Conference 7.286::digital

    Paul Nelson (MRO 1-2) is in charge of keyboards.
    
    							Ann B.

    or...
    
    alot of new devices are coming out of projects in virtual reality
    dealing with multisensory technology (mainly from other vendors).  The 
    PRESENCE notes conference that has alot of discussions about new devices 
    and the like that may interest you.  

        Re:                    <<< Note 2876.0 by TINCUP::VENTURELLA >>>

        Talk to Jon Mysel (MYJAM::MYSEL) in the Corporate Design Group.
        
        Anker

    Some thought in my opinion needs to be put into whether Digital wants
    to stay progressive with new devices.   As of late I have not seen
    Digital really caring about VR market place for example.  So far I do 
    not believe that Digital has noticed a market place with new human 
    interfaces starting to develope.  It is starting to develope...   
     
    5 year from now will there be a marketplace for devices like data
    gloves? head tracking, eye tracking and the like?    Will Digital
    be a part of it?  (I.E) will there be $$$$$$ to make? 
    
    Paper tape went the way of cards and cards when the way of C.R.T.'s. 
    Will Digital find profit to make with new devices?   
    
    We have an excellent start from the CPU side with the Alpha - what's
    next?   

    At one time about 3-4 years back, the VIPPS group was investigating
    something called a desktop bus. The idea was to do something like
    what Apple was doing, in that a daisy-chained serial bus (ala RS232)
    could be used to connect "n" number of pointer/input devices. Since
    the user could only utilize 1 device at a time, there could be a
    mouse, trackball, touchpad etc all connected and presumably selected
    at the users preference. I can't remember the architect's name, but
    I believe it had been specified, proto-typed and demonstrated. This
    would have allowed so much more marketability to our workstations,
    I find it amazing that the desktop bus never became a product. After
    all, we make beans on pointer devices, and I thought the idea was to
    sell Alphas ? Even if we were to just make our pointer bus port spec
    public, it would allow the third parties out there to develop an
    infinet number of input devices for our platforms. When I was in that
    space, I know we lost a ton of process control business. Not only
    for our closed input busses (keyboard, pointer) but also our lack of
    re-packaging flexibility. For example Sun utilizes a number of partners
    who can repackage workstations, monitors, etc, etc. I was fighting
    this battle when DEC (err Digital) was riding that big wave in the
    decadent 80's. But now that we're starving for business, we still
    don't get it. Anyway .0, you may want to explore the Video and printers
    group. I believe that Paul Nelson was the same person I dealt with
    at the time. Good Luck !
    

Re: the last few

Paul Nelson (REGENT::NELSON) is still doing engineering management for
keyboards.

I was the architect (along with Chris Cuenod) for the desktop bus
("ACCESS.bus") referred to in .5

- ACCESS.bus is a product, it is used on the Personal DECstation 5000 (MAXINE).
  Digital has sold some 10K of these systems (you can order an LK501 today).

  ACCESS.bus is open to the public:

- The ACCESS.bus Industry Group (ABIG) is alive and well with over 40
  member companies including Microsoft, SUN, HP, Digital, IBM, Compaq,
  Apple, Phillips, Logitech, Honeywell, etc...

  You can buy A.b keyboards, mice, and an interface card for PC clones
  from Computer Access Technology.  In one demo, you can have 4 mice being
  used simultaneously in a multi-player game.  Players can be added
  by hot plugging additional pointing devices without restarting the
  system or application.

- ACCESS.bus appeared at the recent COMDEX where it was described by some
  as 'best of show'.  Numerous articles have appeard in the trade press.

- Microsoft will soon formally announce their intention to support
  ACCESS.bus in Chicago as part of their "Plug-and-Play" initiative.

ACCESS.bus hardware was designed into the first Alpha Workstations (Flamingo)
but was later removed because the software people at the time were too busy
with other priorities to support it.

There is still time for Digital to be one of the first companies to
widely support ACCESS.bus in its systems.

There's a somewhat outdated description of ACCESS.bus in the
Digital Technical Journal, volume 3 number 4 published Fall 1991.

ACCESS.bus may yet become an industry standard.  It has just taken
a long time, and Digital was too strapped by other problems to drive it.

Feel free to contact me directly for more information - Peter

    RE: .6
    Its nice to see that ACCESS.bus became a reality, but its also
    too bad that the workstations group is "too busy" to implement it
    on our Alpha platforms. Having just spent the last 3 years in
    workstations quality assurance, I can tell you that I cannot
    remember ever seeing or hearing any mention of this capability.
    On the one hand they lamment the fact that Alpha is not gaining
    market share, yet on the other hand it seems nobody cares about
    implementing any inovative change into operator controls. Since
    the Flamingo I/Sandpiper I products, they have productized
    Flamingo II, Sandpiper +, Pelican, Hotpink, and Jensen. They
    apparently are REALLY busy, because none of those platforms
    implement the ACCESS.bus. Peter, do you know what the REAL reason
    is ? Surely we need to consider other features beyond just pure
    speed if we want to succeed in the workstation market. I hope
    someone wakes up soon.

>   Peter, do you know what the REAL reason is ?

At the time Flamingo I was being developed, that was the reason given.
The software folks had their hands full just porting VMS and ancilliary stuff
without having to redo the keyboard driver.  [The driver changes were
implemented successfully for ULTRIX on MAXINE]

Beyond this I can speculate:

  There was some risk aversion.

  Cost is another sticking point.  Keyboard cost is determined primarily
  by volume and secondarily by technology.  Workstations are inherently
  a low volume product, so in order to get a low cost keyboard, they
  have to piggy back on a high volume product such as VTs or PCs.
  A.b technology is comparable in cost to other keyboards, but the
  volumes are not there yet.  Some workstation folks probably reasoned,
  we'll wait and do what the VT folks do.  The VT folks are in a very
  cost competitive business and have less need for multiple pointing devices
  so reasoned, we'll adopt A.b when it becomes a low cost off-the-shelf
  industry standard.  So the group behind A.b set about getting the PC
  industry to adopt it.  The technology is elegant, a generation ahead of
  ADB (Apple Desktop Bus), but who will be first and risk offering an
  initially higher cost non-industry standard keyboard?  It has taken 2+
  years to build momentum and pull together a group of industry leading
  vendors (the ACCESS.bus Industry Group "ABIG") who can do it together.

The question is:  Do we have the vision and willingness to innovate?

Where and under what conditions?

  [Alpha is fighting the last war, to build
   a better CPU like the PDP-11 and VAX before it]

- Peter