On Tue, 30 Aug 2016 05:09:39 -0700 (MST)
L Boyd lboyd@frontiernet.net wrote:

They are constants but they are scoped. No different to C

const x = 1

static void foo(void)
{
const x = 4;

blah;

}

Or perhaps think of them as functions or macros, they are really just an
expansion.

Alan

Alan Cox wrote

Larry

View this message in context: http://forum.openscad.org/What-to-call-variables-tp18260p18262.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

Why NOT make them behave like variables in other languages and allow a = a

• 1 ?

On Wed, Aug 31, 2016 at 12:09 AM, L Boyd lboyd@frontiernet.net wrote:

"Constants within a scope"?

That would describe them.
On Aug 30, 2016 8:10 AM, "L Boyd" lboyd@frontiernet.net wrote:

On 08-30-2016 2:18 PM, Frank van der Hulst wrote:

I second that.

         - Michele

Because there is no assignment in OpenSCAD. When you write:

in OpenSCAD, this is a definition, like in math, not an assignment.
Assignments are the cause of many misleading code and many languages
restrict their effects. Functional programming languages have no variables
neither assignments. We don't need them.

--
View this message in context: http://forum.openscad.org/What-to-call-variables-tp18260p18266.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

On Tue, Aug 30, 2016 at 12:06 PM, Ronaldo rcmpersiano@gmail.com wrote:

excuse me? There are assignments in Functional programming languages.
And yes, we really could use them. Functional programming languages in
simple terms make a
difference between where the variable (in memory) is, and the value that
points
to it.

So, why have true variables only in certain circumstances (like for loops,
arguments in module calls, functions, etc), and not in others.
If a value could be assigned to a "variable" at any time, then the
controversy would just go away.

Jerry

--
Extra Ham Operator: K7AZJ
Registered Linux User: 275424
Raspberry Pi and Openscad developer

The most exciting phrase to hear in science - the one that heralds new
discoveries - is not "Eureka!" but "That's funny...".- Isaac. Asimov

OpenSCAD is a mathematical notation for describing geometric shapes. It's
not a general purpose programming language.

OpenSCAD's use of the term "variable" is consistent with the meaning of
"variable" in mathematics, and in pure functional programming languages.
The word "variable" in mathematics and formal logic has a distinguished
history going back hundreds of years, and I don't think there's a problem
using that word with its traditional meaning.

The terminology I'd get rid of instead is "assignment statement". I
consider x = 1; to be a "definition" of x. It's a mathematical equation
that defines the meaning of x within its scope. The underlying
mathematical reason that x = x + 1; is invalid is because this is not a
valid equation defining x. There is no number x that is equal to itself
plus 1.

On 30 August 2016 at 08:09, L Boyd lboyd@frontiernet.net wrote:

How about having a warning message something like
WARNING Value reassigned: line 6
and a preferences to turn this one warning off.  If the warning
is on by default, it may cause the new user to notice that this
procedure is not expected and to look in the manual.

- Don

On Aug 30, doug moen propounded certain bytes, to wit:

While "variable" may have a long history in math and logic, we are
dealing with a programming language here, and the term "variable" has a
common meaning that is, unfortunately, different than what is intended
in OpenSCAD.  I think calling those beasts "named constants" is the
closest we will come, and it will be much less confusing (to a
programmer) than "variable".  We need to document for the intended audience.

IMHO, of course.  Goes without saying

Jon

On 8/30/2016 4:14 PM, doug moen wrote:

I am far from being an expert in programming languages and I don't want to
start a controversy.

According to  wikipedia
https://en.wikipedia.org/wiki/Assignment_(computer_science)#Single_assignment
:

Named constant seems fair to me.

In none programming language I know, function and procedure arguments are
called variables. They are a way to reference values. In OpenSCAD for
statements, the identifier that index the iteration is just a named
reference of a value in a list or range inside a specific scope. It is not a
true variable.

--
View this message in context: http://forum.openscad.org/What-to-call-variables-tp18260p18271.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

L Boyd wrote

Remember this is not an iterative language, the for() does not loop in the
programmatic sense, it builds a tree of objects one branch for each
representation of the loop 'identifier', inside each of which the 'variable'
is a specific and seperate instantiation of that 'identifier'.

So:

for (i=[0:3])
translate([i*10,0,0])
cube(i+1);

Produces:

group() {
group() {
multmatrix([[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]]) {
cube(size = [1, 1, 1], center = false);
}
multmatrix([[1, 0, 0, 10], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]]) {
cube(size = [2, 2, 2], center = false);
}
multmatrix([[1, 0, 0, 20], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]]) {
cube(size = [3, 3, 3], center = false);
}
multmatrix([[1, 0, 0, 30], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]]) {
cube(size = [4, 4, 4], center = false);
}
}
}

All instances of the for() exist at the same time, they do not iterate
sequentially.

In imperative programming languages 'variable' by definition can vary, thus
they also have 'constants' which can't. {corollary to Murphy's law: variable
won't & constants aren't}
{wikipedia} "In pure functional and logic languages,variables{*1} are bound
to expressions and keep a single value during their entire lifetime due to
the requirements of referential transparency", but it is common to use the
term 'variable' even tho they don't, and they don't generally have the
concept of 'constants'.

So technically, as a declarative language, 'variable' as a term IS
appropriate, but for imperative programmers 'constant' would be better.

{*1} I'd argue that wikipedia definition is wrong, it should be "
'identifiers' are bound to ..."

So how about, in the manual, we use the term 'identifier', and expunge
'variable'? Also any reference to 'assignment' changes to 'binding'.

?

Admin - PM me if you need anything, or if I've done something stupid...

Unless specifically shown otherwise above, my contribution is in the Public Domain; to the extent possible under law, I have waived all copyright and related or neighbouring rights to this work. Obviously inclusion of works of previous authors is not included in the above.

The TPP is no simple “trade agreement.”  Fight it! http://www.ourfairdeal.org/  time is running out!

View this message in context: http://forum.openscad.org/What-to-call-variables-tp18260p18275.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

What's wrong with "scoped constant" . Only uses two words and it describes
the behavior.  Variables is definitely extremely misleading.
On Aug 31, 2016 9:23 AM, "L Boyd" lboyd@frontiernet.net wrote:

Interesting possibility.

MichaelAtOz wrote

Larry

View this message in context: http://forum.openscad.org/What-to-call-variables-tp18260p18283.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

Why not call them symbols, as in symbolic math?

Variables in programming language have an implied time dependency. X=X+1 is
valid because it refers to two different X symbols, old and new.

Anyway, for loop iterative math can be replaced with recursive functions,
just like the classical factorial function. Tested something like this
yesterday, and it seemed to work. While this is a little harder to write, i
am OK with the behavior.
F(n) = n<1? 1 : n*F(n-1)

For loops in rendering work as they should.

Em 31 de ago de 2016 10:28, "Dan Shriver" tabbydan@gmail.com escreveu:

DanS wrote

Excellent idea. It correctly describes their behavior, while flagging the
difference from constants in many imperative programming languages.

That makes 2 votes, how about anyone else?

Larry

View this message in context: http://forum.openscad.org/What-to-call-variables-tp18260p18285.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

Good description, better than what I wrote in the manual.

MichaelAtOz wrote

Larry

View this message in context: http://forum.openscad.org/What-to-call-variables-tp18260p18286.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

lhartmann wrote

+1

$  Runsun Pan, PhD $ libs: doctest , faces ( git ), offline doc ( git ), runscad.py ( 2 , git ), synwrite ( 2 );   $ tips: hash ( 2 ), matrix ( 2 , 3 ), sweep ( 2 ), var ( 2 ), lerp , animation ( gif , prodVid , animlib ), precision ( 2 ), xl-control , type , rounded polygon , chfont , tailRecur ( 2, 3 ), isosphere ( 2 ), area , vol/center , RGB , CurvedImg ; $ Apps: rollApp , blockscad , openjscad , on AWS ( pdf )

View this message in context: http://forum.openscad.org/What-to-call-variables-tp18260p18306.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

I just had a great idea for a name.

Zarfog the Destroyer

Zarfog Destroys all before him, lays waste to the values of previously
declared Zarfogs. None are left in his path.

--
don bright
hmbright@fastmail.com

On Tue, Aug 30, 2016, at 07:09 AM, L Boyd wrote:

Great idea, indeed! It distinguish it from
Joker-you-will-never-know-who-I-am, AKA variables.

2016-09-04 14:39 GMT-03:00 don bright hmbright@fastmail.com:

It seems to me that "Varonstants" would be a good name. That should
alert the people who are quick to make assumptions about languages from
their own reference points, that is, hey! this is something you've
probably not come across before?

Would I be churlish to suggest that it appears that openSCAD was a 3-D
solids tool that began as a quick way into defining solids, required
nice curves, endpoints and ranges, so adopted the mathematics, but never
quite got the programming side of it right, and that the present way
that varonstants are treated is sort of an oversight because openSCAD is
just a "solids defining tool", and not a strictly and ideal programming
language at all?? In other words the refinement of the programming
aspects of the language has been a compromise from the early days when
to render a shape was cool and the ramifications of the script behind it
was just accepted, warts and all.  After all a statement that says
cube([1,2,3]) that does not resemble a cube is somewhat of a giveaway.

On the other hand... I am really pleased with openSCAD mainly because I
am happy with simple stuff and have no desires to push the boundaries
and happy to live within its present limitations. I have completed many
productive projects through to 3D Printed objects as a result. A
fantastic tool.  I just don't think that people have a right to
seriously admonish the work of so many that have created this tool with
too many high minded criticisms.  I admire the work of people who have
volunteered to contribute to this project, and especially to those who
recognise its origins and strive to move us all onwards.

On 05/09/16 04:27, Ronaldo Persiano wrote:

--

Rob Ward
Lake Tyers Beach, 3909
Lake Tyers Beach Website http://www.laketyersbeach.net.au
XP to XUbuntu http://www.laketyersbeach.net.au/XP2XU.html

Why not call them what they really are, PARAMETERS?

Just my $0.02,
Jean-Paul
N1JPL

I like PARAMETERS, that makes sense.

On Tue, Sep 6, 2016 at 10:18 PM, dmcnutt dmcnutt@macnauchtan.com wrote:

variety of outside diameters. Hopefully openSCAD will help me to create
parts that will work over a whole array of outside diameters. I just have to
figure where to put them as I change the requirements. That's a whole bunch
of variables that must change.

That's a good example of the difference between variables and constants.
Variables can and do change between calculations, constants don't. But in
mathematics and OpenSCAD variables keep the same value throughout a program.

There is a variety of computer architecture called "Data-Flow". It removes
the bottle-neck caused by only having a single processor by contructing a
program as a flow of data between operations. When a particular operation
has all its inputs available then it is executed. The program can be easily
distibuted between many processors; it's a very interesting technique. It
also shares OpenSCAD's requirement that variables do not change because
loops have to be unwrapped into a single path of data. Each operation on
that path can only be executed once just as in OpenSCAD each created
polyhedron is unique and unchanging. Our experiences with OpenSCAD could
feed into language design in the data-flow field and vice-versa.

It seems to me that one time "a = a + 1" is desired is within loops, and
that is usually because the loop variable is doing something else. I have
recently been using the following construction when iterating over lists in
C++:

for (element = First(), count = 0; element != NULL && count < 1000; element
= Next(element), count++)

Maybe something similar could be provided in OpenSCAD, or maybe there could
be a function that returned the loop index: 0 on the first iteration, 1 on
the second, and so on, even if the loop was over elements of a list.

The other use of "a = a + 1" is to count occurrences of some event. In that
case the intermediate values of the variable are never used; only the final
value is valid when all the events have been counted. Maybe we could have a
"tally" statement that made such a use explicit. I don't think that would
break OpenSCAD.

I'm considering creating a pre-compiler that mixes OpenSCAD and Python.

GOD is REAL (unless declared INTEGER)

--
View this message in context: http://forum.openscad.org/What-to-call-variables-tp18260p18337.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

If you want to have both an element and its index you can do this:

for(i = [0 : len(list) - 1]) {
element = list[i];
...
}

To count occurrences you can use a recursive function instead of a for
loop.

function count_events(list, i = 0, tally = 0) = i == len(list) ? tally :
count(list, i + 1, tally + is_event(list[i]) ? 1 : 0);

On 7 September 2016 at 11:52, Richard Urwin <
soronlin+openscad@googlemail.com> wrote:

Von: "Richard Urwin" soronlin+openscad@googlemail.com

That already exists as experimental feature:

https://en.wikibooks.org/wiki/OpenSCAD_User_Manual/WIP#for_expression

Please consider contributing to OpenSCAD directly. Yes, it's
much more difficult to change things that are already used by
lots of people who really don't want their scripts broken by
updates (which unfortunately does happen, most of the time
not intentionally).
On the other hand, it is used by lots of people which does
make things interesting and rewarding :-).

While there are probably a number of use cases where external
programs using OpenSCAD as backend make very much sense, it's
still a bit sad to see all those projects started with huge
enthusiasm die after half a year or so when it gets hard to
continue supporting users.

ciao,
Torsten.

tp3 wrote

What I'm thinking about would be a dozen lines or so of code. Basically
search and replace <<...>> with "print """...""" and then execute the
resulting program. There is a bit more stuff around defining the input and
output streams and handling includes, but not much.

I'll need more experience before I'm comfortable suggesting modifications to
OpenSCAD, and plugging an imperative language into it would just make for
confusion. The sorts of things that might be interesting are constructions
like this: http://www2.cmp.uea.ac.uk/~jrwg/Sisal/16.Seq.loops.html

--
View this message in context: http://forum.openscad.org/What-to-call-variables-tp18260p18341.html
Sent from the OpenSCAD mailing list archive at Nabble.com.