Conference 7.286::digital

   >Who'd want DEC?  Not IBM.
    
    Actually, this would probably be the best fit of all.  IBM is making
    it's big profits from mainframes, mainframe downsizing, and service. 
    In fact, they even will sell a VAX or Alpha into a shop, and they are
    going after the servicing of DEC backrooms (and hiring the folks we
    layed off, underpayed, and overworked).
    
    So.  IBM buys DEC, dumps the low-margin losers (maybe to Compaq ;-),
    gets DECs installed base of high margin server backrooms, the service
    revenue, and what is left of our service organization.  Pumps a few
    bucks back into OpenVMS because it's a cash flow monster.  Merges UNIX
    development, picking up the Alpha processor as a hot-box bonus.  They
    also get the additional capacity of our chip fab.
    
    *They* know how to market.  *They* know how to make money.  *They* are
    actually more inline with what we actually do, than a PC maker like
    Compaq.  We get a new lease on life, and a company with a track record
    of being able to market and sell *anything* and make tons of money.
    
    I became a DECcie (a long time ago) to avoid working for big blue, but
    it might be the best thing that could possibly happen.
    
    

    We could always become the NT arm of SUN...

    DIGITAL shareholders voted in a poison pill in 1989 that severely
    dilutes the stock when a buyer is deemed hostile by the board
    of directors.  This was rammed thru when the stock was above 80;
    perhaps shareholders might reconsider now and get a vote on the
    ballot for the next shareholders meeting.  Nevertheless, as is,
    a takeover of DIGITAL is unlikely.  A takeover would mean the BOD
    gets fired, meaning they lose their retaining fees, and since they're
    not large shareholders, they have every incentive to view any offer
    by anybody as hostile.
    .02 Kratz

    Possible Value of DEC Aquisition by HP
    	- 64-bit UNIX O/S
    	- Better RISC architecture
    	- Compatible Intel strategy
    	- Real NT RISC strategy tody, 64-bit NT starting late '97 vs. '99
    	- Synergystic customer base
    	- Complimentary storage products and plans (?)
    	- Our cash could finance half the aquisition
    Of course, if they really are not abandoning PA-RISC, then the Alpha
    benifit would be really diminished. 

This really makes no sense.

1) Since 'Digital knows this.' I dont see why the problems have presisted so    
   long w/o being addressed and corrected. Mr. Bismuth sounds as though nothing 
   can be done. Atleast thats the tone of it.

2) I also dont understand why 'Robert Palmer, chief executive of DEC, has run   
   the company since 1992 and has tried to shake it up, but with limited        
   results.' Does this mean his subordinates do what they want? Doesnt he hold  
   them accountable to a plan? As CEO isnt he in charge of seeing his strategies
   thru?

I know there must be something real obvious that Im missing...so call me
clueless. I really would like to understand it though.

    RE: .4
    
    Jim,
    
    HP (and to a lesser extent Intel) are out there flogging their new
    PA-RISC replacement architecture pretty hard.  It doesn't matter that
    it won't be available until 1999 or later; unlike Digital, they know
    that mindshare will win over technical excellence anyday.  They don't
    need Alpha.  And Intel will certainly bring their muscle to bear on
    anyone "naive enough" to look at a non-Intel architecture.  Look at how
    they recently helped Oracle see the light as far as using Intel chips,
    and not StrongARM, as their primary architecture in their upcoming NCs.
    
    On the other hand, I did see some positives from the article Jon posted
    from Forbes to start this conference.  If nothing else it helped to
    alert potential customers and investors that there might be some life
    left in the old dog afterall.  I just hope it isn't too late.  I know
    there is attrition in the sales force, but 40%?  At that rate you
    pretty much turn over completely in the span of a few years.  So much
    for sowing confidence with customers.  I hope it was not accurate.
    
    Chuck 

    
    After being unable to bring the VAX back into price/performance
    competetiveness, and the ensuing decline, Digital has embarked on a
    course to transform the company from a full line computer manufacturer
    into something else, and something smaller.  The problem is that the
    company was geared to sell high margin systems and service.  We've
    taken a miriad of steps to evolve into something else, but those steps
    have not as yet resulted in return to profitability and growth.  In the
    mean time, our high margin business is in decline, in part due to
    deliberate decisions to get out of the "traditional" business where our
    customer base is.
    
    I don't see what Sun or HP get out of a DEC aquisition except losing a
    competetor.
    
    Poison pills do not always stand up in court.
    
    

>    anyone "naive enough" to look at a non-Intel architecture.  Look at how
>    they recently helped Oracle see the light as far as using Intel chips,
>    and not StrongARM, as their primary architecture in their upcoming NCs.
    
    How did Intel do this?
    

re: .5

>2) I also dont understand why 'Robert Palmer, chief executive of DEC, has run
>   the company since 1992 and has tried to shake it up, but with limited
>   results.' Does this mean his subordinates do what they want? Doesnt he hold
>   them accountable to a plan? As CEO isnt he in charge of seeing his
>   strategies thru?

Well, he's fired most of his subordinates after a year or two, including the
ones he inherited from Ken and the ones he hired himself.  Probably the real
resistance is coming from the managerial levels one or two steps under his
direct subordinates though.

Since he refuses to flatten the management hierarchy, things will probably
continue in this vein.

- paul

    
    It was in the news that Netscape CEO has cut his own salary to
    $1.  If it still won't work after sometime, he might resign.

    re: .10
    
    Piety? I think not...
    
    That same Netscape CEO cashed in a small pile of options for at least 
    $10,000,000 (that's millions o' bucks) this year when the stock was a 
    few multiples higher...

    re .9
    
    The management hierarchy, you misunderstand. That is the poison pill.
        
    Dave

    er, since WHEN did we get management?

    Re: .8
    
    I don't recall where I read the article, but basically the gist
    of what I saw about this was that Mr's. Ellison and Grove were
    talking about some other stuff (dinner meeting?) and Larry brought
    up the fact that his engineers were unable to get the 80x86 based
    systems up and working.  The NEXT DAY, Mr. Grove caused a number of
    engineers to show up at Oracle (something like 10+) until the
    Intel boxes were up and running.
    

    
    .9 hit the nail on the head
    
    When I joined DIGITAL in '90, "frozen layer" was one of the first
    things I had to experience; here in Germany nothing has changed since,
    though the "resources" are down to about 2400 (from about 9000).
    
    uli
    

    	Interesting tidbit.  This would have been at approximately
    	$60/share.  Rats.
    
    	NEW YORK, May 5 (Reuter) - Compaq Computer Corp held
        high-level talks with Digital Equipment Corp in both
        1995 and 1996 about a possible takeover of the company but
        discussions broke down last summer, the Wall Street Journal said
        in Monday's electronic edition. 
    
        According to people familiar with the events, the two firms
        reached a general agreement on a purchase price for Digital of
        between $9 billion and $10 billion but Digital moved back from
        the deal, the newspaper said. 
    
        The Journal said talks broke down over management and
        structural issues. 
    
        Compaq renewed talks in mid-1996 but both sides decided a
        transaction did not make sense by September, the paper said. 
    
        People close to the events say for now, the talks are dead,
        the Wall Street Journal reported. 
    
        Compaq and Digital were quoted as declining to comment. 
    
        The Journal said Digital's $6 billion services organization
        was one reason behind Compaq's interest. 

	I would have loved to have been a fly on the wall at those
	meetings..

	"talks broke down over managment"

	No comment.

							mike

    Mr Foley......
    
    It wouldn't have worked......you, (the fly on said wall) would have
    laughed your head off, and all them eyes would have teared so much from
    laughter, , that it would have RAINED on the talks, and you would have 
    been exposed!  
    
    I still chuckle when I think of it myself.....
    
    I'm sure you will find some reference of this in a future Dilbert
    cartoon!
    
    chet

    RE: .8
    
    Sorry about the lateness of the reply; I've been traveling with
    customers with no times for NOTES.  It was apparent from a recent "Info
    World" article that Oracle was shown the light as far as using the
    Intel chip in their NC boxes.  Ellison was called directly by Grove,
    who explained the futility of not using a Pentium chip in such an
    architecture.  One of Oracle's top VPs was then paraded to the Far
    East, where Intel customers like Acer could "explain" how an NC that
    did not use an Intel chip would not succeed in the marketplace.  Oracle
    then put a marketing front on the whole mess by saying Grove was a
    genius who saw the merits of using a Pentium chip before Ellison did.
    
    Give me a break.  They/Oracle were both bullied and sweet-talked into
    using Intel, and there isn't much we can do about it.  That's what
    having a 96.8% share of the marketplace (Intel) can do for you, as
    opposed to .1% (Alpha).  In case anyone disputes the %, they came from
    a recent Business Week or Forbes article on Digital.
    
    Chuck

    What do you mean there isn't much we can do about it?  Using Ellison's
    argument against himself: obviously if it doesn't say Microsoft on
    the "inside", then nobody will buy it.  Bob Palmer should be jumping
    all over Ellison right now - shaming him in the stupidity of his
    arguements.  I don't deny that Intel should be _part_ of Ellison's
    strategy, but his pet NCI will fail if he doesn't show some clear
    and dramatically adventageuos differentiators to the Microsoft
    alternative.  StrongARM, with its low power, portability and
    performance offers some of those advantages.  But then, (my opinion
    of course) I don't think Digital management has done enough to
    support the NC cause.  The amazing design efforts of the SHARK team 
    were funded from Supnik's research budget, rather than a coherent
    choreographed marketing roll-out.  Yes, Intel can afford to put some
    muscle into this initiative, but Digital can't afford not to!
    
    mark
    
    
    Well, I guess I'm glad that we are porting WindowsCE to StrongARM.
    At least we can get a little more of the crumbs from one monopoly.

    
    Why is StrongArm and the Network Computer continue to be used in the
    same sentence.  What does low power consumption do for a desktop/AC
    powered computer?  Not much if you ask me.  
    
    Again I don't see it as a big deal not having the StrongArm in the NCs,
    the StrongArm was designed for the Newtons/PDAs of the world.  Not this
    NC market, or lack of a NC market.  
    
    I personally/professionally don't think the NCs will fly as far as a 
    high volume/percentage of total sales.  But I've been wrong before.
    
    /art

    >What does low power consumption do for a desktop/AC
    >powered computer?  Not much if you ask me.  
    
    Not as much as for laptops/PDAs etc. but it certainly does have _some_
    meaning. In this context, I guess price is more important.
    

>    Why is StrongArm and the Network Computer continue to be used in the
>    same sentence.  What does low power consumption do for a desktop/AC
>    powered computer?  Not much if you ask me.  
    
    The NC was first talked about has the $500 client.  StrongARM has good
    a good performance to power ratio and a good performance to cost ratio.
    StrongARM seems to do very well as a JAVA processor.
    
>    Again I don't see it as a big deal not having the StrongArm in the NCs,
>    the StrongArm was designed for the Newtons/PDAs of the world.  Not this
>    NC market, or lack of a NC market.
    
    Digital Semiconductor has other StrongARM designs for other markets.
    
>    I personally/professionally don't think the NCs will fly as far as a 
>    high volume/percentage of total sales.  But I've been wrong before.
    
    I read(in Communication Week I think) that 750K NC's have been purchased 
    to date. Not bad for new technology.
    
    -Bruce
     
 
 

    
    re: .19 -
    
    >>>
    Give me a break.  They/Oracle were both bullied and sweet-talked into
    using Intel, and there isn't much we can do about it.  That's what
    having a 96.8% share of the marketplace (Intel) can do for you, as
    opposed to .1% (Alpha).  In case anyone disputes the %, they came from
    a recent Business Week or Forbes article on Digital.
    >>>
    
    Sigh ... it always amazes me how simple articles can generate so
    much discussions ... many who read these articles assumes that
    "it must be true, I saw it in a magazine!"
    
    FWIW, many of these magazines get their info from numerous different
    sources and do not even stop to consider that yes, Intel has a very
    high % of the market - Desktop AND server combined. 
    
    Of course, Alpha does not have a % of the desktop market, but what
    about the server market ? Obviously Intel still has a huge advantage
    here as well, but the Alpha % is more like 5% not .1%...as quoted in
    Windows NT magazine. And this was driven by the fact that Alpha's 
    have always (until recently) been hugely more expensive than Intel.
    
    Now, Alpha's are competively priced with Intel, more vendors are
    jumping on Alpha bandwagon (see recent Borland announcement), the
    internal Sales and BU issues are starting to get addressed, so we
    shall see how the next few months progress..
    
    BTW - also keep in mind that all of the NT marketing for DIGITAL in
    the last year has been controlled by the PCBU. Up until the last few
    weeks, the "Alpha" word did not even appear on the DIGITAL windows 
    home page www.windows.digital.com. Even now, it is only a passing
    reference as being supported in the new NT Cluster release.
    
    Hopefully, this is another area that will shortly be resolved.
    
    Regards,
    
    / Kerry
    

    RE: .24
    
    Kerry,
    
    Regardless of your points, the 96.8% versus .1% is what software
    development houses see when they make plans to port or develop
    software.  It is the proverbial chicken and egg situation.  I can't say
    I blame them, either.  Until the Alpha share of the market is
    increased, particularly in the workstation space, this problem will
    continue.
    
    Personally, I feel that Digital has a fighting chance to make it here. 
    But we have to be hitting on all cylinders as a company to do so.  Just
    having a fast chip in and of itself will not be enough, otherwise we
    would have knocked Intel (and Sun for that matter) for a loop a long
    time ago.
    
    Chuck

    Chuck,
    
    Everyone agree's (I hope) that you need SW app's to sell HW. DIGITAL
    has the HW, but no marketing to present the over 1800 app's that now
    run on Alpha .. reference www.enorex.com (our marketing folks should
    take lessons from these guys..)
    
    However, in the case of the Intel market, they have the opposite
    problem, ie. their 32bit HW is peaking out at a time when users are
    screaming for much more power and yet Intel marketing folks have the
    clout to convince the market that even with 64bit sw (VLM, IPV6) and 
    additional heavy compute environments (thin client, fat server, NT5
    encryption) on it's way, that deploying 32bit servers that are the 
    same speed as their clients really is the way to go ..
    
    My concern is that people need to understand and educate Customers that
    there are server SW types and desktop / ws sw types. Alpha has most of the
    leading app's already ported with more to come. Recent press has
    announced VB5, VC++ and Borland are now doing Alpha NT. WORD, Excel
    is due shortly. PRO/Engineer just did their annual benchmark and Alpha
    NT WS won - against SUN, HP, SGI and numerous PPro's .. why does these
    type of events get so little focus internally to Digital ?
    
    >>>
    Regardless of your points, the 96.8% versus .1% is what software
    development houses see when they make plans to port or develop
    software.  
    >>>
    
    This was true 2 years ago. It is not the story today, but too many
    people within Digital believe the Intel marketing machine and simply
    bow their heads and say oh well ..
    
    It all depends on ones viewpoint, is the glass 1/2 full or 1/2 empty ?
    
    We need more people looking at the 1/2 full glass of water.
    
    / Kerry

    >>  Regardless of your points, the 96.8% versus .1% is what software
        development houses see when they make plans to port or develop
        software.
        >>>
    
        This was true 2 years ago. It is not the story today, but too many
        people within Digital believe the Intel marketing machine and simply
    >>  bow their heads and say oh well ..
    
    Are you suggesting our market share is more than this? So what is it
    now 1% instead of .1%? or maybe even 2%? ( I dont recall us having any
    100% growth numbers)
    
    Unless the share is a reasonable number- maybe 20% or greater, then the
    point still stands. Our market share is much to small to be worth the
    effort to port software.

    
    Larry Ellison did the right thing.  You can lose by trying to beat both
    Microsoft *and* Intel, or you can beat the one who is the real threat -
    Microsoft.  My guess is that Larry really doesn't care a lot about the
    chip that is inside the NC, he wants to sell big a**** servers running
    Oracle database products.
    
    While this may be a bad omen for StrongArm, pushing it into smaller
    niches than the potential volume that a successful NC market would give
    it, this is not a bad thing for DEC and Alpha.  If the BA server +
    cheap NC combo combined works, we could sell a lot of BA servers -
    which is where the high margin is.
    
    The NC doesn't have to be the high-end Pentium, nor does it need to be
    64-bits.  The whole idea is that the *real* work is done in the BA
    servers, the NC is a fancy terminal.
    
    

re Note 5273.23 by YIELD::HARRIS:

>     I read(in Communication Week I think) that 750K NC's have been purchased 
>     to date. Not bad for new technology.
  
        What are they being used for?

        What software do they run?

        Bob

    RE: .26
    
    Kerry,
    
    It is actually worse today than it was two years ago, from a
    marketshare standpoint.  We have pushed 10s of thousands of systems out
    the doors during those two years (servers and workstations.)  During
    the same time Intel has pushed out 10s of millions.  As I said, the
    people at the software houses have to eat, too, so they will develop
    for the products that hold the greatest marketshare, and Alpha is just
    a blip (albeit a noticable one) on the radar scope.
    
    I found it interesting that when you pointed out the software products
    that have been ported, you reiterated that "Word and Excel will be
    ported soon", or something very similar.  If one is looking for volume
    in this marketplace, as Digital desires, they better be near to the
    FIRST applications available.  Digital wants to be a large player. 
    Then give the salesforce the tools they need to be successful in this
    endeavor, such as all the most desirable apps.  It may be too late at
    this point, but we'll see.
    
    Personally, I do view the glass as 1/2 full.  I stated earlier that I
    felt Digital had a fighting chance.  But I cannot bury my head in the
    sand and say that we have the apps needed to make Alpha a heavyweight
    player, when we don't.
    
    I think we are closer in thought on this matter than apart.
    
    Chuck                                    

    re: .24
    
    >    Now, Alpha's are competively priced with Intel, more vendors are
    >    jumping on Alpha bandwagon (see recent Borland announcement), the
    >    internal Sales and BU issues are starting to get addressed, so we
    >    shall see how the next few months progress..
    
    Kerry, 
    you keep trotting out this Borland announcement as a proof point of
    success. In fact, its more the opposite.
    
    The announcement only relates to OLEnterprise, which is a product of
    the OEC arm of Borland. Given that OEC's Entera on NT is based on our
    DCE and the intel product has been around for ages this has been a 
    long time coming. Everyone in Borland considers the OEC line as
    non-core, although they will become part of the Borland development
    environments. Unfortunately NONE of the Borland development
    environments are on Alpha.
    
    When I see Borland core products being delivered on NT Alpha, I'll
    believe you. By that I mean: C++ Builder, Delphi and IntraBuilder.
    To date I have seen no announcements for these products.
    
    - John

re: Note 5273.29 

>>     I read(in Communication Week I think) that 750K NC's have been purchased 
>>     to date. Not bad for new technology.
>  
>        What are they being used for?
>
>        What software do they run?
    
    Bob,  
    
    I made a mistake, the number is 550K for NC's.  The number comes from a
    table of Net Clients with credit for the information going to IDC.  I
    have no other information.
    
    -Bruce
    
    

    John,
    
    >>>
    you keep trotting out this Borland announcement as a proof point of
    success. In fact, its more the opposite.
    
    The announcement only relates to OLEnterprise, which is a product of
    the OEC arm of Borland. Given that OEC's Entera on NT is based on our
    DCE and the intel product has been around for ages this has been a 
    long time coming. 
    >>>
    
    I would suggest that any other marketing company would turn this
    announcement into a big event - even though a technical analysis
    might reveal something else.
    
    Just look at how Microsoft ships Exchange and states with a straight
    face that it is "X.500 like" and everyone assumes that it is X.500
    compliant, because all they hear are the words X.500.
    
    IMHO, for every good announcement DIGITAL gets lately, there are 
    a multitude of people who jump in and say "well, it's not really 
    that great, because if you consider this ... or ..., then it really
    is not that great .." and then all of the past issues come gushing out.
    
    I agree that the Borland announcement is not as great as them porting
    all of their stuff to Alpha, but at least lets acknowledge that it's
    a good start ! It's the proverbial "foot in the door .."
    
    Recent announcements - We now have VB5, VC++, all of the core Oracle 
    products (reference 
    http://www.oracle.com:/corporate/press/html/PR120996.111314.html),
    Lotus Notes, Word, Excel, Borland, some great PRO/Engineer benchmarks
    (www.proe.com, also great sales web page at www.systems3.com/promo.htm), 
    some great reviews in traditional PC Mags, PC Mags
    and industry analysts questioning if Intel is running out of steam
    and press announcements that Klamath (Intel follow-on to Pentium II)
    is delayed until likely summer to fall '98.
    
    Glass is definately 1/2 full and starting to fill up ..
    
    :-)
    
    / Kerry

    Kerry,
    
    I agree that we need to make the most of this externally and to keep
    working them for more. However, internally, I think a little bit of
    reality also does not go astray. The way you stated it, any person not
    following the space, including most of our sales force, could end up
    inadvertedly walking into a a situation.

Re .33:

>    Recent announcements - We now have ... VB5, ... Word, Excel

When/where have these been announced? Where can they be obtained?

PJDM

VB5 Enterprise Edition is available for Alpha as the "RISC" edition, 
available at no extra charge by sending in a form included with the
Windows edition. Can't tell if it's available directly.

You should be able to order VB Enterprise via the standard software
channels, including STREAM. STREAM is even shipping those 
orders now....

Is the fact that IA-64 has 0% share in the NC market segment, 0% of the NT market,
and in fact 0% share of any market useful to this discussion?  I recognize that
Intel probably has the muscle to cross the generation divide more smoothly than
we did our last two times but my guess is it will be the most difficult thing
they've faced in a lot of years.  We should take advantage.

Dennis

    Is it possible to insert a little of our own FUD here? In a customer
    meeting yesterday they asked about using Intel for their future.
    I mentioned some of the scuttlebutt about Intel's architecture
    running out of gas...10% jumps here and there...and speculated
    that in a few years we may be seeing them tell everyone "OK, you
    can run old Intel apps, a little faster than Pentium, but if
    you want them to REALLY scream you can run this new native mode
    we have...". In other words, the kind of conversion that has already
    been done in the Alpha space, and the current Intel app compatibility
    becomes moot because nobody'll want to use it and sacrifice
    performance.
    
    Intel will of course deny this, until they have iron, but their
    numbers might lend credence to such speculation.

    >In a customer meeting yesterday
    >I mentioned some scuttlebutt about Intel's architecture
    >running out of gas
    
    You're going to have to change tactics... One of the Claflin seven:
    "We will never criticize any DIGITAL products, services initiatives,
    organizations or people in front of the customer."
    
    While we're still in the Intel business, this now means you talk
    about the ATTRIBUTES of Alpha.  Besides, the Intel-running-out-of-gas
    argument is getting pretty old by now.
    
    Kratz

    The issue of having to port applications over the next few years
    isn't an attribute of Alpha? (The Intel systems being considered
    were NOT from Digital.) If you're speculating about futures
    then it would seem to me that telling the truth to customers
    is, finally, the only way to go. Giving your best guess about
    where that future will be is part of that.
    
    It was of course clear that this was a guess, and was also clear
    it invited the customer to think about it, in a context in which
    they would make their own choices. They could (and did) see the
    performance numbers...and their slopes...
    
    When arguments about futures are made, though, questions about how
    the 8080 architecture can be extended (again) start becoming 
    highly relevant. 

    re: .38
    re: Intel vs Alpha
    
    First, with Customers that I deal with, I usually preface what I am
    about to say that DIGITAL has a dual platform strategy that scales
    from 32 bit to high end 64bit systems. The recent AIM awards illustrate
    the capabilities of our Intel line to handle CURRENT low-medium
    workloads just as well, if not better, than the competition's Intel 
    based servers.
    
    However, as a Consultant, Customers want me to give them the benefit
    of my experience, cut out the marketing crap and make recommendations
    as to what they should do to prepare for the future, keeping in mind 
    these requirements that need to be met:
    
    1. Their IT budgets and head counts are being slashed at the same time
       as they are expected to deliver higher levels of service.
    2. They now recognize that one of the biggest issues the IT dept has
       is how to bring some control to the mess of distributed desktops
       and servers. 
    3. Servers they deploy today will likely have to do them for at least
       18-24 months as they will not be able to go back to the pot for
       additional money for upgrades... the labour and downtime costs
       associated with doing these upgrades would likely double this cost
       as well.
    
    All of these issues are what is driving most major companies to
    recentralize their computing environments. It is also what is driving
    the interest in thin client / fat server computing.
    
    Ok - here we go -
    
    >>>
        Is it possible to insert a little of our own FUD here? 
    >>>
    
    No need to - simply quote and/or hand the Customer recent PC Week staff 
    editorial:
    http://www8.zdnet.com/pcweek/opinion/0414/14chip.html
    
    To summarize - Intel x86 architecture is running out of gas, and
    Intel's marketing machine is kicking in with rapid fire cpu
    "enhancements" - each of which provide only incremental benefits. Their
    goal is to get Customers thinking Intel only without looking at the
    technologies.
    
    Or recent European magazine analysis:
    http://www.heise.de/ct/english/9705154/
    
    Bottom line, DIGITAL does have a dual platform strategy, but if I
    am asked by a Customer for my recommendation on how they can best meet
    the requirements outlined earlier, then this is what I will say :
    
    1. Deploy 64bit for servers and power users / WS's. Expect power user
       desktops and WS's to become much more 3D oriented.
    2. Deploy 32 bit PC's for medium/light PC users.
    3. Consider and/or start to evaluate thin client / fat server
       technologies for those area's of the company where it makes sense ie.
       perhaps light PC users, area's of high security etc.
    4. Recognize that every NT server deployed requires approxmately 5-8
       additional 3rd party app's to manage it in a production environment
       e.g. backup, virus control, defrag, disk quota, centralized event
       and security mgmt, scheduler, config mgmt (SMS etc), batch (esp for
       larger servers with applic maint requirements). Each applic costs
       approximately $500-$3000 per app per server. This is one of those
       "hidden costs" that most shops fail to take into consideration when
       deploying multiple distributed servers.
    
    As an example, yes, Intel does handle current file serving loads very
    well and it is hard to position Intel VS Alpha (even though now the
    price differences are being eliminated) because there is very little
    cpu intervention required to simply take an IO off disk, over the PCI
    bus to the net card and out to the client desktop.
    
    However, with NT5, there is a huge amount of encryption being
    introduced. This now means the cpu is going to play a much bigger part
    of the traditional file serving environment. Which is better for file
    encrypting/decrypting - 32 or 64 bit?
    
    The discussion of IPV6 is another discussion, but suffice to say that
    it is 64bit aligned. It also has autoconfiguration capabilities - think
    many Customers might be interested in reducing / eliminating their
    dependancies on DHCP/ WINS ?
    
    Bottom line is that everyone will eventually do 64 bit computing in the
    same manner as we have migrated from 16bit to 32 bit. It is NOT a question 
    of "should we do 64bit computing?", but rather "when should we do 64bit 
    computing?".
    
    If you have the Customers long term interests in mind, and a good
    understanding of the technologies, then I really find it hard to
    tell a Customer that they should continue to invest in multiple 32bit 
    servers with the knowledge of so much additional cpu work load coming 
    down the road.
    
    Easy question to ask Customer "With NT5 encryption (much more cpu
    intensive work loads coming), IPV6 (64bit aligned), thin client/fat
    server strategies coming, and VLM capabilities, does it make sense to 
    deploy servers that are the same speed as their PC desktops being 
    deployed ?" 
    
    With Pentium II desktops appearing in the next few months, many Customers 
    will now be faced with desktops running at higher speeds than their 
    current PPRO servers...
    
    Regards,
    
    / Kerry

    re .-1
    
    The PROBLEM is that YOU are telling them this and Digital is not.
    
    The big Digit has a few billion people in marketing each working on new
    programs etc., du jour which never get implemented to fruition.  So the
    field ends up creating/telling the stories that corporate should be
    giving us to tell.
    
    Wonder why we're not doing well?  Folks in the 3M area don't come out
    to the field and understand the impact of what they create and ask us
    to sell.
    
    There is NO strategy, just products and an arrogant "go figure it out
    yourself Mr/Mrs Customer...we built a good product and you're too
    stupid to understand it..."
    
    You would think that after 5 years of that silly "A" chip we'd have
    understood the code by now......it's targeting and selling - NOT
    technology stupid!
    
    

RE: .41

Nicely said.  Thank you for some good talking points...


RE: .42

I agree with most of your points, but cannot agree with the attacks on
the people on the GMA.

I agree that what is needed is targetting and selling, but wish to point
out that the reason we are so frustrated in the field is because we do in
fact have the best technology (Alpha, OpenVMS and VMSclusters, Digital UNIX
and TruClusters, NT on Alpha and NT Clusters, GIGAswitch, AltaVista tools
including the Search Engine and the Tunnel), which makes it even more
agonizing that we can't seem to get the message out.  If we had poor
technology (think Sun and IBM) or huge holes in our product lines (think
SGI and Bay Networks) or were facing massive disruptions in the next few 
years (think HP with Merced), then our poor showing in marketing might be 
understandable.  But instead we have the world's best technology and no 
marketing at all, as opposed to Sun/HP/IBM who have much worse technology 
but marketing which makes the buying public believe they are the leaders.

>    The PROBLEM is that YOU are telling them this and Digital is not.

Uhhmmm, the last time I checked my badge said "Digital Equipment".  If
I am telling them, then Digital is telling them.  So what you are really
complaining about is the lack of high-level, saturation advertising that
would put Digital Equipment on everyone's lips.  Well, guess what?  We
are in total agreement there.  Whenever I see an HP or IBM or Sun or
Bay Networks or Cisco ad, and I *know* that we have either superior or
even competitive products, but I don't see a response from Digital, I
get really frustrated.

But instead of flaming in notes files and making insulting remarks about
"arrogance" and "the customer is too stupid to understand", I go out and
do what I can in my little world.  Can I solve the big problems?  No, but
I can help make my Sales Reps successful, and I can try to boost the
morale and attitudes of my fellow Digits, and I can help spread the word
to at least some customers and VARs that even though Digital doesn't do
fancy advertising, it can help solve their business problems.  So, Mr/Ms
Customer, would you prefer fancy ads in airline magazines, or real 
solutions to your current business problems?

So which do you plan to do, Jon?  Boost your fellow employees morale and
attitudes, or tear them down?  Work to do what you can to solve the 
problems we all acknowledge in your own little area of the world, or whine
about it in notes files?  Find those people in the GMA who can help and
who are as committed to doing the job as you and I are, or throw around
gratuitous insults?

I know what I plan to do, and I know a lot of people in the field and in
the GMA who are equally interested in doing the same thing.

-- Ken Moreau

    RE: .41, .42, and .43
    
    Kerry/Jon/Ken,
    
    First off, let me say thanks to Kerry for your comments.  They were
    some of the clearer points to be brought out in awhile.  I printed them
    off to review them further and likely to plagarize them later.
    
    As for Jon and Ken, I think both of you have the same thoughts around
    marketing, or the lack thereof, in Digital right now.  Jon's are
    tempered by the fact that he has been in the field, like lots of us,
    for some time now and has been taking many arrows from our competitors
    and customers, and doesn't see the cavalry in sight.  His comments tend
    to have a harder edge as opposed to those of someone like Ken.  On the
    other hand, I believe that Ken makes some very valid points.  We are
    all Digital, and small battles won here and there by those of us in the
    field can translate into major victories after awhile.  We need to
    continue pressing for more effective marketing in this company - people
    just might listen, if they are approached rationally.  And at the same
    time, we need to put the best face on to customers that we can,
    insuring we can continue to win some of these little battles.  
    
    Bottom line - continue to press for more effective marketing campaigns
    from Corporate, while continuing to get out the messages to our
    individual customers in the field.  It's a numbers game out there -
    each no or rejection of our points by a customer brings you that much
    closer to one that will say Yes!
    
    Chuck

    Last time I looked, Ken was out here in the trenches with the rest of
    us...
    
    tls

RE: .44 -< You're on the Same Page >-
    
>    As for Jon and Ken, I think both of you have the same thoughts around
>    marketing, or the lack thereof, in Digital right now.  Jon's are
>    tempered by the fact that he has been in the field, like lots of us,
>    for some time now and has been taking many arrows from our competitors
>    and customers, and doesn't see the cavalry in sight.  His comments tend
>    to have a harder edge as opposed to those of someone like Ken.  

Chuck, let me clarify something.  I don't know how long Jon Danzak has
been in the Field, but I have been out here for 7 years.  (thanks tls)
I've won a few and I've lost a few, and the big reason for 2 of those 
huge losses has been Digital's lack of solid, consistent corporate messages 
around it's products.  

Big loss # 1: 3,000 high-end workstation deal, average price per workstation
between $60K and $75K, depending on configuration.  We won the performance
benchmarks, we had the best price, we had 100% compatibility with their
current environment, we ran every one of their applications, we proposed
services to do the swapout, we had a *far* better NT story, and Sun won.
The customer said that the primary reason was that Digital, and especially
Digital UNIX, had no mind-share among the senior executives.

Big loss # 2: Digital owned the manufacturing and testing world of this
major company, 100% OpenVMS.  Average run-rate business of $6M to $10M
per year.  Today the run-rate business is $0, and the reason is that the
customer was convinced that Digital was going to abandon OpenVMS and
leave them with an orphan.

Out here in the trenches it can get pretty lonely sometimes.  My primary
difference of opinion with Jon is how we both react to that loneliness
and feeling that we are getting no help from the GMA.  Jon insults people,
I try to solve the problems in my little world.  Jon makes sure that
everyone sees all of the problems and depresses people, I try to boost the
morale and attitude of my compatriots (and sometimes they help me back
when I am feeling down).  Jon deliberately isolates himself from knowing
even the bare minimum about some of our premier products and chooses to
ridicule those he doesn't care about (Alfer), and I try to know as much
as I can about *all* of Digital's products, so that I can recognize any
opportunities and bring in the right experts to make the sale (often these 
experts are from the GMA, believe it or not :^)).

We both recognize the same issues, but our approach to dealing with this
knowledge is vastly different.

-- Ken Moreau

    Re.: what software do NCs run ?
    
    Digital Semiconductors Digital Network Appliance Reference Design 
    (http://www.digital.com/semiconductor/dna.htm) runs NetBSD and
    hence you can have it execute about anything that you can build
    from source for BSD Unix or Ultrix. We have two of them here and
    they are nice machines.

    My remarks were references to some of the articles mentioned later
    in the thread. As an additional point, you might want to look over
    the site
    
    http://www.chipanalyst.com/mpr/merced
    
    which rather well shows that the caution that a new "native mode" for
    Intel is coming. They'll still execute the x86 mode, but it looks like
    if you want high performance, rewriting your code could be needed.
    Certainly recompilation will. Point customers at this where they need
    to make long term plans and ask "do you want to have a whole new
    migration headache in a couple years?"
    
    This is an area where selling what Digital makes first, telling
    customers the truth, and competing against what Digital does not
    make can go together.
    
    The likely situation is there'll be a "compatibility mode" for Intel
    indefinitely. Wonder if they'll find out, though, that compat mode
    stuff winds up destroying performance of native mode (as I recall
    it doing on early VAXen)?
    
    That's also a potentially useful bit of experience one could share if
    you concur...
    
    

    I never remember any claim that PDP-11 compatibility mode hurt the
    performance of early VAX machines, but then again the early VAX were
    not on a single chip trying to squeeze as much on-chip cache as
    possible.  They were large, multi-board, and microcoded.  Not much
    of a comparison to today's single-chip CPUs.
    
    				-John

    The compat mode slowdown was bruited about DECUS some. The problem was
    that the processor was either in compat mode or not, and compat mode
    code ran slower than native. Therefore a mix of compat and native
    mode stuff would tend to have the compat mode stuff CPU bound,
    which kept the CPU from doing native mode. (This is somebody's
    law as I recall but darn if I remember what name is associated.
    Brooks?)
    
    The effect got really noticeable on the 8600 which still had compat
    mode iron, but whose native mode was much faster.
    
    Back at RCA, some sys managers had a tendency to want to shoot
    people who ran compat mode jobs much interactively...
    
    Glenn Everhart
    

    RE: .46
    
    Ken,
    
    I share your frustration with botched marketing messages, musical
    chairs among the management team, and all the myriad of other problems
    and situations that have caused us to lose business big time.  My
    inference was that Jon shared the same frustrations, and how he shared
    those with the world was perhaps a little "harder edge" than your own.
    
    You have been in the field for 7 years, I believe you stated.  If the
    same Jon Danzak I knew in Syracuse (before the move to Pittsburg), he
    has had 10+ years I would guess.  I have been in field sales with
    Digital for 15 years.  Between the 3 of us we have seen good times and
    bad, so we can pretty effectively comment on historical trends within
    Digital (er, DEC, no darn, DIGITAL.)  I too have lost business big-time
    over the years, just as you have pointed out in your memo.  But we have
    also been winning some business, also, and I see opportunities in the
    NT arena we would have been closed out of, if they had stayed UNIX only. 
    Yes, I can get frustrated at high-priced management that seems to be
    without a clue (some of the time).  But I must have woke up on the right
    side of the bed this morning, since I seem to be in a much more
    positive mood towards our chances of success.
    
    Your points around Jon's style as opposed to your own were
    enlightening.  Personally, while Jon's can be fun to read at times, and
    needed at others, your approach of trying to rationally bring about
    change will likely be more successful in the long-haul.  I appreciate
    all your Notes commentaries, Ken, so don't feel slighted by anything
    said in an earlier Note - I just didn't have your background until you
    stated it.
    
    Chuck    

>    The effect got really noticeable on the 8600 which still had compat
>    mode iron, but whose native mode was much faster.

     (rathole alert) the 86x0 did not have dedicated hardware for
     pdp11 compatibility mode. This was achieved via microcode
     and the only ramification (that I can recall) was flushing
     the istream and too much consumption of control store space.
     

    >    This is somebody's
    >    law as I recall but darn if I remember what name is associated.
    >    Brooks?
    
    The "Brooks' Law" I know is "Adding man-power to an already late
    software project will ownly make it later" (Frederick Brooks, from "The
    Mythical Man-Month), which probably isn't the law you're looking for...
    
    Unless of course is that the reason compat mode was slow was because
    microcode efficiency was sacrificed to meet the delivery deadline...
    ;^) ;^)

    If you want some more FUD, a careful reading of
    http://www.chipanalyst.com/mpr/merced
    
    implies that the processor will take interrupts in x86 mode.
    This is so the system is compatable with old operating systems.
    
    I suspect that will get in the way (as well as the ancient crufty
    interrupt structure used on most x86 systems).
    
    Steveg

re Note 5273.54 by NPSS::GLASER:

>     implies that the processor will take interrupts in x86 mode.
>     This is so the system is compatable with old operating systems.
>     
>     I suspect that will get in the way (as well as the ancient crufty
>     interrupt structure used on most x86 systems).
  
        Sounds like Intel plans to repeat what they did very
        successfully with the transition first to the 286 and then to
        the 386 architectures, i.e., build a chip that can be treated
        as just a faster Pentium by software that hasn't caught up
        with it.

        Bob

    RE: .41, .42, .43, .44, & .51
    
    As someone who has been in the field selling for 16 years -
    all three of you make good points - And I always get something out of
    every note you write, optimistic or pessimistic -
    
    To the issue of marketing - Yes, it is desperately needed - but Bob's
    latest message to the field seems to indicate something will be going
    on shortly in that area - 
    
    What bothers me is the exit of most of the quality sales talent due to 
    the frustration of a broken compensation system - one that has been 
    "broken" since day one - regardless of the software application(s) 
    employed to fix it - 
    
    Sadder still we will not be able to attract quality talent until this
    is fixed - because that talent knows it can go elsewhere "where the
    system works" - so while we are sending messages to the corporation
    please remember this one -
    
    People buy from People they trust -
    This trust must be earned
    The trust must start within the company 
    Only then can it be effectively driven to the customer.
    
    Just my two cents -
    
    
    

>    To the issue of marketing - Yes, it is desperately needed - but Bob's
>    latest message to the field seems to indicate something will be going
>    on shortly in that area - 

I've been hearing that line ever since I joined over 6 years ago.

>    What bothers me is the exit of most of the quality sales talent due to 
>    the frustration of a broken compensation system - one that has been 
>    "broken" since day one - regardless of the software application(s) 
>    employed to fix it - 

Ditto.

PJDM

    re: -1
    
    	   >>>	I've been hearing that line ever since I joined over
    		6 years ago.
    
    If you need some help to get your foot out of your mouth,
    g'us a shout !