Conference 7.286::digital

    This may not be what you want, but here goes:
    
    I used to work for General Electric in Salem, Virginia.  Salem is
    a small town next to Roanoke, which is in the western part of the
    state, between the Blue Ridge and the Alleganies.  GE's plant there
    was Drive Systems: they made motor controls and so on.
    
    While the plant employed two thousand people (about like Spit Brook),
    most of the people working there were on the factory floor, in blue
    overalls.  There were maybe a hundred hardware people, and only
    a few software people--I was one of eight.  Of course, most of the
    hardware people were writing software anyway.

    This job had its good points: the location could hardly be beat,
    as I could look at the Blue Ridge on my way to work and the Alleganies
    on the way home.  I got to see a much more complete picture of what
    I was working on, as I worked with the hardware designers as well.
    And the cost of living was low (houses were about one third the
    current prices in Southern New Hampshire when I lived there in 1980).
    The weather was good, and winters were short.
    
    Other things were not so good:  I couldn't wear jeans.  I had to
    be there from 8 in the morning to 5 in the evening.   Roanoke is
    hardly a cultural center.
    
    Anyway, I wound up getting a job in New England.  DEC is different--
    the best characterization I have for the difference is that working
    in Spit Brook is more like being in graduate school than it is like
    working for GE.
    					-John Bishop

    Digital promotes lateral mobility.  

    Digital is such a large and multi-disciplined company that a person
    can have any career he wants and never leave the company to do it.
    
    My anecdote:  I was at Data General in the early 70s, working as
    a system tech in the Computer Special Systems group.  I tried without
    success to transfer from CSS to field service, even to the point
    of talking to the VP of both organizations with whom I was on a
    first-name basis with.  He said "not now, maybe later."  I went
    to work in field service for a competitor.  
    

    I worked for a competitor between times working for DEC. That is
    to say, I left DEC to work for Perkin-Elmer (now Concurrent Computer),
    and then left there to come back. It's safe for you to assume I
    like it at DEC a lot better.
    
    A few anecdotes. When I worked for DEC in NYC it was a suit and
    tie office. A sport jacket is considered casual in the financial
    district. When I went to Maynard for training or what ever it
    was very casual dress (as it is today). When I worked for P-E in
    Boston it was still pretty dress up. I usually wore three piece
    suits. The first time P-E sent me to training at their HQ I was
    all set with casual clothes. Everyone, even the people who dressed
    casual in their field office had a suit and tie on. I was shocked.
    Seems at P-E one dresses up to go to HQ.
    
    Second thing is that the computer division of P-E is one of a number
    of large divisions. The other divisions build things totally unrelated
    to computers (other then machines to build chips). These divisions
    also have market share that would make IBM blush. 90-98% share were
    quoted for a number of their products. Obviously with that kind
    of market share and product that was so good that no one if their
    right mind would buy elsewhere if they could afford yours, advertising
    was a foreign concept to the corporate parent. If you think product
    managers at DEC have trouble getting their products advertised you'd
    never understand P-E.
    
    Computer division people were second class people because the corporate
    parent never understood them. They also never understood why they
    didn't have the best product on the market. The other divisions,
    including the instruments division who continued to buy VAXs, did
    but the message never seemed to get to the top.
    
    Also the emphasis was on hardware. No one really seemed to care
    that the OS made RT-11 look sophisticated and RSX look friendly.
    The standard story was about a h/w s/w horse and pony show. The
    h/w guy was making the first presentation and was periodically
    interrupted with questions. A common answer was that that would
    be covered when they talked about s/w. Finally at the end of the
    h/w talk one of the customers to be said, "That's great! Why doesn't
    everyone buy P-E computers?" The voice from the back of the room
    called out, "We'll cover that when we talk about s/w." Supposed
    to be a true story. Getting fed up with working with and selling
    garbage s/w is a big part of the reason I came back to DEC.
    
    One small note in P-E favor. The average field office had 2-5 software
    support people who shared a roughly 11/780 (or larger) class machine.
    This was at a time when the 11/780 was as big a VAX as DEC sold.
    How many DEC field offices have a 11/780 per every 3 s/w support
    people even today?
    
    			Alfred

    I worked for Sperry (1100 Series) for 8 years and left Unisys.
    
    Sperry was built up from lots of acquisitions that never totally
    felt like one company, even after 35 years.  As a result,
    your part of the large company was likely to be much smaller.  This
    also meant (senseless) competition and finger-pointing between groups.
    
    Computers for internal use were much more difficult to come by.
    That's probably because, physically, an 8800 would be a *small*
    machine there.  Personal workstations were unheard of (though most
    people had roughly VT100-class terminals), and mainframes were often
    overloaded.  (I understand it was much worse in the Defense Systems
    Division, with keypunches still common as of about a year ago.)
    PC clones were becoming common, though.
    
    Cross-group communication was not as good; the equivalent of the
    "5x5" meeting was unheard of.  There were more layers of management
    for equivalent numbers of people.  There was almost no direct contact
    between customers (or marketing) and engineering.
    
    In general, the press didn't much care what Sperry did (especially
    in software); I think in 8 years I saw maybe 3 articles quoting
    people I knew.  It was frustrating to see Sperry's supercomputer
    largely ignored, since it was (is) a nice system.
    
    Under the better managers, the development environment did have
    the same surface casualness as DEC.

    
    I worked for a non-competitor but found that computer horsepower
    was allways scarce when 'timesharing systems' were used.  This seems
    to be the case still.  It is very hard for most groups in any company
    to justify buying a $150,000 peice of capital equipment for a few
    people..  We worked on an overloaded 11/34 running RSX, we then
    grew to a BIG vax (11/750) with the power of a microvax, it too
    was very ovrloaded.
    
    When we started designing systems based on 11/23-73 computers we
    got the luxury that DEC engineering groups have.. We could go to
    the stock room and just sign out a system..  We all had our own
    computer at that point...
    
    The MicroVAX 2000 series computers starting @4.6K $$ should help
    the common workers (with low $$ signoff for equipmet) get some really
    usefull equipment..

RE: 393.4

Jon, Could you explain the "5x5" meeting?

>    Cross-group communication was not as good; the equivalent of the
>    "5x5" meeting was unheard of.  There were more layers of management

Sam

    What about the pay in other companies?  How was the base pay and raises
    in comparison to DEC?
    
    

    I worked in Honeywell's minicomputer group as a technical writer for
    ten years before coming to Digital.  Things went downhill sharply over
    that period; I hope there was no connection!  Seriously, that period
    corresponds to Honeywell's marketplace defeat on the mainframe front
    (losing to IBM) and the minicomputer front (losing to Digital et al).
    What follows are strictly my own recollections and opinions. 
	
    When I started at Honeywell in 1976, it was a multibillion- dollar
    conglomerate, and one of the top computer vendors (our slogan was "The
    Other Computer Company"), while Digital was just a smaller niche-market
    competitor.  Now, Honeywell is formally leaving the computer business,
    and Digital is Number 3 with a bullet.  Losing is certainly a different
    perspective from what DECcies have known over the past few years. 

    Honeywell Information Systems Division (1955-1987) was always a
    troubled stepchild of Honeywell, Incorporated.  Created through
    acquisitions of castoffs from Raytheon, General Electric, and Xerox, we
    struggled to integrate product lines while attempting to compete across
    the board with IBM.  The corporate offices in Minneapolis never quite
    knew what to do with us.  Honeywell Control Systems has always been the
    market leader; Honeywell Aerospace and Defense is strongly competitive
    in its markets. Honeywell Information Systems barely broke even. 

    Our perspective in the minicomputer group was that we had very good
    hardware, but not enough good software.  We were in the business
    market, so we did things like COBOL, transaction processing, and
    database well.  Digital had the scientific market, and we didn't
    interfere.  We thought we could take over Wang's share of the
    office-automation market, because we could integrate DP and WP better.
    I think that strategy is having mixed success today.  It seemed the
    sales force always found an excuse to sell a mainframe, for a fat
    commission, instead of a mini, for a relatively puny commission.
    Overall, Honeywell's market share stagnated, while Digital's share (for
    example) skyrocketed.  After literally years of rumored pullouts,
    Honeywell pulled the plug last year and sold most of Information
    Systems to NEC and Group Bull. 

    There is a sharp difference in corporate culture between Honeywell
    Information Systems and Digital.  The tangible differences between
    market failure and market success include layoffs, accelerating from
    unexpected shocks to yearly prunings; cancelled employee events (no
    more Family Day at Canobie Lake Park); real hiring freezes, salary
    freezes, and salary-action pushouts; and a constant stream of negative
    comments from the press.  It bred a frustration that there weren't
    enough resources to keep up, a frantic pace as we released products on
    a quickening pace to maintain the illusion that we were viable, and an
    attendant cynicism about what management was doing. 

    Honeywell is over a century old, and some of its managers seem to have
    been founders.  As is the case with older companies, Honeywell is
    top-heavy with managers.  This hurts productivity. Imagine one section
    head for every six workers, one manager for every three section heads,
    one director for every three managers, one vice-president for every
    three directors, one group vice-president for every three
    vice-presidents, and so on up.  With less than 40,000 employees there,
    I had more layers of management between me and the chairman of the
    board than I now have between me and Ken Olson in a company of over
    100,000 employees. 
	
    Having too many managers also clogs action.  Everything has to be
    studied, approved, and implemented.  Sometimes this is superior to the
    Digital style of mass consensus, but sometimes it isn't.  The best
    example is the simplest.  Once upon a time, the mini leader was not
    Digital but--you guessed it--Honeywell. This happened when we acquired
    Computer Controls Corporation, and the rights to its products, in the
    mid-1960's.  Management considered the potential market for
    minicomputers, decided there was none, and continued to sell IBM
    1401-busting OS/2000 systems. 
	
    Finally, management overpopulation promotes savage political
    infighting, as the managers, like oversexed lemmings, competed for
    their share of the scarce turf.  Sometimes managers would infuriate you
    because you knew they were just professional managers, and technically
    incompetent; they would do an equally bad job managing factories,
    people, or inventories.  The Digital credo, "Do the right thing," is
    always thrust at you, and people genuinely try to adhere to it.  There
    was nothing of the sort at Honeywell (probably there's nothing of the
    sort at most companies, to be fair); often, the politically correct
    action was taken.  At least it was acknowledged; sometimes as Digital,
    the same things seem to happen, but everyone denies it. 

    I touched on resource problems before.  You probably think you have it
    tough where you are.  Let me assure you that things could be worse.
    Digital is an engineer's Nirvana, where you can get two or even three
    terminals/workstations/computers in your office.  That is very rare.
    Digital is turning away applicants, even as a salary laggard.
    Honeywell suffered (and still suffers) from a hemmorhage of talent,
    even though it of necessity pays well.  Advertise for a programmer who
    knows VMS, and you'll get a lot of applicants; advertise for a
    programmer who knows MOD 400, and twiddle your thumbs.  It's that way
    for technical writers, support people, and testers, too; that's what
    market share does to you. 
	
    Having so many resources, and having a matrix management style, Digital
    seems to practice a strange (to me) sort of internal competition to
    bring new products to market.  It's safe to assume that there are at
    least two groups at Digital working on new offerings in every product
    line and market segment!  Only the strong survive. At Honeywell, we
    simply didn't have the resources to work that way.  I still don't know
    if the practice is wasteful or not. 

    I don't think the people at Digital are superior in any way to the
    people who were at Honeywell.  In fact, some Digital people seem both
    parochial in their viewpoint and spoiled by the company's success.
    There's an attitude that "we're the best company and the best workers."
    It's understandable; it's just not true. 

    re .6:
    
    The "5x5" (read five by five) is a term used to describe a
    cross-functional team in product development. (Depending on group
    this can be anywhere from a 3x3 to a 6x6 or more, but the concept
    is the same.) I believe that 5 members is now the standard. In this
    case, the members represent:
    	- engineering
    	- product management/marketing/finance
    	- manufacturing
    	- service
    	- quality
    Generally the 5x5 team meets regularly and takes operational decisions
    representing all the functions. The 5x5 usually is the driver of
    the phase reviews as well.

re Note 393.8 by DELNI::JONG:

>     It's safe to assume that there are at
>     least two groups at Digital working on new offerings in every product
>     line and market segment!  Only the strong survive. 

I'm afraid that there are people in this company who think that this is a very
bad thing.  Many organization changes through the past 5 or so years have been
motivated, at least in part, by a desire to eliminate "duplication of effort"
and, in particular, overlapping products.

But I feel that it is one of our strengths.

Bob

    Re: .10:                              
    
    That: having two separate groups working on similar product offerings
    is a strength not a weakness --
    
    I can think of two examples to illustrate this: one is often called
    'engineering redundancy.' As with a navy warship that has two power
    plants complete with its own drive shaft. When one goes out of action,
    the other automatically comes on line. 
    
    The other example comes from a method of reducing data entry errors.
    The same data is entered in two separate operations in separate
    databases. The database are then electronically cross-checked. The
    technique is based on the unlikelyhood of two different fumble-fingered
    typists making exactly the same error in exactly the same place. 
    
    In a broad sense the same principles can apply to any working group.

    
    re: .11:
    
    Both of the aforementioned examples require some level of inter-
    communication and a common goal, which we have historically lacked.
    The development groups tend to compete against each other instead
    of cooperating and sharing.
    

    
    I don't think that two groups who may be engineering similar products
    need to be in contact with each other.  There may be others on high,
    associated with neither group, who are watching the two efforts
    to see how they pan out.   Having a common goal could mean
    heading in the same wrong direction.  This doesn't necessarily
    mean that the decisions made by higher ups are always right - they're
    people too.  But personally I feel better knowing that we've
    got a couple of extra aces up our sleeve (read that: under development).
    
    

Isn't it a case of having the right amount of duplication and
the right kinds of communication and co-ordination?

I don't see that redunancy is always good, nor always, bad, nor
that communications are always required nor never required.

It depends on the market, the products, the people, and other variables.

Sometimes ZERO groups working on the product is the right answer,
especially if it's the wrong product.

- greg

    Duplication of effort in an engineering sense is an interesting
    topic in itself, meriting a separate topic by itself. Keeping with
    the spirit of the first few replies, could we get back to hearing
    what it's like working for Digital as opposed to working for some
    one else please! We haven't heard from any ex-IBM'ers. Is it because
    they don't read this notesfile or they don't see any differences
    at all ;-).
    
    Nari

>We haven't heard from any ex-IBM'ers. Is it because
>they don't read this notesfile or they don't see any differences
>at all ;-).

    .. OR :
    
    .. no Notes option in their DECmail menus ?
    
    .. EVE/EDT is too intimidating ?

    .. can't differentiate between WRITE and REPLY ?
    
    .. won't contribute if they have to unlock the CAP LOCK key
    
    km    

    I worked at Wang my first 5 years out of school, then came here
    last June.
    
    I think the "duplication of effort" issue is a key to some of the
    differences between Wang and DEC. At Wang, the decision making was
    hghly centralized, (at least it appeared to be), and there was little
    duplication of effort. However, there was also very little flexibility
    and slow reactions to rapidly changing business conditions. Here,
    there is much more decision making, responsibility, and accountability
    at lower levels, and thus, there seems to be more flexibility and
    faster reaction. 
    
    It's kind of like a centrally planned economy vs. a democratic society.
    There is more duplication of effort in a democracy, but the increase
    in effort required to survive in a democracy on an individual basis
    more than makes up for it.
    
    Plus it's more fun.
    
    Just an opinion, not a theory :-)
    GREG