 25km of fiber.  They are currently
    using VAX CLUSTERS and have had only two brief down times in 15 years. 
    One was caused by an operator invoking a manual procedure when he
    shouldn;t and the other was caused by a software bug that only
    manifested when a boundary no one knew about was exceeded.
    
    I stated that this customer has begun to move some functions to UNIX
    for various reasons and that those functions seemed to be fine running
    on todays UNIX.  The customer believes that eventually UNIX will be
    worthy of running critical applications and is looking at it very
    closely.   This customer has also put some functions on NT and they
    seem to be running okay as well.  There is enough redundancy that a
    failure in the UNIX or NT areas will not be critical today.
    
    My customer believes that until recently it was impossible to source
    compatible hardware (above PC's) from multiple vendors so was willing
    to take the risk that DEC would always be able to supply his needs. 
    After our miserable performance over the last 5 years he decided that
    that risk was too great and began looking for a solution that would
    allow him to move to other hardware (not easily, but at least
    potentially) if the need ever came up.  He found UNIX to be his primary
    possibility, with NT being a second choice.
    
    If our company had licensed VMS or VAX chips years ago, such as we are
    trying to do with ALPHA and we have done with PDP11, then VMS would
    have been available on other vendors hardware and there would have been
    no reason to look at other operating systems.  I do not know why
    Digital was so unwiling to consider licensing VMS and/or VAX, but there
    will always be customers whose preference is to be able to source
    hardware from different vendors if at all possible.
    
    I know it is possible to write OVMS code that ports to other platforms
    easily, but it is even easier to write portable UNIX code.
    
    And the volume of business we need is not going to come from those few
    trillion dollar a year companies, it is going to have to come from
    across the spectrum.
    
    Our UNIX business grew > 65% last year (from an admittedly small base). 
    That is not because of super salesmanship, it is because it is a good
    product with a lot of potential.
    
    

    Tom, you are obviously enthralled with NeXTstep, but it is a
    non-starter in the marketplace.  OpenVMS is mass market by comparison.
    Embracing NeXTstep as you propose would be suicidal, no matter what
    its technical merits are.
    
    					Steve

Tom - who exactly is going to write the applications for this new UI/ABI?
Where is the market pull to get ISVs interested going to come from?
It's like pushing a rope - particularly for something that isn't radically
(and I do mean *radically*, not incrementally) better than what's out
there.  Even then, "radically" could easily turn off a lot of people
that have "grown up" using either Windows or MacOS.  (At least Alpha/NT
runs Win16/286 apps, and hopefully will eventually run both Win16/386 and
Win32/x86 out of the box as well.  Windows, Mac and X are all "firmly
entrenched", and two of those are ABI, not just source, compatible.)

Note that in many respects you're making a classical Digital Engineering
mistake - looking at technical merits and not on market factors that can
absolutely dwarf even the most technically excellent product (a VERY
common Digital problem over the years).  If you build a better widget,
people will NOT beat a path to your door.  They *might* do you the favor
of paying attention to your product the next time you buy - if you act
like you're really interested in selling them something, and increasingly,
you have to be "open" in the real operational sense - i.e. you're not
trying to "lock them in" somehow...

You're also advocating what I consider another Digital inherent weakness
of the past - re-inventing the wheel (or maybe it's "the grass is greener
on the other widget"...)  Our successful competitors haven't diluted their
efforts between a lot of different incompatible lines, but have focused
on doing a set of things well.  Consider that we're roughly Compaq's size
in terms of annual volume these days (if I remember correctly), but have
2 architectures, several OSes, etc.  Compaq gets to leverage off of ISVs
viewing the whole x86 market (which isn't even just Intel or just MS from
the perspective of Win16 x86/Win32 x86-driven ISVs.)  Contrast this with
just a portion of Digital pushing 3 different ABIs with significantly
different APIs.  (Ironically, not unlike 11's with RSX-11M / IAS / RSTS /
RT / Unix - unfortunately in a radically different marketplace...)

Personally, despite its problems, I think we have to figure out how to make
Alpha/NT fly in the marketplace.  It's far from perfect, but then so were
initial versions of Windows (or the Mac, or...)  It's unfortunate to see
that there were alternatives that we might have done differently, and
possibly smoother roads that could have been taken (though appearently
locally optimal choices can actually be globally non-optimal, and if
Digital had done better and deferred the pain of getting to it's current
size longer, it's quite possible that things would be worse rather than
better - noone will ever know one way or the other.)

Anyway, given where we are, it seems like we need to focus on gaining
acceptance for what we have, rather than trying to put yet more "brands"
on the shelf.  If DEC Unix considers the Next GUI and important thing
for their marketplace, they should do it by all means.  However, pushing
Unix, particularly given higher entry costs along a number of dimensions
as far as consumers are concerned, against Chicago/x86 NT seems like a very
losing proposition even under the most optimistic assumptions (and more
moderate assumptions noting that Digital is a system vendor rather than
a software house, etc.)  For better or worse, we have been on a particular
path, and changing seems even more hazardous until people really get
the message we're serious about what we're already doing.  (We should
by all means *plan* for the future - something else we've been weak
on quite frequently, however, what we ship as products shouldn't send
a confusing message.  Note the situation IBM is in at the moment with PowerPC
and NT vs. OS/2.  It's very, very hard to be a competitor at one level
and cooperative at another, even if you use divisions.  That's one
of the reasons why RDB is no longer part of Digital, and one of the reasons
why so many "upper case" projects at Digital fared so poorly over time
given the pure-software competitors.)

    Tom, if people really wanted the best and the fastest, there wouldn't
    be sixty zillion DOS seats in the world.  People want what works, and
    at the moment, warts and all, DOS (and Windows) works for them.  They
    accept the flaws, because the benefits outweigh them.

    re.-1
    
    Here, here. Annoying, isn't it?
    
    Dan$loved_RISCOS,shame_really ;-)

    
    >The customer believes that eventually UNIX will be
    >worthy of running critical applications and is looking at it very
    >closely.   
    
    Eventually is not today... Until that day occurs and Unix systems have
    a track record most of my customers feel uncomfortable trusting Unix
    for really critical jobs.  
    
    Offload?   Non critical... WNT, PC or Unix.. Whatever is least cost
    and does the job should be the deciding factor...
                          

   > My customer believes that until recently it was impossible to source
   > compatible hardware (above PC's) from multiple vendors so was willing
   > to take the risk that DEC would always be able to supply his needs. 
   > After our miserable performance over the last 5 years he decided that
   > that risk was too great and began looking for a solution that would
   > allow him to move to other hardware (not easily, but at least
   > potentially) if the need ever came up.  He found UNIX to be his primary
   > possibility, with NT being a second choice.
    
    These are business issues and need to be addressed in a business
    discussion.  There are plenty of UNIX vendors out there that are 
    in much worse financial shape then Digital...
    
    Do we consider these as secondary sources and weigh the risk of the 
    Unix market collapsing?
    
>    If our company had licensed VMS or VAX chips years ago, such as we are
>    trying to do with ALPHA and we have done with PDP11, then VMS would
>    have been available on other vendors hardware and there would have been
>    no reason to look at other operating systems.  I do not know why
>    Digital was so unwiling to consider licensing VMS and/or VAX, but there
>    will always be customers whose preference is to be able to source
>    hardware from different vendors if at all possible.
 
     Old history... We've licensed OpenVMS to other integrators, made it 
    available on two architectures... and tried to open up the business 
    practices as best as we can and still serve our installed base.
       
>    I know it is possible to write OVMS code that ports to other platforms
>    easily, but it is even easier to write portable UNIX code.
    
    Pardon me?  You either write portable code or you don't.  If it's
    portable it comes across to OpenVMS just as easy as any other Unix
    Port and that same port goes to WNT too.  Xwindows, DCE, POSIX, XPG3
    and XPG4 branding, the standards are all in there and ready to be used.
    
    You write to OpenVMS's specific interfaces (QIO, ASTs, system services etc)
    and you get locked in but no more than using specific calls in
    Solaris, or AIX, or HPUX.  Have you ever really looked at that so
    called portable code "If AIX uncomment and include, If HPUX ...."
    It's like they wrote every OS's specific includes into every program;-)
    
    You want portable code... Write portable code and it runs on OpenVMS
    too...  OpenVMS isn't locking any door... People lock themselves into 
    an OS (and a coding style) all by themselves...
 
    
    >   And the volume of business we need is not going to come from those few
    >trillion dollar a year companies, it is going to have to come from
    >across the spectrum.
    
    That's why Tandum is folding up it's tent and IBM is going out of
    business too?  There is a very good, P-R-O-F-I-T-A-B-L-E market for 
    customers who want bullet proof computing.  OpenVMS does that today
    better and at less cost than anyone else and will continue to improve 
    it's abilities into the future.
    
    Let's not throw out a 10 billion dollar a year business just because 
    it's on the wrong OS...
    
    
    >Our UNIX business grew > 65% last year (from an admittedly small base). 
    >That is not because of super salesmanship, it is because it is a good
    >product with a lot of potential.
    
    And WNT grew from 0 to 2,000,000 seats in the same time period,
    percentagees can be deceiving...
    
    Digital UNIX should be taking marketshare away from HPUX, SUN's OSes,
    and AIX... After WE do that, I'll begin to help them take over the rest of 
    the Industry...
    
    Alpha speed and performance is pulling a great deal of Digital UNIX 
    Sales which attracts ISVs which completes the circle in it's
    marketplace.  

    
    --
    
    I'm not saying that Digital UNIX clusters or WNT Clusters won't measure
    up to OpenVMS's current Cluster  capablities in the future but for
    right now let's sell what we have.  
    
    But today OpenVMS, Digital UNIX and WNT share our solutions portfolio
    so let's sell all three where appropriate... That's all.. We're all in 
    this Company togeather...
    
    If OpenVMS was a 10 billion dollar a year unit
    and
    If Digital UNIX was a 10 billion dollar a year unit
    and 
    if Digital Alpha/WNT was a 10 billion dollar a year unit
    
    We'd be a 30+ billion dollar a year company.
    
    WE SHOULD HAVE THESE PROBLEMS AT THE EXPENSE OF OUR COMPETITORS NOT OUR
    OWN PRODUCTS.
    
    John Wisniewski

RE: .138

	You must be joking, right Thomas? What applications would 
	drive customers to pick an OpenVMS system over a Windows PC?

	Really Thomas, I don't mean to be harsh, but your notes lately
	are starting to get a little annoying.

							mike

	Really Thomas, I'm serious. Other than the installed base and a
	SMALL number of customers, we probably wouldn't sell an awful
	lot of Alpha VMS systems.

	My annoyance is at the fact that no matter what FACTS are 
	presented to you, you still insist on gunning for things like
	NeXtstep and "VMS on PowerPC running Windows and NeXt apps!" and 
	other non-prudent business moves. That's my annoyance. The
	fact that others half-jokingly wonder what prescription drugs
	you are on should account for something. You DO have some
	really neat suggestions, but geez, we can't do them in DEC 
	right now. We have a business to continue to run and grow.

							mike

    I missed .138:-)
    
>           <<< Note 3860.142 by AXEL::FOLEY "Rebel without a Clue" >>>



>	Really Thomas, I'm serious. Other than the installed base and a
>	SMALL number of customers, we probably wouldn't sell an awful
>	lot of Alpha VMS systems.

>							mike

    
    When you mention the Installed base and a "Small" number of customers
    you really make it sound like our 8-10million OpenVMS Users made a 
    "small" decision in choosing OpenVMS.  The OpenVMS continues to grow
    and not just from the Installed base.
    
    We need to be POSITIVE about OpenVMS, Digital UNIX, and our Windows
    offerings.
    
    Customers don't want to buy into a niche market no matter how much 
    added value there is.  There is plenty of over lap between these 
    platforms with a variety of availble software (in many cases the same 
    product ported to all of our platforms). 
    
    Let's let the Customer, the market, and the price determine where and
    how Customers deploy application strategy...
    
    Give the customers a choice.. It's an advantage many of our
    competitors can't provide.
    
    John Wisniewski
    

RE: .145

	John, I think you know that I'm just as bullish on you about
	VMS. I didn't work in VMS Engineering for 5 years for naught.
	I don't still run AXEL as a VMS system for nothing. (tho
	it WILL change to Windows NT at sometime I'm afraid)

	It's hard to remember EXACTLY what point I was addressing
	due to some missing notes, but I think we are in violent
	agreement. I just didn't want anyone to think that VMS
	was gonna seriously battle Windows which I believe was
	the point that was trying to be made by someone else.
	VMS most definately has its place, but it's not in a 
	PC in Computer City.

	I'm postive John. I'm positive that each of the operating
	system offerings (VMS, Unix, NT, Windows) we sell has
	it's place. I want VMS to OWN its world. (24x7)

							mike

    Mike,
    
    I'm just here to remind you of your roots;-)
    
    And to reiterate my threat to throttle who ever doesn't sell and say
    nice things about all of Digital's fine products in this company;-)
    
    Let's save the war for our competitors... and put our own differences
    aside...That makes us a better player in all of our markets!
    
    Be seeing you,
    
    John Wisniewski