[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference 7.286::digital

Title:The Digital way of working
Moderator:QUARK::LIONELON
Created:Fri Feb 14 1986
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:5321
Total number of notes:139771

227.0. "Does DEC leak?" by RDGENG::CORNE (Take 1000 lines - I must do my backups!) Thu Dec 04 1986 15:55

    	(reprinted from DEC Computing, 29-Nov-86 - a UK trade paper)
    
	
    ...
    
    Standing in the car park of a Berkshire pub after a day in the
    wastelands of DECpark is a good time for Astronomy, and naturally,
    it causes the mind to go back to DEC's habit of using astronomical
    nomenclature (try saying that after a night in a Berkshire pub)
    for its development projects.

    Perhaps we should say its former habit, since the latest attempt
    to produce a RISC machine is going under the name PRISM. Perhaps
    that has some connection with Hewlett_Packard's Spectrum name for
    its RISC development work, because Prism has the similar precondition
    that it must, to all intents and purposes, appear to users, software
    and other machines to be a VAX.
    
    Prism is just one of the many ways which DEC is trying to get the
    VAX performance graph extended up into the area occupied by mainfraims
    like the IBM 3090 machines. Most of this centres on multiprocessor
    solutions, like the Andromeda project, which was intended to link
    up to 128 Micro VAXs. According to the latest reports in the US,
    however, Andromeda didn't work out and has been shelved.
    
    The idea of having VMS running across several processors is still
    the favorite for upping the performance level. That doesn't mean
    doing it in a master-slave configuration like they do in the 8300
    and 8800 dual processors, but in proper symmetry across four or
    mor processors.
    
    According to industry watchers at Gartner Securities, the target
    performance for such systems is 44 MIPS, and the VAXBI bus is the
    means of connection.
    
    Down at the MicroVAX level, multi-processors are also on the menu,
    with one project - Firefly - said to involve joining five CPU's
    together. Interestingly, this is happening under Ultrix, not VMS.
    
    But if VAXBI is the means of linking everything together, does that
    mean that the inevitable BI-based MicroVAX is now on the horizon?
    Good question. The answer depends on whether you think a year from
    now can be considered the horizon or not. Either way, thats a timescale
    a DEcperson indicated to a potential BI licencee with whome we fell
    into conversation.

    Another DECperson told us to forget all the things we read and hear
    about a forthcoming MicroVAX III. No such thing is planned, they
    said. Lots more enhancements , though, but we'd left the thumscrews
    in a Compec hospitality suite so no more was forthcoming.
    
    ...
    
    (the article ends with a phone number to ring to share your gossip)
    
T.RTitleUserPersonal
Name
DateLines
227.1Well?RDGENG::CORNETake 1000 lines - I must do my backups!Thu Dec 04 1986 15:579
    Some time back there was a long discussion about the rights and
    wrongs of someone who intervened in a discussion between two DECpeople.
    
    Any one recognise any of .0?
    
    Any one recognise themself?
    
    Jc
    
227.2HYDRA::ECKERTJerry EckertThu Dec 04 1986 16:422
    Sounds like they've been reading "Digital Review"...
    
227.3Name that pub!STAR::MEREWOODRichard, ZKO1-1/D42, DTN 381-1429Thu Dec 04 1986 20:233
    Being an emigrant from Reading, I'll bet I could name it in one.
    
    Richard.
227.4I don't get paranoid about these thingsDECNA::GOLDSTEINNot Insane / Not ResponsibleThu Dec 04 1986 21:1412
    Charlie Matco should sue.
    
    The English article is even a bad copy of "Charlie's" style
    in Digital Review.  He loves to talk about a pub in Bellevue,
    Washington.  Yeah, right.
    
    Some of his stuff is disinformation which people throw at him. 
    Some of it is stuff which we couldn't pre-announce but really wouldn't
    mind if customers were anxious for.  Some of it is real leaks. 
    Outsiders don't know which is which, either.
    
    I think it's a good funny page. 
227.5RDGENG::CORNETake 1000 lines - I must do my backups!Fri Dec 05 1986 08:306
    re .4,
    
    Yes, Its the best page in the mag - its got no job ads worth reading
    yet...;-)
    
    Jc
227.6INK::KALLISSupport Hallowe'enFri Dec 05 1986 12:0827
    Re "Disniformation":
    
    _Nobody_ should be talking to a publication except a designated
    company representative (Personnel Policies and Procedures 6.28).
    "Disinformation" also hurts the company and sours its legitimate
    relations with the press.
    
    Re leaks:
    
    Yes, the company is very leaky.  Some examples of ways information
    can leak out:
    
    Employee discussions in public and semi-public circumstances
    (restaurants, pubs, stores, trade shows, etc) of sensitive matters.
    
    Sales personnel who might wish to keep a wavering customer by releasing
    information about unannounced products.
    
    Co-Op students who might wish to "make points" with their instructors.
    
    People who, for whatever reason, give outsiders "internal distribution"
    documents such as _Sales Update_.
    
    There are _lots_ of ways for information to get out.
    
    Steve Kallis, Jr.
    
227.7COVERT::COVERTJohn CovertFri Dec 05 1986 14:1311
>    _Nobody_ should be talking to a publication except a designated
>    company representative (Personnel Policies and Procedures 6.28).

The policy states that "Any requests for information about a particular
business group should be directed to the appropriate management in that
group."

From time to time, group managers may issue "Party Lines."  This information
is public knowledge, and may be released by any employee to anyone.

/john
227.8Other disclosure sourceSKYLAB::FISHERBurns Fisher 381-1466, ZKO1-1/D42Fri Dec 05 1986 16:557
    re .6:  Another big source of info:  Customers who have been given
    non-disclosure presentations, or who have field test h/w or s/w.
    We will assume that the customers who talk to the Matco's of the
    world are employees who are not aware of the nondisclosure provisions.
    
    Burns
    
227.9INK::KALLISSupport Hallowe'enFri Dec 05 1986 20:0024
    Re .7:
    
    Please note that I didn't specify who the designated company
    representative should be.  However, from Personell Policies and
    Procedures 6.30:
    
    "Every employee is responsible for maintaining the confidentiality
    of information that could jeopardize Digitral's competitive interests
    or patent filings. ..."
    
    "This policy applies to all information presented at seminars,
    conferences, industry standards committees, and any other public
    forum ....  It does not appluy to communications with the press
    (Electronic News, Boston Globe, Computer World, Computer Business
    Daily, etc.)  Press communications must be handled through Digital's
    Public Relations Department."
    
    That seems pretty clear.
    
    I also call your attention to policy 6.06, pages 7 - 8, on ways
    to handle confidentiality.
    
    Steve Kallis, Jr.
    
227.10COVERT::COVERTJohn CovertFri Dec 05 1986 20:2614
re .9

Well, interesting how someone who studies the P&P manual can find a way to
prohibit anything.

It's also interesting that policy 6.28 and 6.30 are so clearly contradictory
of each other, 6.28 stating that general inquiries about the business of the
company must be sent to PR but that specific inquiries about products must
be referred to the project's management, but 6.30 seeming to say that the
PR office has to handle every dealing with the press.

Maybe there's yet another interpretation in this policy mess.

/john
227.11who do you think I had in mind?DELNI::GOLDSTEINNot Insane / Not ResponsibleFri Dec 05 1986 20:422
    Aw, shucks, Steve, I thought you were the "official" Charlie Matco
    fool-him agent, with your Groucho nose and glasses on. 
227.12Easy to Guess!MOSAIC::GOLDBERGMarshall R. Goldberg, PCSGMon Dec 08 1986 02:3610
    
    One of the things that strikes me most about rumors is that some
    are so self-evident as to be rediculous. Any moron knows Digital
    * must * be working on faster VAX computers and new uVAX computers!
    
    It is easy to guess at a number of approaches any good group of
    engineers would probably take and put them into print. That is
    a good way of 'smoking' some additional information out of folks
    who wrongly think the publication has some 'special' knowledge.
    
227.13We don't need investigationsGOBLIN::MCVAYPete McVay, VRO (Telecomm)Mon Dec 08 1986 13:5013
    re: .12

    Not only is the information easy to guess, but it's easy to print as a
    guess.  The Charlie-Matco-clone talked to some DECcies around a pub,
    which is hardly an authoritative source.  Anyone who reads that article
    and thinks they're getting inside information also must pick stock
    quotes from the "National Enquirer" astrology pages.

    Before we go off looking for leaks, why not make sure they are leaks.
    Julius and Ethel Rosenburg were executed for supposedly passing
    information which later turned out to be (1) useless for building
    a bomb and (2) common knowledge that any competent nuclear scientist
    already knew.
227.14INK::KALLISSupport Hallowe'enMon Dec 08 1986 14:2930
    Re .10:
    
    Maybe it's not so contradictory as one might think.  Suppose there's
    an inquiry about a specific product (say, my favorite mythical PDP-13).
   
     1) Suppose it's unannounced.
    
    If that's the case, and someone wants information about it (presuming
    anyone who's non-Digital) one would hope the management would be
    sensible enough to say something on the order of, "Sorry, no such
    product's been announced."
    
     2) Suppose it's been announced (officially, as in press-released).
    
    Then management, usually through a Marketing Communication Manager
    can refer it to the appropriate proper channel: press inquiries,
    to a Publioc Relations representative; consultant inquiries, to
    Consultant Relations; analysts' inquiries to Financial Relations;
    and customer inquiries to the appropriate sales/service group.
    
     2a) Suppose it's "announced" (i.e., appears in Sales Update or
    equivalent, but hasn't been exposed to the public yet).
    
    If a press inquiry (tracking down a rumor), the best choice is the
    Public Relations department.  Otherwise, as 1.
    
    What's so "contradictory" about that?
    
    Steve Kallis, Jr.
    
227.15Firefly is public knowledgeMOLE::BARKERJeremy Barker - NAC Europe - REO2-G/K3Fri Dec 19 1986 22:165
The existence of the Firefly multiprocessor mentioned in the quoted article
is public knowledge.  I read about how its cacheing system works in an ACM
publication.

jb