On 5/4/2024 1:01 AM, Steve Hayes wrote:
> On Mon, 29 Apr 2024 06:33:38 -0700, "John C." <r9jmg0@yahoo.com>
> wrote:
>
>> Newyana2 wrote:
>
>>>    On Windows it's always been "programs". Only some
>>> programmers say "application". (It could be worse. For
>>> awhile people were talking about their programming
>>> projects as "solutions".)
>>
>> Yes, that wouldn't have been a good thing.
>
> I have seen advertisements for "solutions", but they never tell you
> what problem they are claiming to be able to solve.
>
> Yes, they're all programs, but some are applications, like word
> processors, spreadsheets, databases etc.
>
> But I have a program called "Glary Utilities", which is not an
> application. It just helps the computer to run better. Likewise, the
> operating system is a program, but not an application.

The operating system is not a program. It is an executive.

It loads applications in Ring3.

There is a scheduler giving "execution time slices" to the application.

In Ring 0, lives a kernel and hardware drivers. Applications are
not allowed to access hardware directly, and go through kernel calls.

There is a task scheduler, that allows items to be loaded/executed at
fixed time points. That's similar to CRON in Linux or Unix.

And to further complicate matters, even though the OS is an executive,
it is virtualized via an inverted hypervisor. The diagram of how
an executive works, no longer looks the same as it did in Windows XP.
Windows 10 would be the root partition. The Linux partition would
really exist, if you had installed WSL plus the Linux distro of your
choice. If you had VirtualBox, it would have a position in this diagram
too (not shown of course). VirtualBox is not nested, that I can detect.
Nested has never worked on my computers here. I tried.

https://web.archive.org/web/20111205072921/https://msdn.microsoft.com/en-us/library/cc768520%28v=bts.10%29.aspx

Task Manager was not modified in any way, to make details about this
apparent to the machine operator. This is why I use a *Power Meter*
on the AC line cord, to detect foreign activity (even if it is windows
doing it, and does not show in the list). In Task Manager for example,
try and find "Memory Compressor". Now, use Process Explorer, you will
find Memory Compressor is listed as a process.

[What Task Manager should have been - percent CPU with two digits after decimal, nice!!! ]

https://learn.microsoft.com/en-us/sysinternals/downloads/process-explorer

One of the observations you can make, is via the "ruggedness" of Task Manager.
The modern Task Manager can freeze. In an emergency, it is the patron
saint of "Useless". Can't do a thing with it. You will notice the Task
Manager in Windows XP was not like that. Via statically provided resources,
it always seemed to have resources, and when the OS had gone to hell in
a hand basket, you could still "attempt to do stuff" in Task Manager.
That's all changed. In WinXP you could alt-tab, even when the OS was in
serious trouble. W10/W11 just die instantly, when even a tiny bit of
smothering is applied :-) One of my favorite examples, was using
ImageMagick one day, and OpenMP happened to be enabled (use multiple
cores to open an image for display on the screen). OS froze... instantly.
As instantly as you can envisage "instantly" means. One frame time. Dead.
Oh, the electrons are flying around in there, but "nobody is home".
It's not a crash. It's a deadlock, a form of software deadly embrace.

When an OS has no observational capability, we can only dream as
sheep dream, about what is the matter. Gone are the days of having
dual CRT tubes with critical information recorded on the screen.
In uni, when one of the students did a DOS attack on the mainframe,
he stood by the window and watched the "free disk" counter decrement
over a matter of 30 seconds or so, killing the mainframe. On a timeshare
system, jobs are rotated in and out via that particular disk. Our student
had used a primitive fork bomb, and the system operator (a "stable genius")
had forgotten to enable a policy to prevent that :-) It was kinda a
splash of cold water for the gentleman, to discover he had competition.
The operator used to play chess on one of those CRTs.

https://archive.computerhistory.org/resources/still-image/LLNL%20Computer%20Museum/102675044.03.01.lg.jpg

The power meter on my computer, is the last vestige of observability.
I may not know what is going on, but I know "something, is inside the machine".

Paul

In message <v14jj8$12jqr$1@dont-email.me>, Paul <nospam@needed.invalid>
writes
<big snip>
>The operating system is not a program. It is an executive.

That brings back memories. The computer I used back in the 1970s, an ICL
1900, actually called the core part of its OS "Executive" with a capital
E.
--
John Hall
"Acting is merely the art of keeping a large group of people
from coughing."
Sir Ralph Richardson (1902-83)

On 5/4/2024 5:28 AM, John Hall wrote:
> In message <v14jj8$12jqr$1@dont-email.me>, Paul <nospam@needed.invalid> writes
> <big snip>
>> The operating system is not a program. It is an executive.
>
> That brings back memories. The computer I used back in the 1970s,
> an ICL 1900, actually called the core part of its OS "Executive" with a capital E.

There are some differences.

Mainframes were "ball jugglers". They could support 500 users at once,
by keeping 490 sessions "asleep". A limited number of users would be
in "near orbit" and have access to CPU cycles. Perhaps one drive, or
even, an entire Storage Director, took care of paging of jobs.

This was managed with disk drives (and occasionally, if you were lucky,
with a drum storage device).

The difference on a personal computer, is there is no juggling of processes
in the same way. The processes are stored in memory, and are "ready to run".
The scheduler gives them slices. Some of the things in Task Manager, use
zero cycles, they use nothing at all. They are "mostly silent". Only a few
of the SVCHOSTs are busy little beavers. Some things related to security,
might always have a busy stance.

Modern Windows has taken to "suspending" some items, but it's unclear whether
that ever pans out (actually helps a user). Some of the same states as were
always there, are still present. The "zombie" state for example. A "zombie"
is a process that did not get harvested properly, and might disappear on a reboot.

And while early Windows had pagefile.sys for paging out of virtual memory,
that's hardly ever used on machines potentially using SSD drives for storage.
We really are reliant now, on gobs of main memory, embarrassing excess,
for how the machine works. That's how my browser a few minutes ago, could
be using 7GB of memory, while a web site recorded every line I read on a
web page. They measured my "interest" in each article, my dwell time,
whether I saw the adverts or not. They even interfere with my scroll
bar, until I get annoyed and close the session.

Paul

John Hall <john_nospam@jhall.co.uk> wrote:
> In message <v14jj8$12jqr$1@dont-email.me>, Paul <nospam@needed.invalid>
> writes
> <big snip>
> >The operating system is not a program. It is an executive.
>
> That brings back memories. The computer I used back in the 1970s, an ICL
> 1900, actually called the core part of its OS "Executive" with a capital
> E.

Yep, in the early 70s, I was using/supporting HP's 'mini' computers,
running RTE, Real Time Executive.

Later, the business side of HP had MPE, Multi Programming Excecutive.

On Sat, 4 May 2024 02:11:51 -0400, Paul <nospam@needed.invalid> wrote:

>On 5/4/2024 1:01 AM, Steve Hayes wrote:
>> Yes, they're all programs, but some are applications, like word
>> processors, spreadsheets, databases etc.
>>
>> But I have a program called "Glary Utilities", which is not an
>> application. It just helps the computer to run better. Likewise, the
>> operating system is a program, but not an application.
>
>The operating system is not a program. It is an executive.

In MyEnglish the definition of a program is "a set of instructions to
operate a computer". Is this "Executive" not that?

My first computer had an OS in ROM, with built-in BASIC, with which
you could write other programs and save them and their data on tape,
but all were programs. With a couple of add-ons you could have a DOS
(CP/M), and save stuff on floppy disks, but that by-passed the BASIC
in ROM, so you had to get programming languages that ran under CP/M,
but whether they loaded from ROM, tape or disks, all there sets of
instructions tomake the computer do different things.

--
Steve Hayes from Tshwane, South Africa
Web: http://www.khanya.org.za/stevesig.htm
Blog: http://khanya.wordpress.com
E-mail - see web page, or parse: shayes at dunelm full stop org full stop uk

On 04/05/2024 06:01, Steve Hayes wrote:
> On Mon, 29 Apr 2024 06:33:38 -0700, "John C." <r9jmg0@yahoo.com>
> wrote:
> But I have a program called "Glary Utilities", which is not an
> application. It just helps the computer to run better.

Adding more software to Windows often makes it run slower. Especially if
said software adds yet another background task that runs al the time
regardless of if you are using the software or not.

> I have seen advertisements for "solutions", but they never tell you
> what problem they are claiming to be able to solve.

You have just reminded me of this...

Terence McKenna: "Technologies do not always increase peoples options!
Sometimes they decrease peoples options. One may not want a car, with
that sort of equipment, but one has no choice.

The next point is, technology does not always solve important problems.
We like to think that technological innovation will almost always lead
to an enriched and enhanced life. But very often, technological progress
does not progress it’s self to important problems, but rather to be
worse. And yet we proceed anyway, in spite of the fact that in solving a
trivial problem, we may be creating a greater problem than the problem
we solve.

For example, we now have in America the issue of weather or not we
should spend billions of dollars for something called a super
information highway.

Well, if we ask our authorities on this, the same question I put to the
cartels, we get some curious answers! What is the problem to which this
super information highway will be a solution?

One of the answers you’ll get is ”well, we now have available only 60
television channels. With the super information highway we will have to
have 500, maybe even a 1000”.

Is this a problem that really needs a solution?

We have to ask this question and several others, about technological change.

Because technological change is almost always what I call a faustian
bargain “it giveth, and it taketh away."

Good old Terence...

https://www.sheldrake.org/audios/the-sheldrake-mckenna-abraham-trialogues

On 5/4/2024 12:56 PM, Steve Hayes wrote:
> On Sat, 4 May 2024 02:11:51 -0400, Paul <nospam@needed.invalid> wrote:
>
>> On 5/4/2024 1:01 AM, Steve Hayes wrote:
>>> Yes, they're all programs, but some are applications, like word
>>> processors, spreadsheets, databases etc.
>>>
>>> But I have a program called "Glary Utilities", which is not an
>>> application. It just helps the computer to run better. Likewise, the
>>> operating system is a program, but not an application.
>>
>> The operating system is not a program. It is an executive.
>
> In MyEnglish the definition of a program is "a set of instructions to
> operate a computer". Is this "Executive" not that?
>
> My first computer had an OS in ROM, with built-in BASIC, with which
> you could write other programs and save them and their data on tape,
> but all were programs. With a couple of add-ons you could have a DOS
> (CP/M), and save stuff on floppy disks, but that by-passed the BASIC
> in ROM, so you had to get programming languages that ran under CP/M,
> but whether they loaded from ROM, tape or disks, all there sets of
> instructions tomake the computer do different things.

If you boot a memtest floppy, that is a single program that
runs 100% of the time. Nobody tells it to do anything.
It is the boss. All machine resources are available.

If it wanted to erase your hard drive, nothing would stop it.

The frame buffer for the graphics are at a fixed address. It
takes fixed font pixmaps of characters and writes them to the
frame buffer. That makes the character display.

The memory is linear mapped. The virtual address equals the physical address.
It could be using "Giant Pages", a 1GB mapping, as you might notice the
program has a fixation with 1GB and 2GB chunks as it runs. Loading the
mapper, is an executive-type function.

You can see then, that it is running the restaurant all by itself.
It's taking the orders (from the keyboard), it's running into the
back of the restaurant to the kitchen, it is putting steaks and
veggies on the table and cooking the steak, it's doing all the jobs.
That means, when the guy wrote the program, he had to "think of
all the details", not just "some of the details". Well, that's what
happens when you have No Executive.

*******

When a program runs in an OS, a lot of details have been worked out.
The program "just sits down and eats". It worries not about the
grill, about the dish washer person, about the staff to carry the
orders to the tables and so on. The memory mapper is set defensively
by the executive, to prevent "shenanigans" (self-modifying code is
not allowed, and has not been allowed for a lot of years). This
means the code segment is read only, and after the loader has
loaded it (written to it), the mapping the program sees for its
code is read-only.

Paul

On 05/05/24 02:56, Steve Hayes wrote:
> On Sat, 4 May 2024 02:11:51 -0400, Paul <nospam@needed.invalid>
> wrote:
>
>> On 5/4/2024 1:01 AM, Steve Hayes wrote:
>>> Yes, they're all programs, but some are applications, like word
>>> processors, spreadsheets, databases etc.
>>>
>>> But I have a program called "Glary Utilities", which is not an
>>> application. It just helps the computer to run better. Likewise,
>>> the operating system is a program, but not an application.
>>
>> The operating system is not a program. It is an executive.
>
> In MyEnglish the definition of a program is "a set of instructions to
> operate a computer". Is this "Executive" not that?

It's all software, of course, but a person writing operating systems
code has to be aware of hardware features that the average applications
programmer never comes into contact with. For certain functions, e.g.
thread switching, it might be necessary to descend into assembly language.

There's another distinction that occurs to me. An application program
has a beginning, a middle, and an end. It does it job and then
terminates. The operating system never terminates until the whole
computer is shut down.

The distinction is becoming fuzzier now that some operating systems are
built in layers. The bottom-level operating system is probably small and
does only some basic things. (Typically thread switching and memory
management.) Then another operating system is built on top of that. And
maybe even another on top of that.

--
Peter Moylan http://www.pmoylan.org
Newcastle, NSW

On 2024-05-04 10:56, Steve Hayes wrote:
> On Sat, 4 May 2024 02:11:51 -0400, Paul <nospam@needed.invalid> wrote:
>
>> On 5/4/2024 1:01 AM, Steve Hayes wrote:
>>> Yes, they're all programs, but some are applications, like word
>>> processors, spreadsheets, databases etc.
>>>
>>> But I have a program called "Glary Utilities", which is not an
>>> application. It just helps the computer to run better. Likewise, the
>>> operating system is a program, but not an application.
>>
>> The operating system is not a program. It is an executive.
>
> In MyEnglish the definition of a program is "a set of instructions to
> operate a computer". Is this "Executive" not that?

I agree with you on that one, Steve.

> My first computer had an OS in ROM, with built-in BASIC, with which
> you could write other programs and save them and their data on tape,
> but all were programs.

The first computer I worked on, though it was not called a computer, but
rather an 'Electronic Accounting Machine'. I 'wrote' programs on it by
plugging wires into a board that made contact with relays.

It was definitely a computer, though, stepping through programs, doing
math, branching on tested conditions, and reading and writing to and
from I/O devices.

I worked on that for about two years, at which time I started working on
a different set of machines, and found out that a program could actually
be stored in a memory.

> With a couple of add-ons you could have a DOS
> (CP/M), and save stuff on floppy disks, but that by-passed the BASIC
> in ROM, so you had to get programming languages that ran under CP/M,
> but whether they loaded from ROM, tape or disks, all there sets of
> instructions tomake the computer do different things.

--
A computer won't stop you being an idiot,
but it'll make you a faster, better idiot

On 05/05/24 15:40, lar3ryca wrote:
>
> The first computer I worked on, though it was not called a computer,
> but rather an 'Electronic Accounting Machine'. I 'wrote' programs on
> it by plugging wires into a board that made contact with relays.
>
> It was definitely a computer, though, stepping through programs,
> doing math, branching on tested conditions, and reading and writing
> to and from I/O devices.
>
> I worked on that for about two years, at which time I started
> working on a different set of machines, and found out that a program
> could actually be stored in a memory.

We take memory for granted now, but it took a while to get it right.
Looking back, the ideas of having main memory on a rotating drum or in
columns of mercury sound incredible crude.

To make computer memory practical, the electronics people had to learn
how to put transistors on semiconductor chips. You can also build
flip-flops with vacuum tubes, but vacuum tubes already become awkward
once you have a few hundred of them.

In my student days, and for some time afterwards, I used the plugboard
approach, but that was with analogue computers. Those were very good at
solving differential equations, but they died out through not being
sufficiently scalable.

--
Peter Moylan http://www.pmoylan.org
Newcastle, NSW

Steve Hayes wrote:

> In MyEnglish the definition of a program is "a set of instructions to
> operate a computer". Is this "Executive" not that?

It's mine too. If you operate with a class called "executives" (or
OS's), it's just a subset of the programs.

--
Bertel
Kolt, Denmark

[snip]

> If you boot a memtest floppy, that is a single program that
> runs 100% of the time. Nobody tells it to do anything.
> It is the boss. All machine resources are available.
>
> If it wanted to erase your hard drive, nothing would stop it.

There would be something if you had a real write-protect switch (not one
of those lying ones that software can just ignore).

[snip]

--
Mark Lloyd
http://notstupid.us/

"If there is a supreme being, he's crazy." -- Marlene Dietrich
(1901-1992), quoted in Rave magazine, November 1986

Newyana2 <mayayana@invalid.nospam> writes:
> On Windows it's always been "programs". Only some
> programmers say "application". (It could be worse. For
> awhile people were talking about their programming
> projects as "solutions".)

So Windows 3.0 never had UAEs - Unexpected Application Errors? (And of
course, none of the ".exe" files were ever called "executables" either?)
"Apps" is quite well established in Windows, the oft derided "Hungarian
Notation" had an "Apps Hungarian" flavour used in the Apps Division,
in contrast to "System Hungarian" used in the Systems Division.

Phil
--
We are no longer hunters and nomads. No longer awed and frightened, as we have
gained some understanding of the world in which we live. As such, we can cast
aside childish remnants from the dawn of our civilization.
-- NotSanguine on SoylentNews, after Eugen Weber in /The Western Tradition/

On Sun, 5 May 2024 13:59:57 +1000, Peter Moylan
<peter@pmoylan.org.invalid> wrote:

>On 05/05/24 02:56, Steve Hayes wrote:
>> On Sat, 4 May 2024 02:11:51 -0400, Paul <nospam@needed.invalid>
>> wrote:
>>
>>> On 5/4/2024 1:01 AM, Steve Hayes wrote:
>>>> Yes, they're all programs, but some are applications, like word
>>>> processors, spreadsheets, databases etc.
>>>>
>>>> But I have a program called "Glary Utilities", which is not an
>>>> application. It just helps the computer to run better. Likewise,
>>>> the operating system is a program, but not an application.
>>>
>>> The operating system is not a program. It is an executive.
>>
>> In MyEnglish the definition of a program is "a set of instructions to
>> operate a computer". Is this "Executive" not that?
>
>It's all software, of course, but a person writing operating systems
>code has to be aware of hardware features that the average applications
>programmer never comes into contact with. For certain functions, e.g.
>thread switching, it might be necessary to descend into assembly language.
>
>There's another distinction that occurs to me. An application program
>has a beginning, a middle, and an end. It does it job and then
>terminates. The operating system never terminates until the whole
>computer is shut down.
>
>The distinction is becoming fuzzier now that some operating systems are
>built in layers. The bottom-level operating system is probably small and
>does only some basic things. (Typically thread switching and memory
>management.) Then another operating system is built on top of that. And
>maybe even another on top of that.

Yes, that supports the distinction I am trying to make: an operating
system, whether it is built on top of another one or not, and an app
are both programs, but not all programs are apps. And yes, both are
software as well.

US English is somewhat at a disadvantage here, be cause they use
"program" in a wider sense, so they often have to add "software" to
it, to distinguish a "software program" from other kinds of what the
rest of us would call "programmes", like TV programmes, sports events
programmes etc.

--
Steve Hayes from Tshwane, South Africa
Web: http://www.khanya.org.za/stevesig.htm
Blog: http://khanya.wordpress.com
E-mail - see web page, or parse: shayes at dunelm full stop org full stop uk

On Sun, 5 May 2024 11:28:19 +0200, Bertel Lund Hansen
<gadekryds@lundhansen.dk> wrote:

>Steve Hayes wrote:
>
>> In MyEnglish the definition of a program is "a set of instructions to
>> operate a computer". Is this "Executive" not that?
>
>It's mine too. If you operate with a class called "executives" (or
>OS's), it's just a subset of the programs.

And all programs execute instructions (and then, of course, there are
the executioners in another thread on aue).

In low-level operating systems, the instructions are given by the
programmer, rather than the user. From the point of view of the user,
the most common instruction given by a user to the OS is to run an
app, and to exit the app when the user has finished with it.

--
Steve Hayes from Tshwane, South Africa
Web: http://www.khanya.org.za/stevesig.htm
Blog: http://khanya.wordpress.com
E-mail - see web page, or parse: shayes at dunelm full stop org full stop uk

On 2024-05-05 02:15, Peter Moylan wrote:
> On 05/05/24 15:40, lar3ryca wrote:
>>
>> The first computer I worked on, though it was not called a computer,
>> but rather an 'Electronic Accounting Machine'. I 'wrote' programs on
>> it by plugging wires into a board that made contact with relays.
>>
>> It was definitely a computer, though, stepping through programs,
>> doing math, branching on tested conditions, and reading and writing
>> to and from I/O devices.
>>
>> I worked on that for about two years, at which time I started
>> working on a different set of machines, and found out that a program
>> could actually be stored in a memory.
>
> We take memory for granted now, but it took a while to get it right.
> Looking back, the ideas of having main memory on a rotating drum or in
> columns of mercury sound incredible crude.

How about memory on acoustic delay lines?
When I worked for CDC, I sometimes got called to Allstate (the insurance
folks), to repair the terminals.

The memory on those consisted of modules containing spirals of what
might best be described as 'piano wire'. Bits were input by giving the
wire a quick twist (twist, then reset back to idle). That generated
something like a shock wave that travelled around the spiral to a
transducer on the other end, I can't remember if it twisted one
direction for a '1' and the other direction for a '0', or if it only
used one direction of twist, relying on framing pulses and time between
bit to differentiate between them.

Nor do I remember the amount of data it could hold, probably in the
order of a few hundred bytes.

Of course, once the data got to the end and was read, it had to be
re-sent again, unless that particular packet had to be changed, at which
time it was not re-sent, and a new packet was sent in its place.

> To make computer memory practical, the electronics people had to learn
> how to put transistors on semiconductor chips. You can also build
> flip-flops with vacuum tubes, but vacuum tubes already become awkward
> once you have a few hundred of them.
>
> In my student days, and for some time afterwards, I used the plugboard
> approach, but that was with analogue computers. Those were very good at
> solving differential equations, but they died out through not being
> sufficiently scalable.

The plugboards I spoke of were for programming a digital computer, When
a plugboard was mounted, pressing "Start" sent a 48V pulse out of the
'start hub', and into a 'program step' hub (almost always 'program step
1'. This would pick up a relay, which would cause voltage to be applied
to hubs called 'operation, 'in word 1', 'inword 2', and 'out word'.

That would, in turn, call on the electronics (SMS cards) to perform the
operation.

As for analog computers, my step-brother and I bought one, but it was
pretty simple, consisting og circuits to do math, input via dials
(potentiometers), and output on a voltmeter.

When a step was active,

--
Whose idea was it to put an "S" in the word "lisp"?

Steve Hayes wrote:

> US English is somewhat at a disadvantage here, be cause they use
> "program" in a wider sense, so they often have to add "software" to
> it, to distinguish a "software program" from other kinds of what the
> rest of us would call "programmes", like TV programmes, sports events
> programmes etc.

Danish has the exact same 'problem' - which isn't really a problem since
context usually makes the meaning clear.

--
Bertel
Kolt, Denmark

lar3ryca <larry@invalid.ca> wrote:

> On 2024-05-05 02:15, Peter Moylan wrote:
> > On 05/05/24 15:40, lar3ryca wrote:
> >>
> >> The first computer I worked on, though it was not called a computer,
> >> but rather an 'Electronic Accounting Machine'. I 'wrote' programs on
> >> it by plugging wires into a board that made contact with relays.
> >>
> >> It was definitely a computer, though, stepping through programs,
> >> doing math, branching on tested conditions, and reading and writing
> >> to and from I/O devices.
> >>
> >> I worked on that for about two years, at which time I started
> >> working on a different set of machines, and found out that a program
> >> could actually be stored in a memory.
> >
> > We take memory for granted now, but it took a while to get it right.
> > Looking back, the ideas of having main memory on a rotating drum or in
> > columns of mercury sound incredible crude.
>
> How about memory on acoustic delay lines?
> When I worked for CDC, I sometimes got called to Allstate (the insurance
> folks), to repair the terminals.
>
> The memory on those consisted of modules containing spirals of what
> might best be described as 'piano wire'. Bits were input by giving the
> wire a quick twist (twist, then reset back to idle). That generated
> something like a shock wave that travelled around the spiral to a
> transducer on the other end, I can't remember if it twisted one
> direction for a '1' and the other direction for a '0', or if it only
> used one direction of twist, relying on framing pulses and time between
> bit to differentiate between them.

That's a bit primitive.
One of the first bulk computer memories,
used by Alan Turing himself, was a mercury delay line.
Bits were stored as sound pulses in a column of mercury.

Just as with a hard drive the computer had to wait
for the right part of the bit string to pass by.
I would have to look up how long its bit string could be,

Jan

On Mon, 06 May 2024 05:56:10 +0200, Steve Hayes
<hayesstw@telkomsa.net> wrote:

>US English is somewhat at a disadvantage here, be cause they use
>"program" in a wider sense, so they often have to add "software" to
>it, to distinguish a "software program" from other kinds of what the
>rest of us would call "programmes", like TV programmes, sports events
>programmes etc.

That used to be true, but not so much anymore. It was in 1962, when I
was out of work, and I answered an ad for "Programmer Trainee --
College Graduate, Any Major" for a higher salary than I had ever had.
I didn't know what a programmer was, but I thought it referred to TV
or Radio programming.

I didn't get the job, but I soon started a computer programming course
and got a programming job soon afterward.

On 5/6/2024 7:52 AM, J. J. Lodder wrote:
> lar3ryca <larry@invalid.ca> wrote:
>
>> On 2024-05-05 02:15, Peter Moylan wrote:
>>> On 05/05/24 15:40, lar3ryca wrote:
>>>>
>>>> The first computer I worked on, though it was not called a computer,
>>>> but rather an 'Electronic Accounting Machine'. I 'wrote' programs on
>>>> it by plugging wires into a board that made contact with relays.
>>>>
>>>> It was definitely a computer, though, stepping through programs,
>>>> doing math, branching on tested conditions, and reading and writing
>>>> to and from I/O devices.
>>>>
>>>> I worked on that for about two years, at which time I started
>>>> working on a different set of machines, and found out that a program
>>>> could actually be stored in a memory.
>>>
>>> We take memory for granted now, but it took a while to get it right.
>>> Looking back, the ideas of having main memory on a rotating drum or in
>>> columns of mercury sound incredible crude.
>>
>> How about memory on acoustic delay lines?
>> When I worked for CDC, I sometimes got called to Allstate (the insurance
>> folks), to repair the terminals.
>>
>> The memory on those consisted of modules containing spirals of what
>> might best be described as 'piano wire'. Bits were input by giving the
>> wire a quick twist (twist, then reset back to idle). That generated
>> something like a shock wave that travelled around the spiral to a
>> transducer on the other end, I can't remember if it twisted one
>> direction for a '1' and the other direction for a '0', or if it only
>> used one direction of twist, relying on framing pulses and time between
>> bit to differentiate between them.
>
> That's a bit primitive.
> One of the first bulk computer memories,
> used by Alan Turing himself, was a mercury delay line.
> Bits were stored as sound pulses in a column of mercury.
>
> Just as with a hard drive the computer had to wait
> for the right part of the bit string to pass by.
> I would have to look up how long its bit string could be,
>
> Jan
>

https://www.oxfordreference.com/display/10.1093/oi/authority.20110803095708310

"quartz crystals were used as transducers and the ultrasonic pulses were
passed along a tube of mercury about 5 feet (1.5 meters) in length.
The delay was approximately 1 millisecond but it enabled nearly 1000 pulses
to be stored. Later acoustic memory used magnetostrictive transducers and
nickel-iron wire, with the electrical signals converted into stress waves."

A thousand bits, isn't a lot.

Some of the first SRAM (suitable for home computer projects)
were 256x4 bits and 1024x1 bit static RAM running at 5 volts.
"Beautiful stuff". Compared to the dreadful DRAM of the day.
And just one of those chips, stores the same stuff as a delay
line, and also offers "random" access, so is a lot faster.

My breadboarded home computer used (4) 256x4 chips. As a 256x16 array (16 bit CPU).

Paul

On 07/05/24 04:47, Paul wrote:

> Some of the first SRAM (suitable for home computer projects) were
> 256x4 bits and 1024x1 bit static RAM running at 5 volts. "Beautiful
> stuff". Compared to the dreadful DRAM of the day. And just one of
> those chips, stores the same stuff as a delay line, and also offers
> "random" access, so is a lot faster.
>
> My breadboarded home computer used (4) 256x4 chips. As a 256x16
> array (16 bit CPU).

My first computer had an 8080A processor (very new at the time) and 1k
bytes of RAM. The "motherboard" was a whole lot of wire-wrap sockets,
which took ages to wire up. I was very proud of the metal chassis that I
built as well, with eight switches for input[1] and eight LED lights, plus
a couple of pushbuttons. The switches were mainly for loading a program
into memory, although I later wrote a loader that took the data from an
audio cassette tape. Then, with the aid of a few resistors, I turned the
front panel leds into a D/A converter. With the analogue output
connected to an amplifier, I got the computer to play 3-part music.

[1] Or possibly 24 switches. I've now forgotten whether I had separate
address and data switches.

The front panel bore the logo "IDSFA-80". If anyone asked what IDSFA
stood for, I could tell them it doesn't stand for anything.

--
Peter Moylan http://www.pmoylan.org
Newcastle, NSW

On 2024-05-06 12:47, Paul wrote:
> On 5/6/2024 7:52 AM, J. J. Lodder wrote:
>> lar3ryca <larry@invalid.ca> wrote:
>>
>>> On 2024-05-05 02:15, Peter Moylan wrote:
>>>> On 05/05/24 15:40, lar3ryca wrote:
>>>>>
>>>>> The first computer I worked on, though it was not called a computer,
>>>>> but rather an 'Electronic Accounting Machine'. I 'wrote' programs on
>>>>> it by plugging wires into a board that made contact with relays.
>>>>>
>>>>> It was definitely a computer, though, stepping through programs,
>>>>> doing math, branching on tested conditions, and reading and writing
>>>>> to and from I/O devices.
>>>>>
>>>>> I worked on that for about two years, at which time I started
>>>>> working on a different set of machines, and found out that a program
>>>>> could actually be stored in a memory.
>>>>
>>>> We take memory for granted now, but it took a while to get it right.
>>>> Looking back, the ideas of having main memory on a rotating drum or in
>>>> columns of mercury sound incredible crude.
>>>
>>> How about memory on acoustic delay lines?
>>> When I worked for CDC, I sometimes got called to Allstate (the insurance
>>> folks), to repair the terminals.
>>>
>>> The memory on those consisted of modules containing spirals of what
>>> might best be described as 'piano wire'. Bits were input by giving the
>>> wire a quick twist (twist, then reset back to idle). That generated
>>> something like a shock wave that travelled around the spiral to a
>>> transducer on the other end, I can't remember if it twisted one
>>> direction for a '1' and the other direction for a '0', or if it only
>>> used one direction of twist, relying on framing pulses and time between
>>> bit to differentiate between them.
>>
>> That's a bit primitive.
>> One of the first bulk computer memories,
>> used by Alan Turing himself, was a mercury delay line.
>> Bits were stored as sound pulses in a column of mercury.
>>
>> Just as with a hard drive the computer had to wait
>> for the right part of the bit string to pass by.
>> I would have to look up how long its bit string could be,
>>
>> Jan
>>
>
> https://www.oxfordreference.com/display/10.1093/oi/authority.20110803095708310
>
> "quartz crystals were used as transducers and the ultrasonic pulses were
> passed along a tube of mercury about 5 feet (1.5 meters) in length.
> The delay was approximately 1 millisecond but it enabled nearly 1000 pulses
> to be stored. Later acoustic memory used magnetostrictive transducers and
> nickel-iron wire, with the electrical signals converted into stress waves."
>
> A thousand bits, isn't a lot.
>
> Some of the first SRAM (suitable for home computer projects)
> were 256x4 bits and 1024x1 bit static RAM running at 5 volts.
> "Beautiful stuff". Compared to the dreadful DRAM of the day.
> And just one of those chips, stores the same stuff as a delay
> line, and also offers "random" access, so is a lot faster.
>
> My breadboarded home computer used (4) 256x4 chips. As a 256x16 array (16 bit CPU).

And my breadboard computer used 8 1024x1 chips (2501?), and that was
also my first computer. The memory cost me $85.00 CAD at the time, about
1975 if I remember correctly).

--
Whose idea was it to put an "S" in the word "lisp"?

On 5/6/2024 9:18 PM, lar3ryca wrote:
> On 2024-05-06 12:47, Paul wrote:
>> On 5/6/2024 7:52 AM, J. J. Lodder wrote:
>>> lar3ryca <larry@invalid.ca> wrote:
>>>
>>>> On 2024-05-05 02:15, Peter Moylan wrote:
>>>>> On 05/05/24 15:40, lar3ryca wrote:
>>>>>>
>>>>>> The first computer I worked on, though it was not called a computer,
>>>>>> but rather an 'Electronic Accounting Machine'. I 'wrote' programs on
>>>>>> it by plugging wires into a board that made contact with relays.
>>>>>>
>>>>>> It was definitely a computer, though, stepping through programs,
>>>>>> doing math, branching on tested conditions, and reading and writing
>>>>>> to and from I/O devices.
>>>>>>
>>>>>> I worked on that for about two years, at which time I started
>>>>>> working on a different set of machines, and found out that a program
>>>>>> could actually be stored in a memory.
>>>>>
>>>>> We take memory for granted now, but it took a while to get it right.
>>>>> Looking back, the ideas of having main memory on a rotating drum or in
>>>>> columns of mercury sound incredible crude.
>>>>
>>>> How about memory on acoustic delay lines?
>>>> When I worked for CDC, I sometimes got called to Allstate (the insurance
>>>> folks), to repair the terminals.
>>>>
>>>> The memory on those consisted of modules containing spirals of what
>>>> might best be described as 'piano wire'. Bits were input by giving the
>>>> wire a quick twist (twist, then reset back to idle). That generated
>>>> something like a shock wave that travelled around the spiral to a
>>>> transducer on the other end, I can't remember if it twisted one
>>>> direction for a '1' and the other direction for a '0', or if it only
>>>> used one direction of twist, relying on framing pulses and time between
>>>> bit to differentiate between them.
>>>
>>> That's a bit primitive.
>>> One of the first bulk computer memories,
>>> used by Alan Turing himself, was a mercury delay line.
>>> Bits were stored as sound pulses in a column of mercury.
>>>
>>> Just as with a hard drive the computer had to wait
>>> for the right part of the bit string to pass by.
>>> I would have to look up how long its bit string could be,
>>>
>>> Jan
>>>
>>
>> https://www.oxfordreference.com/display/10.1093/oi/authority.20110803095708310
>>
>>     "quartz crystals were used as transducers and the ultrasonic pulses were
>>      passed along a tube of mercury about 5 feet (1.5 meters) in length.
>>      The delay was approximately 1 millisecond but it enabled nearly 1000 pulses
>>      to be stored. Later acoustic memory used magnetostrictive transducers and
>>      nickel-iron wire, with the electrical signals converted into stress waves."
>>
>> A thousand bits, isn't a lot.
>>
>> Some of the first SRAM (suitable for home computer projects)
>> were 256x4 bits and 1024x1 bit static RAM running at 5 volts.
>> "Beautiful stuff". Compared to the dreadful DRAM of the day.
>> And just one of those chips, stores the same stuff as a delay
>> line, and also offers "random" access, so is a lot faster.
>>
>> My breadboarded home computer used (4) 256x4 chips. As a 256x16 array (16 bit CPU).
>
> And my breadboard computer used 8 1024x1 chips (2501?), and that was also my first computer. The memory cost me $85.00 CAD at the time, about 1975 if I remember correctly).
>

That stuff was a lot easier to work with, than the DRAM of the day.
The little SRAM chips meant that anybody could build a computer. You
didn't need Einstein on the team, to do the DRAM. I think my SRAM were
between $2 and $3, but when you had 64 of those on a memory card,
the money adds up quickly. I only needed a few of the chips to get the
prototype running.

Paul

Paul <nospam@needed.invalid> wrote:

> On 5/6/2024 7:52 AM, J. J. Lodder wrote:
> > lar3ryca <larry@invalid.ca> wrote:
> >
> >> On 2024-05-05 02:15, Peter Moylan wrote:
[-]
> >>> We take memory for granted now, but it took a while to get it right.
> >>> Looking back, the ideas of having main memory on a rotating drum or in
> >>> columns of mercury sound incredible crude.
> >>
> >> How about memory on acoustic delay lines?
> >> When I worked for CDC, I sometimes got called to Allstate (the insurance
> >> folks), to repair the terminals.
> >>
> >> The memory on those consisted of modules containing spirals of what
> >> might best be described as 'piano wire'. Bits were input by giving the
> >> wire a quick twist (twist, then reset back to idle). That generated
> >> something like a shock wave that travelled around the spiral to a
> >> transducer on the other end, I can't remember if it twisted one
> >> direction for a '1' and the other direction for a '0', or if it only
> >> used one direction of twist, relying on framing pulses and time between
> >> bit to differentiate between them.
> >
> > That's a bit primitive.
> > One of the first bulk computer memories,
> > used by Alan Turing himself, was a mercury delay line.
> > Bits were stored as sound pulses in a column of mercury.
> >
> > Just as with a hard drive the computer had to wait
> > for the right part of the bit string to pass by.
> > I would have to look up how long its bit string could be,
> >
> > Jan
> >
>
> https://www.oxfordreference.com/display/10.1093/oi/authority.20110803095708310
>
> "quartz crystals were used as transducers and the ultrasonic pulses were
> passed along a tube of mercury about 5 feet (1.5 meters) in length.
> The delay was approximately 1 millisecond but it enabled nearly 1000
> pulses to be stored. Later acoustic memory used magnetostrictive
> transducers and nickel-iron wire, with the electrical signals
> converted into stress waves."
>
> A thousand bits, isn't a lot.

Maybe it was, in 1949. [1]
Trivia: I happened to remember this particular tidbit of information
from the witty chapter title in Hodges' biography of Alan Turing:
"Mercury Delayed,

Jan

[1] Mercury delay lines had the great advantage
that they were already available, from radar applications