Unisys ? What is Unisys?
Peter Niepel
I'll move to New Zealand and there are running a lot of Unisys. I'm a
developer for COBOL and this stuff on MVS and I wonder if there are some
informations available about Unisys for someone like me, without having the
faintest idea about it.
I can read and I'm able to browse the internet. Therefore, some advices
about books and websites would be very useful.
Thanks a lot
Peter
Peter Niepel
Colin Zealley
I guess a good place to start is the Unisys website -
http://www.unisys.com/.
Unisys offers wide-ranging services, and makes mainframes (two different
architectures, both very different from the S/390 architecture) and also
high-power Intel-based NT servers (and can mostly also run Unixware). And a
bunch of other stuff, including some very powerful application generation
tools.
Check the website out - I'm sure other folks here will give you some
other pointers.
Regards
Colin
Peter Niepel wrote in message <7s5svd$7om$1...@newsread.do.de.uu.net>...
Richard
I work for the Information Services side of Unisys so if you have any
specific questions mail me at richar...@gb.unisys.com and I'll do my
best to get you an answer.
Cheers - Rich
--
Richard Haigh
UNISYS - Print/Publishing Consultant
Richard C. Steiner
>I'll move to New Zealand and there are running a lot of Unisys. I'm a
>developer for COBOL and this stuff on MVS and I wonder if there are some
>informations available about Unisys for someone like me, without having the
>faintest idea about it.
Unisys was formed by the merger of two older mainframe makers: Sperry Rand
(makers of Sperry Univac 1100-series mainframes), and Burroughs (makers of
Burroughs B-series mainframes). Both mainframe lines are still maintained
and sold to their existing customer bases (renamed to Unisys 2200-series
and Unisys A-series respectively).
Unisys also made a/o sold PCs at one point (I'm not sure if they still do
it) and Unix servers, and they seem heavily into larger NT servers now as
well.
--
-Rich Steiner >>>---> rste...@visi.com >>>---> Bloomington, MN
Written online using slrn 0.9.5.4!
The Theorem Theorem: If If, Then Then.
smyers
> >I'll move to New Zealand and there are running a lot of Unisys. I'm a
> >developer for COBOL and this stuff on MVS and I wonder if there are some
> >informations available about Unisys for someone like me, without having
the
> >faintest idea about it.
>
> Unisys was formed by the merger of two older mainframe makers: Sperry Rand
> (makers of Sperry Univac 1100-series mainframes), and Burroughs (makers of
> Burroughs B-series mainframes). Both mainframe lines are still maintained
> and sold to their existing customer bases (renamed to Unisys 2200-series
> and Unisys A-series respectively).
>
> Unisys also made a/o sold PCs at one point (I'm not sure if they still do
> it) and Unix servers, and they seem heavily into larger NT servers now as
> well.
Left out the V Series (Medium Systems Architecture), BTOS Platforms, S/3,
......... there was a lot more history than just the 1100 and A Series.
Steve Myers
do...@77.usenet.us.com
: I'll move to New Zealand and there are running a lot of Unisys. I'm a
: developer for COBOL and this stuff on MVS and I wonder if there are some
: informations available about Unisys for someone like me, without having the
: faintest idea about it.
There has historically been a lot more open Unisys activity in Australia. I
suppose that would spill over to New Zealand as well.
Unisys Australia seemed more willing to accomodate customer variations
than in the US, where one was expected to be a Unisys-only shop, full
of Unisys-supplied accessories.
http://www.unisys.co.nz/
http://www.unisys.co.nz/cube/cubewho.htm
--
---
Clarence A Dold - do...@network.rahul.net
- Pope Valley & Napa CA.
Luko Willms
schrieb smy...@nospam.hotmail.com (smyers)
auf /COMP/SYS/UNISYS
in 1SWF3.6933$LK5.2...@typhoon2.kc.rr.com
ueber Re: Unisys ? What is Unisys?
s> Left out the V Series (Medium Systems Architecture), BTOS Platforms,
s> S/3, ......... there was a lot more history than just the 1100 and A
s> Series.
Sure, but 1100 and BLS (A-Series) are not history, but present.
Lüko Willms
/--------- L.WI...@LINK-F.frankfurt.org -- Alle Rechte vorbehalten --
"Die Interessen der Nation lassen sich nicht anders formulieren als unter
dem Gesichtspunkt der herrschenden Klasse oder der Klasse, die die
Herrschaft anstrebt." - Leo Trotzki (27. Januar 1932)
Richard C. Steiner
>Left out the V Series (Medium Systems Architecture), BTOS Platforms, S/3,
>......... there was a lot more history than just the 1100 and A Series.
I can only mention what I'm familiar with. :-)
Bill McDonald
Richard C. Steiner <rste...@isis.visi.com> wrote in message
news:slrn7ui73u....@isis.visi.com...
hans rutgers
UNISYS is one of the mainframe companies that still exist despite their
management and sales forces.
If you are a COBOL programmaer, then you should not worry since the UNISYS
COBOL is 100% or better compatible with IBM cobol.
JCL is different but very easy to learn.
(I am speaking for the 2200)
Its a top quality product (old fashioned ok ok ) as you will find out when
using them.
Try it out, and be welcome to send me your questions
h a n s
In article <7s5r0n$56r$1...@newsread.do.de.uu.net>,
"Peter Niepel" <PeNi...@niepel.ka.uunet.de> wrote:
>Hi,
>
>I'll move to New Zealand and there are running a lot of Unisys. I'm a
>developer for COBOL and this stuff on MVS and I wonder if there are some
>informations available about Unisys for someone like me, without having the
>faintest idea about it.
>
Malcolm Ball
<rste...@isis.visi.com> writes
>
>and sold to their existing customer bases (renamed to Unisys 2200-series
>and Unisys A-series respectively).
Now both lines are named ClearPath
>Unisys also made a/o sold PCs at one point (I'm not sure if they still do
>it) and Unix servers, and they seem heavily into larger NT servers now as
>well.
HP make the desktops and small servers, Unisys I believe still make the
larger servers.
--
Malcolm Ball
Colin Zealley
Malcolm Ball wrote in message ...
>>Unisys also made a/o sold PCs at one point (I'm not sure if they still do
>>it) and Unix servers, and they seem heavily into larger NT servers now as
>>well.
>HP make the desktops and small servers, Unisys I believe still make the
>larger servers.
manufacturing of larger servers ourselves, currently marketed under the
Aquanta brand name.
Richard C. Steiner
>And, pray tell, how did that differ from an IBM site?.
I don't understand your comment in the context of this thread.
Marc Wilson
|If you are a COBOL programmaer, then you should not worry since the UNISYS
|COBOL is 100% or better compatible with IBM cobol.
110% compatible? Mmmmm.
|
|JCL is different but very easy to learn.
|
|(I am speaking for the 2200)
|
|Its a top quality product (old fashioned ok ok ) as you will find out when
|using them.
If he's moving from an IBM environment, it ain't that old-fashioned! I had
to use MVS[1] once, on a Teradata course- sheesh! A 20-line essay, just to
assign a file! It was the most user-hostile environment I've seen since
pure batch with punch cards.
[1] or was it VMS, or some other anagram?
--
Marc Wilson
Cleopatra Consultants Limited - IT Consultants - MAPPER Associates
2, Langham Grove, Timperley, Altrincham, Cheshire WA15 6DU UK
Tel: (44/0) 161 973 1580 Mobile: (44/0) 70-500-15051
Mail: in...@cleopatra.co.uk Fax: (44/0) 870-164-0054
___________________________________________________________________
MAPPER User Group mailing list: send *SUBSCRIBE to M...@cleopatra.co.uk
Jim McMahon
>If he's moving from an IBM environment, it ain't that old-fashioned! I had
>to use MVS[1] once, on a Teradata course- sheesh! A 20-line essay, just to
>assign a file! It was the most user-hostile environment I've seen since
>pure batch with punch cards.
>[1] or was it VMS, or some other anagram?
>--
>Marc Wilson
MVS = Multiple Virtual Storage = IBM Mainframe (S370)
VAX/VMS = Virtual Adressing Extension/Virtual Management System = DEC
Being ordinary and nothing special is a full-time job.
mcma...@flash.net (Jim McMahon in real life)
Mark Perew
>> Unisys was formed by the merger of two older mainframe makers: Sperry Rand
>> (makers of Sperry Univac 1100-series mainframes), and Burroughs (makers of
>> Burroughs B-series mainframes). Both mainframe lines are still maintained
>> and Unisys A-series respectively).
>>
>> it) and Unix servers, and they seem heavily into larger NT servers now as
>> well.
>
Don't forget the noble B1000 (aka Small Systems) line.
Ian Joyner
I read all the replies and they weren't particularly helpful. In NZ you
will find mostly A Series because NZ was dominated by Burroughs at one
stage (a very enlightened country computing wise).
Burroughs brought to mainframes an ease of use and operations in the
same way that Apple (via Xerox) brought to personal computing. Of course
in any environment, there is a lot to learn and you might hate it at
first, but getting to know A Series is a very fulfilling experience.
Great machines, shame about the management and sales prevention force.
Hardware wise, you will find the A Series is a stack based machine.
Because there are no programmer accessible registers this makes them a
very scalable machine, from the low end that runs emulated on Intel
processors (Isn't it time Unisys did one on the new PowerPC G4
processors? Well perhaps the need isn't quite realized yet, but it will
be!). The high end are single chip processors, but still very powerful machines.
Software wise, there is no assembler. All systems software is written in
derivatives of ALGOL. (Did I mention there is even an Eiffel compiler around;-))
Then of course you have COBOL for all your commercial programming.
For databases you have DMSII with an SQL frontend if needs be.
Equivalents of things like CICS are your MCSs (Message Control Systems).
The one you will use for commercial transaction processing is COMS, and
this is probably the best TP Monitor ever devised.
For program development you will use the CANDE MCS (Command and Edit).
You can run editor on top of this, which is a really nifty editor, which
reads compiler generated files to help you browse through your programs.
This is maybe the best editor in concept that I have used, but the front
end was slightly old fashioned and the real draw back that stopped a lot
of shops using it was the fact that it fired up one process per every
person using it. I think these problems have been fixed lately.
So, you will most likely be moving into one of the nicest programming
environments in computing, once you understand it.
Obviously, you can look up http://www.unisys.com
but even better, you have Gregory's Publishing and independent organizaton:
They have many books and journals that will help.
Alas, there is nothing you can pick up in a bookshop.
Hope this helps.
--
Ian Joyner
i.jo...@acm.org
http://homepages.tig.com.au/~ijoyner/
Objects Unencapsulated
http://www.prenhall.com/allbooks/ptr_0130142697.html
Pete Mullins
Bill McDonald <wmcd...@potomacsys.com> wrote in message
news:7sb6vf$qkt$1...@winter.news.rcn.net...
> And, pray tell, how did that differ from an IBM site?.
Well, for one thing, IBM isn't only about history. Unlike V Series, CTOS,
S/3, B1000's, etc.
> news:slrn7ui73u....@isis.visi.com...
> > In article <1SWF3.6933$LK5.2...@typhoon2.kc.rr.com>, smyers wrote:
> >
> > >Left out the V Series (Medium Systems Architecture), BTOS Platforms,
S/3,
> > >......... there was a lot more history than just the 1100 and A Series.
> >
> > I can only mention what I'm familiar with. :-)
> >
do...@77.usenet.us.com
: I read all the replies and they weren't particularly helpful.
I'll be sure to send replies to you for value checking prior to posting.
ShivkuMAR Visvanathan
confusing for a beginner.
Richard C. Steiner
>I read all the replies and they weren't particularly helpful. In NZ you
>will find mostly A Series because NZ was dominated by Burroughs at one
>stage (a very enlightened country computing wise).
This, by itself, is more information about the situation than most of us
had to work with. I do 2200-series stuff in the states, man. :-) But I
honestly believed that a few general statements about the origins of the
two main Unisys mainframe liness might be of some interest to him.
Sorry you disagree. <shrug>
Chad A. Fernandez
what they are, or where they originated. I just get by with used Unisys PC hardware :-)
Chad A. Fernandez
Michigan, USA
Peter Niepel
matter who gives the best information at all, because all information helps.
As I wrote, I will move to New Zealand at 2000-02-05 and the chances are
good, that I'll work with a Unisys System, because I hope that I could stay
with my current employer (EDS). They use a Unisys and I'll try to get the
information,what sort of it they use. I think when I'll sitting in front of
it, I'll have enough questions to ask.
I'll keep contact with some of you if you don't mind.
Thanks a lot.
Yours
Peter
Don Macpherson
I guess comparisons are in order, since both companies have some
history.
Unisys
CTOS - An Intel-based microcomputer OS. (migration path to NT)
IBM
OS/2 - An Intel-based microcomputer OS. (where to?, its dwindling
market presence (0.5% of 1998 sales according to IDC) is hardly reassuring)
What about IBM S/34 and S/36 systems?
Are they still around or have they been "retired" like the S/3, B1000
and V Series systems?
Don Macpherson
"Pete Mullins" <pmu...@ibm.net> wrote in message
news:37eb...@news2.prserv.net...
Luko Willms
schrieb nearly_donmac@almost_erols.com (Don Macpherson)
auf /COMP/SYS/UNISYS
in 7sjh0v$3fi$1...@winter.news.rcn.net
ueber Re: Unisys ? What is Unisys?
DM> What about IBM S/34 and S/36 systems?
DM> Are they still around or have they been "retired" like the S/3,
DM> B1000 and V Series systems?
Replaced by AS/400.
Don Macpherson
"replaced" by the A Series.
Don Macpherson
"Luko Willms" <L.WI...@LINK-F.frankfurt.org> wrote in message
news:7Pc6e...@lf155.link-f.rhein-main.de...
Edward Reid
> You can run editor on top of this, which is a really nifty editor, which
I agree. I'm constantly amazed and horrified by the number of people I see
using CANDE page mode, which is truly ghastly compared with SYSTEM/EDITOR
(aka U ED).
I used a pre-release version of EDITOR in 1982 for a two-week project which
involved patching the COBOL compilers. I estimated at the time that it cut my
development time by 50%, even with the learning curve.
> reads compiler generated files to help you browse through your programs.
> This is maybe the best editor in concept that I have used, but the front
> end was slightly old fashioned
Yes, because it operated on the oldest terminals still in use at the time and
thus has never taken advantage of the possibilities of having a PC on the end
of the wire.
There was an effort to rectify this in the mid-1980s, resulting in IDE
(Interactive Distributed Editor), which did the actual editing on a
workstation closely communicating with a mainframe process. I'm not sure what
happened to IDE, though I recall it had some performance ... um, issues. I
think there's a newer distributed editor available now, but I can't recall
the name or any other info.
> and the real draw back that stopped a lot
> of shops using it was the fact that it fired up one process per every
> person using it. I think these problems have been fixed lately.
That was an early drawback, but dropping memory prices fixed that soon after
it was released. It is still a bit slow to start up -- perhaps ten seconds
when I'm running it on an A14 with two modems between me and the A14; only a
second or two on a faster A-Series with a LAN connection.
The biggest problem was always that the Burroughs/Unisys sales force didn't
know what it was or how to sell it, so it only sold to customers who already
knew about it and insisted on getting it.
The other problem is that it takes an extra hour to learn, and most
programmers won't take the time unless their supervisors insist. Sad when
programmers contribute so much to inertia.
Edward Reid
RDF
will never replace either..
Fetch
Martyn Richard Jones
Dear Sir,
> Left out the V Series (Medium Systems Architecture), BTOS Platforms, S/3,
> ......... there was a lot more history than just the 1100 and A Series.
> Steve Myers
Left out ...? All there ever was in historical terms was the Univac which
grew into the 1100 and eventually the 2200 to ClearPath etc.
Credit where credit is due, the following weren´t so bad either:
The Sperry PC
The BC/7
The UDS 2000
The SperryLink DOPS
The V77 - even though it was almost a bigger dog than the DEC equivalent
9000 and 90/30 and 90/40
The Uniscope 200 to the UTS 60
MAPPER 10
MAPPER 5 (well that was a bit iffy)
SYSTEM 11
The rest was just a pure result of its own
Sorry Hysterical Information Technology Era ... A Series V Series huh!
S.H.I.T.E.
Tying a sinking rock to a rising star brought ten years of trouble and
strife and a company no better than just Sperry itself was in 1985.
B*st*rd ex US Treasurers and their f*ck*d up ideas!
Regards
Martyn R Jones
John G Dobnick
>
> B*st*rd ex US Treasurers and their f*ck*d up ideas!
I still think that the best name for the merged [1] Burroughs/Sperry
entity should have been MCC -- Mikee's Computer Company.
[1] Merged? Did I say "merged"? How about "Burroughs takover"!
--
John G Dobnick "Knowing how things work is the basis
Information & Media Technologies for appreciation, and is thus a
University of Wisconsin - Milwaukee source of civilized delight."
j...@uwm.edu ATTnet: (414) 229-5727 -- William Safire
Ken McNeill
NX/EDIT is the PC based update to EDITOR. It is available on A Series as
well as NX and LX systems.
Happiness,
Ken McNeill
Edward Reid <edw...@paleo.org> wrote in message
news:01HW.B41528DF0...@news.talstar.com...
Ken McNeill
I don't/can't argue with your assessment of the merger and Mike Blumenthal.
Personally, I think he should have ended up in court for conflict of
interest issues when he left Unisys and moved to Lazzard Freres to supervise
the Unisys sell off, but that is a different story.
As for your dismissal of A Series, BTOS, etc. I think you are being a little
too one-eyed. You list fine equipment/software no doubt but the A Series
(and its predecessors have been very innovative in their own right). Note
that although IBM 'invented' virtual memory in 1972 it had been running in a
multi-processor multi-tasking environment on the B5000 line since 1963 and
the B6000 line since 1968. Architecturally it is also interesting because
the systems have only ever (except for a very brief bootstrap process for
the B5000) supported high level languages. I could wax lyrical, but no one
really needs this.
As for BTOS/CTOS, it was a wonderful architecture far ahead of its time.
Workstation and server implementations were released with built in
networking, resource sharing and client/server functionality to name but a
few of the features. Its word processing software was only surpassed in the
last few years by Word and its other PC competitors. Indeed, NT looks a lot
like CTOS with a GUI front end.
Ah well, another defeat snatched from the jaws of victory. For a while
there it looked like BTOS/CTOS would make a go of it in spite of the sales
force but eventually they managed to kill it.
Martyn Richard Jones <inici...@ibm.net> wrote in message
news:37f1...@news2.prserv.net...
> B*st*rd ex US Treasurers and their f*ck*d up ideas!
>
> Regards
>
> Martyn R Jones
>
>
>
>
Samuel Tan
loved the BTOS (still have one stashed away for a museum some day).
If "Product Management" had the foresight to go ahead with implementing
CTOS/BTOS with GUI, we would have had NT at least 10 years ahead of
Microsoft..
Ken McNeill <Ken...@Leibnitz.com> wrote in message
news:%0HI3.4383$ea5....@dfw-read.news.verio.net...
Steve Schneider
"superior" to those of Sperry Univac, Burroughs was in decline and had to find a
company that had sufficient cash and a customer base to suck off of to try and
stay alive. I, too, liked the BTOS machines. They were extremely advanced for
what was available at the time, but also very costly and proprietary, which
ultimately turned a lot of customers off.
As far as Blumenthal is concerned, he should have been thrown in jail to rot for
the misery he brought on many employees and their families. He fled to France
to enjoy the good life while a lot of employees watched as their retirement
funds that Unisys invested in the defunct Executive Life and Mutual Benefit
Insurance Contracts went south. (He didn't have any of his retirement in these
funds.) I know of one case where a guy was laid off, so he committed suicide.
Other employees looked for jobs for a year or more, scraping and just trying to
get by. Then we get James Unruh, and his book on how to "Customerize".
Marvelous! Where I work now, "Unisys" is a dirty word.
Edward Reid
> although IBM 'invented' virtual memory in 1972 it had been running in a
> multi-processor multi-tasking environment on the B5000 line since 1963 and
> the B6000 line since 1968.
The first virtual-memory machine was the ICL Atlas, around 1955.
However, I don't know much else about the Atlas.
> Architecturally it is also interesting because
> the systems have only ever (except for a very brief bootstrap process for
> the B5000) supported high level languages.
Sort of vague here ... neither ever required the user to use assembler.
The B5000 allowed an escape into assembler. I wasn't there and so don't
remember the details. I think the escape was via a POLISH function, but
I don't know whether it was available to user programmers or only for
the MCP.
There has never been any such escape on the B6700 and successors. On
the rare occasions where the MCP must execute arbitrarily generated
machine code, it has to play memory games to get there. I'm aware of
three situations where code must have been written as machine code.
These cases are short and I suspect the code was hand-assembled. (There
are other places where I'm sure the programmer first figured out the
machine instructions needed, then figured out the Algol or Espol or
Newp constructs needed to generate that code. But those are actually
coded and maintained in the HLL.)
1) the two-card loader used on early systems. In the early days,
cold-starting required loading SYSTEM/LOADER from a binary card deck.
The two-card loader did the actual reading of cards and setup. Needless
to say it was very tight code.
2) the disk bootstrap stored in sector zero of a halt/load disk unit.
The hardware cleared some registers, did a 180-byte read from the
selected halt/load unit and branched to memory location zero. The code
had to set up a working environment, read more code, and branch to it
-- all in 180 bytes.
3) the code to implement ??PHL. I don't remember the details, and it's
a bit surprising that with all the stuff the MCP does, this bit of code
required machine code. I don't know why. I just remember that the code
was in a value array, and the MCP converted it into a code segment and
branched to it.
Notably, the loadable microcode used on all A-Series systems since
about 1984 (started on the B5900 in about 1980) is also written in a
high-level language. AFAIK the same language, OHNE, is still in use. I
have no idea whether it allows escaping to a different level. In a
sense, the microcode is itself an escape into the bowels of the
architecture compared with the user languages.
Edward Reid
Colin Zealley
Steve Schneider wrote in message <37F42273...@amdyne.net>...
[huge snip]
>Where I work now, "Unisys" is a dirty word.
That's a real pity. The Unisys of Larry Weinbach is a whole different place
to the Blumenthal/Unruh era. Maybe you should take another look - I do
believe we should not be seen in the asme way as a couple of years ago.
Colin
Edward Reid
> NX/EDIT is the PC based update to EDITOR. It is available on A Series as
> well as NX and LX systems.
That's it. Have you used it? Any opinions?
Edward Reid
Louis Krupp
Edward Reid wrote:
>
<snip>
> Sort of vague here ... neither ever required the user to use assembler.
> The B5000 allowed an escape into assembler. I wasn't there and so don't
> remember the details. I think the escape was via a POLISH function, but
> I don't know whether it was available to user programmers or only for
> the MCP.
>
> There has never been any such escape on the B6700 and successors. On
> the rare occasions where the MCP must execute arbitrarily generated
> machine code, it has to play memory games to get there. I'm aware of
> three situations where code must have been written as machine code.
> These cases are short and I suspect the code was hand-assembled. (There
> are other places where I'm sure the programmer first figured out the
> machine instructions needed, then figured out the Algol or Espol or
> Newp constructs needed to generate that code. But those are actually
> coded and maintained in the HLL.)
POLISH (usually abbreviated P, presumably with a DEFINE) was a B5500
ESPOL feature, as I recall. This would have restricted its use to
the MCP and INTRINSICS.
I vaguely remember that B5500 ALGOL stream procedures made it
possible to hand-code and execute machine language instructions. I
am not saying I did this or that in doing it I may have inadvertantly
hung the system.
The stream procedure was an efficient way of doing a lot of things,
and trashing memory links was one of them. The B5500 MCP had a
simple, concise algorthim for dealing with bad links and other
unexpected anomalies:
DO UNTIL FALSE;
B6700 ESPOL had the TOPOFSTACK variable, usually abbreviated TOS.
"TOS := <something>" meant LITC or VALC; "<something> := TOS"
compiled to "NAMC, STOD." It seemed to be an efficiency thing
designed to reduce the need for local variables and the resulting
memory references.
B6700 NEWP didn't have TOPOFSTACK. It made the ESPOL -> NEWP
conversion interesting.
Louis Krupp
Ken McNeill
architecture. I don't remember there being an assmbler escape function on
the B5500/B5700, but I do remember, and have used, the issue of converting
an array of operators and operands/values into a code stream (diddle the
tag) and then FORKing the resultant code stream. We did, however, put a
patch into the ALGOL compilers to permit us to pass op codes in as 'source'
language and have the compiler spit out the binary. This was particularly
useful before DMALGOL supported DISALLOW and ALLOW and meant that we didn't
have to resort to ESPOL in a transaction processing environment.
As for the mini loader, my memory tells me it was three (3) cards. Whatever
it was, it was incredibly smart and compact code. I remember studying it
for several days and being constantly amazed at how cunningly it was put
together.
OHNE is still certainly supported on the architecture, but to my knowledge
has never been released to the field.
Edward Reid <edw...@paleo.org> wrote in message
news:01HW.B4199BEE0...@news.talstar.com...
> On Thu, 30 Sep 1999 6:36:41 -0400, Ken McNeill wrote
> > although IBM 'invented' virtual memory in 1972 it had been running in a
> > multi-processor multi-tasking environment on the B5000 line since 1963
and
> > the B6000 line since 1968.
>
> The first virtual-memory machine was the ICL Atlas, around 1955.
> However, I don't know much else about the Atlas.
>
> > Architecturally it is also interesting because
> > the systems have only ever (except for a very brief bootstrap process
for
> > the B5000) supported high level languages.
>
> Sort of vague here ... neither ever required the user to use assembler.
> The B5000 allowed an escape into assembler. I wasn't there and so don't
> remember the details. I think the escape was via a POLISH function, but
> I don't know whether it was available to user programmers or only for
> the MCP.
>
> There has never been any such escape on the B6700 and successors. On
> the rare occasions where the MCP must execute arbitrarily generated
> machine code, it has to play memory games to get there. I'm aware of
> three situations where code must have been written as machine code.
> These cases are short and I suspect the code was hand-assembled. (There
> are other places where I'm sure the programmer first figured out the
> machine instructions needed, then figured out the Algol or Espol or
> Newp constructs needed to generate that code. But those are actually
> coded and maintained in the HLL.)
>
Ken McNeill
allow operation with ACCESSCODEs and the primary site I work at requires
them. The problems has been fixed (very speedily) but I haven't found the
inclination/energy to try it again.
I'll let you know when I give it a whirl.
Edward Reid <edw...@paleo.org> wrote in message
news:01HW.B41A369D0...@news.talstar.com...
Lyle Craver
Dobnick) wrote:
> [1] Merged? Did I say "merged"? How about "Burroughs takover"!
Interesting - if you had been in the Canadian Head Office staff
in 1986/87 you would have reached precisely the opposite
conclusion.
------------------------------------
To reply to me remove 1 from address
Juha Veijalainen
> > [1] Merged? Did I say "merged"? How about "Burroughs takover"!
>
> Interesting - if you had been in the Canadian Head Office staff
> in 1986/87 you would have reached precisely the opposite
> conclusion.
If I remember correctly, technically it was a Burroughs takeover.
Locally, however, the effects depended on whether the Sperry or Burroughs
side was 'stronger'
--
Juha Veijalainen, Helsinki, Finland, http://www.iki.fi/juhave/
Some random words: bomb,steganography,cryptography,reindeer
** Mielipiteet omiani ** Opinions personal, facts suspect **
John G Dobnick
>
> Interesting - if you had been in the Canadian Head Office staff
> in 1986/87 you would have reached precisely the opposite
> conclusion.
Within a year almost all the (former) top level Sperry execs were
gone. Smells like "takeover" to me.
RDF
Colin Zealley wrote:
Sounds great Colin, But tell that to the thousands that Blumenthal/Unruh
axed to make themselves look good and more important increase their wallet
size. Larry has good deals but still has to much of the old still in control.
If I was still with Unisys I would always be looking for the knife to appear
from nowhere.. I still think that the company has the best in technology and
new ideas but must get handle on the idiots the have running marketing..
Fetch
Colin Zealley
RDF wrote in message <37F8BA74...@bellatlantic.net>...
[snip]
> Sounds great Colin,
It is! :-)
>But tell that to the thousands that Blumenthal/Unruh
>axed to make themselves look good and more important increase their wallet
>size.
No disagreement here. And Unruh still owns more shares than the entire
current board, according to the last Annual Report that I saw. I guess he
must be happy he's been replaced!
>Larry has good deals but still has to much of the old still in control.
>If I was still with Unisys I would always be looking for the knife to
appear
>from nowhere..
And for the first year or so after his arrival, I and my friends did
likewise. But it hasn't happened; instead we're growing again on all scales.
Of course, there are still holdouts of the "old guard" in place here and
there, but they *are* being flushed out. They have to be if Unisys is to
achieve the goals that Weinbach and Brust have set for us, and thus far we
seem to be on course towards those. Even Larry can't fix the whole thing
overnight. But each change we see points in the same direction, and it's the
right direction IMHO.
>I still think that the company has the best in technology and
>new ideas but must get handle on the idiots the have running marketing..
>
Well, we have a new head marketing guy just come on board. Maybe he'll make
a difference. :-/
But feedback from friends and relations in other IT companies suggests to me
that just about *everybody* thinks their own marketing organization is cr*p;
personally, I think this is perhaps more of a perception problem than a real
one. We'll see, I guess.
Colin
Bert Hyman
Colin Zealley <zi...@technologist.com> wrote:
> ...
>But feedback from friends and relations in other IT companies suggests to me
>that just about *everybody* thinks their own marketing organization is cr*p;
>personally, I think this is perhaps more of a perception problem than a real
>one. We'll see, I guess.
>
appreciated tells me that the problems ARE both widespread and real.
--
--
The opinions expressed in this message are my own and are not the opinions of
anyone who does not hold those opinions.
Bert Hyman | St. Paul, MN | be...@visi.com
Colin Zealley
Bert Hyman wrote in message ...
>In article <7taeha$t2d$1...@mail.pl.unisys.com>,
>Colin Zealley <zi...@technologist.com> wrote:
>> ...
>>But feedback from friends and relations in other IT companies suggests to
me
>>that just about *everybody* thinks their own marketing organization is
cr*p;
>>personally, I think this is perhaps more of a perception problem than a
real
>>one. We'll see, I guess.
>>
>The fact that the "Dilbert" comic strip is universally understood and
>appreciated tells me that the problems ARE both widespread and real.
Understood, Bert. I agree entirely.
What I meant to criticise was the perception that Unisys somehow has the
worst ones. As you say, Dilbert's success underscores the fact that most
corporate problems are common across most companies.
Personally, I think the reason that Dilbert is so popular in IT as well as
"pure" engineering (when actually he's a "real" engineer, not a computer
bod, in the first place) is simply that the hierarchically-minded type of
people who are attracted (quite properly) to management and business
consultancy positions are often quite genuinely unaware that people with an
engineering mindset simply don't march to the same drum. They honestly fail
to realise that not everybody is motivated by wanting an increase in power
and influence over others as the natural course of career progression.
Pointed-haired-boss's problem with Dilbert & his colleagues is not so much
that PHB doesn't understand them; it's that he doesn't even understand that
he doesn't understand them.
If more managers understood what motivates engineers, a lot more technology
would get developed a lot more efficiently. Of course, whether that would
actually improve the world or not is another question entirely ...
All IMHO, of course.
Colin