On Mon, 8 Oct 2007 10:09:49 -0700, "Joel Kolstad"
wrote:

"Joel Kolstad" wrote in message
...
XEQ P (the parallel resistance program :-) )

^^^ BTW, while everyone knows that the parallel impedance formula is 1/Result
= 1/Z1 + 1/Z2, for those of us who can get geeky with respect to calculators
and numerical methods, using Result = (Z1*Z2)/(Z1+Z2) is more accurate when Z1
is significantly larger or smaller than Z2... hence some of the motivation to
write a program to do it each time.


When I have anything worth programming, I do it in PowerBasic. It's
portable and archivable, and I can use double floats if needed. PB
even has 80-bit floats!

Is the new 35S okay? It looks complicated to me. I still have a couple
of HP35's that work, and they are ideal: simple, clean, and pi is in
plain sight.

They should have precisely cloned the 35!

John

John Larkin posted to
alt.binaries.schematics.electronic:

On Mon, 8 Oct 2007 10:09:49 -0700, "Joel Kolstad"
wrote:

"Joel Kolstad" wrote in message
...
XEQ P (the parallel resistance program :-) )

^^^ BTW, while everyone knows that the parallel impedance formula is
1/Result = 1/Z1 + 1/Z2, for those of us who can get geeky with
respect to calculators and numerical methods, using Result =
(Z1*Z2)/(Z1+Z2) is more accurate when Z1 is significantly larger or
smaller than Z2... hence some of the motivation to write a program
to do it each time.


When I have anything worth programming, I do it in PowerBasic. It's
portable and archivable, and I can use double floats if needed. PB
even has 80-bit floats!

Is the new 35S okay? It looks complicated to me. I still have a
couple of HP35's that work, and they are ideal: simple, clean, and
pi is in plain sight.

They should have precisely cloned the 35!

John

Each to their own, my prince of breed is the HP45.

I have a couple of surviving slide rules, from high school and college
days.

On Mon, 08 Oct 2007 23:43:09 -0700, ChairmanOfTheBored
wrote:

On Sun, 07 Oct 2007 14:46:48 -0700, John Larkin
wrote:



Idiot, you can't make a product, then state "distribute freely under
standard GPL rules", without declaring the entire GPL standard license.


Can't? I just did.

John

"John Larkin" wrote in message
...
Is the new 35S okay?

Based on what you've said about the 35, I'd say that the 35S is mostly OK.
Things they could have done better on:

-- STO is shifted-RCL (it should have been its own key!)
-- Using (and converting between) hex/octal/binary is questionably
implemented. Regardless of the display mode, numbers entered are always
considered decimal unless you suffic them with "h" "o" or "b" -- which
requires two keystrokes. Every other implementation I've used assumed that,
if you're displaying things it hex, it implies you want to then enter numbers
in hex as well.
-- For programming, there's no REAL(x) or IMAG(x). There is ARG(x) and
MAG(x), so you can extract the real and imaginary parts of a number using
MAG(x)*cos(ARG(x)) etc., but numerically that's not quite as desirable. To
make matters worse, there's a bug that makes the transcendental functions far
less accurate than they should be for angles near 90 degrees.
-- Oddly, taking the square root of a complex number doesn't work, but you can
raise it to the 0.5 power instead.
-- The display routine isn't particularly "smart" for complex numbers.
Something like -j0.1592 displays, in engineering mode, as
"0.000e0i-159.2e-3" -- with the "3" cut off the end of the display (you press
the right arrow to scroll).

Still, it's definitely HP's best effort in a long time. Definitely qualifies
as "good" in my book -- just not "excellent" like the older machines. For
quite some time HP's calculators definitely seemed to be getting worse, after
the HP/Agilent split caused the shutdown of the Corvallis, OR calculator
division. Calculators then bounced around between Singapore and Australia a
bit, and now they're finally back in Boise... mostly. The guys working on
them now definitely seem to have their heads screwed on right, and from what
was said at the conference it appears that there really is a chance HP will
return to "excellent" calculators in the next couple of years.

Here's one data point: My boss's last HP calculator was the HP-32S, and he's
been quite happy with the new 35S... so much so he bought two.

A more in-depth review of the 35S:
http://www.hpcc.org/datafile/V26Special/the35s.pdf

They should have precisely cloned the 35!

Well, HP didn't, but there are many dozens of HP-35 emulators out there these
days (for PCs, PDAs, etc.). There's even one guy (Eric Smith) who wrote a CPU
emulator for the old HP calculators and built actual hardware (off-the-shelf
LCD and ARM CPU running the emulator) to recreate some of the old models. One
snag was getting a ROM dump... for the HP-35, they did it optically! --
http://www.pmonta.com/calculators/hp-35/

---Joel

On Mon, 08 Oct 2007 23:54:12 -0700, ChairmanOfTheBored
wrote:

On Mon, 08 Oct 2007 18:48:32 -0700, John Larkin
wrote:

On Mon, 8 Oct 2007 10:09:49 -0700, "Joel Kolstad"
wrote:

"Joel Kolstad" wrote in message
...
XEQ P (the parallel resistance program :-) )

^^^ BTW, while everyone knows that the parallel impedance formula is 1/Result
= 1/Z1 + 1/Z2, for those of us who can get geeky with respect to calculators
and numerical methods, using Result = (Z1*Z2)/(Z1+Z2) is more accurate when Z1
is significantly larger or smaller than Z2... hence some of the motivation to
write a program to do it each time.


When I have anything worth programming, I do it in PowerBasic. It's
portable and archivable, and I can use double floats if needed. PB
even has 80-bit floats!

Is the new 35S okay? It looks complicated to me. I still have a couple
of HP35's that work, and they are ideal: simple, clean, and pi is in
plain sight.

They should have precisely cloned the 35!


Whimp.

Bad speller!

John

John Larkin wrote:

On Mon, 8 Oct 2007 10:09:49 -0700, "Joel Kolstad"
wrote:


"Joel Kolstad" wrote in message
...

XEQ P (the parallel resistance program :-) )

^^^ BTW, while everyone knows that the parallel impedance formula is 1/Result
= 1/Z1 + 1/Z2, for those of us who can get geeky with respect to calculators
and numerical methods, using Result = (Z1*Z2)/(Z1+Z2) is more accurate when Z1
is significantly larger or smaller than Z2... hence some of the motivation to
write a program to do it each time.



When I have anything worth programming, I do it in PowerBasic. It's
portable and archivable, and I can use double floats if needed. PB
even has 80-bit floats!

Is the new 35S okay? It looks complicated to me. I still have a couple
of HP35's that work, and they are ideal: simple, clean, and pi is in
plain sight.

They should have precisely cloned the 35!


They did come out with an anniversary edition of the 12C. Doesn't quite
have the robust feel of the old ones and unfortunately they did not do
the 11C. So I have to baby the 11C, make sure it lasts until retirement.

--
Regards, Joerg

http://www.analogconsultants.com

ChairmanOfTheBored wrote:

On Tue, 9 Oct 2007 10:35:18 -0700, "Joel Kolstad"
wrote:


Well, HP didn't, but there are many dozens of HP-35 emulators out there these
days (for PCs, PDAs, etc.). There's even one guy (Eric Smith) who wrote a CPU
emulator for the old HP calculators and built actual hardware (off-the-shelf
LCD and ARM CPU running the emulator) to recreate some of the old models. One
snag was getting a ROM dump... for the HP-35, they did it optically! --
http://www.pmonta.com/calculators/hp-35/

---Joel



10 microns! Now that's cool!


HP made the best calculator mankind ever saw, and may never see again.
My HP-11C ran a whopping 15 years on the first set of battery. $6 later
it had a new set and that's on its fourth year now. Modern equipment
goes like this: BIG LiIon Battery. Turn on ... hard disk grinding ... HD
grinding ... some more HD grinding. Battery is now down to 97%. Fire up
WP, write "Hello World" ... down to 95%. So yeah, my new processor here
is 65nm but what good does it do in a portable device if they can't
figure out a power efficient design anymore?

--
Regards, Joerg

http://www.analogconsultants.com

Fred Abse wrote:

On Mon, 08 Oct 2007 07:52:52 -0700, Joerg wrote:


I still use it. It's the Aristor-Scholar 0903VS


I still use my Faber-Castell 1/54 Darmstadt.


Why were they all made in Germany? Ok, I came from there myself so in my
case that's different. But even in the US 95% plus of the slide rules I
saw were Aristo, Boettcher, Faber-Castell, Nestler but rarely a Pickett
oder Fredericks Post.

--
Regards, Joerg

http://www.analogconsultants.com

Eeyore wrote:


John Larkin wrote:


Do you think a right-angle trace is more likely to break than a curve
or bevel? I can't see why.


Concentration of mechanical stress AIUI.


Jim would say that it's the left-angle traces that cause all the problems.

--
SCNR, Joerg

http://www.analogconsultants.com

"Joerg" wrote in message
...
My HP-11C ran a whopping 15 years on the first set of battery. $6 later it
had a new set and that's on its fourth year now.

You'll be pleased to know that the guy who's sort of the defacto head design
engineer for HP calculators these days actually cares about power consumption
and gave a presentation demonstrating that things are getting better in the
sense of future calculators that are comparably featured to those vintage ones
(like the 11C) will use the same or less power than before. He had a story
about how he went and verbally kicked some overseas engineer's butt for
replacing a switching regulator with a linear and significantly shortening
battery life on a certain model; the guy said, "but a linear is cheaper!" to
which he replied, "Yes, but it's stupid!"

That's the kind of engineer I want designing calculators -- one who knows his
customers' priorities. :-)

(I know that if it were you, Joerg, the switcher would end up being cheaper as
well, of course. :-) )

Unfortunately details of this (specific models and features and hardware) are
under NDA.

---Joel

Joerg wrote:
Fred Abse wrote:

On Mon, 08 Oct 2007 07:52:52 -0700, Joerg wrote:


I still use it. It's the Aristor-Scholar 0903VS


I still use my Faber-Castell 1/54 Darmstadt.


Why were they all made in Germany? Ok, I came from there myself so in my
case that's different. But even in the US 95% plus of the slide rules I
saw were Aristo, Boettcher, Faber-Castell, Nestler but rarely a Pickett
oder Fredericks Post.

Around here in Maryland, they were all Picket, Sterling, or K&E.

The Picketts were very popular, but being aluminum, easily damaged.
The Sterlings were cheap plastic trash, and the K&Es were a true
piece of art.

-Chuck

Joel Kolstad wrote:
"Joerg" wrote in message
...

My HP-11C ran a whopping 15 years on the first set of battery. $6 later it
had a new set and that's on its fourth year now.


You'll be pleased to know that the guy who's sort of the defacto head design
engineer for HP calculators these days actually cares about power consumption
and gave a presentation demonstrating that things are getting better in the
sense of future calculators that are comparably featured to those vintage ones
(like the 11C) will use the same or less power than before. He had a story
about how he went and verbally kicked some overseas engineer's butt for
replacing a switching regulator with a linear and significantly shortening
battery life on a certain model; the guy said, "but a linear is cheaper!" to
which he replied, "Yes, but it's stupid!"

That's the kind of engineer I want designing calculators -- one who knows his
customers' priorities. :-)

(I know that if it were you, Joerg, the switcher would end up being cheaper as
well, of course. :-) )


Well, I'd have to see what they've go so far ;-)


Unfortunately details of this (specific models and features and hardware) are
under NDA.


Thanks for sharing this, Joel. So there is hope in case my old 11C
croaks. Actually it's my wife's.

--
Regards, Joerg

http://www.analogconsultants.com

On Tue, 9 Oct 2007 10:35:18 -0700, "Joel Kolstad"
wrote:

"John Larkin" wrote in message
.. .
Is the new 35S okay?

Based on what you've said about the 35, I'd say that the 35S is mostly OK.
Things they could have done better on:

-- STO is shifted-RCL (it should have been its own key!)
-- Using (and converting between) hex/octal/binary is questionably
implemented. Regardless of the display mode, numbers entered are always
considered decimal unless you suffic them with "h" "o" or "b" -- which
requires two keystrokes. Every other implementation I've used assumed that,
if you're displaying things it hex, it implies you want to then enter numbers
in hex as well.
-- For programming, there's no REAL(x) or IMAG(x). There is ARG(x) and
MAG(x), so you can extract the real and imaginary parts of a number using
MAG(x)*cos(ARG(x)) etc., but numerically that's not quite as desirable. To
make matters worse, there's a bug that makes the transcendental functions far
less accurate than they should be for angles near 90 degrees.
-- Oddly, taking the square root of a complex number doesn't work, but you can
raise it to the 0.5 power instead.
-- The display routine isn't particularly "smart" for complex numbers.
Something like -j0.1592 displays, in engineering mode, as
"0.000e0i-159.2e-3" -- with the "3" cut off the end of the display (you press
the right arrow to scroll).

Still, it's definitely HP's best effort in a long time. Definitely qualifies
as "good" in my book -- just not "excellent" like the older machines. For
quite some time HP's calculators definitely seemed to be getting worse, after
the HP/Agilent split caused the shutdown of the Corvallis, OR calculator
division. Calculators then bounced around between Singapore and Australia a
bit, and now they're finally back in Boise... mostly. The guys working on
them now definitely seem to have their heads screwed on right, and from what
was said at the conference it appears that there really is a chance HP will
return to "excellent" calculators in the next couple of years.

Here's one data point: My boss's last HP calculator was the HP-32S, and he's
been quite happy with the new 35S... so much so he bought two.

A more in-depth review of the 35S:
http://www.hpcc.org/datafile/V26Special/the35s.pdf

They should have precisely cloned the 35!

Well, HP didn't, but there are many dozens of HP-35 emulators out there these
days (for PCs, PDAs, etc.). There's even one guy (Eric Smith) who wrote a CPU
emulator for the old HP calculators and built actual hardware (off-the-shelf
LCD and ARM CPU running the emulator) to recreate some of the old models. One
snag was getting a ROM dump... for the HP-35, they did it optically! --
http://www.pmonta.com/calculators/hp-35/

Cool.

My first programmable calculator was the astonishing desktop HP9100,
all discrete transistors with a reverse Polish stack displayed on a
7-segment vector-stroke CRT! Among other things, I used it to simulate
steamship propulsion/control systems, and got into the marine
automation business as a result. An old salt looked at my
throttle-position vs time graphs and said "that's the way a good
operator would work the valve manually" and we got the job.

I have two 9100's, both dead, and HP refuses to release the
schematics, after over 30 years!

John

On Tue, 09 Oct 2007 14:14:28 -0700, ChairmanOfTheBored
wrote:

On Tue, 09 Oct 2007 08:25:03 -0700, John Larkin
wrote:

On Mon, 08 Oct 2007 23:43:09 -0700, ChairmanOfTheBored
wrote:

On Sun, 07 Oct 2007 14:46:48 -0700, John Larkin
wrote:



Idiot, you can't make a product, then state "distribute freely under
standard GPL rules", without declaring the entire GPL standard license.


Can't? I just did.

John



Not without breaking the GPL rules, dumb****.

Oh, and that "copyright" horse****... NOT!

Well, post some of your source code and show us how it's done.

John

On Tue, 09 Oct 2007 22:13:07 -0700, ChairmanOfTheBored
wrote:

On Tue, 09 Oct 2007 20:52:11 -0700, John Larkin
wrote:

I have two 9100's, both dead, and HP refuses to release the
schematics, after over 30 years!


Well trace back through it, and draft one up.

Show us how it's done!

That would be a monstrous amount of work, and most of the logic is in
little epoxy-dipped sip-type boards, like the things you used to see
in old TV sets.

The 9100 used core memory to store the program; if it was running a
program, you could shut it off and turn it back on, and it would keep
going. It had three different roms: diode matrix, resistor matrix, and
a multilayer pcb that used magnetic coupling between layers.

The HP archivist has all the docs, but won't let me see them. But they
spent over a million bucks to buy "the garage."

John

Chuck Harris wrote:

Joerg wrote:

Fred Abse wrote:

On Mon, 08 Oct 2007 07:52:52 -0700, Joerg wrote:


I still use it. It's the Aristor-Scholar 0903VS



I still use my Faber-Castell 1/54 Darmstadt.


Why were they all made in Germany? Ok, I came from there myself so in
my case that's different. But even in the US 95% plus of the slide
rules I saw were Aristo, Boettcher, Faber-Castell, Nestler but rarely
a Pickett oder Fredericks Post.


Around here in Maryland, they were all Picket, Sterling, or K&E.

The Picketts were very popular, but being aluminum, easily damaged.
The Sterlings were cheap plastic trash, and the K&Es were a true
piece of art.


I think many of the German ones were melamine resin. My Aristo is really
sturdy. But yeah, in school I had a simple one that always got stuck
when it was warm and the old trick with "soap greasing" didn't work. I
don't dare to do that with a piece that cannot be replaced but I always
wonder what Teflon spray would do.

--
Regards, Joerg

http://www.analogconsultants.com

John Larkin wrote:

On Tue, 09 Oct 2007 22:13:07 -0700, ChairmanOfTheBored
wrote:


On Tue, 09 Oct 2007 20:52:11 -0700, John Larkin
wrote:


I have two 9100's, both dead, and HP refuses to release the
schematics, after over 30 years!


Maybe they are still afraid the Russians would copy it ;-)


Well trace back through it, and draft one up.

Show us how it's done!


That would be a monstrous amount of work, and most of the logic is in
little epoxy-dipped sip-type boards, like the things you used to see
in old TV sets.

The 9100 used core memory to store the program; if it was running a
program, you could shut it off and turn it back on, and it would keep
going. It had three different roms: diode matrix, resistor matrix, and
a multilayer pcb that used magnetic coupling between layers.

The HP archivist has all the docs, but won't let me see them. But they
spent over a million bucks to buy "the garage."


Wasn't that accidentally bulldozed away?

--
Regards, Joerg

http://www.analogconsultants.com

John Larkin wrote:

On Mon, 08 Oct 2007 23:54:12 -0700, ChairmanOfTheBored
wrote:

On Mon, 08 Oct 2007 18:48:32 -0700, John Larkin
wrote:

On Mon, 8 Oct 2007 10:09:49 -0700, "Joel Kolstad"
wrote:

"Joel Kolstad" wrote in message
...
XEQ P (the parallel resistance program :-) )

^^^ BTW, while everyone knows that the parallel impedance formula is 1/Result
= 1/Z1 + 1/Z2, for those of us who can get geeky with respect to calculators
and numerical methods, using Result = (Z1*Z2)/(Z1+Z2) is more accurate when Z1
is significantly larger or smaller than Z2... hence some of the motivation to
write a program to do it each time.


When I have anything worth programming, I do it in PowerBasic. It's
portable and archivable, and I can use double floats if needed. PB
even has 80-bit floats!

Is the new 35S okay? It looks complicated to me. I still have a couple
of HP35's that work, and they are ideal: simple, clean, and pi is in
plain sight.

They should have precisely cloned the 35!


Whimp.

Bad speller!

John


If IQ was rounded to the nearest multiple of ten, dimBulb would be a
zero.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

"John Larkin" wrote
in message ...
On Mon, 8 Oct 2007 10:09:49 -0700, "Joel Kolstad"
wrote:

"Joel Kolstad" wrote in
message
...
XEQ P (the parallel resistance program :-) )

^^^ BTW, while everyone knows that the parallel impedance
formula is 1/Result
= 1/Z1 + 1/Z2, for those of us who can get geeky with respect
to calculators
and numerical methods, using Result = (Z1*Z2)/(Z1+Z2) is more
accurate when Z1
is significantly larger or smaller than Z2... hence some of
the motivation to
write a program to do it each time.


When I have anything worth programming, I do it in PowerBasic.
It's
portable and archivable, and I can use double floats if needed.
PB
even has 80-bit floats!


if you need 80 bit floats ....
I am surprised you guys are not using some functional programming
language like scheme (lisp/Lambda calc variants) where your reals
are number abstractions with no language or data type imposed
limit on the number size or precision and of course no numerical
methods issues/errors from typical float/double data type
limitations
just wondering,
robb

"ChairmanOfTheBored" wrote in message
...

One more thing for the OP.

The word you are looking for is CRITIQUE.

thanks , im still gettin my edumacation ...

however, i think the context implies the use of criticism over
critique as in the expression "constructive criticism" where
critique alone would have been appropriate. At the time of
posting i could not remeber how to spell
criticism/critisicm/critisism and i did not want to look dumb or
invite comments on spelling errors so i went with critic

oh well i guess i failed on both counts

robb

On Thu, 11 Oct 2007 10:23:27 -0400, "robb" wrote:


"John Larkin" wrote
in message ...
On Mon, 8 Oct 2007 10:09:49 -0700, "Joel Kolstad"
wrote:

"Joel Kolstad" wrote in
message
...
XEQ P (the parallel resistance program :-) )

^^^ BTW, while everyone knows that the parallel impedance
formula is 1/Result
= 1/Z1 + 1/Z2, for those of us who can get geeky with respect
to calculators
and numerical methods, using Result = (Z1*Z2)/(Z1+Z2) is more
accurate when Z1
is significantly larger or smaller than Z2... hence some of
the motivation to
write a program to do it each time.


When I have anything worth programming, I do it in PowerBasic.
It's
portable and archivable, and I can use double floats if needed.
PB
even has 80-bit floats!


if you need 80 bit floats ....
I am surprised you guys are not using some functional programming
language like scheme (lisp/Lambda calc variants) where your reals
are number abstractions with no language or data type imposed
limit on the number size or precision and of course no numerical
methods issues/errors from typical float/double data type
limitations
just wondering,
robb


I've done several digital delay generators, with a uP inside,
programmed in assembly. Internally, we store time as a 64-bit integer,
with lsb of 1 picosecond, and max range of 2000 seconds. I couldn't
find any calculator program that would compute my scaling constants
correctly, much less express them as binary fractionals, so I wound up
fine-tweaking several factors iteratively until they gave the right
results. What a pain. Most PC calculators just use the standard
floats.

John

On Mon, 08 Oct 2007 23:47:44 -0700, ChairmanOfTheBored
wrote:

On Sun, 07 Oct 2007 22:14:29 GMT, Joerg
wrote:

John Larkin wrote:

On Sun, 07 Oct 2007 21:14:53 GMT, Joerg
wrote:



I wrote a little LC filter de-normalization program (free for the
asking) that makes it fairly easy. Look up a prototype in Williams,
plug that in, and fiddle with terminations and cutoffs until you
stumble onto a set of values that you can get. Then run LTspice and
see how it will look.


I'd be interested. "jsc AT ieee DOT org" is shorter to type than my biz
email.

I use routines like Aade but mostly just the old slide rule. Goes like
"Ok, we've got 1.2uH, 2.7uH and 4.7uH available here plus the E12 series
for caps, lets see how we can get into the ballpark with that". I am a
bit worried about my Williams, the pages are beginning to turn yellow.


Heck, here it is:

John


Ah, a good old Basic program and under DOS. I wonder why so many people
are dissing Basic. It works and under DOS the execution time seems to be
measurable only in microseconds. No hourglass and stuff like that.

But I won't part with my old slide rule ;-)


One more thing for the OP.

The word you are looking for is CRITIQUE.

And the word you are looking for is PARROT.

John

John Larkin wrote:

On Thu, 11 Oct 2007 10:23:27 -0400, "robb" wrote:


"John Larkin" wrote
in message ...

On Mon, 8 Oct 2007 10:09:49 -0700, "Joel Kolstad"
wrote:


"Joel Kolstad" wrote in

message

...

XEQ P (the parallel resistance program :-) )

^^^ BTW, while everyone knows that the parallel impedance

formula is 1/Result

= 1/Z1 + 1/Z2, for those of us who can get geeky with respect

to calculators

and numerical methods, using Result = (Z1*Z2)/(Z1+Z2) is more

accurate when Z1

is significantly larger or smaller than Z2... hence some of

the motivation to

write a program to do it each time.


When I have anything worth programming, I do it in PowerBasic.

It's

portable and archivable, and I can use double floats if needed.

PB

even has 80-bit floats!


if you need 80 bit floats ....
I am surprised you guys are not using some functional programming
language like scheme (lisp/Lambda calc variants) where your reals
are number abstractions with no language or data type imposed
limit on the number size or precision and of course no numerical
methods issues/errors from typical float/double data type
limitations
just wondering,
robb



I've done several digital delay generators, with a uP inside,
programmed in assembly. Internally, we store time as a 64-bit integer,
with lsb of 1 picosecond, and max range of 2000 seconds. I couldn't
find any calculator program that would compute my scaling constants
correctly, much less express them as binary fractionals, so I wound up
fine-tweaking several factors iteratively until they gave the right
results. What a pain. Most PC calculators just use the standard
floats.


Doesn't SciLab offer a 64bit integer library?

--
Regards, Joerg

http://www.analogconsultants.com

On Thu, 11 Oct 2007 10:23:27 -0400, robb wrote:
"John Larkin" wrote
On Mon, 8 Oct 2007 10:09:49 -0700, "Joel Kolstad"
"Joel Kolstad" wrote in
message
...
XEQ P (the parallel resistance program :-) )

^^^ BTW, while everyone knows that the parallel impedance
formula is 1/Result
= 1/Z1 + 1/Z2, for those of us who can get geeky with respect
to calculators
and numerical methods, using Result = (Z1*Z2)/(Z1+Z2) is more
accurate when Z1
is significantly larger or smaller than Z2... hence some of
the motivation to
write a program to do it each time.

When I have anything worth programming, I do it in PowerBasic.
It's
portable and archivable, and I can use double floats if needed.
PB
even has 80-bit floats!

if you need 80 bit floats ....
I am surprised you guys are not using some functional programming
language like scheme (lisp/Lambda calc variants) where your reals
are number abstractions with no language or data type imposed
limit on the number size or precision and of course no numerical
methods issues/errors from typical float/double data type
limitations
just wondering,

Loosely-typed or untyped data is a mare's nest of bugs just
waiting to happen.

Cheers!
Rich

On Thu, 11 Oct 2007 09:06:46 -0700, John Larkin wrote:

so I wound up
fine-tweaking several factors iteratively until they gave the right
results.

Kind of like the warmingists do with their atmospheric "models"? ;-)

Cheers!
Rich

Richard The Dreaded Libertarian wrote:

On Thu, 11 Oct 2007 09:06:46 -0700, John Larkin wrote:


so I wound up
fine-tweaking several factors iteratively until they gave the right
results.


Kind of like the warmingists do with their atmospheric "models"? ;-)


What warming? We already ran half a dozen loads through the wood stove
and it's not even mid-October! Never happened in 10 years out here.

--
Regards, Joerg

http://www.analogconsultants.com

On Thu, 11 Oct 2007 18:21:02 GMT, Rich Grise wrote:

On Thu, 11 Oct 2007 10:23:27 -0400, robb wrote:
"John Larkin" wrote
On Mon, 8 Oct 2007 10:09:49 -0700, "Joel Kolstad"
"Joel Kolstad" wrote in
message
...
XEQ P (the parallel resistance program :-) )

^^^ BTW, while everyone knows that the parallel impedance
formula is 1/Result
= 1/Z1 + 1/Z2, for those of us who can get geeky with respect
to calculators
and numerical methods, using Result = (Z1*Z2)/(Z1+Z2) is more
accurate when Z1
is significantly larger or smaller than Z2... hence some of
the motivation to
write a program to do it each time.

When I have anything worth programming, I do it in PowerBasic.
It's
portable and archivable, and I can use double floats if needed.
PB
even has 80-bit floats!

if you need 80 bit floats ....
I am surprised you guys are not using some functional programming
language like scheme (lisp/Lambda calc variants) where your reals
are number abstractions with no language or data type imposed
limit on the number size or precision and of course no numerical
methods issues/errors from typical float/double data type
limitations
just wondering,

Loosely-typed or untyped data is a mare's nest of bugs just
waiting to happen.

There are no types in assembly. We don't need no stinkin' types!

John

"Rich Grise" wrote in message
news
Loosely-typed or untyped data is a mare's nest of bugs just
waiting to happen.

For relatively simple programs the productivity gains of not dealing with
typed data (and explicit data conversions) often outweighs the potential for
problems. Especially when there's at least the appearance of only a couple of
core data types, such as strings and floating point numbers.

For complex programs I think more people agree that stronger typing does a lot
of good. Exactly where the dividing line between "simple" and "complex" lies
is largely a function of just how skilled the individual programmer is!

"Rich Grise" wrote in message
news
On Thu, 11 Oct 2007 10:23:27 -0400, robb wrote:
[snip]
even has 80-bit floats!

if you need 80 bit floats ....
I am surprised you guys are not using some functional
programming
language like scheme (lisp/Lambda calc variants) where your
reals
are number abstractions with no language or data type imposed
limit on the number size or precision and of course no
numerical
methods issues/errors from typical float/double data type
limitations
just wondering,

Loosely-typed or untyped data is a mare's nest of bugs just
waiting to happen.

Cheers!
Rich


I should have known better than to bring up software in hardware
group :}

perhaps you meant flexible or dynamic instead of loosely and un
typed ?

it is very strongly typed as there are a few predefined basic
prmitives and "everything" is a defined type even operators and
and functions are all user defined types (or library of other
users types) some say scalable

i am sure they said the same things about basic when it was first
introduced along with a host of other seemingly magical things
that illicited suspicion at first and then gave way to acceptance
and trust.

a scheme implementation is what ? basic with some datastructures
{stacks,lists,etc}

and look at that google reveals 'BIT' a scheme implementation
for microcontroller running in 4k ram and stored in 13k rom among
others

not trying to start any wars i was curious why hrdw engy types
still use basic as opposed to other very useful tools that exist
that have eliminated many problems associated with those old
tools,
robb

robb wrote:
"Rich Grise" wrote in message
news
On Thu, 11 Oct 2007 10:23:27 -0400, robb wrote:

[snip]

even has 80-bit floats!

if you need 80 bit floats ....
I am surprised you guys are not using some functional

programming

language like scheme (lisp/Lambda calc variants) where your

reals

are number abstractions with no language or data type imposed
limit on the number size or precision and of course no

numerical

methods issues/errors from typical float/double data type
limitations
just wondering,

Loosely-typed or untyped data is a mare's nest of bugs just
waiting to happen.

Cheers!
Rich



I should have known better than to bring up software in hardware
group :}

perhaps you meant flexible or dynamic instead of loosely and un
typed ?


Amongst us hardware dudes flexible means like that hose that's around
cables that connect to a pump. Dynamic is some kind of memory, right
after ferrite cores was gitten short shrift.

[...]

--
SCNR, Joerg

http://www.analogconsultants.com

On Thu, 11 Oct 2007 20:57:59 -0700, ChairmanOfTheBored
wrote:

On Thu, 11 Oct 2007 09:07:33 -0700, John Larkin
wrote:

On Mon, 08 Oct 2007 23:47:44 -0700, ChairmanOfTheBored
wrote:

On Sun, 07 Oct 2007 22:14:29 GMT, Joerg
wrote:

John Larkin wrote:

On Sun, 07 Oct 2007 21:14:53 GMT, Joerg
wrote:



I wrote a little LC filter de-normalization program (free for the
asking) that makes it fairly easy. Look up a prototype in Williams,
plug that in, and fiddle with terminations and cutoffs until you
stumble onto a set of values that you can get. Then run LTspice and
see how it will look.


I'd be interested. "jsc AT ieee DOT org" is shorter to type than my biz
email.

I use routines like Aade but mostly just the old slide rule. Goes like
"Ok, we've got 1.2uH, 2.7uH and 4.7uH available here plus the E12 series
for caps, lets see how we can get into the ballpark with that". I am a
bit worried about my Williams, the pages are beginning to turn yellow.


Heck, here it is:

John


Ah, a good old Basic program and under DOS. I wonder why so many people
are dissing Basic. It works and under DOS the execution time seems to be
measurable only in microseconds. No hourglass and stuff like that.

But I won't part with my old slide rule ;-)


One more thing for the OP.

The word you are looking for is CRITIQUE.

And the word you are looking for is PARROT.

John


The REAL, NON-imaginary question is: how can you go from displaying a
modicum of gray matter, and jump to this utter horse****? You REALly are
one sad piece of ****, boy.

But I'm having fun, and you're not.

John

On Thu, 11 Oct 2007 21:00:21 -0700, IAmTheSlime
TheSlimeFromYourVideo@oozingacrossyourlivingroomf loor.org wrote:

On Thu, 11 Oct 2007 18:22:32 GMT, Richard The Dreaded Libertarian
wrote:

On Thu, 11 Oct 2007 09:06:46 -0700, John Larkin wrote:

so I wound up
fine-tweaking several factors iteratively until they gave the right
results.

Kind of like the warmingists do with their atmospheric "models"? ;-)


That's OK. I heard the dumb****s are thinking about giving Gore the
Peace Prize for his spew. Reminds me of the year they gave it to that
terrorist, nose picking ******* Arafat!

Don't forget Jimmy Carter, who bribed the Egyptians and Israelis to
make peace, when they were already at peace.

The Nobel committee has really degraded itself. The Emmy awards has
more integrity.

Bill Clinton was agressively lobbying for the Peace Prize; I bet he's
****ed that Al got it.

John

On Thu, 11 Oct 2007 13:13:59 -0700, "Joel Koltner"
wrote:

"Rich Grise" wrote in message
news
Loosely-typed or untyped data is a mare's nest of bugs just
waiting to happen.

For relatively simple programs the productivity gains of not dealing with
typed data (and explicit data conversions) often outweighs the potential for
problems. Especially when there's at least the appearance of only a couple of
core data types, such as strings and floating point numbers.

For complex programs I think more people agree that stronger typing does a lot
of good. Exactly where the dividing line between "simple" and "complex" lies
is largely a function of just how skilled the individual programmer is!


There's a secret, seldom-used, nearly foolproof way to avoid program
bugs, which I will now reveal to the world:

Every time you write a line of code, think about it.

John

On Thu, 11 Oct 2007 18:13:28 -0400, "robb" wrote:


"Rich Grise" wrote in message
news
On Thu, 11 Oct 2007 10:23:27 -0400, robb wrote:
[snip]
even has 80-bit floats!

if you need 80 bit floats ....
I am surprised you guys are not using some functional
programming
language like scheme (lisp/Lambda calc variants) where your
reals
are number abstractions with no language or data type imposed
limit on the number size or precision and of course no
numerical
methods issues/errors from typical float/double data type
limitations
just wondering,

Loosely-typed or untyped data is a mare's nest of bugs just
waiting to happen.

Cheers!
Rich


I should have known better than to bring up software in hardware
group :}

perhaps you meant flexible or dynamic instead of loosely and un
typed ?

it is very strongly typed as there are a few predefined basic
prmitives and "everything" is a defined type even operators and
and functions are all user defined types (or library of other
users types) some say scalable

i am sure they said the same things about basic when it was first
introduced along with a host of other seemingly magical things
that illicited suspicion at first and then gave way to acceptance
and trust.

a scheme implementation is what ? basic with some datastructures
{stacks,lists,etc}

and look at that google reveals 'BIT' a scheme implementation
for microcontroller running in 4k ram and stored in 13k rom among
others

not trying to start any wars i was curious why hrdw engy types
still use basic as opposed to other very useful tools that exist
that have eliminated many problems associated with those old
tools,

I write engineering apps in Basic because it's fast, it's easy, and it
works. And because the resulting programs are easy to read and
understand. "Windows" programming has far too much overhead... I just
want the answers.

Engineers are nearly unique in that we are a class of people who
program, but are not programmers. The closest the rest of the
non-programmer population gets to actual coding is to use
spreadsheets, or maybe design web pages.

Here's a simulation of a CRT vector character generator, for a
heads-up display on a military airplane. It takes into account the
character generator, DAC response, deflectiion amp response, and
phosphor characteristics. All done in a day, in PowerBasic.

John

"John Larkin" wrote
in message ...
On Thu, 11 Oct 2007 13:13:59 -0700, "Joel Koltner"
wrote:

There's a secret, seldom-used, nearly foolproof way to avoid
program
bugs, which I will now reveal to the world:

Every time you write a line of code, think about it.

John

Damn John i wish i had not read that !

all those years of using software debugging tools of various
complexities, tedium, exorbitant costs in money and time and all
i had to do was use the John methodolgy of code construction the
"T.A.I. method"

now if you could just work some traceability in to that method
along with peer review ideas, some QA stability and QC measures
then you could sell it

yes it was a joke (partially)
robb

On Fri, 12 Oct 2007 16:14:32 -0400, "robb" wrote:


"John Larkin" wrote
in message ...
On Thu, 11 Oct 2007 13:13:59 -0700, "Joel Koltner"
wrote:

There's a secret, seldom-used, nearly foolproof way to avoid
program
bugs, which I will now reveal to the world:

Every time you write a line of code, think about it.

John

Damn John i wish i had not read that !

all those years of using software debugging tools of various
complexities, tedium, exorbitant costs in money and time and all
i had to do was use the John methodolgy of code construction the
"T.A.I. method"

now if you could just work some traceability in to that method
along with peer review ideas, some QA stability and QC measures
then you could sell it

yes it was a joke (partially)
robb


If you're spending more than 15% of your effort debugging, you should
adopt the T.I.A. methodology.

John

Fred Abse wrote:
On Wed, 10 Oct 2007 17:00:51 +0000, Joerg wrote:

I think many of the German ones were melamine resin.

My Faber-Castell is wood, reinforced with brass strips.


Wow, must be an older one or some kind of luxury edition. It could be
worth quite some money by now.

--
Regards, Joerg

http://www.analogconsultants.com/

Fred Abse wrote:
On Tue, 09 Oct 2007 16:56:32 -0700, Joerg wrote:

Why were they all made in Germany? Ok, I came from there myself so in my
case that's different. But even in the US 95% plus of the slide rules I
saw were Aristo, Boettcher, Faber-Castell, Nestler but rarely a Pickett
oder Fredericks Post.

Back in those days, most "serious" photographers had German cameras, too.


They still do (Leica).

--
Regards, Joerg

http://www.analogconsultants.com/

Fred Abse wrote:
On Mon, 08 Oct 2007 08:38:14 -0700, John Larkin wrote:

t
1/x
p
1/x
-
1/x


If you want a good laugh, lend a reverse Polish calculator to a bean
counter.


I did that! Not out of malfeasance, I plain forgot that there are people
that use non-RPN calculators. Since nobody else in that meeting had a
calculator he asked me to run the numbers after after having given up.
Then at the end one guy let off a comment along the lines of "Ah, these
MBA types just don't know, lemme show you how it's done." A minute later
he sat there totally red in the face.

--
Regards, Joerg

http://www.analogconsultants.com/

"John Larkin" wrote
in message ...
On Fri, 12 Oct 2007 16:14:32 -0400, "robb"
wrote:


"John Larkin"
wrote
in message ...
On Thu, 11 Oct 2007 13:13:59 -0700, "Joel Koltner"
wrote:

There's a secret, seldom-used, nearly foolproof way to avoid
program
bugs, which I will now reveal to the world:

Every time you write a line of code, think about it.

John

Damn John i wish i had not read that !

all those years of using software debugging tools of various
complexities, tedium, exorbitant costs in money and time and
all
i had to do was use the John methodolgy of code construction
the
"T.A.I. method"

now if you could just work some traceability in to that method
along with peer review ideas, some QA stability and QC
measures
then you could sell it

yes it was a joke (partially)
robb


If you're spending more than 15% of your effort debugging, you
should
adopt the T.I.A. methodology.

oh, i was only involved in churn/burn/test/fix/repeat company
once while still in school, because the owner/prez/sales manager
was a play hard and work twice as hard adrenaline junkie.
(results yesterday)

the debugging occured during the legacy systems era, we mostly
debugged third party legacy software
addins/plugins/libraries/toolboxes/etc the ones we were forced to
use and integrate by the pennymeisters

if we were spending 15% debugging our code there would have
been some F.I.A and A.O.D.
(foot in azz / azz out door) methods practiced

i still like the T.A.I. (think about it) method

robb