Conference 7.286::digital

    
    	Great question - I, too, would like to see that. But we won't.

The answer to the question in .0 is, yes, somebody has those figures.  
If you have a business need to know, you can get access to them.  The 
OpenVMS Product Manager would be a logical person to ask about this 
sort of thing.

--PSW

    I'm sure I read somewhere (internal note) that around 50% of Alpha
    sales were to our installed base. If you have access to these two
    figures (size of installed base, # Alpha systems sold) maybe you
    can work out what's happening.
    
    BP said (DVN plus Annual report) that we will 'defend our installed
    base'. Are we our own worst enemy ?. Discuss.
    
    
    AW

    In the latest edition of INFORM (News Magazine for VAX & Alpha Users),
    are the results of an upgrade survey held at DECUS '94 in New Orleans,
    394 customers were asked about migrating to Alpha.  The results shown
    are:
    
    54% Yes, within 1 year
    30% Yes, in 1-2 years
    10% No Decision Yet
    6% No Plans
    
    The operating system that will be important in the future:
    
    59% OpenVMS
    21% UNIX
    20% WindowsNT
    
    Not a factual report, but interesting for reference.
    
    Mark
    

Thanks for the input.  It helps.  So if we combine the results
from the Alpha Customer Survey (in a previous Note) we find that

80% of our current VAX user make up some 90% of the new Alpha
users (since 9% were not Digital users for at least 2 years.)

Out of these 80% only 21% will actually use DEC OSF/1? (This
may be fuzzy since they could only pick 1 OS of future importance)

I'm assuming that the INFORM users included non-Digital systems
in their shop when they thought of migration to AXP (since we
saw in the Customer Survey that 13% was replacement of cometitors'
systems.)

And what I was looking for to begin with was some estimate on how
many OSF/1 systems/sites we'll have in about 3 years.

I'm sure this has as many holes as a Jarslberg, but it's something.

    Many of my customers are planning to migrate to Alpha...unfortunately,
    I can't even get a reply from Engineering on how the hell we're
    supposed to migrate our product, data intact, from here to there.
    
    I'm concerned.  Especially since I've got a 10am con all tomorrow to
    explain "How Digital will migrate your existing system to Alpha". 
    Anybody got tap shoes?
    
    								Tex

What's the issue?

	Steve

RE: .5

Based on what I've seen of the real information, your observations 
are way off base.  I strongly suggest that if you have a real 
business need for this sort of information, as opposed to idle 
curiosity, that you contact the product management groups for those 
business units and get the real figures.  We've already been damaged 
pretty badly by well-intentioned but ill-informed DECcies bandying 
about incorrect stats about sales percentages.

--PSW

    re -1
    	Wall Street analysts that do surveys pretty much are saying
    that no one outside our installed base is buying AXP.  This public
    information is either:
    	
    	true: your secret facts are wrong
    
    	false: someone in DEC responsible for our perception on 
    	       Wall Street is not doing their job
    
    
    	If you think Wall Street doesn't matter to our overall success,
    	guess again.  Anyone making a major $$$ investment in product
    	would be foolish not to check out their vendors current and long
    	term financial health. 
    
    Brad.

re .6

We already have two successful binary translators that move almost all
applications, data intact, from OpenVMS VAX to OpenVMS AXP and from
MIPS ULTRIX to DEC OSF/1 on Alpha. The products are:

DECmigrate for OpenVMS AXP  (VEST)
DECmigrate for DEC OSF/1 AXP (mx)

These products can make make most applications executable on AXP systems
within hours. There's also documentation available on porting source from
one platform to the next. Check out the CONOLD CD-ROMs for the OpenVMS and
OSF/1 operating systems.

The Alpha Migration Tools group, part of Digital Semiconductor, engineered
these products and we're working on other binary translators and emulators
as well. Depending on what platforms you're dealing with, I suggest contacting
OSF/1 or OpenVMS product management for information on migrating applications.

RE: .9

The Wall Street analyst's perception is wrong, and, according to a 
memo from upper management that I received on the subject, it is 
to a great extent due to bad information-passing and spin-doctoring 
of DECcies when presenting sales percentage figures to the analysts. 
According to this memo, we are in the future going to exercise a lot 
more caution in disclosing information such as Alpha migration rates 
and the percentage coming out of the installed base to those outside 
the company.  That is why I urge those with a business need to know 
to seek the correct figures from those who know them.  Idle 
speculation on this subject has been and will continue to damage the 
company.

--PSW

>The Wall Street analyst's perception is wrong, and, according to a 
>memo from upper management that I received on the subject, it is 
>to a great extent due to bad information-passing and spin-doctoring 
>of DECcies when presenting sales percentage figures to the analysts. 
    
    Let's take a deep breath and think about this for a minute and
    put ourselves in the shoes of the potential customer:
    obviously (as a customer) I want to develop an informed opinion of
    the chances of Alpha's survival in the market place. One key element
    of that is actual and projected migration rates. 
    
    Now do I go and take Digital's info on this as my only input? Hey,
    how stupid do you think I look?? No, I go nose around and find
    out what the independent analysts are projecting. 
    
    So now I'm an independent analyst and some big bobo comes asking
    my analysis on projected migration rates. Now do I say "Sorry,
    I don't know because Digital hasn't released that information
    and is being very cautious about giving out"? Hey, how stupid do
    you think I look? I've got kids to feed. No, I go give an analysis, 
    on the spot and out of thin air if I have to. No freely available 
    detailed info from Digital? Hey, tough luck if my perception is wrong
    because Digital hasn't given me any info, I'm only the piano player.
    
>According to this memo, we are in the future going to exercise a lot 
>more caution in disclosing information such as Alpha migration rates 
>and the percentage coming out of the installed base to those outside 
>the company.  
    
    There's a strong public need to know the information (see above). 
    If the information isn't available you can bet your cute behind 
    it will be guessed, second guessed and failing that made up
    (and our competitors will be only to happy to give a helping hand). 
    
    Apparently, the actual figures are _better_ than the 
    perceptions flying around and so to combat those perceptions, what
    we are going to do is...keep the information tighter to our chests. 
    Heh, heh, caught you out there :-). We don't want our competitors
    to know how well we're doing, do we now? Stroke of genius, what? 
     
    >That is why I urge those with a business need to know 
    >to seek the correct figures from those who know them.  Idle 
    >speculation on this subject has been and will continue to damage the 
    >company.

    And the second stroke of genius is to lay the blame for inadequately
    forseeing and reacting to this information need at the feet of those 
    Deccies out in the field in front of the customer by trying to shift the
    blame to them for "bad information passing and spin doctoring". Hey,
    best form of defense is the offense... 
    
    re from_someone_who's_been_out_there_in front_of_customers_and_had_to_make
    a_lot_of_snap_decisions_about_what_to_say_and_what_not_to say_and_
    _saying_I_don't_know_lemme_get_back_to_you_on_that_is_not_always_a_valid
    _option_because_there_might_be_nobody_to_get_back_to...
    
    Flame Off                                                

re .9

Brad, and others,

I am the Product Manager for the Rackmount AlphaServer 2100; or, as my 
customers call it, RackMount Sable.

Most of our customers are in the installed base, not suprisingly.  Many, 
however, are purchasing their first equipment from Digital.  Perhaps I could 
convince some "Wall Street analysts" to answer my phone.  I certainly could use 
the help.

Dave

    I deleted  the  memo  -.2  I  believe  makes  reference  to, but if I
    remember correctly, the issue  in  that  memo  was not that the total
    rate of conversion was wrong (this was never released), but that each
    SBU group was releasing it's own growth rates and when adding them up
    they didn't make sense, or they  signaled that the group that had not
    released numbers was declining at a very fast rate.
    
    In any case, I believe that most of it was internal politics.  how we
    divide the pie to most analyst doesn't matter as much, same with what
    operating  system  they  replace  their  base  with, as long  as  the
    customer  retention rate is high.  In other words, what  analyst  and
    most customers want to know is what % of base customers are replacing
    VAX systems with Alpha systems, regardless of OS.
    
    FWIW, MCS has launched a program called LOYALTY specifically targeted
    to retain the installed base.

    >FWIW, MCS has launched a program called LOYALTY specifically targeted
    >to retain the installed base.
    
    Too bad they haven't launched a similar program specifically targeted
    to retain MCS. :-(
    
    Harry
    

In a number of Wall Street analysts minds DIGITAL = VMS (guess who is to
blame for this).  And in my field geography, the VMS AXP sales are going
to installed base almost exclusively (sort of like PowerPC Macs sell
into Apple installed base almost exclusively).  Well VMS constitutes
less than 50% of Alpha sales.  The OSF and WNT sales are going offbase
to a much larger extent than VMS.

    

thanks for all the interest in this topic.  I didn't
quite anticipate the breadth of the discussions, but I
found it very interesting.

Yes, I'd like to get some better (= more accurate) numbers,
just like several other Noters: any chance to get some
more precise direction on who to ask (product management
is kind of "wide"?)

And I sure agree that perception many times is taken as
reality (hey, I've been involved in the now-gone database
business, and we learned that bitterly)

I'd still say that the ALpha Customer Survey seemed respectable;
is that a dangerous assumption?
Based on that survey's results, is it fair to guess that most of
the "replacing other Digital systems" is going to OpenVMS AXP?
("Most" meaning some 90+%?)  Then what about "new applications";
is it reasonable to expect some 2/3 being DEC OSF/1 systems?
And "competitor replacement"; does some 90% OSF/1 seem OK?

If the Mod's feel that I'm really damaging Digital's interests
by this exercise, go ahead - delete this reply.  I just want to 
get some reasonable understanding of the market.

Thx,

- Lars

Bruce,
	there's no doubt that the majority of migrating
systems go to OpenVMS, the question (in my mind) was how
big is that majority? (Put another way: how many will go
to OSF/1?)

thx,

- Lars 

    re -.11
    
    	Re-read my note.  Wall Street analysts are supposed to give
    investors good advice on the soundness of various companies.  Most
    analysts follow certain markets (computers for example).  Now some
    analyst could ring up the product manager in .13 and ask, "Are you
    selling loads of AXPs outside your installed base?"  He could reply
    "Tons!!!"  Analysts could swallow this and say "Thanks", or more
    likely would say in our case "Prove it!".  Other analysts may wish
    more objective information, so they go talk to MIS directors, you
    know, the people that actually buy stuff.  They ask questions like,
    
    Are you a DEC customer?
    	YES 	Are you planning to migrate to AXP in...
    		12 months 	XX%
    		18 months	YY%
    		never		ZZ%
    	NO	Have you considered AXP to purchase in...
    		12 months 	XX%
    		18 months	YY%
    		never		ZZ%
    
    This is not 'spin doctoring', and as a matter of fact, quite
    eye-opening.  These surveys cover most major desktop, mini and
    mainframe companies.  Want to guess where we end up?  It's not
    pretty.  I would suggest taking off those rose colored glasses.
    
    re -.13
    
    	Keep advertising your product, work hard with our VARs and
    distributors, meet as many customers as you can, set agressive
    ship numbers, stay operating system independent, and prove how the 
    AXP2100 can do things Pentium platforms can't, and they will call.
    Oh, and one more thing, please make it easy to configure your box, not with
    15000 choices.  Put yourself in a salesmans shoes, if it is too damn
    complicated to configure, he/she will find something easier to sell.
    
    Brad.

 Good News in Latest CI Digital Installed Base Survey


The most comprehensive survey of the buying intentions of Digital's large
VAX customer base is CI's survey of  18,000 sites in the US.  Their new
survey was just released, and there is good news for both Digital's 
OpenVMS and UNIX Systems businesses.

Of the active installed base, 15% of the base upgrades to newer technology
each year.   Of this 15% that upgrades, now more than 86% upgrade to new
Digital technology, predominantly Alpha today:
   	
			80% to OpenVMS  (including Alpha and new VAX's)
			 6% to OSF/1    (obviously Alpha
			 ?% to Digital Microsoft based Intel/Alpha PC-based 	
				technology

This is an increase of 9% for Digital only within the past year!

According to the CI survey, this means we retain 98% of the base!

( or 85% continuing to run existing system + 13% (86% upgrades to 
Digital OpenVMS or OSF of the 15% upgrading),
   
This does not even take into account new business from new applications and
mainframe downsizing. 

There is a very positive trend since sales of our Alpha systems are
taking off both for OpenVMS and OSF/1 in our base.

Customers can now get the functionality and applications from partners and
Digital they need on Alpha, in addition to the superior performance and
price/performance.
 
You can copy the survey and other customer data in a public directory:

	VMSMKT"OpenVMS_info"::IBSURVEY.PPT


Mitch Hall
Systems Upgrade Group

RE: .12

Turn off the flame-thrower.  I didn't lay any blame for anything on 
anybody.  The "bad information and spin-doctoring" I was talking 
about is from the folks who talk to the industry analyists who hail 
from "corporate", not the field.  The market analyists get their 
information mostly from the vendors, then they draw their 
conclusions.  What they conclude depends a lot on what spin you put 
on the information you present to them.  You have to be very careful 
what you say and especially how you say it.  For example, if you put 
a lot of emphasis on how many new customers Alpha is attracting and 
how much Unix we're selling on Alpha, the analyst writes "DEC is 
abandoning its installed base of VMS customers".  On the other hand, 
if you emphasize how well Alpha is being accepted by the installed 
base, the anaylist writes "Alpha only appeals to existing customers 
and isn't attracting any new business for DEC".  We've managed to 
elicit both these responses from analysts so far.

--PSW

    Re earlier 'Are we our own worst enemy'
    
    If a customer has a Vax, running VMS, and we migrate him to
    Alpha, running OpenVMS (rather than providing an upgrade to
    his existing system, say) is that the 'right' thing ?
    
    1) MCS revenue takes a hit immediately
    
    2) Margin on Alphas lower than Vaxes.
    
    If customer is happy with Vaxes, sell him another [Note recent
    product announce for Vax systems for customer 'not quite ready to make
    the move to Alpha'].
    
    Views ?
    
    
    
    AW

    Re .23: Yes, here is my view: If you're not prepared to eat your own
    lunch, someone else will do it for you.
    
    Al
     

    re: .23
    
    I believe .24 is correct.  Also consider that as the current {nasty,
    painful} downsizing of MCS continues, it is in Digital's best interest 
    to place as many easily-maintainable systems as possible.  I keep
    hearing stories of MCS fighting to keep up with the workload.  If we
    continue this course (another discussion in itself) without replacing
    harder-to-maintain units with easier-to-maintain ones, customer
    satisfaction is likely to decline, throwing the entire account into 
    jeapardy (not just the MCS portion of it).
    
    We picked this game.  When they decided to shrink MCS, they didn't
    leave much room for any game plan other than eating our own lunch.
    
    If we don't eat our own lunch, we'd better be prepared to live on crumbs.
    
    -- Russ

    
    	Let me see if I can put this into a definable context since this
    question is probably the most important issue to be raised to date.
    
    	Digital will most likely end this fiscal year (FY95) with revenues
    of $12.5 Billion and a gross margin of 30.5% (actual numbers may vary
    depending on sources of information). This gives us $3.8125 Billion of
    actual spending money.
    
    	If people costs constitute 60% of our "spending money", we have
    $2.2875 Billion for employees to get paid on. Quick caveat - do NOT
    flame me for making this little exercise reasonably simplistic, but
    the basics are tried and true methodology, only the author is using
    license to shorten his typing.
    
    	If we take an average "cost" of $45,000 CASH per employee (paycheck
    $$ plus benefits, etc. across total worldwide population) which is
    about normal for a computer manufacturer, we have funds for 50,833
    employees. At the end of the first quarter, Digital had 72,000
    employees plus contract workers. You all can do the rest of the math.
    
    	The point? If we are not securing at LEAST 90% of our install base,
    and GROWING our business with new customers buying Alpha boxes, we,
    dear readers, are in big doggy doo-doo. PCs do not cut it; printers
    and storage do not cut it; only manufacturing and servicing proprietary
    products saves us. Make your own determinations, do your own
    caculations, but the fact is irrefutable - if we do not maintain our
    proprietary VMS and ULTRIX install base, we are as dead as Thanksgiving
    turkeys.
    
    	So knowing what we are maintaining in our install base, for this
    worker bee, is actually the determinent on whether I stay or I go.
    Because I've been through all this before a decade ago, and I remember
    how the story goes.
    
    			the Greyhawk

>    	The point? If we are not securing at LEAST 90% of our install base,
>    and GROWING our business with new customers buying Alpha boxes, we,
>    dear readers, are in big doggy doo-doo. PCs do not cut it; printers
>    and storage do not cut it; only manufacturing and servicing proprietary
>    products saves us. Make your own determinations, do your own
>    caculations, but the fact is irrefutable - if we do not maintain our
>    proprietary VMS and ULTRIX install base, we are as dead as Thanksgiving
>    turkeys.

I agree with this for the most part, however I think you have overstated
the case and understated the PC/peripherals case.

At the momment, the Computer Industry is riding the wave of the biggest
boom there has ever been.It is crossing over from an industrial to a 
consumer base. Bill Gates is worried that he cant sustain his 40% growth 
rate, he's worried what the press will say if it gets cust back to a
measly 25% growth rate. Compaq has revenues/employee of over $800k.

Meanwhile Digital, selling to the installed base, is struggling to get
any growth at all. 

I could be mistaken but I think there's a valuable lesson there somewhere...

Mark

    

    re some previous.
    
    I thought the objective was to either:
    
    1) Go eat someone else's lunch
    
    2) Create new sources of lunches (new markets -( read Prahalad..))
    
    By focussing on selling to our installed base, we spiral down, unless
    we 'renew' the entire infrastructure in our installed base very fast,
    very often. If the customer is loyal, happy with VMS etc, then *manage
    the margin and revenue stream from him*, then use the money to go do
    something new, preferably off base.
    
    Does anyone have a cost/revenue model (necessarily simple, for me :-))
    on what happens when we sell low cost, low maint kit to an existing
    (happy, loyal, VMS) customer ? 
    
    AW
    
    

I pulled the following off of DECUServe, as it shows what some customers are
being told about migration rates.  The OSF/1 percentage sounds a bit high, but
maybe they were talking specifically about the workstation market or something.

         <<< EISNER::$2$DIA7:[NOTES$HIVOL]BUSINESS_PRACTICES.NOTE;1 >>>
                            -< Business Practices >-
================================================================================
Note 382.69      Greater Boston LUG - Why is Digital Killing VMS        69 of 69
EISNER::MEZEI_JF "Jean-Francois Mezei"               29 lines  23-NOV-1994 22:12
                  -< One good way to handle VMS in speeches >-
--------------------------------------------------------------------------------
    I was at a DEC/Microsoft presentation today. and the local office
    spokesperson did a very good job at handling the VMS/OSF/NT image
    problem:
    
    "VMS has a huge installed base, and it is growing by 14% and represents
    43% of DEC's revenus, DEC is very committed to this operating system
    and continues to add improvements.
    
    OSF now represents the biggest growth area at Digital. Who would have
    though a few years ago that Digital would have become a major player in
    the UNIX market. OSF now represents about 53% of DEC's revenues.
    
    NT now represents some 4% of ths share, but because NT had only been
    available for a few months when those statistics were produced, the 4%
    is not really reprensative. Recent numbers place NT at about 10%."
    
    Now, this is a much more "diplomatically" correct way of presenting
    these systems. It doesn't relegate VMS to "the installed base only"
    yet does give the message that DEC is serious about UNIX and NT.

    In the November 21 issue of Digital Today on page 7, Scott Roeth, VP of
    the SBU, states that in Q1 "New business development signed up 425 new
    business accounts (55 percent growth from Q4)".
    
    Sounds like new business to me.  
    
    Quite being so negative!
    

re: .30 What's "Prahalad?"

re: .31 "VMS has a huge installed base, and it is growing by 14%"  
                                                             ^^^ 
                                                             huh?
Could someone provide a little backup for this?

Bruce

    Percentages can lie...
    when you sell 7 licenses one month, and 8 the next month, that's
    growing at 14%.

    re .33
    
    Sorry, 'Prahalad' = C.K. Prahalad, who with Gary Hamel wrote
    'The core competence of the corporation' (Harvard Business Review), and
    recently published 'Competing for the future' (also, Harvard Business
    School Press).
    
    please don't read these and think 'aha ! *THE ANSWER*'.  There have
    been IMHO too many cack-handed attempts to 'fix' Digital based on some
    insights gleaned from some management text or other.
    
    but they are useful reads.
    
    [potential rathole averted, maybe..]
    
    
    AW

re .26

>        Digital will most likely end this fiscal year (FY95) with revenues
>    of $12.5 Billion and a gross margin of 30.5% (actual numbers may vary
>    depending on sources of information). This gives us $3.8125 Billion of
>    actual spending money.
    
>        If people costs constitute 60% of our "spending money", we have
>    $2.2875 Billion for employees to get paid on...
    
>        If we take an average "cost" of $45,000 CASH per employee...
>   we have funds for 50,833 employees. At the end of the first quarter,
>   Digital had 72,000 employees...

	The problem with this analysis is that gross margin is what's left 
over after such items as cost of sales (including the cost of everybody in 
manufacturing, distribution, consulting, services).  Gross margin then has 
to pay for R&E and SG&A.  The former is well under control, the latter is 
not, but it's getting there as we move to the indirect selling model which 
serves people like Compaq so well.

	So it will be interesting if you go back to the balance sheet and 
see where the various people (and other) costs are really accounted for, 
and repeat this analysis on that basis.

		David

    We have seen in through a survey here in our group, a fair amount of
    activity where mass deployment of NT 3.5 is happening in small/mid
    sized businesses... So we are committed to it from the PC side.
    JR
    

>    	The point? If we are not securing at LEAST 90% of our install base,
>    and GROWING our business with new customers buying Alpha boxes, we,
>    dear readers, are in big doggy doo-doo. PCs do not cut it; printers
>    and storage do not cut it; only manufacturing and servicing proprietary
>    products saves us. Make your own determinations, do your own
>    caculations, but the fact is irrefutable 
    
    >- if we do not maintain our
    >proprietary VMS and ULTRIX install base, we are as dead as Thanksgiving
    ^^^^^^^^^^^^^^^^
    >turkeys.
    
    Don't take this wrong Greyhawk... We changed the name of this product
    over 3 years ago to represent our commitment to open standards and
    the Open Systems environment.  OpenVMS, if taken as a single business
    is 1 Billion (that's 1 with a B) dollars larger than any single Unix
    Competitor in the marketplace and more compliant to Spec 1170 and other
    industry standards than many of our Unix competitors.
    
    Please, even in casual notes stop refering to OpenVMS as "Proprietary",
    or "Legacy" or any other words to those effects.
    
    If we do in casual converstions and notes think what the rest of the
    world will does...
    
    Journeys start with the first step... Let's all togeather say:
    
    OpenVMS ... Not just for the installed based anymore;-)
    
    
    And yes.. don't forget our installed base either...
    
    >			the Greyhawk
    
    John Wisniewski
    

    
        

    re: .38  Proprietary and legacy and OpenVMS newspeak ...
    
   >Don't take this wrong Greyhawk... We changed the name of this product
   >over 3 years ago to represent our commitment to open standards and
   >the Open Systems environment.  OpenVMS, if taken as a single business
   >is 1 Billion (that's 1 with a B) dollars larger than any single Unix
   >Competitor in the marketplace and more compliant to Spec 1170 and other
   >industry standards than many of our Unix competitors.
    
   >Please, even in casual notes stop refering to OpenVMS as "Proprietary",
   >or "Legacy" or any other words to those effects.
    
    John, don't take this wrong, but ...
    	1) Who else supplies OpenVMS as software?
    	2) Who else supplies hardware that runs OpenVMS?
    
    Any time someone gets up and says that OpenVMS is more UNIX than UNIX
    I see lots of eyes roll. Do we actually have any case studies where we
    have won business away from a UNIX box based on POSIX standards?
    
    Finally, I thought that both SUN and HP did more than $1B (much more)
    in UNIX systems every year. 
    
    In truth, I've worked on VMS for over 14 years, and I consider it the 
    best multi-tasking production-worthy operating system ever created. 
    UNIX is a distant second unless it has lots of customizations and 
    optional S/W bundled in. But I see our only hope is to continue 
    selling OpenVMS for situations that it is best suited for, 
    not as a UNIX also-ran.
    
    Geoff
    

RE: .39

>    John, don't take this wrong, but ...
>    	1) Who else supplies OpenVMS as software?
>    	2) Who else supplies hardware that runs OpenVMS?

By definition #1, Solaris, Oracle, Sybase, HPux and AIX are all proprietary, 
since you can only buy those products from a single vendor.  And quite a 
few companies (ISVs and VARs) take OpenVMS (and Solaris etc), add their own 
products on top, and sell the entire system as a single entity.  So by your
own definition and questions, OpenVMS is as open/proprietary as any of the
other systems I named above...  Now you may claim that by the above definition
that all of these systems are proprietary, and you would have good reasons
to make that statement.  Marketing people would claim that by the above
definition that all of these systems are open, and they would have good
reasons to make *that* statement.  But they are the same, by your definition.

And as for definition #2, who else supplies hardware for HPux, AIX or Solaris?

Now, you may be stating your definition of open: the exact same product 
available from many different vendors.  Well, that rules out all operating
systems available today (even MS-DOS is only available from Microsoft, and
the other DOS systems are not the exact same product, at least according
to the advertising literature claiming they are so much better than MS-DOS).
And that also rules out every processor available today, with the possible
exception of the 386/486 line, where AMD and Cyrix make decent clones.  It
is certainly true in the Pentium case, which is proprietary, and the way I
read the literature, Intel is not shy about making other people back off of
Pentium clones.

So again, you are stating a valid definition of open, but it is not one that
is in common usage around the industry.  "Open" means whatever you are
selling, and "proprietary" means whatever your competition is selling.  I
for one state that OpenVMS and DEC OSF/1 are open!
    
>   >                                OpenVMS, if taken as a single business
>   >is 1 Billion (that's 1 with a B) dollars larger than any single Unix
>   >Competitor in the marketplace and more compliant to Spec 1170 and other
>   >industry standards than many of our Unix competitors.
>    
>    Finally, I thought that both SUN and HP did more than $1B (much more)
>    in UNIX systems every year. 
    
No, what John said was that OpenVMS business is $1B larger than any other
single vendor's UNIX business.  That does not mean that Sun (for example)
only did $1B in UNIX business last year, it means that we did $1B *MORE*
in OpenVMS business than they did in UNIX business (and that is a subject
best left to each of the companies annual reports).

-- Ken Moreau

    I have yet to find anyone that can say 'OpenVMS' in front of a customer
    and keep a straight face. It always gets a rude snicker from customers.
    
    Another dumb attempt to delude ourselves. 

    And of course I still run into many people, Digital and Customer who
    will swear up and down that OpenVMS is the Alpha version and VMS is
    what runs on VAX. 

    
    I really don't want to get involved in the streaming of each others
    shoes; however....
    
    OpenVMS is NOT VMS. Ask any of our VARs who are longtime VMS shops. 
    And to compare OpenVMS (tm) to UNIX is like comparing beer to champagne
    because they both have bubbles. Give me a break!
    
    Now, as stated before, the problem here is a lack of focus by Digital
    on what the marketplace PERCEIVES Digital to be - which is a
    proprietary systems house that now happens to make PCs also. Until
    that is fixed nothing more is going to happen beyond our sales force
    brute selling systems to our installed base.
    
    Until OpenVMS and OSF/1 run on RS/6000s or HP 9000 or Pentium boxes,
    etc. it will continue to be labeled proprietary, just like its
    competitors. Why do you think everyone is buying PCs, anyway? And
    Intel-based servers?
    
    		the Greyhawk
    

OpenVMS is VMS - by definition.

Can you run MS-DOS on an HP-9000 or RS/6000?

			Steve

RE: .41

>    I have yet to find anyone that can say 'OpenVMS' in front of a customer
>    and keep a straight face. It always gets a rude snicker from customers.
>    
>    Another dumb attempt to delude ourselves. 

Do you enjoy (semi)publicly trashing that which pays a large majority of your 
salary?  Do you get some pleasure out of denigrating the hard work, dedication,
skill, talent, and commitment of thousands of people?  

If you worked for Ford Motor Company, would you go around ridiculing the Taurus
in (semi)public forums?  If you worked for Lotus Corporation, would you state
that 1-2-3 is not a good product?  But yet you feel that it is perfectly proper
for you to do that to one of Digital's products, which is supporting this 
company on it's way back to profitability.

By the way, your base statement is wrong.  There are **LOTS** of us who say 
"OpenVMS" with a *perfectly* straight face, because we have a definition of 
"open" which is accepted by the majority of customers and industry analysts.  
And with that definition, OpenVMS is "open".  And also, I have *never* had
a customer snicker when I said it.  Questions concerning how compatible VMS
and OpenVMS are, those I have had many times, but customers are usually not as
narrow-minded, prejudiced, bigoted, or otherwise as foolish as your comment is.

Now, if you talk to our competitors, they will say what you are saying: that
"OpenVMS is another dumb attempt [by Digital] to delude customers".  And I 
will agree that there are customers who buy that line of garbage.  These 
people are died-in-the-wool HP/IBM/Sun/SGI bigots, who will *never* accept
another solution from *any* other vendor than the one they chose.  Fine, there
are people everywhere who prefer to stick to their position rather than take 
a chance of entertaining a new idea, or otherwise thinking rationally.

But for anyone who is looking at the situation from a neutral perspective, who
is using the word "open" in the way it makes sense to a CIO/CEO who wants to
get the most bang for the buck/pound/lire/etc, then OpenVMS (and DEC OSF/1) 
competes very nicely in the "open" systems market, and I (and many other 
people) have the sales figures to prove it.

-- Ken Moreau

> Can you run MS-DOS on an HP-9000 or RS/6000?

Not directly, but you can run DOS apps OSes from a number of different
vendors (MS-DOS, DR-DOS, IBM DOS, OS/2, Windows/NT, more if you count
SoftPC), and you can run those OSes on hardware from different vendors
(any "IBM compatible PC" - even those with non-Intel-manufactured x86
CPUs - ignoring emulation).

DOS is probably the only truly "open" system (if by open you mean
"can I take my app and replace *everything* under it with no vendor
linkage.)  Windows is "close" given that you can still pick and chose
the dominant cost piece of the system (hardware), while not having
to spend any money on the OS to upgrade (since you can move your OS
to a new binary-compatible CPU free of charge - or, for the moment,
run your Win-16 apps on Alpha NT or MIPS NT, until Win32 becomes
more common.)

Being able to swap the dominant costs of your system without "linkage"
between components is the "real" definition of open - and one of the reason
pseudo-"open systems" (both U*x derivatives and VMS) have been taking
such a beating in the marketplace relative to PCs.  (Beating in this
case being defined as market share - a lot of things are hidden
in the fact that the market as a whole is growing so stagnant $
volumes are really shrinking market shares.)

Standards conformance is OK, but a very large part of the market is
interested in *binary* compatibility (since that typically implies
no need to re-license application software to replace hardware,
no worries about upgrades or applications being available, etc.)

    re .45                                  
    I imagine Lotus has had to go thru some soul searching with 123.
    The parallels between 123 and VMS are eerie: both were once top
    dog in their market, but have since fallen by the wayside in the
    face of competition.  Lotus could have renamed it ("Open123"?),
    dumped lots of money into trying to convince customers that it
    still was good at what it did, etc.  Instead, they've more or
    less admitted to 123's fate and have moved on to other products
    with spectacular growth, namely Notes.
    
    So, if you were on a committee at Lotus that was trying to decide
    whether Notes or 123 received the next infusion of big bucks in terms
    of ads or R&D, then I'd say yes, there's been some internal bashing of
    123... and quite appropriately so.  kb

    
    OpenVMS is still Top-Dog in it's markets.
    
    Show me a Unix server in 24x7x365 operation with more than
    150 GIGabytes of storage  (I can show you 10 OpenVMS systems in Dallas)
    
    Show me a Mainframe system that's doing Point of Sale in lots of 
    stores (How about Block Buster or Toys 'R Us again running on OpenVMS)
    
    Bet your business?  Bet the Farm?  Bet on OpenVMS...
    
    And as to the term OpenVMS(tm)  I wasn't pleased with the name change
    either but I will tell you this:  Everyone at Digital should 
    be singing from the same hymn book on all our products and their
    virtues -- that includes marketing directions and naming conventions
    like OpenVMS.
    
    Digital has no bad products in front of a customer -- Only Bad 
    Employees who denigrate one Vs the other.  
    
    I'm proud of OSF/1's achievements, extolled WNT's and PCs virtues
    and Spoken with Pride about OpenVMS -- All because they come from 
    the finest computer company on the planet -- Digital Equipment
    
    JMHO
    
    John W.
    
    

    Actually, I equate product pride with being _more_ critical of
    it than my customers are and I certainly don't equate critical with
    negative.
    
    Note that 7x24x365 performance is indeed a key differentiator
    of the product but that doesn't show in the name.
    
    RobustVMS, BetYourButtVMS, MissionVMS, YouNameItVMS would
    have all been better names because they wouldn't have been
    the red herring that the name OpenVMS is.
    
    Marketing _did_ fail because this was in essence a move to try
    to deflect a (strong) negative criticism. In Marketing 201
    (this wise group has got past Marketing 101 :^) we all
    learn that you play to your strong points not your (percieved)
    weak ones.
    
    You would then have seen Ad campaigns like this:
    
    Open? We're Open alright. Open for business. 24 hours a day.
    7 days a week. 365 days a year. Bla bla bla...
    
    Come to think of it we still can redefine the Open in OpenVMS to
    mean exactly this: Open for Business 24 Hours a Day :-) :-)
    
    re roelof

It seems that we along with many others are still struggling
to agree on what "Open" really means.  Okay, I can accept that.  Let's
talk about what it means from different perspectives.  Let's try and
really understand how we can make our systems *more* Open from the
customers point of view.

Changing the name by itself doesn't do anything to make VMS more open.
Perhaps if we had effectively communicated a new set of more open business
practices and strategies, it would make sense.  But without it this,
it comes accross as a shallow attempt to manipulate perceptions.
Critisizing us for speaking the truth as we see it isn't very
persuasive or constructive.  Tell us why VMS really is more open.
Tell us what we've missed or overlooked.

To me, Open means no artificial barriers to mixing and matching the
components I want.  It means no locking me in to paying for things I
don't need.  It means the systematic elimination of low value
differentiation.

  When we choose not to port, or otherwise prevent our critical applications
  and protocols from working on other peoples hardware is that open?

  When we make it more difficult or neglect to support industry standard
  protocols along side or in place of our own, is that open?

  When we don't allow customers to transfer or sell software licenses
  in a manner consistent with fair payment for fair use, is that open?

Perhaps we could recognize that we each see truth from our own
limited perspective instead of assuming:

  "I'm the truth.  Since you're different, you must not be the truth.
   Therefore, it's my duty to eliminate your point of view by whatever
   means necessary."

Just a thought...  - Peter

>       <<< Note 3520.50 by LEEL::LINDQUIST "Luke 2:4; Patriots 200:1" >>>

>>    <<< Note 3520.49 by DPDMAI::WISNIEWSKI "ADEPT of the Virtual Space." >>>
>>                      -< OpenVMS(tm) is the correct term >-
    
>>    And as to the term OpenVMS(tm)  I wasn't pleased with the name change
>>    either but I will tell you this:  Everyone at Digital should 
>>    be singing from the same hymn book on all our products and their
>>    virtues -- that includes marketing directions and naming conventions
>>    like OpenVMS.

>    That's what we need, more blind following of every digital
>    marketing whim.   

    As I had said I and many of the OpenVMS partners were not pleased
    with the change.  It was done, it's over, get over it.  And more
    important, start using the real name of our flagship product...
    
    
>    But, I'm right with you, I'll read only Pravda (aka
>    DigitalToday), and where can I sign up for the
>    Palmer-youth-league?

    No one can accuse an OpenVMS Partner of being part of the 
    Palmer-youth-league.  I have committed more CLMs by expressing
    my opinions in closed Digital meetings than I care to remember.
    
    But when the meeting's over -- You close ranks, and get behind
    the program and make it work.  
    
    
>    I'd try harder, but the rules keep changing.  One week it's
>    alphageneration, the next week it's not.  One week DEC, the
>    next Digital.  I can't remember if AXP is in or out.  Maybe
>    there should be a quiz - whip those 'Bad Employees' into
>    shape!!!

     It's been three years, the name of the product was changed 
     from VMS to OpenVMS.  This is not a week to week waffling...
    
    The product has a proper name and we should be refering to it
    and not snickering when we say it.  Snickering or backhanded
    compliments doesn't help us convince the customers of our 
    seriousness about the product and contributes to it's perception
    of decline.
    
>    On the other hand, I do know the difference between it's and its.

    Bully for you...
    
>    	Openly.

    I certainly hope so...
    
    John W.
    

    If the Customer snickers I snicker to.  If the customer ask me a
    straight question like whats the difference, I tell them there is no
    difference.  It's just a name change and that it's still the best damn
    operating system ever written.  NO FLAMES HERE PLEASE!  Las week I just
    shut down a server that was up for over 270 days.  Only powered down to
    move it to a new location.  Befaore that it was up another 175 days,
    only rebooted to do an OS upgrade.  I'm sure the hardware has something
    to do with that.  Oh, and by the way it's been up for over 2 weeks
    since the move.
    
    - Mike
    
    

I have yet to lay hands on an OpenVMS system that has earned its "Open stripes"
(i.e. that has the POSIX subsystem installed on it);  my OpenVMS partner has
never laid hands on such a system and I have never seen such a system at a
customer site; and I have never talked with a person who has claimed to have
laid hands on one.  The point I am making is, that when hardly anyone uses the
"Open" features, the fact the system is Open is purely academic.  I would also
be interested in knowing about applications that were ported to OpenVMS mainly
because of the Open interfaces provided.  I suspect that this number is small
because the assumptions used in real applications in coding to these
interfaces (e.g. fork is cheap) don't hold up when these semantics are
emulated on OpenVMS.  Please correct me if I am wrong above.

In any case, I would greatly appreciate the ability to "test drive" an OpenVMS
system.  Here is your chance to set me straight!

Ram Rao
UNIX Partner, Indianapolis, USA
DTN 443-3341
ram@ini.dec.com
indyx::ram
Ram Rao @INI

    
    

    	I know a lot of trouble was put into making fork on VMS a lot
    faster than spawn (though spawn was obviously used in feasibility
    prototypes). I don't remember how it compares with a Unix spawn, but I
    believe there were some figures in the POSIX notes file.

>       <<< Note 3520.55 by INDY50::ram "Ram Rao, SPARCosaurus hunter" >>>
>                             -< Whither OpenVMS? >-

>I have yet to lay hands on an OpenVMS system that has earned its "Open stripes"
>(i.e. that has the POSIX subsystem installed on it);  my OpenVMS partner has
>never laid hands on such a system and I have never seen such a system at a
>customer site; and I have never talked with a person who has claimed to have
>laid hands on one.  
    
    Look in comp.os.vms.. folks are using POSIX on OpenVMS everyday
    for a wide variety of Unix-OpenVMS interoperablities.
    
    
    Also note the X/Open base branded OpenVMS with XPG3 branding in 
    July of 1992
    
    Also note that the X/Open consortium base branded OpenVMS with
    the XPG4 base brand in July of 1993.
    
    X/Open are the folks that certify "Unix"(tm) in the Industry...
    
    Your OpenVMS partner may be focused on other areas (as you may be)
    but OpenVMS has earned it's stripes in providing the Industry
    standard APIs and certifications.
    
    My DECUS LUG has been running on an OpenVMS/POSIX BBS for the 
    last two years (POSIX made the Unix folks feel much more comfortable
    about the file system, access to the systems and editors;-)
    
    Shell scripts, Unix tools, and much more work fine under OpenVMS
    as you would expect them to under any Unix system.   
    
    I have customers that rave about TAR on OpenVMS (saying it's better
    than other industry versions), and have system managers that are
    greatful that they can offer a Unix environment to their Unix
    trained users without compromising their production environments.
    
    
    Take your alpha, put another system disk on it boot OpenVMS
    and load POSIX if you've never used it.  Other wise your 
    opinion is heresay.
    
    
    >The point I am making is, that when hardly anyone uses the
>"Open" features, the fact the system is Open is purely academic.  I would also
>be interested in knowing about applications that were ported to OpenVMS mainly
>because of the Open interfaces provided.  I suspect that this number is small
>because the assumptions used in real applications in coding to these
>interfaces (e.g. fork is cheap) don't hold up when these semantics are
>emulated on OpenVMS.  Please correct me if I am wrong above.

    Yes it is expensive to implement fork/spawn given the OpenVMS scheduler 
    and need for a full process context, but how much does it really cost?
    You only suspect that because you haven't used it or benchmarked it.
    
    Many OpenVMS users are using POSIX to provide Unix features with OpenVMS
    every day and to help protect their users investment in Unix training.  
    
    To provide and tune POSIX APIs will take a little longer then just 
    implementing them.
    
    But OpenVMS on Alpha has a lot of extra cycles to burn and there is no
    margin in touting the difference performance between OSF/1 and OpenVMS
    (because if there was we'd be talking about SMP, TCP/IP, production
    Databases, benchmarks and all the other system issues and not just fork...)
    
    
    btw I would like to see benchmarks between HPUX, SUNOS, AIX and 
    OpenVMS/POSIX.
    
    
>In any case, I would greatly appreciate the ability to "test drive" an OpenVMS
>system.  Here is your chance to set me straight!

    Go take an OpenVMS course if you want to truely appriciate the 
    OS.    
    
    You do no one at Digital a service by passing along half truths, rumors
    and the page from the SUN Salesman's guide to discrediting OpenVMS
    (even if it's in a Digital Notesfile.)  And if you want to argue the 
    technical merits go to the POSIX notesfile.  
    
    If you really need a public OpenVMS system 214-270-3313 is the number 
    of the DFWLUG DECUS BBS  we should be on the Internet by end of the 
    year at dfwlug.decus.org and you'll be able to telnet to it.
    
    Of course you could just build an OpenVMS/POSIX SYSTEM ON YOUR
    ALPHA...If I boot my demo systems off the net they boot OpenVMS
    if I boot stand alone they boot OSF/1 or WNTAS.. The promise of the 
    Universal platforms delivered -- From Digital.
    
>Ram Rao
>UNIX Partner, Indianapolis, USA
>DTN 443-3341
>ram@ini.dec.com
>indyx::ram
>Ram Rao @INI

    
    John Wisniewski
    OpenVMS Partner North Texas Oklahoma
    DTN 483-4138
    wisniewski@dfwlug.decus.org     Home Internet
    wisniewski@dpdmai.enet.dec.com  Work Internet
    dpdmai::wisniewski              Work VAXmail
    John Wisniewski  @SCA           Work ALL-IN-1

    

There is at least one current Tier 1 partner who is working on porting to
VMS with POSIX as we speak (of the vendors/partners we are working with).  
We've had a total of two vendors that we have talked to from the Palo Alto 
ISVETS organization over the last 6 months about POSIX on VMS.

Your mileage may vary.

Ed

    
>>>    John, don't take this wrong, but ...
>>>    	1) Who else supplies OpenVMS as software?
>>>    	2) Who else supplies hardware that runs OpenVMS?

    Aspen Systems has just announced the Alpine 275x workstation complete
    with your choice of either NT or OpenVMS.  See Digital News & Review,
    November 21, 1994, page 1, "VMS: Operating system gets open at last". 
    This is just the first of the third-party vendors to supply OpenVMS
    with their Alpha-clone systems.



    							-Paul

    But who do you call with a VMS problem, Aspen or Digital?
    I wouldn't be surprised that Aspen will point you at Digital,
    and Digital will tell you to that it isn't our machine. kb
    

>    But who do you call with a VMS problem, Aspen or Digital?
>    I wouldn't be surprised that Aspen will point you at Digital,
>    and Digital will tell you to that it isn't our machine. kb

Digital *shouldn't* ask whose machine it is.  A software provider should
provide software support (e.g. Microsoft doesn't ask what brand of system
you happen to be running on, other than to figure out of there's a hardware
problem in some circumstances.)  Digital presumably got the revenue from
the system sale, and would presumably charge if things weren't under
software warranty any more.

RE: Note 3520.59 by STAR::JACOBI

>    This is just the first of the third-party vendors to supply OpenVMS
>    with their Alpha-clone systems.

Raytheon also has home grown VAX's that run OpenVMS.  They sell these
to customers who have special needs (like going to space, etc.)

	- mark

    They'll call Digital anyways, so who cares?  I would be more interested
    in how tech. people can recognize the Aspen machines.  What does
    F$GETSYI("HW_NAME"), HW_MODEL, etc. return on these things.  Who is the
    engineering resource in the VMS group for Aspen?
    
    Mark

>>>    	2) Who else supplies hardware that runs OpenVMS?

Many years ago I used several systems built by Systime that ran VMS.

Peter

    and two years ago, when I was at Cominius University in Bratislava, I
    was shown a "clone" VAX running VMS....
    
    tony

    	When DEC opened up its subsidiary in Hungary, the first problem
    they had to face was that VAX and VMS were the most popular
    architecture and operating system in the country. Unfortunately all of
    the VAX machines were Hungarian made clones, and the VMS operating
    systems were pirated since the U.S. government had forbidden export of
    the technology to (communist) Hungary.

Re .61:

>>    But who do you call with a VMS problem, Aspen or Digital?
>>    I wouldn't be surprised that Aspen will point you at Digital,
>>    and Digital will tell you to that it isn't our machine. kb
>
>Digital *shouldn't* ask whose machine it is.  A software provider should
>provide software support (e.g. Microsoft doesn't ask what brand of system
>you happen to be running on, other than to figure out of there's a hardware
>problem in some circumstances.)  Digital presumably got the revenue from
>the system sale, and would presumably charge if things weren't under
>software warranty any more.

Well, let's just log in to our handy-dandy Alpha/VMS system and see what a
system interface file claims:

$ TYPE SYS$LIBRARY:STARLET.REQ

! Version:      'X-6'
!
!****************************************************************************
!*                                                                          *
!*  COPYRIGHT (c) 1978, 1990, 1992                                          *
!*  DIGITAL EQUIPMENT CORPORATION, MAYNARD, MASSACHUSETTS.                  *
!*  ALL RIGHTS RESERVED.                                                    *
...
!*  DIGITAL ASSUMES NO RESPONSIBILITY FOR THE USE  OR  RELIABILITY OF ITS   *
!*  SOFTWARE ON EQUIPMENT WHICH IS NOT SUPPLIED BY DIGITAL.                 *
!*                                                                          *
!*                                                                          *
!****************************************************************************
...
$

One shudders to think what dire exclusions existed back when VMS wasn't open.
				/AHM

RE: .68

Yep - Digital has a split personality.

It is pushing:

    - Alpha, the architecture (in that it should give Digital a good
      position in a larger market if it is ever adopted on a larger
      scale beyond Digital - e.g. "first among several" as Intel is
      today with the x86.)

    - 21064[A]/21066[A]/21164/etc. manufactured by Digital semi

    - Workstations and servers manufactured by the SBU

Digital has put VMS in the SBU because it's important to sell systems
(Digital's "traditional" business.)  However, if Digital ever expects
the first two items to make independent inroads, then VMS should be
in a *software* business unit that can provide products and services
to other Alpha vendors (initially using Digital-manufactured chips).
(Note that that might lead to other interesting behavior, like that
group providing "VMS toolkits" on other platforms to make money, other
ports, etc., if they represented profitable enough opportunities.)
Digital has repeatedly cannabalized software to support hardware over
the years, and it's inability to treat software as a business lead
to VMS being "closed", and the eventual dismantling of quite of bit
of DEC software development.  The net result is we've pushed our software
business (in many cases) into commodity markets (e.g. operating systems)
while we are at the same time unwilling to engage in commodity business
practises (i.e. V O L U M E to amortize development costs.)  3 of the 4
components in Digital's systems (chip / box / OS / appl. software)
are VERY volume sensitive, and with feedback loops (both positive and
negative) based on volume (e.g. OS investment can't be sustained at low
volumes, 3rd party appl devos will be uninterested at low volumes,
chip architectures can't generally be sustained at low volume).
Only the "box business" can get by with low volumes, BUT that seems
to be what "Digital understands" and how we (still) try to do business.

Also, note that depending on how you interpret "equipment" in the copyright
notice - that could refer to the *chip*, in which case there isn't a problem -
yet (and may never be given our business practises.)

Re .69:

>Also, note that depending on how you interpret "equipment" in the copyright
>notice - that could refer to the *chip*, in which case there isn't a problem -
>yet (and may never be given our business practises.)

Agreed.  (Although, if a non-chip component somehow caused a software
compatibility problem, I'm suspect that the corporation would backpedal at
Warp 8).

And in retrospect I do note in particular that the word is "equipment", not
"system".  Somewhere in one of the Marketing conferences is the old rationale
from an LMF star that goes something like, "Digital licenses software for a
particular *system*.  If a customer changes the configuration [this was a
cluster or SMP issue], then it's not the same system and they need a different
license".  (I assume that particular money tree's fallen by the wayside of
late).
				/AHM