Non- geek talks to a friend about Y2K
art welling
Howdy all.
Just wanted to relate a short conversation I had today.
In our local bird cage liner there was an AP story on the IRS and Y2K on
page
three. FIRST I have ever seen in our paper. It mentioned most of the
problem
and the 8-10% attrition rate of programmers fleeing Govt service.
I read this over breakfast in a restuarant, sitting next to the guy who
runs the local Radio Shack. It happens my wife works there part time so
we became sorta friends. First I asked him if he was on the net. He
said
"no". I was floored speechless for a moment, kinda like finding out a
friend can't read.
Then I asked him if he had read the piece. "Yes". I asked " Did you
notice
one major thing they left out of the story ?". He said " What did they
leave
out ?". I replied " what happens if the DON'T get it fixed ?".
He asked " So what happens ?". I said to him "your a smart guy,
extropolate
and answer your own question.". After a moment he said " Everything
collapses".
HOORAY ! Easiest conversion I have ever had. Who's next ? Next in
line, please
step up............
The Goobers
art welling wrote:
>
[snippage of relating Y2K difficulty in-a-nutshell to a Radio Shack
manager]
> After a moment he said " Everything collapses".
> HOORAY ! Easiest conversion I have ever had. Who's next ? Next in
> line, please
> step up............
Ahh, but you see, you started with Good Material... this fellow ran a
Radio Shack, an organization which comes from the tradition of the
2-meter dipole; imagine what the response would have been were the
fellow to be an employee of, say, Lechter's or Nobody Beats The Wiz!
DD
Jim Rivera
David Ray wrote:
<previous snipped>
>IRS, like everyone else,
> has some systems that are more critical than others. In fact, I can't
> imagine that the vast majority of their systems are even that important in
> the overall scheme of collecting taxes. For example, the "CAF", which
> tracks powers of attorney, is a frequently used system which hardly
> functions as it is. What if it quits working? Not much. Just doesn't
> matter. Now, if BMF or IMF (Business or Individual Master File) systems
> collapse, THAT could be a problem. But IRS is going to fix those first.
> So, the ancillary systems like CAF or even the dread DIF (Discriminant
> Function scoring system) were to collpase, it's really not going to be that
> big of a problem.
>
> If everyone could be a little less excitable about this problem, it would
> really help us keep a more accurate perspective.
>
> David Ray, CPA
>
This post assumes that the triage will be on a rational basis. And on
time. This might or might not be the case. They may be trying to fix
it all, or in the wrong order. And what happens if their personnel loss
rate doubles from its already-high 8-9% yearly to around 20%? IMHO this
is at least possible. And I don't see any kind of retention bonuses
even being proposed in the current US Congress.
JR
Eric Turner
David Ray wrote:
>
> >
> > Then I asked him if he had read the piece. "Yes". I asked " Did you
> > notice
> > one major thing they left out of the story ?". He said " What did they
> > leave
> > out ?". I replied " what happens if the DON'T get it fixed ?".
> > He asked " So what happens ?". I said to him "your a smart guy,
> > extropolate
> > collapses".
> > HOORAY ! Easiest conversion I have ever had. Who's next ? Next in
> > line, please
> > step up............
>
> has some systems that are more critical than others. In fact, I can't
> imagine that the vast majority of their systems are even that important in
> the overall scheme of collecting taxes. For example, the "CAF", which
> tracks powers of attorney, is a frequently used system which hardly
> functions as it is. What if it quits working? Not much. Just doesn't
> matter. Now, if BMF or IMF (Business or Individual Master File) systems
> collapse, THAT could be a problem. But IRS is going to fix those first.
> So, the ancillary systems like CAF or even the dread DIF (Discriminant
> Function scoring system) were to collpase, it's really not going to be that
> big of a problem.
>
> If everyone could be a little less excitable about this problem, it would
> really help us keep a more accurate perspective.
>
> David Ray, CPA
>
Look Dave - this isn't a matter of how I or anyone else 'feel' about the
situation. The 'developed' world is what I would call 'heavily
leveraged'. These societies have bet on the fact that computers can run
the whole show. Not a bad bet considering the past history of
technology. Turns out it may have been something more along the lines
of the Tulip Mania. Instead of things as usual, it has turned into the
most speculative event of all time.
It is difficult for me to believe things will be remediated to such a
degree that we end up perceiving this to be a bump in the road, but
that's my opinion today. Our societies are based on computers, like it
or not, and we are likely to gain or suffer accordingly. If you have
some evidence that the systemic nature of our society is not at risk
here I would be interested in hearing it. The 'system' is the issue as
far I can tell, but my opinion is subject to change.
ET
David Ray
cory hamasaki
In <01bd0f4a$70ad8b20$6776b4cc@sys>, "David Ray" <dr...@cei.net> writes:
>
>Well, except that's really not quite the answer. IRS, like everyone else,
>has some systems that are more critical than others. In fact, I can't
>imagine that the vast majority of their systems are even that important in
>the overall scheme of collecting taxes. For example, the "CAF", which
>tracks powers of attorney, is a frequently used system which hardly
>functions as it is. What if it quits working? Not much. Just doesn't
>matter. Now, if BMF or IMF (Business or Individual Master File) systems
>collapse, THAT could be a problem. But IRS is going to fix those first.
Are they? Are you sure? I know a guy (not Frank) who worked on an IRS
system that:
1) ran for 3 weeks non-stop every month.
2) was written in S/390 assembly language.
3) had no MVS macros in the source.
4) keeps dates in two digit year/three digit day, packed decimal without
the sign nibble. 99365F is stored as 99365 and unaligned.
Item three was the shocker. Since they were getting close to having a
monthly job that took longer than a month to run, they squeezed every
instruction possible out of the machine. One trick was to assemble the IOS
macros and hand tune the macro generated code, keeping values in registers,
resequencing instructions, eliminating unnecessary or redundant code.
Item four is a surprise too. They were trading off the time to align and OI
a sign against the time to move the data though the channel. Amdahl's Law
says that they are doing the right thing.
>So, the ancillary systems like CAF or even the dread DIF (Discriminant
>Function scoring system) were to collpase, it's really not going to be that
>big of a problem.
>
>If everyone could be a little less excitable about this problem, it would
>really help us keep a more accurate perspective.
If you didn't understand my surprise and horror at what the IRS programmers
are doing, you are not qualified to comment on the scale of the problem.
>
>David Ray, CPA
>
As an exercise for the student, quote Amdahl's law and use it to prove that
the IRS programmers made the right choice in item 4. No help from the know
assembler crankers.
Cory Hamasaki
Gary L. Smith
cory hamasaki (kiy...@ibm.net) wrote:
: 4) keeps dates in two digit year/three digit day, packed decimal without
: the sign nibble. 99365F is stored as 99365 and unaligned.
That's two and a half bytes. What are they doing with the reamining half
byte?
--
Gary L. Smith g...@infinet.com
Columbus, Ohio Gary_...@oclc.org
cory hamasaki
Unaligned! My geek agent didn't say but I'm guessing that they do
something like position two date fields next to each other:
99365F98364F and squeeze two dates in five bytes
9936598364.. Where the .. represents a savings in the
filesize.
I'm certain they don't move with offset entire character strings just to
save space.
When he told me what the IRS was doing, and they've been doing this for
decades, I was astonished. What's unsettling is that, on one level it
makes a lot of sense, but on another it's so odd that that I know that
managment's 'happy-talk' (tm) is just more baloney... no matter how you
slice it.
If they take 3 weeks to pass this file once, they don't have time to
convert. Oops maybe I just had a senior moment and it was Brooks law.
It's the one about balancing IO and MIPS and the tradeoff in S/360. I'm
starting to natter... haven't had my morning coffee... grey stubble on my
chin, what's left of my hair sticking out at odd angles.
I'm still waiting for a non-geek to explain why the IRS's choice was a good
one. It wasn't great choice but this is an essay question.
Cory Hamasaki
Michael Dodas
David Ray wrote:
>
> >
> > Then I asked him if he had read the piece. "Yes". I asked " Did you
> > notice
> > one major thing they left out of the story ?". He said " What did they
> > leave
> > out ?". I replied " what happens if the DON'T get it fixed ?".
> > He asked " So what happens ?". I said to him "your a smart guy,
> > extropolate
> > and answer your own question.". After a moment he said " Everything
> > collapses".
> > HOORAY ! Easiest conversion I have ever had. Who's next ? Next in
> > line, please
> > step up............
>
> has some systems that are more critical than others. In fact, I can't
> imagine that the vast majority of their systems are even that important in
> the overall scheme of collecting taxes. For example, the "CAF", which
> tracks powers of attorney, is a frequently used system which hardly
> functions as it is. What if it quits working? Not much. Just doesn't
> matter. Now, if BMF or IMF (Business or Individual Master File) systems
> collapse, THAT could be a problem. But IRS is going to fix those first.
> Function scoring system) were to collpase, it's really not going to be that
> big of a problem.
>
> If everyone could be a little less excitable about this problem, it would
> really help us keep a more accurate perspective.
>
>
Systems like the "CAF" and other that you mention, which you say are not
as important and will probably take a lower fix priority, are probably
far less complex than the ones you say are important. If you think
IRS's systems are repairable within the necessary time frames, you do
not understand their computing environment. If most of their systems are
not important, as you mention, then why are we wasting taxpayer money on
them? I would think those systems might help keep you employed. No?
Read IRS's "Request For Comments" to modernize their systems. If you
think what IRS wants to accomplish in time is possible, then I've got
some land in Florida I'd like to sell you.
--
Mike Dodas
(Remove cutthejunk for e-mail replies)
Gary L. Smith
cory hamasaki (kiy...@ibm.net) wrote:
: In <ZyKn.204$Ue.5425400@news2>, g...@infinet.com (Gary L. Smith) writes:
: >cory hamasaki (kiy...@ibm.net) wrote:
: >
: >: 4) keeps dates in two digit year/three digit day, packed decimal without
: >: the sign nibble. 99365F is stored as 99365 and unaligned.
: >
: >That's two and a half bytes. What are they doing with the reamining half
: >byte?
: Unaligned! My geek agent didn't say but I'm guessing that they do
: something like position two date fields next to each other:
: 99365F98364F and squeeze two dates in five bytes
: 9936598364.. Where the .. represents a savings in the
: filesize.
OIC. We did something similar to that in a system I worked on back in the
sixties. In that case, we packed three or four numeric fields together,
but didn't go so far as stripping the remaining zone nibble. That saved
a couple of bytes per record, which seemed important at the time.
Chris King
David Ray wrote in message <01bd0f4a$70ad8b20$6776b4cc@sys>...
>If everyone could be a little less excitable about this problem, it would
>really help us keep a more accurate perspective.
>
>
devaluations. Or someone who doesn't receive his food stamps on time.
I've found the folks in this NG to be quite sober in their appreciation of
the problem.
--
Reply to king at the domain of weirdness dot com.
Chris King
>cory hamasaki wrote in message <349f2...@news3.ibm.net>
>...As an exercise for the student, quote Amdahl's law and use it to prove
that
>the IRS programmers made the right choice in item 4. No help from the know
>assembler crankers.
Cory,
I can find Amdahl's law, but what do "OI" and "unaligned" mean?
Chris
(Someone who still pauses, with head cocked, to think about the differences
between a "do...while" and a "while...do"...)
cory hamasaki
I'll put it in a future DC WRP.
OI is Or Immediate. Can be used to flip on a sign nibble. '0F'X
Unaligned means, well, ah, in the old days, harumph, IN the good old days,
the electronics that drove memory was incapable of performing some
operations unless they were on addresses divisible by 2 or 4.
So some data had to be aligned on these half word and full word boundaries.
Data not aligned would, in the old days, not be readable by some
instructions.
Some programming languages, PL/I comes to mind, gave the programmer the
option of aligning the data to the natural boundaries or leaving it
unaligned. The idea of aligned, again in PL/I, extended downward below the
byte level to the bit level.
Of course nibbles don't exist but programmers have vivid imaginations and I
was imagining data unaligned on the nibble boundary... ah, so to speak.
Listen, in this case, it's not you... there're two obscure concepts at
work here. Storage boundaries which haven't existed for about 30 years,
and nibbles, which never existed.
Cory Hamasaki
David K. Bryant
kiy...@ibm.net (cory hamasaki) writes:
>Unaligned! My geek agent didn't say but I'm guessing that they do
>something like position two date fields next to each other:
> 99365F98364F and squeeze two dates in five bytes
> 9936598364.. Where the .. represents a savings in the
> filesize.
OK, taking horizontal compression a little further,
since only values of 0 - 9 are used in each nybble
capable of storing 0 - F, a full three eighths of the storage
capacity wasted. Wouldn't it be possible to, with a little
Creative Coding (tm), pack those bits in a little closer?
Oh, oh, and by applying Run Length Encoding, get even more
compression by looking for runs of three or more 1s or 0s
and encode them into just two bits. Oh, this is getting
exciting.
(just having fun with y'all)
Sign of the times:
This morning on BART I wake up and notice that the person
ahead of me is reading thru perhaps ten pages of COBOL
interview questions...
What are some of the differences between VS COBOL and COBOL II?
What utility is used to create, display, and edit files?
Is it proper etiquette to laugh in a PHM's face?
Crib notes for the Pepsi Generation.
David K. Bryant
"Chris King" <j...@microsoft.com> writes:
>Excitable is someone who faces losing his life's savings in currency
>devaluations. Or someone who doesn't receive his food stamps on time.
Excitable is Bill Gates finding out that the value of his stock
holdings has dropped by $4 BILLION in a week.
The Goobers
David K. Bryant wrote:
>
[snippage of Moderate Geekery]
>
> This morning on BART I wake up and notice that the person
> ahead of me is reading thru perhaps ten pages of COBOL
> interview questions...
>
> What are some of the differences between VS COBOL and COBOL II?
> What utility is used to create, display, and edit files?
> Is it proper etiquette to laugh in a PHM's face?
How interesting and serendipitous... just the other day in
comp.lang.cobol I gave some Big Laffs with a posting that mentioned
newbies who are still figuring out the difference between an index and a
subscript.
DD
Francis A. Ney, Jr.
In article <349f2...@news3.ibm.net> kiy...@ibm.XOUT.net writes:
> >If everyone could be a little less excitable about this problem, it would
> >really help us keep a more accurate perspective.
>
> If you didn't understand my surprise and horror at what the IRS programmers
> are doing, you are not qualified to comment on the scale of the problem.
Big time not qualified.
> >David Ray, CPA
I'm a Certified Public Assassin, too. But my nine years working for the most
hated government agency in the United States makes mine the better opinion.
And Cory's multiple decades working this hardware makes his the even better
opinion.
> As an exercise for the student, quote Amdahl's law and use it to prove that
> the IRS programmers made the right choice in item 4. No help from the know
> assembler crankers.
You're no fun. I'm not a a-cranker, but I have to DQ myself because I've
heard the answer before, from the King-Geek speaking ex cathedra from his
belly button.
---
Frank Ney WV/EMT-B VA/EMT-A N4ZHG LPWV NRA(L) GOA CCRKBA JPFO
Sponsor, BATF Abuse page http://www.access.digex.net/~croaker/batfabus.html
West Virginia Coordinator, Libertarian Second Amendment Caucus
NOTICE: Flaming email received will be posted to the appropriate newsgroups
- --
"It's nice to know that the MBAs of the 80's are still gainfully employed at
what they do best. . .making a mess of things so I can fix them."
- David K. Bryant dbr...@netcom.com
Francis A. Ney, Jr.
In article <01bd0f4a$70ad8b20$6776b4cc@sys> dr...@cei.net writes:
> Well, except that's really not quite the answer. IRS, like everyone else,
> has some systems that are more critical than others. In fact, I can't
> imagine that the vast majority of their systems are even that important in
> the overall scheme of collecting taxes. For example, the "CAF", which
> tracks powers of attorney, is a frequently used system which hardly
> functions as it is. What if it quits working? Not much. Just doesn't
> matter. Now, if BMF or IMF (Business or Individual Master File) systems
> collapse, THAT could be a problem. But IRS is going to fix those first.
Except for one thing: They haven't even started.
As of my last day at the IRS, 21 Oct 97. IS:AD:SP:E-SJ (If you don't get
that last, don't worry about it, it's just to establish my bona fides)
They have lots of handwaving, lots of CYA, lots of smoke-blowing, lots of
meaningless reports, lots of neat Powerpoint slides, but no remediation.
This after they spent $4.5G trying to go C/S.
That's why I resigned. Among other reasons.
Then there's the manual input routines in the service centers. All that
Autocoder that *NO ONE* has the source code for, nor the expertise to fix it
if they did. There is a reason why IRS is trying to push Telefile and machine
readable paperwork, and electronic filing. . .
RandallBart
cory hamasaki wrote:
> Listen, in this case, it's not you... there're two obscure concepts at
> work here. Storage boundaries which haven't existed for about 30 years,
> and nibbles, which never existed.
This is a typically parochial Big Blue attitude. On Burroughs Medium
System (later Unisys V Series) memory addresses pointed to the nibble
(we called them digits). On the Unisys A Series (formerly Burroughs
Large System), memory is organized in 48 bit words, but descriptors can
index it with a framesize of 4, 8, 48, or 96 bits. (At one time, it
could be indexed with a framesize of 3 or 6 bits, as well).
On Unisys, when we declare a field as PIC 9(6) PACKED it occupies three
bytes. I would consider it a compiler bug if it occupied any more.
--
I |\ Randall Bart mailto:Bart...@usa.spam.net
L |/
o |\ Bart...@worldnet.att.spam.net Bart...@hotmail.spam.com
v | \ 1-818-985-3259 Please reply without spam
e |\
Y |/ Panic in the Year Zero Zero: http://members.aol.com/PanicYr00
o |\
u |/ Have you solved http://members.aol.com/PanicYr00/Sequence.html
cory hamasaki
In <681p93$o...@bgtnsc02.worldnet.att.net>, RandallBart <Bart...@worldnet.att.spam.net> writes:
>cory hamasaki wrote:
>
>> Listen, in this case, it's not you... there're two obscure concepts at
>> work here. Storage boundaries which haven't existed for about 30 years,
>> and nibbles, which never existed.
>
>This is a typically parochial Big Blue attitude. On Burroughs Medium
>System (later Unisys V Series) memory addresses pointed to the nibble
>(we called them digits). On the Unisys A Series (formerly Burroughs
>Large System), memory is organized in 48 bit words, but descriptors can
>index it with a framesize of 4, 8, 48, or 96 bits. (At one time, it
>could be indexed with a framesize of 3 or 6 bits, as well).
>
>On Unisys, when we declare a field as PIC 9(6) PACKED it occupies three
>bytes. I would consider it a compiler bug if it occupied any more.
>--
>I |\ Randall Bart mailto:Bart...@usa.spam.net
Parochial Big Blue??? Does Burroughs still exist? IBM mainframes are
available from:
Hitachi
Amdahl
Data General
Comparex
Fundamental Software
Commercial Data Servers
and, and, there is seventh company but I keep forgetting its name, I
think it's HQ'ed in New York.
All these vendors produce true blue S/390, ESA systems.
Right now, I'm blowing my honker into blue Kleenix, my car is blue, and so
is the cheese on my salad.
Cory Hamasaki
E. N. Landers
> Now, if BMF or IMF (Business or Individual Master File) systems
> > collapse, THAT could be a problem. But IRS is going to fix those first.
>
> Except for one thing: They haven't even started.
Not true
>
>
> As of my last day at the IRS, 21 Oct 97. IS:AD:SP:E-SJ (If you don't get
> that last, don't worry about it, it's just to establish my bona fides)
My last day was August 2, 1996 IS:AD:T:B-B
>
>
> They have lots of handwaving, lots of CYA, lots of smoke-blowing, lots of
> meaningless reports, lots of neat Powerpoint slides, but no remediation.
Mostly true, but I (and others) did a whole bunch of Y2K work starting 1Q/1996.
Without management permission.
>
>
> This after they spent $4.5G trying to go C/S.
I thought it was closer to 6G. We did get some good hardware out of it, though.
>
>
> That's why I resigned. Among other reasons.
>
> Then there's the manual input routines in the service centers. All that
> Autocoder that *NO ONE* has the source code for, nor the expertise to fix it
> if they did. There is a reason why IRS is trying to push Telefile and machine
> readable paperwork, and electronic filing. . .
SC level code was beyond the scope of my position. No comment.
RandallBart
cory hamasaki wrote:
> Parochial Big Blue??? Does Burroughs still exist?
Burroughs is now called Unisys, and they still make computers.
> IBM mainframes are
> available from:
> <snippage>
> All these vendors produce true blue S/390, ESA systems.
There are seven companies that build computers to work exactly like Big
Blue, and their computers work exactly like Big Blue. Well Duh! If any
of those manufacturers built computers which accessed nibbles properly,
then either their computers wouldn't be compatible, or they would
contain op codes that could never be accessed by the existing
compilers.
But you made the preposterous claim that nibbles never existed, when in
fact they exist on all mainframes designed to handle Cobol.
--
I |\ Randall Bart mailto:Bart...@usa.spam.net
cory hamasaki
In <688v8i$m...@bgtnsc03.worldnet.att.net>, RandallBart <Bart...@worldnet.att.spam.net> writes:
>
>But you made the preposterous claim that nibbles never existed, when in
>fact they exist on all mainframes designed to handle Cobol.
They do? I can give bytes names:
dcl 1 mastr_rec unaligned,
2 my_first_byte char(01),
2 other_bytes char(06);
I can also name bits:
dcl 1 mastr2_rec unaligned,
2 my_first_bits(0:7) bit(01),
2 other_bytes char(06);
But I've never worked in a high level language that explicitly allowed
naming and referencing nibbles. Although I suppose C++ allows some kind of
lashup.
If I can't touch it, it doesn't exist.
>
>--
>I |\ Randall Bart mailto:Bart...@usa.spam.net
>L |/
>o |\ Bart...@worldnet.att.spam.net Bart...@hotmail.spam.com
>v | \ 1-818-985-3259 Please reply without spam
>e |\
>Y |/ Panic in the Year Zero Zero: http://members.aol.com/PanicYr00
>o |\
>u |/ Have you solved http://members.aol.com/PanicYr00/Sequence.html
Cory Hamasaki
Ben Rosenberg
cory h <kiyoinc at ibm dot net> reported to us that:
: But I've never worked in a high level language that
: explicitly allowed naming and referencing nibbles.
***
So, unlike myself, you never worked in FORTRAN II on an IBM 1620.
(stretching somewhat the words "high level" and "nibble")
--
Ben
mailto: change domain name from netkom.kom to netcom.com
"If you never stall, then you're riding your clutch." / Car-Talk
RandallBart
cory hamasaki wrote:
>
> But I've never worked in a high level language that explicitly allowed
> lashup.
>
I can touch nibbles, therefore they exist. On Unisys, we access them
through a high level language called Cobol.
Consider the following:
01 GROUP-FIELD.
05 MMDDYY PIC 9(6) PACKED.
05 REDEFINES MMDDYY.
10 MM PIC 99 PACKED.
10 DD PIC 99 PACKED.
10 YY PIC 99 PACKED.
05 YYMMDD PIC 9(6) PACKED.
05 REDEFINES YYMMDD.
10 YY PIC 99 PACKED.
10 MM PIC 99 PACKED.
10 DD PIC 99 PACKED.
...
MOVE CORR MMDDYY TO YYMMDD.
This works on Mainframes Designed For Cobol™ but fails miserably on Big
Blue. Why? On Blue there's no way to access subfields of a packed
field, because every packed field gets a sign, even though there is no S
in the PIC. Blue further confuses things by adding an extra digit when
I specify an even number of digits.
There is one small benefit of this: Users who declared their dates as
PIC 9(6) PACKED find that Blue has given them seven digits, so they may
reinterpret their date format as CYYMMDD, where C=0 means 19, C=1 means
20, etc. Curiously, whenever this one potential advantage of Big Blue's
misimplementation of packed data is mentioned, Cory starts into a rant
about how all dates *must* be expanded.
Consider another example:
01 ACCOUNT-RECORD.
05 WHOLE-ACCOUNT-NUMBER PIC 9(14) PACKED.
05 REDEFINES WHOLE-ACCOUNT-NUMBER.
10 INST PIC 9(3) PACKED.
10 APPL PIC 9 PACKED.
10 BRANCH PIC 9(3) PACKED.
10 ACCOUNT PIC 9(7) PACKED.
This exact format is used by at least 1000 banks. How does one
implement an account number like this on Big Blue? Should I just
declare the group as PIC X(9) DISPLAY (two bytes more than I was asking
for)? Then I can't move it to an edited output field. Should I leave
the redefinition out (only one byte more than I was asking for)? Then I
need to move the account number to a DISPLAY work area to pull out the
sub-fields. Should I just declare all these fields DISPLAY (7 bytes
more than I was asking for)? This is the easiest to use, and therefore
what lazy programmers will do.
Quote Amdahl's Law, and compare Unisys packed data format (nibble
organized, unsigned unless you specify S) with Big Blue's half-assed
packed implementation (byte organized, always signed whether you need it
or not).
Jay McInvale
Cory Hamasaki wrote:
>
> They do? I can give bytes names:
>
> dcl 1 mastr_rec unaligned,
> 2 my_first_byte char(01),
> 2 other_bytes char(06);
>
> I can also name bits:
>
> dcl 1 mastr2_rec unaligned,
> 2 my_first_bits(0:7) bit(01),
> 2 other_bytes char(06);
>
> naming and referencing nibbles. Although I suppose C++ allows some kind of
> lashup.
>
> If I can't touch it, it doesn't exist.
>
DCL EIGHTBYTES CHAR(8);
DCL NIBBLES (16) BIT (4) BASED (ADDR(EIGHTBYTES));
each nibble is addressable.
Kevin Martin
In article <34a85...@news3.ibm.net>,
cory hamasaki <kiy...@ibm.XOUT.net> wrote:
>
>This is an amazing thing. Who uses Unisys machines? I've never seen one,
>never worked in a shop with them. Must make for some hard work in doing
>data exchange.
I bet you've worked with 'em -- it's just the name that's throwing you.
They weren't called Unisys back then; they were Burroughs.
And they were neat. My first hands-on with networking was something
called NDL, but I made one talk SNA once... hacked together something to
tickle the right registers (in COBOL!), hooked it up to a leased line, and
shazam, it worked. Cool thing is that they used EBCDIC, just like your
"real" computers. Took a while to make the shift to ASCII when the shop
went to VAXes.
RandallBart
cory hamasaki wrote:
>
> This is an amazing thing. Who uses Unisys machines?
People who have had Unisys machines for a long time. Historically,
Burroughs was big in banks and hospitals. Unfortunately, Unisys sells
very few computers to new customers.
> I've never seen one,
> never worked in a shop with them. Must make for some hard work in doing
> data exchange.
Data exchange is done in display format, usually in EBCDIC, but
sometimes in ASCII. There's a file attribute to automatically convert
tape output from EBCDIC to ASCII. Just like MVS, it applies to tape
files only, and an entire multi-file reel must be all ASCII.
Louis Krupp
Burroughs Large Systems (forerunners of the Unisys A-Series) were
used by the US Customs Service, the Detroit Police Department,
Nabisco, the National Security Agency, several universities
(UC San Diego, U of Denver, U of Delaware, Queens University in
Ontario), and lots of other places. I'm not sure who all is using
the A-Series now (I'd be surprised if any were still used in
education).
Burroughs had some good ideas - Extended ALGOL makes it easy to
write threaded code (and has since about 1970), as does the job
control language. The equivalent of a DLL was introduced in about
1980. Array bounds are enforced by the machine architecture.
Exception handlers are easy to write. The database access routines
are custom compiled to match the schema.
Other computer vendors had their users. Burroughs had its fanatics.
Unisys still has its believers, although I don't know any
personally at the moment (I guess I could say I still believe,
although I'm not practicing).
To the best of my knowledge, Y2K will hit Unisys systems just as
hard as the others. If I recall correctly, the 48-bit-word was
just the perfect way to return YYMMDD in EBCDIC.
Louis Krupp
(I've dropped comp.lang.cobol from this followup, and
added comp.sys.unisys.)
cory hamasaki wrote:
> This is an amazing thing. Who uses Unisys machines? I've never seen one,
> never worked in a shop with them. Must make for some hard work in doing
> data exchange.
<snip>
RandallBart
Gary Lee wrote:
> Oh, and as an earlier poster said, you don't pack it unless you are going
> to do arithmetic with it. Only an idiot packs keys.
What's wrong with packing keys? On Burroughs, account numbers, dates,
phone numbers and other numeric data were always packed. Why should I
use eight bits where four will do?
cory hamasaki
This is an amazing thing. Who uses Unisys machines? I've never seen one,
never worked in a shop with them. Must make for some hard work in doing
data exchange.
In <6898n7$n...@bgtnsc03.worldnet.att.net>, RandallBart <Bart...@worldnet.att.spam.net> writes:
>cory hamasaki wrote:
>>
>> But I've never worked in a high level language that explicitly allowed
>> naming and referencing nibbles. Although I suppose C++ allows some kind of
>> lashup.
>>
>> If I can't touch it, it doesn't exist.
>
>--
>I |\ Randall Bart mailto:Bart...@usa.spam.net
>L |/
>o |\ Bart...@worldnet.att.spam.net Bart...@hotmail.spam.com
>v | \ 1-818-985-3259 Please reply without spam
>e |\
>Y |/ Panic in the Year Zero Zero: http://members.aol.com/PanicYr00
>o |\
>u |/ Have you solved http://members.aol.com/PanicYr00/Sequence.html
Yes, unless doing computation on the digits, S/390 programmers then to keep
the data in zoned decimal. This is standard practice.
Cory Hamasaki
cory hamasaki
In <MPG.f11e57ee...@news.teleport.com>, webm...@mcinvale.com (Jay McInvale) writes:
> Cory Hamasaki wrote:
>>
>> They do? I can give bytes names:
>>
>> dcl 1 mastr_rec unaligned,
>> 2 my_first_byte char(01),
>> 2 other_bytes char(06);
>>
>> I can also name bits:
>>
>> dcl 1 mastr2_rec unaligned,
>> 2 my_first_bits(0:7) bit(01),
>> 2 other_bytes char(06);
>>
>> naming and referencing nibbles. Although I suppose C++ allows some kind of
>> lashup.
>>
>> If I can't touch it, it doesn't exist.
>>
>
>DCL EIGHTBYTES CHAR(8);
>DCL NIBBLES (16) BIT (4) BASED (ADDR(EIGHTBYTES));
>
>each nibble is addressable.
>
Yes, you and Barticus are right but I still can't do what I want which is:
nibble(4) = 'F'X;
I think PL/AS, PL/X, whatever has this kind of construct but S/390
addresses bytes... even at the assembly language level. So, yes, in PL/AS
there is a good solution
This thread, which for some reason is being scattered around, will
hopefully help someone in the next couple years. Lots of low level, data
access issues to consider.
Cory Hamasaki
Gary Lee
Kevin Martin <can...@nic.com> wrote in article <689u2v$klr$1...@nic.com>...
one,
> >never worked in a shop with them. Must make for some hard work in doing
> >data exchange.
>
> They weren't called Unisys back then; they were Burroughs.
Actually, I thought they were called Univacs. Burroughs were those other
guys who made those big glorified typewriters you used to see in little
Savings and Loans in tiny towns in the middle of nowhere. They'd put your
Passbook in, it would shake around for about ten minutes, and then print
your new balance.
Seriously though, COBOL guys should like Burroughs, and Univacs, and Vax's,
and all those other things COBOL ran on (Data Points, Osbornes, Apple IIs,
Alpha Micros...all those things that are no longer made). It proves just
what a High Level Portable language should do. Try that with Visual Basic!
Oh, and as an earlier poster said, you don't pack it unless you are going
to do arithmetic with it. Only an idiot packs keys.
-----------------
"Twenty years of schoolin' and they put you on the day shift."
Bob Dylan
---------------- snip --------------------------------
Judson McClendon
cory hamasaki <kiy...@ibm.net> wrote:
>
> This is an amazing thing. Who uses Unisys machines? I've never seen one,
> never worked in a shop with them. Must make for some hard work in doing
> data exchange.
Cory, one of the things we non-IBM'ers laugh about is how IBM folks seem
totally unaware that there is a whole world of computers out here that Big
Blue never made. The Burroughs/Unisys machines (I use that term to signify
the Burroughs machines now made by Unisys) are *significantly* easier to use
and support than comparable IBM machines. System Programmers are almost
never needed, and the position hardly exists in B/U shops. Every programmer
I know who has significant experience in both IBM and B/U shops laughs or
complains (depending on which kind of shop they're currently working in)
about how many gyrations IBM programmers have to go through to do things
which are absurdly simple on B/U machines. This sign issue is a typical
example. On any of the B/U machines, if you want a sign, you can have a
sign, if you don't you don't. You can even have two 1 digit comp fields in a
single byte, or a 3 digit field and a 5 digit field can share 4 bytes. The
IBM 'solution' is a true kludge that has been there for decades. What idiot
came up with the idea that you should allocate a sign even for unsigned
fields? Have you considered how many billions (trillions?) of bytes of
memory and disk have been wasted by this?
Data exchanges are no problem, because on B/U machines, you can easily read
or create files just about any way you want. Sign left/right/separate,
display, packed, decimal, binary, no problem. The fact that you view this
issue as a serious problem illustrates my point. Data conversion and
interfacing is a long time area of expertise for me, and usually my worst
problem is finding out from the IBM users exactly what their data format is.
--
Judson McClendon This is a faithful saying and worthy of all
Sun Valley Systems acceptance, that Christ Jesus came into the
judm...@mindspring.com world to save sinners (1 Timothy 1:15)
(please remove zzz from email id to respond)
Dave Eastabrook
on Mon, 29 Dec 1997 Jay McInvale <webm...@mcinvale.com> wrote
<snip = Cory's eg.
>>
>Cory, try this
>
>DCL EIGHTBYTES CHAR(8);
>DCL NIBBLES (16) BIT (4) BASED (ADDR(EIGHTBYTES));
>
>each nibble is addressable.
>
Yes, this works nicely, but technically Cory is 100% correct for the IBM
360 etc. architecture, since the lowest addressable unit of store is a
byte. Quote from my browned, battered and (ah aaah aaahh-chooii) dusty
360 POPS (principles of operations):
<QUOTE>
"Information Formats"
"The system transmits information between main storage and the CPU in
units of eight bits, or a multiple of eight bits at a time. Each eight-
bit unit of information is called a byte, the basic building block of
all formats. A ninth bit, the parity or check bit, is transmitted with
each byte and carries ..."
</quote>
Bits or non-byte sized groupings of bits are not directly addressable.
Operations can be carried out on an individual bit, or combinations of
bits only by using a mask to address a whole byte or number of bytes, as
in NI SWITCH,X'7F' or XI DEC+L'DEC-1,X'0F' or OC DEC,=XL4'F0F0F0F0'
There are also operations such as MVO (Move with Offset) which cause
operations to be performed on data 4 bits at a time (the senior,
leftmost or high-order; and/or junior, rightmost or low-order 4 bits of
a byte). These operations, however, are still addressing and
manipulating data by the byte, not the bit (or nibble).
High-level languages, being translated to low-level ones, are subject to
the same physical limitations. They can of course have a logical
terminology of their own. A "nibble" therefore, is not a physical
description, but a logical one, and physically addresses the whole byte
in the same way as in assembler or machine code.
In Jay's example, an operation on NIBBLE(5), will therefore physically
affect the whole third byte of EIGHTBYTES, though the low-order 4 bits
will remain unchanged after the operation (if the compiler works!).
:Dave [I've reinstated c.l.c since part of this thread was there]
--
"They can take our 029's, they can take our 2314's, but they can never take...
OUR GEEKDOM!!!!!" Ron Kenyon quotes Braveheart Eastabrook. That'll do for me.
Dave Eastabrook. Elmbronze Ltd. Y2K Consultant. 24+ years IBM Mainframes.
http://www.elmbronze.demon.co.uk/year2000/ also /IBM/ or /telework/
Joseph Cotton
>
> Consider another example:
>
01 Date-field.
05 YY-MM-DD PIC S9(6) COMP-3 VALUE 971229.
05 C-YY-MM-DD REDEFINES YY-MM-DD PIC S9(7) COMP-3.
This is the most common type of date field in our system.
But, in reality (machine interpretation), this field looks like the
following:
09 71 22 9C
The field is four bytes long, and the left hand nibble is zero. Try it
yourself. You can check for this nibble by comparing to a large number;
IF C-YY-MM-DD > 1000000 AND < 2000000
You can set the nibble from 0 to 1 by adding.
ADD 1000000 TO C-YY-MM-DD.
And unset it by subtracting.
SUBTRACT 1000000 FROM C-YY-MM-DD.
BUT... is this too elementary? If you want to, you may use the left
most nibble as a year 2000 century marker, as IBM did in their CICS date
routines.
Mauricio Freitas
Strange... I read a lot of people asking things like "Who uses Unisys
machine?". Other guy wrote sometime ago "I used to work with Unisys PC at
high school... I didn't know Unisys is still in business" (this guy probably
is 25 - 30 years old). I have been working with Unisys A Series since '85
(few years, compared with some persons I know), and it's strange how a lot
of persons don't realize what Unisys did and still does!
In Brazil, until 2 years ago, Unisys had some 500 - 600 A Series machines.
On the world you can find they on Communication, Transportation and mainly
in Financial services.
Louis Krupp wrote in message <34A89AAF...@netONE.com>...
>Burroughs Large Systems (forerunners of the Unisys A-Series) were
>used by the US Customs Service, the Detroit Police Department,
>Nabisco, the National Security Agency, several universities
>(UC San Diego, U of Denver, U of Delaware, Queens University in
>Ontario), and lots of other places. I'm not sure who all is using
>the A-Series now (I'd be surprised if any were still used in
>education).
>
>Louis Krupp
>
>(I've dropped comp.lang.cobol from this followup, and
>added comp.sys.unisys.)
>
Gary Lee
RandallBart <Bart...@worldnet.att.spam.net> wrote in article
<68a7lc$1...@mtinsc05.worldnet.att.net>...
> Gary Lee wrote:
>
> > Oh, and as an earlier poster said, you don't pack it unless you are
going
> > to do arithmetic with it. Only an idiot packs keys.
>
> phone numbers and other numeric data were always packed. Why should I
> use eight bits where four will do?
One general reason which keeps popping up again and again in systems I work
on is, what happens when the user decides that they need to put alpha
characters into their key? This happens with all sorts of situations, like
1) We ran out of numbers
2) We merged with another company, and they use alphas
3) We got a new regulation/industry standard/inspiration/communication from
God/yada, yada, yada
4) We got bad data from an outside source, and had some poor flunky ram it
into the system because we didn't have time to clean it
5) Letters are easier to remember than numbers, so we want to change
6) The new manager used to work for our competitor, and they had alphas
there, so he wants them here
7) We got sued by a member of a religious group which considers
identification numbers a sign of Satan
etc., etc., etc.
A key is just a string of bytes to distinguish one
record/row/occurrence/tuple from another. A number is a representation of
a magnitude, on which I can do arithmetic. While I agree that it may be
more efficient to pack keys with numeric characters in them, when I compare
the cost of disk (roughly 5 cents/megabyte) to the cost of converting a
file, the programs which use it, etc., I don't think you can justify
packing a key anymore.
Oh, and don't forget 8) The user decided they want to be able to move
around the period/dash/slash/colon/tilde in the key, so they can handle
the foreign telephone/postal/diagnosis/social security codes which we might
use if we ever do business in a foreign country.
All of these are things which have happened where I've been. You're very
lucky if they haven't happened to you.
David K. Bryant
"Judson McClendon" <judm...@mindspring.com> writes:
>Cory, one of the things we non-IBM'ers laugh about is how IBM folks seem
>totally unaware that there is a whole world of computers out here that Big
>Blue never made. The Burroughs/Unisys machines (I use that term to signify
>the Burroughs machines now made by Unisys) are *significantly* easier to use
>and support than comparable IBM machines. System Programmers are almost
>never needed, and the position hardly exists in B/U shops. Every programmer
>I know who has significant experience in both IBM and B/U shops laughs or
>complains (depending on which kind of shop they're currently working in)
>about how many gyrations IBM programmers have to go through to do things
>which are absurdly simple on B/U machines.
All the above applies to Wangs, too.
Every IBMer I've ever know that has moved to a Wang refuses to go back.
Jeff York
"Judson McClendon" <judm...@mindspring.com> wrote:
> Have you considered how many billions (trillions?) of bytes of
>memory and disk have been wasted by this?
Big Blue never seemed too concerned about saving resources.. Remember
the old comment, "an elephant is a mouse with an IBM operating
system"..? :-)
--
Jeff.
je...@jakfield.xu-netx.com (remove the x..x round u-net for return address)
... There's pleasure sure in being mad
That none but madmen know...
Dryden
RandallBart
Gary Lee wrote:
>
> RandallBart <Bart...@worldnet.att.spam.net> wrote in article
> <68a7lc$1...@mtinsc05.worldnet.att.net>...
> > What's wrong with packing keys? On Burroughs, account numbers, dates,
> > phone numbers and other numeric data were always packed. Why should I
> > use eight bits where four will do?
> --------------- snip ---------------------------
> One general reason which keeps popping up again and again in systems I work
> on is, what happens when the user decides that they need to put alpha
> characters into their key?
Specious argument. Date and time will never contain non-numerics.
Phone numbers will grow longer before they become non-numeric. Bank
account numbers are defined to be numeric in all the interchange
formats. By the same argument you should leave four bytes of slop in
every field in case the user decides he wants it bigger.
> While I agree that it may be
> more efficient to pack keys with numeric characters in them, when I compare
> the cost of disk (roughly 5 cents/megabyte) to the cost of converting a
> file, the programs which use it, etc., I don't think you can justify
> packing a key anymore.
"Anymore". Processing time and disk space have become so cheap that I
would advocate storing everything in a database structure where you can
expand fields without recompiling. But most systems were designed for
the data economics of 10 or more years ago. What would you think of
packing account numbers, dates, phones numbers, etc in 1980?
> All of these are things which have happened where I've been. You're very
> lucky if they haven't happened to you.
I've run into most of them. I've also run into "we need four more
bytes".
The Goobers
RandallBart wrote:
>
[snip]
> By the same argument you should leave four bytes of slop in
> every field in case the user decides he wants it bigger.
>
[snip]
>
> I've run into most of them. I've also run into "we need four more
> bytes".
As my first programming instructor told us:
Q: What is the first thing a programmer does when given an 80-byte
record?
A: Reserve 10 bytes FFU (For Future Use)
... and that reminds me of a time, years back, when a co-worker stopped
by my cube and asked me to take a glance at some design-specs; I noticed
that nothing was expandable and commented on this rather...
derogatorily. As luck would have it the PHM (not an Anderoid but
close... a PM'er, before they became PMM, before they became KMPG)
responsible for that project strolled by and, seeing that His Worker was
talking with someone from Another Project (and a wretched independent
consultant, at that!), stopped by... and the following interchange
ensued:
Me: 'Ummmm, Mr Jones... I've noticed that this system you're working on
here doesn't allow for expansion.'
Mr J: 'That's right, we've really slimmed it down to the metal... it
won't ever need to be expanded!'
Me: 'Mr Jones, I am confused.'
Mr J: 'Why are you confused? All ya gotta do is keep in mind that...'
Me: 'No, no, Mr Jones... all I keep in mind is that in my limited
experience I've seen two kinds of projects... those which get expanded
and those that fail. You've allowed no room for expansion... such a
Public Declaration of your faith in this system is very brave.'
Mr J: 'Yup, that's what we do here... we make new paradigms!'
DD
Judson McClendon
Jeff York <je...@jakfield.xu-netx.com> wrote:
>
> "Judson McClendon" <judm...@mindspring.com> wrote:
> >
> > Have you considered how many billions (trillions?) of bytes of
> > memory and disk have been wasted by this?
>
> Big Blue never seemed too concerned about saving resources.. Remember
> the old comment, "an elephant is a mouse with an IBM operating
> system"..? :-)
Very true, Jeff. After all, the more resources you waste, the more Big Blue
hardware you must buy.
kre...@rocketmail.com
In article <dbryantE...@netcom.com>,
And it applies to other platforms as well. After 27 years in the non-IBM
mainframe world (NCR, Honeywell/Bull, Unisys, etc), I made the move this
past year to an IBM shop. I still cannot believe how difficult and time
consuming it is to accomplish various tasks that were no-brainers in my
previous experience. Perhaps the various operating systems and systems
utilities on these platforms were designed by ex-IBM-ers who said "never
again!" :-)
Kreskin
-------------------==== Posted via Deja News ====-----------------------
http://www.dejanews.com/ Search, Read, Post to Usenet
kre...@rocketmail.com
> Louis Krupp wrote in message <34A89AAF...@netONE.com>...
> >Burroughs Large Systems (forerunners of the Unisys A-Series) were
> >used by the US Customs Service, the Detroit Police Department,
> >Nabisco, the National Security Agency, several universities
> >(UC San Diego, U of Denver, U of Delaware, Queens University in
> >Ontario), and lots of other places. I'm not sure who all is using
> >the A-Series now (I'd be surprised if any were still used in
> >education).
FYI, Queen's University still uses it.
> >
> >Louis Krupp
Bill Lynch
RandallBart wrote:
>
> Gary Lee wrote:
> >
> > RandallBart <Bart...@worldnet.att.spam.net> wrote in article
> > <68a7lc$1...@mtinsc05.worldnet.att.net>...
>
> > > What's wrong with packing keys? On Burroughs, account numbers, dates,
> > > phone numbers and other numeric data were always packed. Why should I
> > > use eight bits where four will do?
> > --------------- snip ---------------------------
> > One general reason which keeps popping up again and again in systems I work
> > on is, what happens when the user decides that they need to put alpha
> > characters into their key?
>
> Specious argument.
Specious is as specious does. Despite their names, account nos. (from
your own example) aren't always numeric. Most businesses use account
nos. and most businesses are not banks. Pool nos. in mortgage backed
securities were numeric for several years, then changed in 1991, or so,
when Freddie Mac started issuing "gold" pool nos.
Gary's for instance about sometime wanting to do business globally is an
excellent example, IMHO. Domestic broker (FINS) nos. are numeric. Global
broker nos. are alpha. It's possible to have negative interest on a
global trade, unlike a domestic one (speaking USA-centric, here). This
is obviously not limited to keys, only.
Have you ever worked on a system which uses (US) taxpayer ID no. (this
is the same as Social Security no. for individual taxpayers)? Many of us
have. How would you distinguish a corporate TIN from an individual one
(hint, they are in different formats and contain characters other than 0
thru 9))?
I worked for a major investment banking firm in the 80's which was
converting from being a Burroughs shop to an IBM one (gee, I wonder
why?). What you, and others, have said about Burroughs equipment is
true. It's *much* easier to use (along with CDC, and probably others) -
people who were familiar with Burroughs' Work Flow Language were stunned
when they got a hint of MVS's Job Control Language. Burroughs had
virtual storage years before IBM (maybe from the 50's, maybe from the
5500, my memory isn't clear here). Unfortunately, their marketing people
had all the market savvy of the Edsel group, while IBM could sell sand
to the Arabs. It's not really surprising, is it, that so many mainframe
people are IBM-centric? Maybe you could save your vitriol for something
more appropriate.
Bill Lynch
(balance snipped)
RonKenyon
DD wrote:
>Me: 'No, no, Mr Jones... all I keep in mind is that in my limited
>experience I've seen two kinds of projects... those which get expanded
>and those that fail. You've allowed no room for expansion... such a
>Public Declaration of your faith in this system is very brave.'
>
>Mr J: 'Yup, that's what we do here... we make new paradigms!'
Yup, and where I come from a buck-eighty-five and a couple shiny new paradigms
will buy you a cup o' latte.
-- Ron "Shift Happens" Kenyon
Peter Hesketh
In article <dbryantE...@netcom.com>, "David K. Bryant"
<dbr...@netcom.com> writes
>Every IBMer I've ever know that has moved to a Wang refuses to go back.
Who makes Wangs nowadays then? Are you talking VS, OIS or 2200?
--
Regards, Peter Hesketh
Mynyddbach, Monmouthshire, UK
"Just because I'm moody doesn't mean you're not irritating."
Judson McClendon
Bill Lynch <WBL...@worldnet.att.net> wrote:
>
> I worked for a major investment banking firm in the 80's which was
> converting from being a Burroughs shop to an IBM one (gee, I wonder
> why?). What you, and others, have said about Burroughs equipment is
> true. It's *much* easier to use (along with CDC, and probably others) -
> people who were familiar with Burroughs' Work Flow Language were stunned
> when they got a hint of MVS's Job Control Language. Burroughs had
> virtual storage years before IBM (maybe from the 50's, maybe from the
> 5500, my memory isn't clear here). Unfortunately, their marketing people
> had all the market savvy of the Edsel group, while IBM could sell sand
> to the Arabs. It's not really surprising, is it, that so many mainframe
> people are IBM-centric? Maybe you could save your vitriol for something
> more appropriate.
Bill, what you say is true. Sadly, Burroughs and now Unisys, don't seem to
have a clue in the marketing department. However, the discussion was on
technical issues. Had the previous comments been made about sales, your last
sentence would have been more apropos. Do you suggest IBM is to be
congratulated because they have foisted an inferior, inefficient technology
on customers through slick marketing and 'creative deception'? In this
context, "being IBM-centric" is quite similar to "being brainwashed". All
the hordes of IBM-centric folks can stand up and proudly say "Yes, we know we
waste huge amounts of time, energy and money doing the same things, but IBM
has a great sales force!" ;-)
Ron Martell
RandallBart <Bart...@worldnet.att.spam.net> wrote:
<snip>
>"Anymore". Processing time and disk space have become so cheap that I
>would advocate storing everything in a database structure where you can
>expand fields without recompiling.
<snip>
That was the feature that I just loved about the old Revelation PC
database (based largely on PICK).
A data field was simply defined as the space between two field marks
(ASCII 253 if memory serves me correctly) with an absolute limit of
64K per field.
Ron Martell Duncan B.C. Canada
Microsoft MVP
On-Line Help Computer Service
http://www.islandnet.com/~rmartell/online.htm
"When the only tool you own is a hammer, every problem you encounter
resembles a nail."
Bill Lynch
Judson McClendon wrote:
>
> Bill Lynch <WBL...@worldnet.att.net> wrote:
> >
>
(snippage)
> Bill, what you say is true. Sadly, Burroughs and now Unisys, don't seem to
> have a clue in the marketing department. However, the discussion was on
> technical issues. Had the previous comments been made about sales, your last
> sentence would have been more apropos.
I was trying to suggest why so many mainframe programmers are
IBM-centric, it's all they've ever seen, been trained for, and have the
opportunity to work on.
> Do you suggest IBM is to be congratulated because they have foisted an inferior, inefficient technology
> on customers through slick marketing and 'creative deception'?
God, no! (Don't forget "Account Control") Do keep in mind that our
society rewards the IBMs and their executives with praise, money and
status. They are "successes" and the Honeywells & Burroughs are
"failures". Neither you (AFAIK) nor I made these rules. Microsoft might
not make the best products, but if you want to earn a living doing PC
software, I think you'd better know some MS products. Both IBM & MS
gained considerable muscle in their industries and they flexed them. How
else would you have our system work?
> In this context, "being IBM-centric" is quite similar to "being brainwashed".
Why? If the auto business were structured like the mainframe business,
with GM playing the IBM role, most people wouldn't know much about
Fords, would they? I don't see where that make them brainwashed.
> All the hordes of IBM-centric folks can stand up and proudly say "Yes, we know we
> waste huge amounts of time, energy and money doing the same things, but IBM
> has a great sales force!" ;-)
They could, but why would they? <g> Are you suggesting that the people
who've been trained on IBM equipment should refuse to work until a
Burroughs slot came up? I wouldn't. Ultimately, IMHO, IBM is like Mt.
Everest, "because it's there".
Bill Lynch
David K. Bryant
Peter Hesketh <p...@phesk.demon.co.uk> writes:
>In article <dbryantE...@netcom.com>, "David K. Bryant" writes
>>Every IBMer I've ever know that has moved to a Wang refuses to go back.
>Who makes Wangs nowadays then? Are you talking VS, OIS or 2200?
Is this a trick question? Wang makes Wangs.
Yes, they've recovered enough to introduce some new CPUs.
Specifically, my comment was about the VS product line.
I think the OIS & 2200 line are now out of production.
Tony Toews
The Goobers <docd...@erols.com> wrote:
>Me: 'No, no, Mr Jones... all I keep in mind is that in my limited
>experience I've seen two kinds of projects... those which get expanded
>and those that fail. You've allowed no room for expansion... such a
>Public Declaration of your faith in this system is very brave.'
>
>Mr J: 'Yup, that's what we do here... we make new paradigms!'
Paradigms. I despise that term. It's indicative of everything PHM's
stand for. YUCK!
Tony
----
Message posted to newsgroup and emailed.
Tony Toews, Independent Computer Consultant
The Year 2000 crisis: Will my parents or your grand parents still be receiving
their pension in January, 2000? See www.granite.ab.ca/year2000 for more info.
Microsoft Access Hints, Tips & Accounting Systems at www.granite.ab.ca/accsmstr.htm
Mark0Young
I hope I got the author right (I am quoting from a quote):
In article <01bd15b4$e2f6e5d0$1885aac7@sysserve01>, "Gary Lee"
<nsp....@nsp.millcomm.nsp.com> writes:
> What's wrong with packing keys? On Burroughs, account numbers,
> dates, phone numbers and other numeric data were always packed.
> Why should I use eight bits where four will do?
One problem we had with packed keys was that one application language would use
preferred signs (X'C' for "positive" or zero) and another application language
on the same platform would "strip" the sign (X'F' for "positive" or zero), so
records written by one application language could not be randomly read by the
other application language. We have had no such problems with "unsigned" zone
decimal (i.e., digit characters).
We are still a multi-languaged platform so we are still cautious about packing
fields that are in keys.
Tony Toews
"Gary Lee" <nsp....@nsp.millcomm.nsp.com> wrote:
>> What's wrong with packing keys? On Burroughs, account numbers, dates,
>> phone numbers and other numeric data were always packed. Why should I
>> use eight bits where four will do?
>--------------- snip ---------------------------
>One general reason which keeps popping up again and again in systems I work
>on is, what happens when the user decides that they need to put alpha
>characters into their key? This happens with all sorts of situations, like
>1) We ran out of numbers
<snip>
>7) We got sued by a member of a religious group which considers
>identification numbers a sign of Satan
All valid points.
>While I agree that it may be
>more efficient to pack keys with numeric characters in them, when I compare
>the cost of disk (roughly 5 cents/megabyte) to the cost of converting a
>file, the programs which use it, etc., I don't think you can justify
>packing a key anymore.
Anymore. However, obviously, there is a lot of software out there at
sites which depend on the key being packed. Difficult task to change
the database and recompile all related software. Hmmm, sounds a lot
like expansion when doing Y2K.
>Oh, and don't forget 8) The user decided they want to be able to move
>around the period/dash/slash/colon/tilde in the key, so they can handle
>the foreign telephone/postal/diagnosis/social security codes which we might
>use if we ever do business in a foreign country.
Ah yes, the assumptions present in so much software produced in the
U.S. 5-4 zip codes 3-3-4 telephone codes, 3-3-3 Social security codes
and so on and so forth.
Mind you the Americans are getting better at this. I don't see
anywhere near so many Americanisms.
Oh, and one of the stupidest things done in North America was to have
the phone number format hard coded at 3-3-4. Look at all the crap
which is going on in terms of new area codes and so forth. A variable
phone number system would have saved those problems. Idiots. The
rest of the world had that one right!
Judson McClendon
Bill Lynch <WBL...@worldnet.att.net> wrote:
>
> Judson McClendon wrote:
> >
> > Do you suggest IBM is to be congratulated because they have foisted
> > an inferior, inefficient technology on customers through slick
> > marketing and 'creative deception'?
>
> God, no! (Don't forget "Account Control") Do keep in mind that our
> society rewards the IBMs and their executives with praise, money and
> status. They are "successes" and the Honeywells & Burroughs are
> "failures". Neither you (AFAIK) nor I made these rules. Microsoft might
> not make the best products, but if you want to earn a living doing PC
> software, I think you'd better know some MS products. Both IBM & MS
> gained considerable muscle in their industries and they flexed them.
> How else would you have our system work?
I'm not suggesting that the system should work differently. I am suggesting that the IBM programming community show significant naiveté in their view toward the technical merits of IBM computers and software vs. other computing systems. In a field as highly technical and demanding as ours, I hardly think this is wise. Surely our culture should have learned by now to take sales hype less seriously, and check for ourselves?
> > In this context, "being IBM-centric" is quite similar to "being
> > brainwashed".
>
> Why?
What else would you call holding wrong opinion by reason of believing sales propaganda? The Microsoft Bookshelf dictionary gives this as the #2 definition of "brainwashing":
"The application of a concentrated means of persuasion, such as an
advertising campaign or repeated suggestion, in order to develop a
specific belief or motivation."
I think "brainwashed" hits it right on the nose.
> If the auto business were structured like the mainframe business,
> with GM playing the IBM role, most people wouldn't know much about
> Fords, would they? I don't see where that make them brainwashed.
Actually, if the auto industry were like that, Ford would be the ONLY auto manufacturer, because GM would have been stillborn. :-)
> > All the hordes of IBM-centric folks can stand up and proudly say
> > "Yes, we know we waste huge amounts of time, energy and money doing
> > the same things, but IBM has a great sales force!" ;-)
>
> Are you suggesting that the people who've been trained on IBM equipment
> should refuse to work until a Burroughs slot came up?
Of course not, that was irony. ;-) What I am suggesting is that they should, for their own knowledge and understanding if nothing else, look beyond IBM's hype and learn the real truth about these matters. Who benefits by lack of knowledge? None of this information is classified, it's freely available. But like sheep, they are being unwittingly and willingly controlled and indoctrinated by IBM. Sure, the programmer in the trenches has little to do with the kind of computer his company buys, but many of those programmers will be making purchase decisions someday, and even more of them will be influencing those decisions. And, if anything, management is more brainwashed than the techies. Management may not have the background to understand, but the techies don't have that excuse. Only a few of the 42 computer languages and dialects I have used were learned formally. I learned most of them by studying on my own. Now I make more money than the managers who hire me as consultant. It is an axiom that "knowledge is power". I just think it behooves the IBM user community to be more knowledgeable about the computer systems they are using relative to those available. Is that bad? :-)
Michael Dodas
Tony Toews wrote:
>
> The Goobers <docd...@erols.com> wrote:
>
> >Me: 'No, no, Mr Jones... all I keep in mind is that in my limited
> >experience I've seen two kinds of projects... those which get expanded
> >and those that fail. You've allowed no room for expansion... such a
> >Public Declaration of your faith in this system is very brave.'
> >
> >Mr J: 'Yup, that's what we do here... we make new paradigms!'
>
> Paradigms. I despise that term. It's indicative of everything PHM's
> stand for. YUCK!
>
> Tony
> ----
It's also the cutsie word thrown out by the software industry with
regularity. Want to start a trend on your product or try to bring
attention to a past release that wasn't so hot? Let's call it a
paradigm shift.
--
Mike Dodas
(Remove cutthejunk for e-mail replies)
Dave Eastabrook
on Thu, 1 Jan 1998 Tony Toews <tto...@telusplanet.net> wrote
>The Goobers <docd...@erols.com> wrote:
>
>>Me: 'No, no, Mr Jones... all I keep in mind is that in my limited
>>experience I've seen two kinds of projects... those which get expanded
>>and those that fail. You've allowed no room for expansion... such a
>>Public Declaration of your faith in this system is very brave.'
>>
>>Mr J: 'Yup, that's what we do here... we make new paradigms!'
>
>Paradigms. I despise that term. It's indicative of everything PHM's
>stand for. YUCK!
>
:Dave:)
Daniel P. B. Smith
In article <dbryantE...@netcom.com>,
David K. Bryant <dbr...@netcom.com> wrote:
>Peter Hesketh <p...@phesk.demon.co.uk> writes:
>
>>In article <dbryantE...@netcom.com>, "David K. Bryant" writes
>>>Every IBMer I've ever know that has moved to a Wang refuses to go back.
>
>>Who makes Wangs nowadays then? Are you talking VS, OIS or 2200?
>
>Is this a trick question? Wang makes Wangs.
>Yes, they've recovered enough to introduce some new CPUs.
Remember the user-society T-shirts when they introduced the VS-5000 (which
was about the size of a PC on steroids)? "It's not how big your Wang is,
it's what you do with it."
--
Daniel P. B. Smith
dpbs...@world.std.com
Daniel P. B. Smith
In article <01bd16b9$1150a2e0$8cad45cf@juddesk>,
Judson McClendon <judm...@mindspring.com> wrote:
>> >
>> > Do you suggest IBM is to be congratulated because they have foisted
>> > an inferior, inefficient technology on customers through slick
>> > marketing and 'creative deception'?
>>
>> God, no! (Don't forget "Account Control") Do keep in mind that our
>> society rewards the IBMs and their executives with praise, money and
>> status. They are "successes" and the Honeywells & Burroughs are
>> "failures".
>I'm not suggesting that the system should work differently. I am
>suggesting that the IBM programming community show significant naiveté in
>their view toward the technical merits of IBM computers and software vs.
>other computing systems. In a field as highly technical and demanding as
>ours, I hardly think this is wise. Surely our culture should have learned
>by now to take sales hype less seriously, and check for ourselves?
One of the reasons I can't bring myself to hate Microsoft as much as I
hate IBM is that IBM used its power to foist software on the computing
world that was technically quite bad and set back progress considerably.
Microsoft uses it's power just as objectionably, but IMHO their software
is not as bad as IBM's. It's usually pretty usable. Often it is
half-decent, sometimes even three-quarters or nine-tenths decent. At it's
worst, it may be half-assed, sometimes quarter-assed, but rarely less. A
chain is as strong as its weakest link, and Microsoft makes very good
chains. Their software is stale, bloated, derivative. They have an
attitude toward bugs and software quality that I detest. They care very
little about either forward or backward compatibility (my colleagues still
send me files that Microsoft's software claims are Word 6 files that, in
reality, are RTF files... what a mess).
But they haven't been nearly as much of a force for _technical_ evil as
IBM has.
Peter Hesketh
In article <EM48A...@world.std.com>, "Daniel P. B. Smith"
<dpbs...@world.std.com> writes
>Remember the user-society T-shirts when they introduced the VS-5000 (which
>was about the size of a PC on steroids)? "It's not how big your Wang is,
>it's what you do with it."
Many years ago, when they re-vamped their Customer Service approach,
they named the new offering "WangCare". In the UK they changed it to
"Wanguard" after a month. Odd, that.
Bill Lynch
Hmmm, okay. My Webster's shows the following for "brainwashing":
(vt.) [Colloq.] to indoctrinate so intensively and thoroughly as to
effect a radical transformation of beliefs and mental attitudes
(C)1995 Zane Publishing, Inc. (C)1994, 1991, 1988 Simon & Schuster,
Inc.
I was taking brainwashing in this context (as I first encountered the
concept in "The Manchurian Candidate"). Given your definition, it would
seem that anyone who buys a product or service, or casts a vote, etc.,
has been brainwashed. This doesn't seem to be a very useful definition
in the modern world, with all the advertising we're subjected to. BTW,
did you consciously have your definition in mind when you posted, or did
you look it up after the fact & find a suitable definition (this is not
meant as a putdown, I am just wondering)?
> > If the auto business were structured like the mainframe business,
> > with GM playing the IBM role, most people wouldn't know much about
> > Fords, would they? I don't see where that make them brainwashed.
>
> Actually, if the auto industry were like that, Ford would be the ONLY
> auto manufacturer, because GM would have been stillborn. :-)
No
>
> > > All the hordes of IBM-centric folks can stand up and proudly say
> > > "Yes, we know we waste huge amounts of time, energy and money
> doing
> > > the same things, but IBM has a great sales force!" ;-)
> >
> > Are you suggesting that the people who've been trained on IBM
> equipment
> > should refuse to work until a Burroughs slot came up?
>
> Of course not, that was irony. ;-) What I am suggesting is that they
> should, for their own knowledge and understanding if nothing else,
> look beyond IBM's hype and learn the real truth about these matters.
Why. If you expect people to devote a substantial portion of their
professional life to doing something, that something better have some
sort of pay back for making the investment. I don't see it from a
practical POV.
> Who benefits by lack of knowledge? None of this information is
> classified, it's freely available. But like sheep, they are being
> unwittingly and willingly controlled and indoctrinated by IBM.
"... like sheep .." I'm coming to the conclusion that you have a fairly
low opinion of those of us who toil in the world of IBM. :-) Seems to me
that this is like blaming the programmers for the Y2K problem.
> Sure, the programmer in the trenches has little to do with the kind of
> computer his company buys, but many of those programmers will be
> making purchase decisions someday, and even more of them will be
> influencing those decisions. And, if anything, management is more
> brainwashed than the techies. Management may not have the background
> to understand, but the techies don't have that excuse. Only a few of
> the 42 computer languages and dialects I have used were learned
> formally. I learned most of them by studying on my own. Now I make
> more money than the managers who hire me as consultant. It is an
> axiom that "knowledge is power". I just think it behooves the IBM
> user community to be more knowledgeable about the computer systems
> they are using relative to those available. Is that bad? :-)
> --
No. Speaking of definitions, Webster's says of "behooves":
(vt.) to be necessary for or incumbent upon [it behooves you to think
for yourself]; (C)1995 Zane Publishing, Inc. (C)1994, 1991, 1988 Simon
& Schuster, Inc.
I don't think it "behooves" us at all. It certainly isn't necessary, is
it? The IBM world and its minions roll merrily along. I can understand
your belief that it would be a better world if we were more
knowledgeable about lots more things, but that's more along the line of
wish fulfillment than being a practical consideration for the real
world.
Also, it also seems to me that if I were an independent contractor and
wanted to maximise my financial opportunities, I'd concentrate on the
most expensive skills on the most common platforms, e.g.,
PC's/Microsoft/C++, IBM/MVS/CICS/DB2. As Barry Goldwater said so many
years ago, "Go hunting where the ducks are".
Anyway, it would certainly be nice if people in general learned more
about what's going on and what's available in their respective fields,
but I don't expect to see it in my lifetime. <g>
Bill Lynch
Bill Lynch
Peter Hesketh wrote:
>
> In article <EM48A...@world.std.com>, "Daniel P. B. Smith"
> <dpbs...@world.std.com> writes
> >Remember the user-society T-shirts when they introduced the VS-5000 (which
> >was about the size of a PC on steroids)? "It's not how big your Wang is,
> >it's what you do with it."
>
> Many years ago, when they re-vamped their Customer Service approach,
> they named the new offering "WangCare". In the UK they changed it to
> "Wanguard" after a month. Odd, that.
ROTFLMAO - Peter, that is great! I know it's only 1998/001, but you're
clearly in the lead for the "Best Quip of the Year" award. Looking
forward to your sebsequent posts.
Bill Lynch
RandallBart
Mark0Young wrote:
>
> I hope I got the author right (I am quoting from a quote):
No, you're quoting me.
>
> In article <01bd15b4$e2f6e5d0$1885aac7@sysserve01>, "Gary Lee"
> <nsp....@nsp.millcomm.nsp.com> writes:
>
> > What's wrong with packing keys? On Burroughs, account numbers,
> > dates, phone numbers and other numeric data were always packed.
> > Why should I use eight bits where four will do?
>
> One problem we had with packed keys was that one application language would use
> preferred signs (X'C' for "positive" or zero) and another application language
> on the same platform would "strip" the sign (X'F' for "positive" or zero), so
> records written by one application language could not be randomly read by the
> other application language. We have had no such problems with "unsigned" zone
> decimal (i.e., digit characters).
>
> We are still a multi-languaged platform so we are still cautious about packing
> fields that are in keys.
Clearly the solution is to use unsigned packed decimal. Who would store
a sign with a packed account number? If you declare a packed field
without a sign and your computer stores it with a sign, throw away that
computer and get one that works.
David K. Bryant
dpbs...@world.std.com (Daniel P. B. Smith) writes:
>David K. Bryant <dbr...@netcom.com> wrote:
>>Peter Hesketh <p...@phesk.demon.co.uk> writes:
>>
>>>Who makes Wangs nowadays then? Are you talking VS, OIS or 2200?
>>
>>Is this a trick question? Wang makes Wangs.
>>Yes, they've recovered enough to introduce some new CPUs.
>Remember the user-society T-shirts when they introduced the VS-5000 (which
>was about the size of a PC on steroids)? "It's not how big your Wang is,
>it's what you do with it."
I recall some coffee cups that (Wang's) Federal Systems Division
made up for a trade show. They were unofficial and only handed
out "under the counter"...
Ask me what I do with my Wang.
The VS-5000 comes in two cabinet sizes: the 5300 case is
the size of a PC AT (6x16x22) and can be arranged as a
tower or a desktop. The 5600 case is only available as a
tower because it is about 10" wide (vs 6"). Each carries
the same CPU card set but the 5600 has twice as many I/O slots.
I have a 5300 at home that can handle a hundred users.
The 5600 at work is running 126 users at the moment but
we're adding another 25-30 soon.
Gary Labowitz
Just couldn't resist....
I worked for IBM in 1962, plugging boards, moved on to 305 RAMAC, 1401,
1410, 1620, etc...
I then worked in Poughkeepsie on Emulators, simulators, sort/merge, OS JCL,
etc...
Let me assure you, we worked as hard as possible to be efficient in space
and time; i.e. we optimised the hell out of code to save cycles and core.
Every release I ever heard of had at least 1000 bugs in it.
There were over 600 programmers in about 10 buildings when I was there.
I also coded LGP-30 (Bendix), Univac I, and other dinosaurs.
Point: Without IBM and what it did for the industry, there wouldn't be any
industry in the USA. Unfortunately, IBM's success nipped the English and
French development.
I have made the switch to PC's (in 1989) because that's what was there.
When I was doing so, a System Administrator at a large site told me "Forget
it, those PC's are just toys."
Point: The PC's and the work Intel and Microsoft have done with them is
pretty much a redevelopment (albeit very slowly and in fits and starts) of
IBM's vision of MVS. Almost all the concepts I have had to dig out of MS's
verbiage describing Win32 maps decently to MVS. There are a few minor
differences.
There just aren't too many different ways to turn a bit on and off. Or do
DAT. Or keep track of jobs/processes. Or keep track of tasks/threads. Or
handle interruptions/messages. etc.
Oh well, in another 500 working days it won't make much difference.
--
Gary (MCT, MCPS, MCSD)
www.enter.net/~garyl/ for references to good books.
Contribute to ga...@enter.net
<<GIANT snip>>
Dr John Stockton
JRS: In article <68h1j3$f...@mtinsc03.worldnet.att.net> of Thu, 1 Jan
1998 16:24:36 in comp.software.year-2000, Bill Lynch
<WBL...@worldnet.att.net> wrote:
>ROTFLMAO - Peter, that is great! I know it's only 1998/001, but you're
>clearly in the lead for the "Best Quip of the Year" award. Looking
>forward to your sebsequent posts.
I don't know whether Peter might know of the tale about the man who many
years ago was asked to paint the initials on the sign for the gate of the
new Admiralty Radar & Signals Establishment, and raised some objection?
--
John Stockton, Surrey, UK. j...@merlyn.demon.co.uk Turnpike v1.12 MIME.
Web URL: http://www.merlyn.demon.co.uk/ -- includes FAQqish topics and links.
Y2k is Leap. My Y2k for beginners http://www.merlyn.demon.co.uk/year2000.txt
My list of Critical Dates is : http://www.merlyn.demon.co.uk/misctime.htm#CDs
Tom Schmidt
David Ray <dr...@cei.net> wrote in article
<01bd0f4a$70ad8b20$6776b4cc@sys>...
> >
> > Then I asked him if he had read the piece. "Yes". I asked " Did
you
> > notice
> > one major thing they left out of the story ?". He said " What did
they
> > leave
> > out ?". I replied " what happens if the DON'T get it fixed ?".
> > He asked " So what happens ?". I said to him "your a smart guy,
> > extropolate
> > and answer your own question.". After a moment he said "
Everything
> > collapses".
> > HOORAY ! Easiest conversion I have ever had. Who's next ? Next
in
> > line, please
> > step up............
>
> Well, except that's really not quite the answer. IRS, like everyone
else,
> has some systems that are more critical than others. In fact, I
can't
> imagine that the vast majority of their systems are even that
important in
> the overall scheme of collecting taxes. For example, the "CAF",
which
> tracks powers of attorney, is a frequently used system which hardly
> functions as it is. What if it quits working? Not much. Just
doesn't
> matter. Now, if BMF or IMF (Business or Individual Master File)
systems
> collapse, THAT could be a problem. But IRS is going to fix those
first.
> So, the ancillary systems like CAF or even the dread DIF
(Discriminant
> Function scoring system) were to collpase, it's really not going to
be that
> big of a problem.
>
> If everyone could be a little less excitable about this problem, it
would
> really help us keep a more accurate perspective.
>
>
> David Ray, CPA
"But IRS is going to fix those first."
If that is true, why aren't those fixed already? Did the IRS get
ambushed by the calendar or something?
Yes, I'm aware that IRS has tried (several times) to replace those
aging systems... and it has failed with each attempt (some bigger
failures than others.) I presume that IRS is aware of its failures,
too.
So if IRS is so darned aware of the issues, why isn't this a done deal
already? Or don't the IRS-issued calendars all say:
"Dates on calendar are closer than they appear"
on the bottom?
The 'accurate perspective' is that the U. S. Internal Revenue Service
has a bad track record implementing massive changes to its data
processing systems... as bad as any company or any other government
agency. They've left too little time to remediate their software suite
and, by their own admission, will not be able to remediate many things
that are used daily. They simply waited until the last minute and then
found that someone had other plans for them during (and after) that
'last minute'. No contingency plan.
But that's not the problem...
The problem is that their largest, most sensitive software systems are
currently stretched to the maximum and simply cannot be taken offline
long enough to swap in remediated code. They need a miracle now.
Personally, I don't blame the IRS itself for the problems ... I believe
that the IRS management over the years is partially responsible, but
the majority of the cause may be due to the micromanagers on Capital
Hill (over the years.)
--
Tom Schmidt
Madison, WI
<jts...@itis.com>
("So much data, so little time.")
Peter Hesketh
In article <mPELHPAi...@merlyn.demon.co.uk>, Dr John Stockton
<j...@merlyn.demon.co.uk> writes
>I don't know whether Peter might know of the tale about the man who many
>years ago was asked to paint the initials on the sign for the gate of the
>new Admiralty Radar & Signals Establishment, and raised some objection?
Thank you for that gem; no, I hadn't seen it before.
Did you know that GEC-Plessey Telecommunications were going to call
themselves GPT throughout the EU until somebody pointed out that in
France it would be pronounced J'ai pe'te'.
The Goobers
Gary Labowitz wrote:
>
> Just couldn't resist....
>
[snip of CV and other credentials]
>
> Point: The PC's and the work Intel and Microsoft have done with them is
> pretty much a redevelopment (albeit very slowly and in fits and starts) of
> IBM's vision of MVS. Almost all the concepts I have had to dig out of MS's
> verbiage describing Win32 maps decently to MVS. There are a few minor
> differences.
>
> There just aren't too many different ways to turn a bit on and off.
As we used to say in Latin: 'ayn chadash tachat hashemesh', literally
'no/not new/a new thing under the sun'.
DD
The Goobers
Dr John Stockton wrote:
>
> JRS: In article <68h1j3$f...@mtinsc03.worldnet.att.net> of Thu, 1 Jan
> 1998 16:24:36 in comp.software.year-2000, Bill Lynch
> <WBL...@worldnet.att.net> wrote:
> >ROTFLMAO - Peter, that is great! I know it's only 1998/001, but you're
> >clearly in the lead for the "Best Quip of the Year" award. Looking
> >forward to your sebsequent posts.
>
> years ago was asked to paint the initials on the sign for the gate of the
> new Admiralty Radar & Signals Establishment, and raised some objection?
Match you and raise you one... what almost got printed on the T-shirts
proposed for the team at Traveller's Insurance coding the Traveller's
Interactive Testing System?
DD
Steve Myers
Louis Krupp <lkr...@netONE.com> wrote in article
<34A89AAF...@netONE.com>...
> Burroughs Large Systems (forerunners of the Unisys A-Series) were
> used by the US Customs Service, the Detroit Police Department,
> Nabisco, the National Security Agency, several universities
> (UC San Diego, U of Denver, U of Delaware, Queens University in
> Ontario), and lots of other places. I'm not sure who all is using
> the A-Series now (I'd be surprised if any were still used in
> education).
Hmmm, that would be interesting news to San Diego Public Schools, Dallas
Public Schools, Denver Public Schools, Livermore Public Schools, Cities of
Miami, Jacksonville, and Tallahassee, McDonnell-Douglas, States of New
York, Texas, and Florida..................
> Unisys still has its believers, although I don't know any personally at
the moment (I guess I could say I still believe, although I'm not
practicing).
>
There are only about 200-300 companies (probably 100 excusively) that sell
and service products in the Unisys arena, that would indicate that they
believe. About 10000 individual sites around the world have supported
Unisys by buying systems. Data Exchange between systems makes a couple of
companies a hundred million or so reasons to do it.
> Louis Krupp
Charles M. Kozierok
That is *not* Latin, boychik. :)
cheers,
-*-
Charles M. Kozierok (mailto:i...@PCGuide.com)
Webslave, The PC Guide - http://www.PCGuide.com
Comprehensive PC reference, troubleshooting, optimization and procedures...
The Goobers
Charles M. Kozierok wrote:
>
> In article <34AE67...@erols.com>, The Goobers <docd...@erols.com> wrote:
> } Gary Labowitz wrote:
> } >
> } > Just couldn't resist....
> } >
> } [snip of CV and other credentials]
> } >
> } > Point: The PC's and the work Intel and Microsoft have done with them is
> } > pretty much a redevelopment (albeit very slowly and in fits and starts) of
> } > IBM's vision of MVS. Almost all the concepts I have had to dig out of MS's
> } > verbiage describing Win32 maps decently to MVS. There are a few minor
> } > differences.
> } >
> } > There just aren't too many different ways to turn a bit on and off.
> }
> } As we used to say in Latin: 'ayn chadash tachat hashemesh', literally
> } 'no/not new/a new thing under the sun'.
>
> That is *not* Latin, boychik. :)
It *isn't*? No *wonder* my using it always gets me a puzzled look from
our Neighborhood Centurion, then!
DD
Dr John Stockton
JRS: In article <34AE69...@erols.com> of Sat, 3 Jan 1998 11:37:31
in comp.software.year-2000, The Goobers <docd...@erols.com> wrote:
>Dr John Stockton wrote:
>>
>Match you and raise you one... what almost got printed on the T-shirts
>proposed for the team at Traveller's Insurance coding the Traveller's
>Interactive Testing System?
DD : please see Web
<URL: http://www.merlyn.demon.co.uk/news-use.htm#N&M>
--
John Stockton, Surrey, UK. j...@merlyn.demon.co.uk Turnpike v1.12 MIME.
Web <URL: http://www.merlyn.demon.co.uk/> - FAQqish topics, acronyms & links.
Correct 4-line sig separator is as above, a line comprising "-- " (SoRFC1036)
Before a reply, quote with ">" / "> ", known to good news readers (SoRFC1036)
The Goobers
Dr John Stockton wrote:
>
> JRS: In article <34AE69...@erols.com> of Sat, 3 Jan 1998 11:37:31
> in comp.software.year-2000, The Goobers <docd...@erols.com> wrote:
> >Dr John Stockton wrote:
> >>
> >Match you and raise you one... what almost got printed on the T-shirts
> >proposed for the team at Traveller's Insurance coding the Traveller's
> >Interactive Testing System?
>
> DD : please see Web
> <URL: http://www.merlyn.demon.co.uk/news-use.htm#N&M>
Okey dokey, now that I've seen it... what next?
DD
Dr John Stockton
JRS: In article <34B042...@erols.com> of Sun, 4 Jan 1998 21:16:26
in comp.software.year-2000, The Goobers <docd...@erols.com> wrote:
>Dr John Stockton wrote:
>>
>> JRS: In article <34AE69...@erols.com> of Sat, 3 Jan 1998 11:37:31
>> in comp.software.year-2000, The Goobers <docd...@erols.com> wrote:
>> >Dr John Stockton wrote:
>> >>
>> >Match you and raise you one... what almost got printed on the T-shirts
>> >proposed for the team at Traveller's Insurance coding the Traveller's
>> >Interactive Testing System?
>>
>> DD : please see Web
>> <URL: http://www.merlyn.demon.co.uk/news-use.htm#N&M>
>
>Okey dokey, now that I've seen it... what next?
You have yet again replied both in News and Mail, and without marking
that you have done so.
URL reads :-
News and Mail
It should rarely be necessary to respond by both News and Mail, which
both imposes an unnecessary load on dial-up users and may lead to
confusion. If you do choose to do so, it is important that you state at
the beginning of both responses that you are doing so; mail is normally
read before news, and the recipient can then decide in time whether to
respond in one or both media.
Transgressors who E-mail me are liable to see a reply in News along the
lines of :
You have responded in both Mail and News, without marking them as
such: please read http://www.merlyn.demon.co.uk/news-use.htm#N&M .
or
In http://www.merlyn.demon.co.uk/news-use.htm#N&M , there are
strong words against those who respond in both Mail and News but
do not mark the duplication.
or even
I consider it quite unnecessary to send by unsolicited E-mail a
copy of a message which is being posted in News, and most
discourteous to do so without marking the beginning of both copies
clearly to this effect.
Nick Upson
In a previous article Dr John Stockton wrote:
> You have yet again replied both in News and Mail, and without marking
> that you have done so.
In my experience replying in both media is done as a courtesy to ensure
the recipient does see the response and they see it quicker (news
propagation delays being measured in days for some people).
In order to make sure that I don't (even by accident) transgress your
unannounced laws (I'm sure everyone will modify usenet just because
you want it to be your way) PLONK
Nick
-----We Solve your Computer Problems---
The Goobers
Dr John Stockton wrote:
>
> JRS: In article <34B042...@erols.com> of Sun, 4 Jan 1998 21:16:26
> in comp.software.year-2000, The Goobers <docd...@erols.com> wrote:
> >Dr John Stockton wrote:
> >>
> >> JRS: In article <34AE69...@erols.com> of Sat, 3 Jan 1998 11:37:31
> >> in comp.software.year-2000, The Goobers <docd...@erols.com> wrote:
> >> >Dr John Stockton wrote:
> >> >>
> >> >Match you and raise you one... what almost got printed on the T-shirts
> >> >proposed for the team at Traveller's Insurance coding the Traveller's
> >> >Interactive Testing System?
> >>
> >> DD : please see Web
> >> <URL: http://www.merlyn.demon.co.uk/news-use.htm#N&M>
> >
> >Okey dokey, now that I've seen it... what next?
>
> that you have done so.
>
>
> News and Mail
>
> It should rarely be necessary to respond by both News and Mail, which
> both imposes an unnecessary load on dial-up users and may lead to
> confusion. If you do choose to do so, it is important that you state at
> the beginning of both responses that you are doing so; mail is normally
> read before news, and the recipient can then decide in time whether to
> respond in one or both media.
>
> Transgressors who E-mail me are liable to see a reply in News along the
> lines of :
>
> You have responded in both Mail and News, without marking them as
> such: please read http://www.merlyn.demon.co.uk/news-use.htm#N&M .
>
> or
>
> In http://www.merlyn.demon.co.uk/news-use.htm#N&M , there are
> strong words against those who respond in both Mail and News but
> do not mark the duplication.
>
> or even
>
> I consider it quite unnecessary to send by unsolicited E-mail a
> copy of a message which is being posted in News, and most
> discourteous to do so without marking the beginning of both copies
> clearly to this effect.
>
Please stop wasting precious bandwidth by quoting the web-page which I
told you I have already read; some people pay bitwise for their
downloads and you might well be forcing them into bankruptcy.
DD
Chris Westbury
In article <VA.0000029a.00594c99@localname>,
Nick Upson <up...@cableol.co.uk> wrote:
>
> In article <Abj+7aAh...@merlyn.demon.co.uk>,
> Dr John Stockton <j...@merlyn.demon.co.uk> wrote:
> >
> > You have yet again replied both in News and Mail, and without marking
> > that you have done so.
>
> the recipient does see the response and they see it quicker (news
> propagation delays being measured in days for some people).
Courtesy does not consist of absolute rules in a vacuum; it exists only
in relation to another person.
The acid test is not your intention, which no one will deny is good, but
the result. Is it courteous to help someone cross a street that he does
not want to cross?
The onus is on the one who would be courteous to determine if the
intended courtesy is welcome to the recipient. Treating everyone exactly
the same whether they like it or not is the very opposite of courtesy.
In addition, if you reply and followup to the same article without any
indication that you have done so, the result is not a courtesy.
> In order to make sure that I don't (even by accident) transgress your
> unannounced laws (I'm sure everyone will modify usenet just because you
> want it to be your way) PLONK
It is not unannounced. It is not a modification to Usenet (just grep
your newsspool for "posted and emailed"). Judging by the fact that you
do not even know how to plonk, it it you who have come late to the party.
--
Christopher Westbury, Midtown Associates, 15 Fallon Place, Cambridge, MA 02138
Dr John Stockton
JRS: In article <34B18A...@erols.com> of Mon, 5 Jan 1998 20:36:01>....
>
>DD
Once again, both news and mail.
*Plonk*.
I'll just hope that, if DD writes anything useful, it gets quoted.
--
John Stockton, Surrey, UK. j...@merlyn.demon.co.uk Turnpike v1.12 MIME.
Don't Mail News. Y2k for beginners http://www.merlyn.demon.co.uk/year2000.txt
My list of Critical Dates is : http://www.merlyn.demon.co.uk/misctime.htm#CDs
Web <URL: http://www.merlyn.demon.co.uk/> -- includes FAQqish topics & links.
The Goobers
Dr John Stockton wrote:
>
> JRS: In article <34B18A...@erols.com> of Mon, 5 Jan 1998 20:36:01
> in comp.software.year-2000, The Goobers <docd...@erols.com> wrote:
> Dr John Stockton wrote:
> >
> >....
> >
> >DD
>
> Once again, both news and mail.
>
> *Plonk*.
>
> I'll just hope that, if DD writes anything useful, it gets quoted.
Well, then, what is it... I'm good enough to read quoted but not
directly? Terribly unscholarly to rely on secondary sources but, as you
will
... as for the rest of you... feel free to make up quotes from me, just
make them *fascinating*, don'cha know.
DD
Phil Edwards
On Tue, 06 Jan 1998 20:09:50 -0500, The Goobers <docd...@erols.com>
wrote:
>Dr John Stockton wrote:
>>
>> Once again, both news and mail.
>>
>> *Plonk*.
>>
>> I'll just hope that, if DD writes anything useful, it gets quoted.
>
>... as for the rest of you... feel free to make up quotes from me, just
>make them *fascinating*, don'cha know.
>
>By the way, let me just add that I consider Dr John Stockton a
>*poopie-head* of the worst order - a blot on the face of that
>once-great nation which gave us Ada Lovelace, the LEO and writers of
>the calibre of Phil Edwards.
Why thanks, Doc.
Phil
but no follow-ups posted and mailed, please...
--
Phil Edwards ph...@news400.com
Editor, NEWS/400.uk +44 (0)161 929 0777
"Hell, get me drunk enough and I'll eat tofu." - David Bryant
docd...@clark.net
In article <34b39946...@news.dircon.co.uk>,
Phil Edwards <new...@dircon.co.uk.nojunk> wrote:
>On Tue, 06 Jan 1998 20:09:50 -0500, The Goobers <docd...@erols.com>
>wrote:
>
>>Dr John Stockton wrote:
>>>
>>> Once again, both news and mail.
>>>
>>> *Plonk*.
>>>
>>> I'll just hope that, if DD writes anything useful, it gets quoted.
>>
>>... as for the rest of you... feel free to make up quotes from me, just
>>make them *fascinating*, don'cha know.
>>
>>By the way, let me just add that I consider Dr John Stockton a
>>*poopie-head* of the worst order - a blot on the face of that
>>once-great nation which gave us Ada Lovelace, the LEO and writers of
>>the calibre of Phil Edwards.
>
>Why thanks, Doc.
... and people actually *wonder* why this kind of stuff isn't admissable
in court.
... and Us Amurricans spell that 'caliber', old boy.
DD
Phil Edwards
On 8 Jan 1998 14:55:18 GMT, docd...@clark.net () wrote:
>In article <34b39946...@news.dircon.co.uk>,
>Phil Edwards <new...@dircon.co.uk.nojunk> wrote:
>>On Tue, 06 Jan 1998 20:09:50 -0500, The Goobers <docd...@erols.com>
>>wrote:
>>
>>>... as for the rest of you... feel free to make up quotes from me, just
>>>make them *fascinating*, don'cha know.
>>>
>>>By the way, let me just add that I consider Dr John Stockton a
>>>*poopie-head* of the worst order - a blot on the face of that
>>>once-great nation which gave us Ada Lovelace, the LEO and writers of
>>>the calibre of Phil Edwards.
>>
>>Why thanks, Doc.
>
>... and people actually *wonder* why this kind of stuff isn't admissable
>in court.
>
>... and Us Amurricans spell that 'caliber', old boy.
Rumbled! What can I say but...
Heck!
Phil
and I would have gotten away with it, too, if it hadn't been for those
pesky ki... er, where was I...
--
Phil Edwards ph...@news400.com
Editor, NEWS/400.uk +44 (0)161 929 0777
"on the net nobody nows your a dog" - binky