6 Jan
2014
6 Jan
'14
1:07 a.m.
On 01/05/2014 10:01 AM, Liam Proven wrote:
So perhaps this explains the fondness and elitism about
HP calculators
and RPN: that they are good for programmers.
Probably 75% of the people I went to business school with (at a University
not know for programming or technology, and there was no programming
requirement) had HP-12C's, and most of the professors would pointedly not
bother to accommodate those that cheaped out and got a non-RPN calc
(usually some ridiculous TI graphing model). My brother went to a
different business school a couple of years later; same thing. Even though
I have stupid amounts of computing available to me, there are all sorts
little financial calculations that are essentially muscle memory on my 12C.
The 2 mortgage closings I've been to in my life featured lawyers double
checking the financial details using HP calculators (one old school HP-37
or -38, one a tricked out HP-41 with the finance ROM). Lawyers, not
programmers.
My last 2 CFO's kept a 12C within arms length, and I've watched both pick
them up and check the details of some deal. Not programmers.
The love of my life in my 20s was a radiation physicist, and she and every
single one of her colleagues carried an HP-41 with the nuclear medicine
pack. Not a programmer amongst them. My current wife is a medical
practice manager. Their lead radiologist has the same setup in his office,
and he uses it to double check what the computers are telling him about
exposures and such. Not something you want to screw up. Also not a
programmer.
I had to have my lot line professionally surveyed recently, and the guy who
did it took all his measurements and then whipped out an HP calculator.
Who knew there's a surveyor pack? He said that the surveying gear has all
the smarts to do the calculations built in, but he liked to double check
because he's found bugs in the gears calculations but never the HP, and he
doesn't like to get screwed by filing incorrect reports. Need I say, not a
programmer?
I could go on. How about an alternative world view: people like RPN and HP
calculators because they're the right tool for the right job.
All anec-data, of course, much as you've offered. But I understand your
personal perspective; when you're half-assing your way through a degree (or
task, or career, whatever) you don't much care about and probably aren't
much good at, you use tools that are the shortest short-cut to getting some
task done, minimal effort, instead of taking time to learn tools that up
your game. The world is full of second (third?) rate talent. But taking
what you're doing seriously and learning the tools needed to exceed isn't
always "elitism", it can just as well be professionalism. Just like it
isn't "elitism" to make an effort to be accurate, factual and impartial in
other professions, instead of sloppy, biased and condescending. YMMV.
KJ
6 Jan
6 Jan
1:35 a.m.
On 6 January 2014 01:07, Ken Seefried <seefriek at gmail.com> wrote:
But I understand your
personal perspective; when you're half-assing your way through a degree (or
task, or career, whatever) you don't much care about and probably aren't
much good at, you use tools that are the shortest short-cut to getting some
task done, minimal effort, instead of taking time to learn tools that up
your game. The world is full of second (third?) rate talent. But taking
what you're doing seriously and learning the tools needed to exceed isn't
always "elitism", it can just as well be professionalism. Just like it
isn't "elitism" to make an effort to be accurate, factual and impartial in
other professions, instead of sloppy, biased and condescending. YMMV.
So, an ad-hominem attack is your idea of a balanced reply, is it?
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
3:18 a.m.
On Jan 5, 2014, at 17:35, Liam Proven <lproven at
gmail.com> wrote:
Even though I agree that RPN calculators are superior to algebraic ones, I think that
Ken's message was thoroughly condescending and rude. RPN vs algebraic is a matter of
opinion, not a law of nature or an indicator of personal worth.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
Web page: http://www.nf6x.net/
On 6 January 2014 01:07, Ken Seefried
<seefriek at gmail.com> wrote:
But I understand your
personal perspective; when you're half-assing your way through a degree (or
task, or career, whatever) you don't much care about and probably aren't
much good at, you use tools that are the shortest short-cut to getting some
task done, minimal effort, instead of taking time to learn tools that up
your game. The world is full of second (third?) rate talent. But taking
what you're doing seriously and learning the tools needed to exceed isn't
always "elitism", it can just as well be professionalism. Just like it
isn't "elitism" to make an effort to be accurate, factual and impartial in
other professions, instead of sloppy, biased and condescending. YMMV.
So, an ad-hominem attack is your idea of a balanced reply, is it? 
9:14 p.m.
Even though I agree that RPN calculators are superior
to algebraic
ones, I think that Ken's message was thoroughly condescending and rude.
RPN vs algebraic is a matter of opinion, not a law of nature or an
indicator of personal worth.
I wonder if part of it is confusing 'Easy to learn' with 'Easy to use'
The two are not the same. And which is more important depends on you, the
problem , and how many tiems you will have to solve it.
I will agree that in a sense RPN calculators are not 'easy to learn'.
You odn't just read a normal maths expression and press the keys in the
same order. You have to think about it.
But they are easy to use in that once you know how to use one, you don't
have to worry about the order that the machine will do operations, you
don't have to worry whether or not you need brackets, and so on. The
machie nalways behaves the same way.
And for me (and jsut about everyone elase I know), that is important. The
time taken to learn RPN is regaiend many times over in not having to think
aobut the machine rahter than the problem.
-tony
8 Jan
8 Jan
9:04 p.m.
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-
bounces at classiccmp.org] On Behalf Of Tony Duell
Sent: Monday, January 06, 2014 1:14 PM
To: cctalk at classiccmp.org
Subject: Re: cctalk Digest, Vol 125, Issue 13
When my daughter started doing serious math in school, I bought her an HP-48SX. I had
already introduced her to RPN on my 48 (which I bought when I did my Masters) and she
'got it' right off. She finds the inside-out way of evaluating algebraically
expressed equations to be intuitive and less error-prone. And she loves to show off her
48 to her teachers, many of whom are too young to have been part of the generation when
RPN and algebraic were truly head-to-head.
By the way, she'll be 14 in ten days. She's a straight-A student in math (and
other subjects). I'll leave the chicken-and-egg argument as an exercise for the
student. -- Ian
Even though I agree that RPN calculators are
superior to algebraic
ones, I think that Ken's message was thoroughly condescending and rude.
RPN vs algebraic is a matter of opinion, not a law of nature or an
indicator of personal worth.
I wonder if part of it is confusing 'Easy to learn' with 'Easy to use'
The two are not the same. And which is more important depends on you, the
problem , and how many tiems you will have to solve it.
I will agree that in a sense RPN calculators are not 'easy to learn'.
You odn't just read a normal maths expression and press the keys in the
same order. You have to think about it.
But they are easy to use in that once you know how to use one, you don't
have to worry about the order that the machine will do operations, you don't
have to worry whether or not you need brackets, and so on. The machie
nalways behaves the same way.
And for me (and jsut about everyone elase I know), that is important. The
time taken to learn RPN is regaiend many times over in not having to think
aobut the machine rahter than the problem.
7 Jan
7 Jan
2:32 a.m.
Even though I agree that RPN calculators are superior
to algebraic ones, I think that Ken's message was thoroughly condescending and rude.
RPN vs algebraic is a matter of opinion, not a law of nature or an indicator of personal
worth.
I will not comment about the niceness of Ken's reply, but I will
comment that some of it is just hopelessly outdated. People in
business - lawyers, bankers, real estate agents, etc. - do not really
use calculators anymore, RPN or otherwise. They have not for the past
10 or 15 years. The laptop computer has well matured, and these days
the first thing they do when meeting is take out their machines and
get all appropriate documents, spreadsheets, and applications
launched. The calculators stay behind, buried in the bottom drawer of
their desks.
Personally, if I need to calculate something, I will just use what is
handy. Some of use can actually use RPN or algebraic, and not feel the
need to preach.
--
Will
6:25 a.m.
I hope that I have not made anyone feel
that I was making any less of them.
If one were to right down the steps they
used to solve a problem with several levels
of parenthesis, I think they'd find that it
was the same as RPN.
I learned to solve problems with pencil and paper.
The calculator I used was a slide ruler.
When exposed to RPN, I found it quite clear
that it was just what I was already doing.
I would suppose that anyone used to a calculator
the had built in brackets and the person hadn't
spent a large amount of time unraveling the
correct order to solve equations would not find
RPN that friendly.
For me it seemed more natural because it was
just what I'd been taught to do to solve problems.
Later when I learned BASIC, I wondered why,
when telling a computer what to do, I was back
to spelling out equations that I only found
the format useful for reduction of problems,
not for solving problems.
When first exposed to Forth I could understand
why people wanted to mix there programming
languages up with mixed things. Part of the
was to tell the computer what and when to do
something while part was to scramble it up and
ask the computer to waste its time figuring it
out.
I knew then that to tell a computer what and when
to do something without making mistakes took
doing just that. No mixing.
Forth may not be the perfect language but it is
closer than any other that I've seen.
Dwight
7:45 p.m.
I will not comment about the niceness of Ken's
reply, but I will
comment that some of it is just hopelessly outdated. People in
business - lawyers, bankers, real estate agents, etc. - do not really
use calculators anymore, RPN or otherwise. They have not for the past
10 or 15 years. The laptop computer has well matured, and these days
How can you possibly make such a sweeping statement unless you know every
businessman in the world, or at least know the practices in every major
country of the world?
_Every_ professional scientist ant enginener that I know uses a
calcualtor. Of course they use PCs too -- they use the msot approprate
tool for the job.
I am told by a friend who works for a banking company in the City of
London (anthough not as a banker) that HP12Cs are very common in such
estalishments. Actually,m it's not uncommon to see a cashier or manager
in a high street bank with an HP12C on their desk.
-tony
8 Jan
8 Jan
12:24 a.m.
On Tue, Jan 07, 2014 at 07:45:30PM +0000, Tony Duell wrote:
[...]
I am told by a friend who works for a banking company
in the City of London
(anthough not as a banker) that HP12Cs are very common in such estalishments.
Actually,m it's not uncommon to see a cashier or manager in a high street
bank with an HP12C on their desk.
And yet I wrote some share trading software for a City firm and didn't see one
anywhere in the office. Microsoft Excel was pretty popular, though. Aren't
anecdotes wonderful?
I suspect the difference is that your friend's office is mainly full of baby
boomers who figured out how to use a HP-12C to do their job back in the day,
and now don't want to learn how to do it on one of those newfangled computer
thingies.
12:50 a.m.
On Jan 7, 2014, at 7:24 PM, Peter Corlett <abuse at cabal.org.uk> wrote:
On Tue, Jan 07, 2014 at 07:45:30PM +0000, Tony Duell
wrote:
[...]
Or they might be people like me ? I?ve been a professional programmer for 40 years and I?m
most likely to type ?python? when I need to do some quick calculation. But I also keep a
slide rule in my office. (Also an abacus ? but unfortunately that one is still only
decoration, I have not yet learned how to use it.)
paul
I am told by a friend who works for a banking
company in the City of London
(anthough not as a banker) that HP12Cs are very common in such estalishments.
Actually,m it's not uncommon to see a cashier or manager in a high street
bank with an HP12C on their desk.
And yet I wrote some share trading software for a City firm and didn't see one
anywhere in the office. Microsoft Excel was pretty popular, though. Aren't
anecdotes wonderful?
I suspect the difference is that your friend's office is mainly full of baby
boomers who figured out how to use a HP-12C to do their job back in the day,
and now don't want to learn how to do it on one of those newfangled computer
thingies.
1:27 a.m.
On Tue, Jan 07, 2014 at 07:50:10PM -0500, Paul Koning wrote:
[...]
Or they might be people like me ? I?ve been a
professional programmer for 40
years and I?m most likely to type ?python? when I need to do some quick
calculation. But I also keep a slide rule in my office. (Also an abacus ? but
unfortunately that one is still only decoration, I have not yet learned how
to use it.)
The Python REPL is also my usual calculator.
I used to own a slide rule as a gag item to keep on my desk, but it went
walkies in the late 1990s. It, like a RPN calculator, is a tool of a bygone era
in that while I have no difficulty understanding the theory of operation and
could figure out how to use them from first principles, I'd rather not actually
do so when there are easier and more accurate tools around.
1:47 a.m.
New subject: RPN calculators; was: cctalk Digest, Vol 125, Issue 13
I think it's remarkable that HP is still making the things, albeit with
a modern processor (ARM, I believe). Based on the reviews from Amazon,
there are issues with power consumption and sticky keys.
But there *must* still be a market for the things.
--Chuck
2:36 a.m.
New subject: RPN calculators; was: cctalk Digest, Vol 125, Issue 13
I think it's remarkable that HP is still making
the things, albeit with a
modern processor (ARM, I believe). Based on the reviews from Amazon, there
are issues with power consumption and sticky keys.
But there *must* still be a market for the things.
They make great gifts.
--
Will
4:29 a.m.
On Jan 7, 2014, at 5:27 PM, Peter Corlett <abuse at cabal.org.uk> wrote:
On Tue, Jan 07, 2014 at 07:50:10PM -0500, Paul Koning
wrote:
[...]
Since we're on this topic...my usual calculator is Macsyma (yes, the giant
LISP
program that can do symbolic computation). I mainly use it for a few reasons:
1) it does "infinite" precision math, so I don't *ever* worry about long
computations
having errors due to intermediate rounding
2) I can get it to present results as the ratio of two integers (sometimes very big
integers) which is useful for a lot of what I do at times
3) it does symbolic computation
4) it's written in LISP (sorry had to add that one). ;-)
TTFN - Guy
Or they might be people like me ? I?ve been a
professional programmer for 40
years and I?m most likely to type ?python? when I need to do some quick
calculation. But I also keep a slide rule in my office. (Also an abacus ? but
unfortunately that one is still only decoration, I have not yet learned how
to use it.)
The Python REPL is also my usual calculator.
I used to own a slide rule as a gag item to keep on my desk, but it went
walkies in the late 1990s. It, like a RPN calculator, is a tool of a bygone era
in that while I have no difficulty understanding the theory of operation and
could figure out how to use them from first principles, I'd rather not actually
do so when there are easier and more accurate tools around.
3:41 a.m.
Or they might be people like me ? I?ve been a
professional programmer for 40 years and I?m most likely to type ?python? when I need to
do some quick calculation. But I also keep a slide rule in my office. (Also an abacus ?
but unfortunately that one is still only decoration, I have not yet learned how to use
it.)
There are RPN options at the console, too. My first
instinct is to type "dc" (or "gforth" if I need to do some
hex conversion). There are some things for which I'll
use Python's interactive mode, but I find them few and
far between. I think, for the most part, I find leaving
things on the stack a lot easier than explicitly saving to
variables. I don't pretend that anyone who prefers algebraic
representations is in some way deficient, though; it's just
how my mind works.
- Dave
5:07 a.m.
I suspect the difference is that your friend's
office is mainly full
of baby boomers who figured out how to use a HP-12C to do their job
back in the day, and now don't want to learn how to do it on one of
those newfangled computer thingies.
(Wow, speaking of feelings of smug superiority!)
I'm sure that happens. But there are plenty of other reasons. For
example, I'm a computer geek by both vocation and avocation, yet I use
calculators. For some tasks I even prefer them.
I have no particular bias any direction on the question of prefix
operators ("Polish") versus postfix operators ("reverse Polish")
versus
infix operators ("algebraeic"). (None that's independent of the task,
that is; there certainly are tasks for which I prefer one or another,
though among calculator tasks they're relatively rare.)
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
7 Jan
7 Jan
10:50 p.m.
On 6 January 2014 03:18, Mark J. Blair <nf6x at nf6x.net> wrote:
I would have to agree.
But then again, it does fit perfectly with the smug, condescending,
patronizing tone of many other RPN users I've met online. They are not
pleasant people, in my general experience. Of course, there are
exceptions, but Ken here is not one.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
So, an
ad-hominem attack is your idea of a balanced reply, is it?
Even though I agree that RPN calculators are superior to algebraic ones, I think that
Ken's message was thoroughly condescending and rude. RPN vs algebraic is a matter of
opinion, not a law of nature or an indicator of personal worth.
8 Jan
8 Jan
1:13 a.m.
On Tue, Jan 07, 2014 at 10:50:59PM +0000, Liam Proven wrote:
[...]
But then again, it does fit perfectly with the smug,
condescending,
patronizing tone of many other RPN users I've met online. They are not
pleasant people, in my general experience. Of course, there are exceptions,
but Ken here is not one.
It's the same kind of superior smugness you get from Lisp zealots, who also use
ten-dollar words like "homoiconcity" and claim that it's the only tool that
can
bring us artificial intelligence, despite the fact that it has yet to actually
produce any AI of note despite having been around for 56 years. It does however
make sense once you realise that Lisp source code is also Polish Notation.
Incidentally, "homoiconicity" just means that they wrote the language back-end
and then knocked off without bothering with the front-end, so you have to enter
your program as a raw AST instead. The apologists claim that the lack of a
front-end is a feature because it makes macros easier to write. That real
programs use macros rarely or not at all, but do contain a lot of arithmetic
that would benefit from being presented in a more conventional form, has
apparently passed them by.
5:31 a.m.
New subject: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
It's the same kind of superior smugness you get
from Lisp zealots,
Zealots will be zealots.
[...], despite the fact that [Lisp] has yet to
actually produce any
AI of note despite having been around for 56 years.
I disagree, except for definitions of "AI" for which it's a tautology.
(There's a tendency I've noticed in myself to act as though "real" AI
is always the stuff that hasn't been done; once it's done it's just
algorithms, not AI.)
As a simple example, back in the '80s someone I knew wrote a Rubik's
Cube solver that developed its own macros. I'd call that AI. He wrote
in Lisp, and IMO with good reason.
Incidentally, "homoiconicity" just means
that they wrote the language
back-end and then knocked off without bothering with the front-end,
so you have to enter your program as a raw AST instead.
Careful. Not only is this statement wrong, it comes awfully close to
exhibiting zealotry of your own. (If you don't understand why or how
it's wrong, I refer you to the Wikipedia page "Homoiconicity".)
The apologists claim that the lack of a front-end is a
feature
because it makes macros easier to write.
That _is_ one of the respects in which it is a feature, but it's hardly
the only one.
That real programs use macros rarely or not at all,
but do contain a
lot of arithmetic that would benefit from being presented in a more
conventional form, has apparently passed them by.
Oddly enough, you are not alone in noticing this weakness of Lisp. The
most "serious" (FWVO...) Lisp I used had a syntax for entering
arithmetic in traditional infix notation. Using @ as the marker
character (it was actually a diamond-shaped thingy), you could write,
for example, @(len*3)+slop@ instead of (+ (* len 3) slop).
As for this passing people by, well, zealots will be zealots; you can
tell a zealot, but you can't tell one much.
And as for the statement itself? It exhibits a remarkable blind spot,
in that it dismisses as non-"real" programs that do symbolic, rather
than numerical, computation - such as the Macsyma someone mentioned.
Such software can hardly be written without building up a
domain-specific language for the task, and, in Lisp, a great deal of
that will generally be done with macros. (There are also macros that
casual users may not realize are macros, such as setf.)
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
2:41 p.m.
New subject: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
Date: Wed, 8 Jan 2014 00:31:14 -0500
From: mouse at rodents-montreal.org
To: cctalk at classiccmp.org
Subject: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
---snip---
It's the same kind of superior smugness you
get from Lisp zealots,
Zealots will be zealots.
[...], despite the fact that [Lisp] has yet to
actually produce any
AI of note despite having been around for 56 years.
I disagree, except for definitions of "AI" for which it's a tautology.
(There's a tendency I've noticed in myself to act as though "real" AI
is always the stuff that hasn't been done; once it's done it's just
algorithms, not AI.)
As a simple example, back in the '80s someone I knew wrote a Rubik's
Cube solver that developed its own macros. I'd call that AI. He wrote
in Lisp, and IMO with good reason.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
An interesting observation. I recall a project to make a maintenance
program to assist in fixing GE diesel engines. It seems they were
all made a little different.
They'd written an AI program in LISP to help with the querying
and matching engines with designs.
After it was up and running, it was translated into Forth in
order to fit into a computer that could be rolled around on a cart.
Of course now days it would just need a cloud.
Dwight
6:21 p.m.
New subject: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-
bounces at classiccmp.org] On Behalf Of dwight
Sent: Wednesday, January 08, 2014 6:41 AM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: RE: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
An interesting observation. I recall a project to make a maintenance program
to assist in fixing GE diesel engines. It seems they were all made a little
different.
They'd written an AI program in LISP to help with the querying and matching
engines with designs.
After it was up and running, it was translated into Forth in order to fit into a
computer that could be rolled around on a cart.
Of course now days it would just need a cloud.
Dwight
"Translated into Forth"? Now that's interesting! -- Ian
2:50 p.m.
New subject: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
On 8 January 2014 05:31, Mouse <mouse at rodents-montreal.org> wrote:
I think you misunderstand what Peter was alluding to.
Remember that S-expressions were only meant to be an internal
representation and the original plan was that programmers would
actually use M-expressions with a more traditional notation.
This never happened because some programmers found that it was
perfectly possible to work in S-expressions, and proficiency in doing
so has become something of a shibboleth among Lispers: "real men work
in S-expressions".
This has led to the neglect of Lisp-like languages or extensions with
higher-level syntax, such as Dylan, CGOL or Sweet Expressions, and
erects a huge and forbidding barrier to the adoption of Lisp. The
problem being that Lispers are mostly fine with this: they learned, so
why shouldn't everyone else? Which means that it becomes a vicious
circle. Lisp has an offputting and peculiar syntax, but Lispers have
climbed this wall and learned it; and that syntax does have
advantages, so Lispers disdain any attempt to fix it; meaning that
Lisp remains a niche language and always will.
Yes, prefix notation makes macros easier, because Lisp handles lists
and Lisp code is made of lists. Thus, prefix notation and macros
together form a unique strength of Lisp. However, prefix notation
makes /everything else/ harder and stops non-Lispers considering Lisp.
Most Lispers seem to be fine with this. Result: marginalisation.
Languages which replace prefix notation and S-expressions and use
algebraic infix notation, such as Dylan and David Moon's PLOT, still
support syntactic macros, but are also readable by non-specialists...
but they are shunned by Lispers, and thus they too remain obscure and
marginal.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
Incidentally,
"homoiconicity" just means that they wrote the language
back-end and then knocked off without bothering with the front-end,
so you have to enter your program as a raw AST instead.
Careful. Not only is this statement wrong, it comes awfully close to
exhibiting zealotry of your own. (If you don't understand why or how
it's wrong, I refer you to the Wikipedia page "Homoiconicity".)
3:56 p.m.
New subject: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
On Jan 8, 2014, at 9:50 AM, Liam Proven <lproven at gmail.com> wrote:
On 8 January 2014 05:31, Mouse <mouse at
rodents-montreal.org> wrote:
I think you misunderstand what Peter was alluding to.
Remember that S-expressions were only meant to be an internal
representation and the original plan was that programmers would
actually use M-expressions with a more traditional notation.
This never happened because some programmers found that it was
perfectly possible to work in S-expressions, and proficiency in doing
so has become something of a shibboleth among Lispers: "real men work
in S-expressions?.
I thought it was ?real programmers work in machine code? (see ?The story of Mel, a real
programmer?). :-)
Incidentally, "homoiconicity" just means that they wrote the language
back-end and then knocked off without bothering with the front-end,
so you have to enter your program as a raw AST instead.
Careful. Not only is this statement wrong, it comes awfully close to
exhibiting zealotry of your own. (If you don't understand why or how
it's wrong, I refer you to the Wikipedia page "Homoiconicity".)
This has led to the neglect of Lisp-like languages or extensions with
higher-level syntax, such as Dylan, CGOL or Sweet Expressions, and
erects a huge and forbidding barrier to the adoption of Lisp.
Then again, the only AI language I?ve run into that isn?t Lisp is just as odd
syntactically in a very different way ? POP-2, brought to you by the AI folks at U of
Edinborough (UK). Just like APL, it has the rule that all operators associate in the same
direction. Unlike APL, it solved the conflict between assignment and the rest by making
assignment associate left to right, so you end up with statements that look like this:
a+1 -> a
I haven?t seen it since college; I wonder if there are any implementations or users still
left.
paul
4:38 p.m.
New subject: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
>>> Incidentally, "homoiconicity"
just means that they wrote the
>>> language back-end and then knocked off without bothering with the
>>> front-end, so you have to enter your program as a raw AST instead.
>> Careful. Not only is this statement wrong, [...]
> I think you misunderstand what Peter was alluding to.
Oh, this is interesting. The Wikipedia page actually contradicts
itself. At one point, it says that "If a language has homoiconicity,
it means that the language text has the same structure as its abstract
syntax tree (i.e. the AST and the syntax are isomorphic)", which _is_
what the quote I was replying to claims, but is not a particularly
interesting property. But the very next sentence says that "In a
homoiconic language the primary representation of programs is also a
data structure in a primitive type of the language itself", which is
not really related to the syntax programmers see and is the aspect
that's relevant to the conversation (or at least to what I understand
the conversation to be about).
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
4:46 p.m.
New subject: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
It was thus said that the Great Mouse once stated:
But the very next sentence says that "In a
homoiconic language the primary representation of programs is also a
data structure in a primitive type of the language itself", which is
not really related to the syntax programmers see and is the aspect
that's relevant to the conversation (or at least to what I understand
the conversation to be about).
By that definition, Forth is homoiconic then.
-spc (Or nearly so ... )
6:35 p.m.
New subject: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
On Jan 8, 2014, at 11:46 AM, Sean Conner <spc at conman.org> wrote:
It was thus said that the Great Mouse once stated:
Yes, and similarly TECO. And in some sense Python (because Python byte code is an object
that has a number of semi-interesting methods).
paul
But the very next sentence says that "In a
homoiconic language the primary representation of programs is also a
data structure in a primitive type of the language itself", which is
not really related to the syntax programmers see and is the aspect
that's relevant to the conversation (or at least to what I understand
the conversation to be about).
By that definition, Forth is homoiconic then. 
6:36 p.m.
New subject: In defense of Lisp [was Re: cctalk Digest, Vol 125, Issue 13]
>>>
Then again, the only AI language
I've run into that isn't Lisp is just as
odd syntactically in a very different way - POP-2, brought to you by the AI
folks at U of Edinborough (UK). Just like APL, it has the rule that all
operators associate in the same direction. Unlike APL, it solved the
conflict between assignment and the rest by making assignment associate left
to right, so you end up with statements that look like this:
a+1 -> a
I haven't seen it since college; I wonder if there are any implementations
or users still left.
<<<<
Look up "Poplog" on Google.
It has extended and is available for Linux and Windows - looks like the OSX
version is unlikely to run on current versions (as it is compiled for PPC)
I wonder if anyone has a copy of the old ICL1900 version of POP2 anywhere?
(That was the version I used more than 40 years ago! I assume it would run
on Dave Holdsworth's George 3 emulator and possibly on the other 1900 series
emulator which apparently exists but seems impossible to obtain)
Andy
9 Jan
9 Jan
3:22 a.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
On 07/01/14 8:13 PM, Peter Corlett wrote:
On Tue, Jan 07, 2014 at 10:50:59PM +0000, Liam Proven
wrote:
[...]
It's not. Lisp does NOT use RPN.
----------------------
Lisp uses prefix ordering: the SAME conventional, familiar form for
function applications that Fortran and Fortran derivatives, including C
and every C derivative, including EVERY mainstream language in current
use, also use:
Head followed by parameters. f(a, b, c) or in Lisp (f a b c)
Some languages (notably Haskell) also allow infix placement of any
function or operator, according to what the author deems most readable.
(This flexibility isn't available in most popular languages, unfortunately.)
There are languages that use RPN (I won't bother listing them) but Lisp
is not among them.
--Toby
But then again, it does fit perfectly with the
smug, condescending,
patronizing tone of many other RPN users I've met online. They are not
pleasant people, in my general experience. Of course, there are exceptions,
but Ken here is not one.
It's the same kind of superior smugness you get from Lisp zealots, who also use
ten-dollar words like "homoiconcity" and claim that it's the only tool that
can
bring us artificial intelligence, despite the fact that it has yet to actually
produce any AI of note despite having been around for 56 years. It does however
make sense once you realise that Lisp source code is also Polish Notation.
Incidentally, "homoiconicity" just means ...
5:55 a.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
once you
realise that Lisp source code is also Polish Notation.
It's not. Lisp does NOT
use RPN.
3:56 p.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
On Jan 9, 2014, at 12:55 AM, Mouse <mouse at rodents-montreal.org> wrote:
RPN stands for Reverse Polish Notation.
You may note the claim does not include the word "reverse?.
Indeed. (+ 1 2) certainly is classic ?ukasiewicz notation, prefix rather than postfix.
paul
once you realise that Lisp source code is also Polish
Notation.
It's not. Lisp does NOT use RPN.
4:59 p.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-
bounces at classiccmp.org] On Behalf Of Paul Koning
Sent: Thursday, January 09, 2014 7:56 AM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
On Jan 9, 2014, at 12:55 AM, Mouse <mouse at rodents-montreal.org> wrote:
RPN stands for Reverse Polish Notation.
You may note the claim does not include the word "reverse?.
Indeed. (+ 1 2) certainly is classic ?ukasiewicz notation, prefix rather than
postfix.
OK, let's sort this out. Lisp uses prefix notation. What is often called
'algebraic' entry is formally infix notation. And RPN employs postfix notation,
as does Forth. That's one reason I think it must have been interesting to translate a
Lisp program into Forth. :-) -- Ian
once you realise that Lisp source code is also Polish
Notation.
It's not. Lisp does NOT use RPN. 
5:59 p.m.
New subject: Lisp - Re: cctalk Digest, Vol 125, Issue 13
If anyone wants to play with a modern lisp, with advanced math and network
support
look at NEWLISP.ORG it is free and reasonably documented.
The other Bob
6:35 p.m.
New subject: Lisp - Re: cctalk Digest, Vol 125, Issue 13
If anyone wants to play with a modern lisp, with
advanced math and
network support look at NEWLISP.ORG it is free and reasonably
documented.
But note, it does a number of things which, coming form a traditional
Lisp perspective, I find utterly bizarre, bordering onj the point where
I'd call them broken. For example, nil and () are not the same thing,
and lambda lists are a magic kind of list, not just lists whose first
element is the symbol lambda. Consing two atoms together returns a
list, not a dotted pair. Obviously they've managed to make it all
hang together, but it is a substantial deviation from tradition and,
while I'd have to do more reading and thinking to offer more than
speculation, might break the clean design that tradtional Lisp is
notable for.
Also, anyone considering using it should read the license; it's one of
the most confusing licenses I've ever seen and I'm not sure I'm even
allowed to run the code. (I feel sure it's supposed to allow that; I'm
just not sure it actually does.)
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
6:07 p.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
On Jan 9, 2014, at 11:59 AM, Ian King <IanK at LivingComputerMuseum.org> wrote:
Jan ?ukasiewicz originally came up with prefix notation. Since he was a Polish
mathematician and non-Poles are likely to have trouble with ??ukasiewicz? it became known
as ?Polish notation?.
At a later time, a variant was invented where the operator is at the end, which works
great with stack machines and as a way to do infix expression parsing. That?s the reverse
of conventional ?ukasiewicz notation so it became ?Reverse Polish Notation? (or sometimes
?Polish Postfix Notation?).
Both have the property that you don?t need parentheses; execution order is evident by the
placement of the operators. Mapping between classic and reverse forms is trivial, simply
move the operator from one end to the other.
Forth is postfix notation, because the Forth execution model is a stack machine with the
Forth ?words? corresponding directly to the execution of primitive operations that act on
the stack. Forth is really an assembly language rather than a high level language; those
familiar with Burroughs mainframe machines will see the analogy with the machine language
of those systems.
paul
-----Original
Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-
bounces at classiccmp.org] On Behalf Of Paul Koning
Sent: Thursday, January 09, 2014 7:56 AM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
On Jan 9, 2014, at 12:55 AM, Mouse <mouse at rodents-montreal.org> wrote:
Indeed. (+ 1 2) certainly is classic ?ukasiewicz notation, prefix rather than
postfix.
OK, let's sort this out. Lisp uses prefix notation. What is often called
'algebraic' entry is formally infix notation. And RPN employs postfix notation,
as does Forth. That's one reason I think it must have been interesting to translate a
Lisp program into Forth. :-) -- Ian > once
you realise that Lisp source code is also Polish Notation.
It's not. Lisp does NOT use RPN.
RPN stands for Reverse Polish Notation.
You may note the claim does not include the word "reverse?.
11:21 p.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-
bounces at classiccmp.org] On Behalf Of Paul Koning
Sent: Thursday, January 09, 2014 10:07 AM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
Forth is postfix notation, because the Forth execution model is a stack
machine with the Forth ?words? corresponding directly to the execution of
primitive operations that act on the stack. Forth is really an assembly
language rather than a high level language; those familiar with Burroughs
mainframe machines will see the analogy with the machine language of
those systems.
I respectfully disagree that it is appropriate to refer to Forth as 'an assembly
language'. Forth is a threaded interpretive language, written in any of several
assembly languages. Burroughs was interesting in that it performed a great deal of
language processing in hardware - we have a Unisys V380 (a rebadged/remicrocoded B4955)
whose machine instructions map almost one for one onto COBOL verbs. I'm aware of
processors that were likewise designed to do much of the lifting of parsing Forth in
hardware, but Forth itself is at a higher level (although not much higher!). -- Ian
6:09 p.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
From: IanK at LivingComputerMuseum.org
---snip---
OK, let's sort this out. Lisp uses prefix notation. What is often called
'algebraic' entry is formally infix notation. And RPN employs postfix notation,
as does Forth. That's one reason I think it must have been interesting to translate a
Lisp program into Forth. :-) -- Ian
The other part is that in LISP, the operation is over a list.
(+ 4 5 6 ) in LISP is not same as 4 5 6 + in Forth.
Still, Forth can easily be expanded to handle somethings
like list processing. It can even make an efficient (+
and do prefixed. I've often created a '[ x x x ],' like operation
when I've needed to compile a long string of data,
like from a PROM dump. Like real LISP, being able to control
all aspects of the interpretation and compilation make for
unlimited ways of dealing with each specific problem.
I'm not sure what level the translation was done at but
I was told it was done, mostly, automatically by a Forth to
LISP predecessor written in Forth.
Dwight
7:13 p.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
The other part is that in LISP, the operation is over
a list.
(+ 4 5 6 ) in LISP is not same as 4 5 6 + in Forth.
Yes. Lisp does not quite use (non-reverse) Polish notation; it
surrounds calls with parens so as to support variable-arity functions.
(For example, most Lisps support not only (+ 1 2) but (+ 1 2 3) and
(+ 4 5 6 7) and the like, whereas FORTH does the analog with 1 2 + and
1 2 3 + + and 4 5 6 7 + + +). If all Lisp operations had fixed arity,
the parens could in principle be omitted in code, though that would
have two problems: (1) they are useful enough to humans that it would
arguably be a good idea to provide them anyway for error checking and
(2) it would break homoiconicity to some extent - even if the +
function always takes exactly two args, the list (+ 1 2 3 4 5) still
makes sense as a list, just one that errors if evaluated.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
7:33 p.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
From: mouse at rodents-montreal.org
My understanding is the LIST recursively interprets the string.
It parses out the first part and the rest of the string. If the rest
is not atomic for add, it nest and repeats until it gets to two atomic
parts that it can apply + to. As it unwinds, it does the add of two
at a time.
Dwight
The other part is that in LISP, the operation is
over a list.
(+ 4 5 6 ) in LISP is not same as 4 5 6 + in Forth.
Yes. Lisp does not quite use (non-reverse) Polish notation; it
surrounds calls with parens so as to support variable-arity functions.
(For example, most Lisps support not only (+ 1 2) but (+ 1 2 3) and
(+ 4 5 6 7) and the like, whereas FORTH does the analog with 1 2 + and
1 2 3 + + and 4 5 6 7 + + +). If all Lisp operations had fixed arity,
the parens could in principle be omitted in code, though that would
have two problems: (1) they are useful enough to humans that it would
arguably be a good idea to provide them anyway for error checking and
(2) it would break homoiconicity to some extent - even if the +
function always takes exactly two args, the list (+ 1 2 3 4 5) still
makes sense as a list, just one that errors if evaluated.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
10 Jan
10 Jan
12:23 a.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
[...] even if
the + function always takes exactly two args, the list
(+ 1 2 3 4 5) still makes sense as a list, just one that errors if
evaluated.
My understanding is the LIST recursively interprets the string. It parses out the first part and the rest of the
string. If the rest
is not atomic for add, it nest and repeats until it gets to two
atomic parts that it can apply + to. As it unwinds, it does the add
of two at a time.
This is implementation-dependent. There's no reason it couldn't be
done as you outline. I'm not familiar with any that work that way, but
that means little, as the only one I really know that sort of detail
about is my own.
In my Lisp engine, evaluating (+ 1 2 3 4) first looks at the function
cell for the + symbol. Assuming it's the default, then, based on the
type of the object found there (a builtin function object of type
BIK_LAMBDA), the engine evaluates the arguments individually, then
calls the implementation code corresponding to that builtin function
object; that code loops through the evaluated arguments, keeping a
running total which is returned when it's done. There's no recursion
such as you describe involved.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
9 Jan
9 Jan
7:55 p.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
make sense
once you realise that Lisp source code is also Polish Notation.
It's not. Lisp does NOT use RPN.
----------------------
Lisp uses prefix ordering: the SAME conventional, familiar form for
function applications that Fortran and Fortran derivatives, including C
and every C derivative, including EVERY mainstream language in current
use, also use:
Head followed by parameters. f(a, b, c) or in Lisp (f a b c)
YEs, but from what I rememeber there are no infix operators in LISP. So
even things like addition are writen (+ 2 5) rather than 2 + 5
-tony
10 Jan
10 Jan
4:01 a.m.
New subject: Lisp does NOT use RPN - Re: cctalk Digest, Vol 125, Issue 13
On Thu, Jan 9, 2014 at 2:55 PM, Tony Duell <ard at p850ug1.demon.co.uk> wrote:
Lisp doesn't really have "operators" in the sense of Algol-ish languages;
the
closest analogue are the special operators which are primitive forms. But
whereas '+' signifying addition is an operator in many languages, in Lisp
it's
(often) just a function. E.g., (apply #'+ (list 1 2 3)) yields 6 in a
Lisp-2.
But depending on the dialect, there are a few things that apply in an
infix-ish
way. Consider the '.' token in an improper list; one can sometimes think
of this as an infix stand-in for a CONS call. In Common Lisp:
CL-USER> '(a . (b . (c . nil)))
(A B C)
CL-USER> (cons 'a (cons 'b (cons 'c nil)))
(A B C)
CL-USER>
Or in Scheme:
guile> '(1 . (2 . (3 . ())))
(1 2 3)
guile>
In Common Lisp, the LOOP macro defines a language with lots of infix
operators:
CL-USER> (loop for (k . v) in '((1 . a) (2 . b) (3 . c))
do (format t "~A maps to ~A~%" k v))
1 maps to A
2 maps to B
3 maps to C
NIL
CL-USER>
Etc.
Further, if you *really* wanted to in Common Lisp, you could change
the read tables and redefine the syntax to essentially anything you
wanted, including an infix language. :-)
So while Lisp is *mostly* written using prefix notation, there are escape
hatches.
- Dan C.
Head followed by parameters. f(a, b, c) or in
Lisp (f a b c)
YEs, but from what I rememeber there are no infix operators in LISP. So
even things like addition are writen (+ 2 5) rather than 2 + 5 
27 Jan
27 Jan
7:59 p.m.
On Wed, Jan 08, 2014 at 01:13:21AM +0000, Peter Corlett wrote:
On Tue, Jan 07, 2014 at 10:50:59PM +0000, Liam Proven
wrote: [...]
Well, it doesn't help that the definition of "AI" is about as solid and
nailed down as quicksilver and changing over time. The complex expert
systems realized in Lisp would at one point in time have counted as
being pretty much AI, but then we understood better how to build those
systems and the target moved on.
But then again, it does fit perfectly with the
smug, condescending,
patronizing tone of many other RPN users I've met online. They are
not pleasant people, in my general experience. Of course, there are
exceptions, but Ken here is not one.
It's the same kind of superior smugness you get from Lisp zealots, who
also use ten-dollar words like "homoiconcity" and claim that it's the
only tool that can bring us artificial intelligence, despite the fact
that it has yet to actually produce any AI of note despite having been
around for 56 years. It does however make sense once you realise that Lisp
source code is
also Polish Notation.
What does _that_ have to do with it?
Incidentally, "homoiconicity" just means
that they wrote the language
back-end and then knocked off without bothering with the front-end, so
you have to enter your program as a raw AST instead. The apologists
claim that the lack of a front-end is a feature because it makes
macros easier to write.
It does. Comparing C macros with Lisp macros is like sticks and stones
versus a machine shop ;-)
Adding that flexibily to a C style syntax would be an interesting
challenge.
That real programs use macros rarely or not
at all, but do contain a lot of arithmetic that would benefit from
being presented in a more conventional form, has apparently passed
them by.
You seem to be out of touch with reality when you state that "real
programs use macros rarely or not at all". Would QPX count as a "real"
program for you? It does handle non-trivial real world problems: it is
the core behind the flight search engine that (among IIRC others) powers
orbitz.com and sees quite a bit of traffic and is written in Lisp (with
some non-Lisp glue/helper code) . Oddly enough, from what I read about
it on the the net, it appears to be about half macros ...
And even in my own, far less complex and powerful Lisp programs, I find
myself using defmacro quite few times - because it is useful and makes
things easier, for instance by reducing boilerplate code quite bit.
Kind regards,
Alex.
--
"Opportunity is missed by most people because it is dressed in overalls and
looks like work." -- Thomas A. Edison
8 Jan
8 Jan
2:42 a.m.
So, an ad-hominem attack is your idea of a balanced
reply, is it?
Even though I agree that RPN calculators are superior to algebraic
ones,
3767
days inactive
3788
days old
44 comments
20 participants
participants (20)
-
abuse＠cabal.org.uk -
als＠thangorodrim.de -
andyh＠andyh-rayleigh.freeserve.co.uk -
ard＠p850ug1.demon.co.uk -
caveguy＠sbcglobal.net -
cclist＠sydex.com -
cisin＠xenosoft.com -
crossd＠gmail.com -
dkelvey＠hotmail.com -
fraveydank＠gmail.com -
ggs＠shiresoft.com -
IanK＠LivingComputerMuseum.org -
lproven＠gmail.com -
mouse＠Rodents-Montreal.ORG -
nf6x＠nf6x.net -
paulkoning＠comcast.net -
seefriek＠gmail.com -
spc＠conman.org -
toby＠telegraphics.com.au -
wdonzelli＠gmail.com