Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
The fact that it was done -- or, done by a woman -- isn\'t the impressivethat put men on the moon. I don\'t know how such things are done but
Margaret Hamilton led the team that did. The article says she had her
little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
On 4/29/2023 1:02 AM, Dean Hoffman wrote:
that put men on the moon. I don\'t know how such things are done but
Margaret Hamilton led the team that did. The article says she had her
little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
The fact that it was done -- or, done by a woman -- isn\'t the impressive
aspect.
Rather, the fact that such a thing could be done with the sorts of
*tools* that were available back then.
I recall writing i4004 code and \"assembling\" it with a little index
card (cheat sheet) that I kept folded in my wallet.
Moving to the i8080/85 was pure luxury -- an assembler AND a linkage editor
(instead of writing big blocks of monolithic code).
Nowadays, writing code is a piece of cake; compilers even alert you
to errors in your code, support versioning, simulation, etc.
On 4/29/2023 1:02 AM, Dean Hoffman wrote:
that put men on the moon. I don\'t know how such things are done but
Margaret Hamilton led the team that did. The article says she had her
little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
The fact that it was done -- or, done by a woman -- isn\'t the impressive
aspect.
Rather, the fact that such a thing could be done with the sorts of
*tools* that were available back then.
I recall writing i4004 code and \"assembling\" it with a little index
card (cheat sheet) that I kept folded in my wallet.
Moving to the i8080/85 was pure luxury -- an assembler AND a linkage editor
(instead of writing big blocks of monolithic code).
Nowadays, writing code is a piece of cake; compilers even alert you
to errors in your code, support versioning, simulation, etc.
On Saturday, April 29, 2023 at 3:16:47â¯AM UTC-5, Don Y wrote:
On 4/29/2023 1:02 AM, Dean Hoffman wrote:
that put men on the moon. I don\'t know how such things are done but
Margaret Hamilton led the team that did. The article says she had her
little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
The fact that it was done -- or, done by a woman -- isn\'t the impressive
aspect.
Rather, the fact that such a thing could be done with the sorts of *tools*
that were available back then.
I recall writing i4004 code and \"assembling\" it with a little index card
(cheat sheet) that I kept folded in my wallet.
Moving to the i8080/85 was pure luxury -- an assembler AND a linkage
editor (instead of writing big blocks of monolithic code).
Nowadays, writing code is a piece of cake; compilers even alert you to
errors in your code, support versioning, simulation, etc.
CBS was the only tv network we could get way back when. I remember some of
the shots of the ground crew sitting behind what looked like long tables.
All white guys, all with ties and white shirts. No jackets. A woman
must\'ve felt like a duck in a desert.
On 4/29/2023 1:02 AM, Dean Hoffman wrote:
that put men on the moon. I don\'t know how such things are done but
Margaret Hamilton led the team that did. The article says she had her
little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
The fact that it was done -- or, done by a woman -- isn\'t the impressive
aspect.
Rather, the fact that such a thing could be done with the sorts of
*tools* that were available back then.
I recall writing i4004 code and \"assembling\" it with a little index
card (cheat sheet) that I kept folded in my wallet.
Moving to the i8080/85 was pure luxury -- an assembler AND a linkage editor
(instead of writing big blocks of monolithic code).
Nowadays, writing code is a piece of cake; compilers even alert you
to errors in your code, support versioning, simulation, etc.
On Sat, 29 Apr 2023 01:16:28 -0700, Don Y
blockedofcourse@foo.invalid> wrote:
On 4/29/2023 1:02 AM, Dean Hoffman wrote:
that put men on the moon. I don\'t know how such things are done but
Margaret Hamilton led the team that did. The article says she had her
little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
The fact that it was done -- or, done by a woman -- isn\'t the impressive
aspect.
Rather, the fact that such a thing could be done with the sorts of
*tools* that were available back then.
I recall writing i4004 code and \"assembling\" it with a little index
card (cheat sheet) that I kept folded in my wallet.
Moving to the i8080/85 was pure luxury -- an assembler AND a linkage editor
(instead of writing big blocks of monolithic code).
Nowadays, writing code is a piece of cake; compilers even alert you
to errors in your code, support versioning, simulation, etc.
That\'s yesterday. You can now get AI to write the entire program for
you. That\'s going to put an awful lot of people out of a job!
On Sat, 29 Apr 2023 14:21:05 +0100, Cursitor Doom <cd@notformail.com
wrote:
On Sat, 29 Apr 2023 01:16:28 -0700, Don Y
blockedofcourse@foo.invalid> wrote:
On 4/29/2023 1:02 AM, Dean Hoffman wrote:
that put men on the moon. I don\'t know how such things are done but
Margaret Hamilton led the team that did. The article says she had her
little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
The fact that it was done -- or, done by a woman -- isn\'t the impressive
aspect.
Rather, the fact that such a thing could be done with the sorts of
*tools* that were available back then.
I recall writing i4004 code and \"assembling\" it with a little index
card (cheat sheet) that I kept folded in my wallet.
Moving to the i8080/85 was pure luxury -- an assembler AND a linkage editor
(instead of writing big blocks of monolithic code).
Nowadays, writing code is a piece of cake; compilers even alert you
to errors in your code, support versioning, simulation, etc.
That\'s yesterday. You can now get AI to write the entire program for
you. That\'s going to put an awful lot of people out of a job!
Do you just ask the ai thing \"Write me a billion dollar video game\" ?
The fact that it was done -- or, done by a woman -- isn\'t the impressive
aspect.
Remember that historically it was often the women who were the manual computors
doing the detailed calculations behind the scenes. They seldom got promoted out
of that rut but a handful did. There is a rather nice film about three such
women in the Apollo programme. Hidden Figures.
Rather, the fact that such a thing could be done with the sorts of
*tools* that were available back then.
I recall writing i4004 code and \"assembling\" it with a little index
card (cheat sheet) that I kept folded in my wallet.
I recall someone who still coded 6502 by hand in hex long after assemblers were
available and routine. I could never understand why...
We are not that far off having AI\'s that can take a set of specifications for a
program and churn out code that will implement whatever the specification asks
for. The problem then becomes one of specifying unambiguously what you want
well enough that you don\'t get unwelcome surprises when the edge cases that you
haven\'t thought of arise in use.
It pushes the problem up another level to the
domain experts to say *exactly* what they want.
The generation after that AI\'s
will ask leading questions of the domain expert(s) to catch all the edge cases.
Some static analysis tools are already quite good at it.
I don\'t expect to see software development cease to be a viable career in the
next decade but after that quite a few white collar jobs could well fall to AI.
Lawyers look particularly vulnerable to being out AI\'d.
that put men on the moon. I don\'t know how such things are done but Margaret Hamilton led the team that did. The article says she had her little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
On 4/29/2023 7:59 AM, Martin Brown wrote:
I recall someone who still coded 6502 by hand in hex long after assemblers were
available and routine. I could never understand why...
(sigh) I had an employer who routinely started arguments about whether
or not Z80 code should be expressed in (split-)octal or hex. I just
looked at him, bewildered, and said \"Is there something wrong with
SYMBOLIC??\"
[Classic case of a guy developing a skillset when it was needed...
and never realizing that the need no longer persists and his
skillset has little value!]
On Sat, 29 Apr 2023 08:15:25 -0700, Don Y
blockedofcourse@foo.invalid> wrote:
On 4/29/2023 7:59 AM, Martin Brown wrote:
I recall someone who still coded 6502 by hand in hex long after assemblers were
available and routine. I could never understand why...
(sigh) I had an employer who routinely started arguments about whether
or not Z80 code should be expressed in (split-)octal or hex. I just
looked at him, bewildered, and said \"Is there something wrong with
SYMBOLIC??\"
8080/Z80 are essentially octal machines and as such very easy to
program in octal (or binary). I still do not understand why later
tools went into hexadecimal.In octal, there is only a few opcodes you
have to remember, the register numbering was straight forward.
Branch
targets were absolute, so you could get the branch target directly
from the coding sheet.Inserting/removing instructions between the
branch instruction and branch target did not need borsch
recalculations.
With processors like 6800 used PC relative branch offsets, so
calculating the branch instruction offset was a mess and had to be
redone if instructions were added/removed between branch instruction
and target.
Then the question of debugging. In many cases the debugging had to be
done in octal using computer front panel switches or some EPROM
routine reading octal digits from some sort of keypad. No use of
symbolics.
It is not that self evident when to switch to symbolic. Depends on
processor type and tools available.
[Classic case of a guy developing a skillset when it was needed...
and never realizing that the need no longer persists and his
skillset has little value!]
A good macro assembler can be more versatile than some high level
language source manipulation in e.g. C.
On 4/30/2023 1:54 AM, upsidedown@downunder.com wrote:
8080/Z80 are essentially octal machines and as such very easy to
program in octal (or binary). I still do not understand why later
tools went into hexadecimal.In octal, there is only a few opcodes you
have to remember, the register numbering was straight forward.
Why should you have to remember something as banal as an opcode map?
That\'s like committing to memory the integrals of all the oddball
functions instead of just looking them up (or, letting a symbolic
tool do that on your behalf).
Why would you *want* to write code in ANY numeric form?
Why would you want to *recognize* opcodes in numeric form?
There were tools available that let you use symbolic instructions
and would display dumps as such.
His adherence to (split-)octal derived from the fact that he had
built hardware tools to let you monitor and patch running code
using 7-segment displays -- hexadecimal would have been required
buying four \"more expensive\" displays instead of relying on
six run-of-the-mill 7-segment displays.
And, for machines (minis) with front panels and \"bit switches\",
they could just as easily be grouped in sets of 4 as sets of 3
(e.g., Nova); and the \"display\" was still binary so does it
matter if you read 0177777 off a coding sheet or 0xFFFF?
(in each case, you have to mentally map a non-binary character
into a set of 3 (or 4) bits.
It is not that self evident when to switch to symbolic. Depends on
processor type and tools available.
*I* used a cheap ICE with a serial port to my workstation.
With processors like 6800 used PC relative branch offsets, so
calculating the branch instruction offset was a mess and had to be
redone if instructions were added/removed between branch instruction
and target.
This was even harder with the i4004 as you needed to know where
(within a page) the target was located cuz the branch simply
replaced (PCm,PCl) keeping PCh from the executing opcode.
If the target was on some other page, then you had to use a
long jump.
[Conditional branches were always within the same page as the
bits in the instruction that would have specified PCh for
a long jump were used to select the condition of interest!]
On 30/04/2023 11:02, Don Y wrote:
On 4/30/2023 1:54 AM, upsidedown@downunder.com wrote:
8080/Z80 are essentially octal machines and as such very easy to
program in octal (or binary). I still do not understand why later
tools went into hexadecimal.In octal, there is only a few opcodes you
have to remember, the register numbering was straight forward.
Why should you have to remember something as banal as an opcode map?
To be fair in the early days on hardware that was as primitive as hell it
helped a lot if you could recognise bits of code in an environment where
disassemblers were still running on mainframes.
That\'s like committing to memory the integrals of all the oddball
functions instead of just looking them up (or, letting a symbolic
tool do that on your behalf).
Again it helps to know some of the basic tricks and standard forms so that you
know what you are looking for. MAxima and its ilk are getting better but they
are only usable as a tool and sometimes oddly baulk at relatively simple
problems. I was probably in the last generation taught some of the ways to do
physics problems analytically that would today just be thrown at a fluid in
cell simulation without a second thought.
Why would you *want* to write code in ANY numeric form?
Why would you want to *recognize* opcodes in numeric form?
There were tools available that let you use symbolic instructions
and would display dumps as such.
Not on the most basic processors and environments there weren\'t. I have vague
recollections of SC/PM and 1802 aka COSMAC being available early on. There were
some expensive breakout boards and ICE\'s. We had the one for 9995 and 9900 as
well as various mainframe style Logitech tools (of mouse fame but long before
they had popularised it they sold compilers).
His adherence to (split-)octal derived from the fact that he had
built hardware tools to let you monitor and patch running code
using 7-segment displays -- hexadecimal would have been required
buying four \"more expensive\" displays instead of relying on
six run-of-the-mill 7-segment displays.
I think octal dates back much further to using mass produced decimal nixie
tubes but only using digits 0-7 in each one. I recall having to bootstrap the
PDP7 from the front panel to load the tape reader to then load a programme. It
was a very nifty machine with a type 340 vector graphics display and could play
a mean game of asteroids and dodgy chess where it would sometimes cheat if you
backed it into a corner.
The bootstrap reader tape would get very dog eared and have to be repunched
from time to time.
And, for machines (minis) with front panels and \"bit switches\",
they could just as easily be grouped in sets of 4 as sets of 3
(e.g., Nova); and the \"display\" was still binary so does it
matter if you read 0177777 off a coding sheet or 0xFFFF?
(in each case, you have to mentally map a non-binary character
into a set of 3 (or 4) bits.
There were a couple of Novas in the same room but they didn\'t have anything
like the appeal of the PDP-7 with its fancy vector graphics.
It is not that self evident when to switch to symbolic. Depends on
processor type and tools available.
*I* used a cheap ICE with a serial port to my workstation.
Very early on (as students) we pretty much only had bare metal CPUs with not
much memory to play with and native cunning to get it all working. Data entry
was painful especially on the Sinclair MK14 implementation which assumed a
terminal and required manual pressing of term/mem/term as each byte was entered
in hex. OTOH it was a bargain for £40 a kit.
6502 based Acorn atom and their much nicer BBC Micro were amongst the first
truly successful mass market home computers in the UK.
Sinclair reacted by launching their rival Z80 based ZX Spectrum which also had
its adherents (it was also cheaper and so more popular).
On 29/04/2023 09:02, Dean Hoffman wrote:
that put men on the moon. I don\'t know how such things are done but Margaret Hamilton led the team that did. The article says she had her little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
What about Katherine Johnson, Dorothy Vaughan, and Mary Jackson?
Oh wait, they\'re black. Fox probably doesn\'t count them as American.
On Sat, 29 Apr 2023 08:15:25 -0700, Don Y
blockedofcourse@foo.invalid> wrote:
On 4/29/2023 7:59 AM, Martin Brown wrote:
I recall someone who still coded 6502 by hand in hex long after assemblers were
available and routine. I could never understand why...
(sigh) I had an employer who routinely started arguments about whether
or not Z80 code should be expressed in (split-)octal or hex. I just
looked at him, bewildered, and said \"Is there something wrong with
SYMBOLIC??\"
8080/Z80 are essentially octal machines and as such very easy to
program in octal (or binary). I still do not understand why later
tools went into hexadecimal.In octal, there is only a few opcodes you
have to remember, the register numbering was straight forward. Branch
targets were absolute, so you could get the branch target directly
from the coding sheet.Inserting/removing instructions between the
branch instruction and branch target did not need borsch
recalculations.
With processors like 6800 used PC relative branch offsets, so
calculating the branch instruction offset was a mess and had to be
redone if instructions were added/removed between branch instruction
and target.
Then the question of debugging. In many cases the debugging had to be
done in octal using computer front panel switches or some EPROM
routine reading octal digits from some sort of keypad. No use of
symbolics.
It is not that self evident when to switch to symbolic. Depends on
processor type and tools available.
[Classic case of a guy developing a skillset when it was needed...
and never realizing that the need no longer persists and his
skillset has little value!]
A good macro assembler can be more versatile than some high level
language source manipulation in e.g. C.
that put men on the moon. I don\'t know how such things are done but Margaret Hamilton led the team that did. The article says she had her little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
On Saturday, April 29, 2023 at 4:02:11?AM UTC-4, Dean Hoffman wrote:
that put men on the moon. I don\'t know how such things are done but Margaret Hamilton led the team that did. The article says she had her little one with her at work now and then.
https://www.foxnews.com/lifestyle/meet-american-who-wrote-moon-landing-software-margaret-hamilton-computer-whiz-mom
Primitive little nothing of a minimalist computer. Article is just more redacted history. Not quite as bad as that previous fiction about a woman programmer, who worked in data reduction on GEOSAT, inventing the concept of GPS. Turns out the publicity campaign for that bs originated with a totally unqualified sorority \"sister.\"
These kinds of stories originate with activities dependent on government funding. Listen to these lunatics and they\'ll have you believing Sally Hemmings wrote the original draft to the Declaration of Independence.
On 4/30/2023 1:54 AM, upsidedown@downunder.com wrote:
On Sat, 29 Apr 2023 08:15:25 -0700, Don Y
blockedofcourse@foo.invalid> wrote:
On 4/29/2023 7:59 AM, Martin Brown wrote:
I recall someone who still coded 6502 by hand in hex long after assemblers were
available and routine. I could never understand why...
(sigh) I had an employer who routinely started arguments about whether
or not Z80 code should be expressed in (split-)octal or hex. I just
looked at him, bewildered, and said \"Is there something wrong with
SYMBOLIC??\"
8080/Z80 are essentially octal machines and as such very easy to
program in octal (or binary). I still do not understand why later
tools went into hexadecimal.In octal, there is only a few opcodes you
have to remember, the register numbering was straight forward.
Why should you have to remember something as banal as an opcode map?
That\'s like committing to memory the integrals of all the oddball
functions instead of just looking them up (or, letting a symbolic
tool do that on your behalf).
Why would you *want* to write code in ANY numeric form?
Why would you want to *recognize* opcodes in numeric form?
There were tools available that let you use symbolic instructions
and would display dumps as such.