F�rster vom Silberwald wrote:
>
>
> It happend once that I have been exploring Forth on the Mac; it was
> PowerMops Forth which had some substantial additions to pure insane
> standard Forth.
>
> However, I mean comparing Forth to Lisp hurts. Surely, one should have
> a better knowledge of Forth than what I call my knowledge of Forth. But
> have you ever seen numerical algorithms implemented in standard Forth
> (study the Forth scientific library)? I mean what are all the chemical
> substances which have wrecked havoc in all the brains of Forth users?
> Nobody on earth can tell me that a human beeing voluntarily choses
> Forth for writing numerical code. Although, the bad side: such humans
> exist - really.
>
Before you assume that intelligent people must be idiots because they
disagree with one or another of your prejudices, have a look at the web
page from a course I taught for many years:
http://Galileo.phys.Virginia.EDU/classes/551.jvn.fall01/
There are lots of good things on it, including a Primer of Forth.
I will now tell you why I, in your phrase, "voluntarily choose Forth for
numerical
work". Having programmed for many years in Fortran, with a little C, Lisp,
Pascal
and others, I turned to Forth when I got tired of trying to debug big programs,
or of waiting out long compile/debug cycles when testing algorithms. My
experience
has been that once I learned to program in Forth I could
a) write much simpler, much clearer programs;
b) get them working almost instantly;
c) modify them rapidly if I wanted to change something;
d) drop down into assembler whenever I needed to speed up a
bottleneck subroutine---Forth makes assembler much easier
than in most integrated environments.
I think that what bugged you about numerical analysis in Forth was
the lack of a FORmula TRANslator. Well, I found that I was writing a
formula as a comment before decomposing it into postfix notation, so I
would be able to understand it again. At some point I realized I was just
doing the work of a FORmula TRANslator, so I wrote one to take this
burden off me. That actually helps a lot. Probably it would help Lisp
to have one also, and of course it would be child's play to write it.
Lots of people use FORmula TRANslators in Forth now, mine or Wil Baden's
or their own version. There is no Standard for this, so feel free to
experiment if you like. Mine is written in Standard Forth, hence should
run on almost any Forth that supposrts the Standard (optional) floating
point wordlist.
You will find code examples in various sub-pages of the above course
home page. Also comparisons with C, etc. Perhaps after having a look
you will be less inclined to rash statements.
--
Julian V. Noble
Professor Emeritus of Physics
···@lessspamformother.virginia.edu
^^^^^^^^^^^^^^^^^^
http://galileo.phys.virginia.edu/~jvn/
"As democracy is perfected, the office of president represents, more and
more closely, the inner soul of the people. On some great and glorious
day the plain folks of the land will reach their heart's desire at last
and the White House will be adorned by a downright moron."
--- H. L. Mencken (1880 - 1956)
"Julian V. Noble" <···@virginia.edu> writes:
> I think that what bugged you about numerical analysis in Forth was
> the lack of a FORmula TRANslator. Well, I found that I was writing a
> formula as a comment before decomposing it into postfix notation, so I
> would be able to understand it again. At some point I realized I was just
> doing the work of a FORmula TRANslator, so I wrote one to take this
> burden off me. That actually helps a lot. Probably it would help Lisp
> to have one also, and of course it would be child's play to write it.
There is the Infix package
http://www.cs.cmu.edu/afs/cs/project/ai-repository/ai/lang/lisp/code/syntax/infix/infix.cl
CL-USER> (let ((x 5))
#I(2*3+sin(x)))
5.0410757
Nicolas.
On Sat, 05 Mar 2005 10:35:58 +0100, Nicolas Neuss
<·······@iwr.uni-heidelberg.de> wrote:
> "Julian V. Noble" <···@virginia.edu> writes:
>
>> I think that what bugged you about numerical analysis in Forth was
>> the lack of a FORmula TRANslator. Well, I found that I was writing a
>> formula as a comment before decomposing it into postfix notation, so I
>> would be able to understand it again. At some point I realized I was
>> just
>> doing the work of a FORmula TRANslator, so I wrote one to take this
>> burden off me. That actually helps a lot. Probably it would help Lisp
>> to have one also, and of course it would be child's play to write it.
>
> There is the Infix package
>
> http://www.cs.cmu.edu/afs/cs/project/ai-repository/ai/lang/lisp/code/syntax/infix/infix.cl
>
> CL-USER> (let ((x 5))
> #I(2*3+sin(x)))
> 5.0410757
>
> Nicolas.
Ugh. The allowence or skipping whitespace between operators means you have
to *fight*
the reader the whole way making this code quite lengthy and complex.
--
Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/
John Thingstad wrote:
> On Sat, 05 Mar 2005 10:35:58 +0100, Nicolas Neuss
> <·······@iwr.uni-heidelberg.de> wrote:
>
>> "Julian V. Noble" <···@virginia.edu> writes:
>>
>>> I think that what bugged you about numerical analysis in Forth was
>>> the lack of a FORmula TRANslator. Well, I found that I was writing a
>>> formula as a comment before decomposing it into postfix notation, so I
>>> would be able to understand it again. At some point I realized I was
>>> just
>>> doing the work of a FORmula TRANslator, so I wrote one to take this
>>> burden off me. That actually helps a lot. Probably it would help Lisp
>>> to have one also, and of course it would be child's play to write it.
>>
>>
>> There is the Infix package
>>
>> http://www.cs.cmu.edu/afs/cs/project/ai-repository/ai/lang/lisp/code/syntax/infix/infix.cl
>>
>>
>> CL-USER> (let ((x 5))
>> #I(2*3+sin(x)))
>> 5.0410757
>>
>> Nicolas.
>
>
> Ugh. The allowence or skipping whitespace between operators means you
> have to *fight*
> the reader the whole way making this code quite lengthy and complex.
>
What do you mean? The above can be written as
CL-USER 3 > (let ((x 5))
#I(2 * 3 + sin(x)))
5.0410757
Cheers
--
Marco
On Mon, 07 Mar 2005 12:37:39 -0500, Marco Antoniotti <·······@cs.nyu.edu>
wrote:
>
> What do you mean? The above can be written as
>
> CL-USER 3 > (let ((x 5))
> #I(2 * 3 + sin(x)))
> 5.0410757
>
>
> Cheers
> --
> Marco
I said allowance skipping whitespace for not requirement for. I think you
misunderstand me.
I was looking through the sourcecode for infix.cl and found it lengthy and
complicated.
It would have been much simpler to write a parser for "2 * 3 + sin ( x )"
and it would be only slightly more annoying to use.
(I think a coder should strike a balance between the time it takes to
develop
the tool vs. the time it saves..)
If you are willing to use it as a black box it seems fine.
--
Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/
John Thingstad wrote:
> On Mon, 07 Mar 2005 12:37:39 -0500, Marco Antoniotti
> <·······@cs.nyu.edu> wrote:
>
>>
>> What do you mean? The above can be written as
>>
>> CL-USER 3 > (let ((x 5))
>> #I(2 * 3 + sin(x)))
>> 5.0410757
>>
>>
>> Cheers
>> --
>> Marco
>
>
> I said allowance skipping whitespace for not requirement for. I think
> you misunderstand me.
> I was looking through the sourcecode for infix.cl and found it lengthy
> and complicated.
> It would have been much simpler to write a parser for "2 * 3 + sin ( x )"
> and it would be only slightly more annoying to use.
> (I think a coder should strike a balance between the time it takes to
> develop
> the tool vs. the time it saves..)
Well, sometimes it is fun to develop tools and code. I do it now in my
spare time.
> If you are willing to use it as a black box it seems fine.
I have used INFIX as a black (and not so black) box for over 10 years
now. Mark Kantrowitz's code looks dated as it was usually written in a
pre-CLtL2 time. However, I find it very very good and remarkably stable.
Cheers
--
Marco
Julian V. Noble Mar 4, 1:28 pm wrote:
==
I will now tell you why I, in your phrase, "voluntarily choose Forth
for
numerical
work". Having programmed for many years in Fortran, with a little C,
Lisp,
Pascal
and others, I turned to Forth when I got tired of trying to debug big
programs,
or of waiting out long compile/debug cycles when testing algorithms. My
experience
has been that once I learned to program in Forth I could
a) write much simpler, much clearer programs;
b) get them working almost instantly;
c) modify them rapidly if I wanted to change something;
d) drop down into assembler whenever I needed to speed up a
bottleneck subroutine---Forth makes assembler much easier
than in most integrated environments.
==
I think the discussion goes into the wrong direction. I am not
questioning your experience, though.
But help you God once you start developing larger applications in
Forth. I mean with larger application not numerical recipies functions.
Once you start: managing huge data sets, reading and writing to files
and at the same time steering external legacy C or Fortran programs
then you will need a better programming language.
The problem in Forth I gather are not so much related to writing a spot
on program which runs from the very first start on. I believe you when
you say so. But the hell will open if you try to change and adopt that
particular program in Forth.
In Scheme it is often so much easier to change a pre-defined program
structure. You can often write very general programs in Scheme - thanks
to its dynamic nature - where a new array lets say will not drive you
into a debugging hell.
You want to have some more parameters out of a Scheme function? Okay,
lets add them in a list without any additional programming.
I am not saying that Forth is useless. But honestly speaking: I like to
go down-hill by a full suspension mountainbike as opposed to a bike
which is just pure steel.
I really wonder what your programs are about and which kinds of
problems you try to solve. I by all honesty cannot understand why you
once upon a time felt into that Forth trap. I mean you were using Lisp
- right? Sorry I cannot understand why people chose a lesser
programming language. Have you ever wondered why people resort themself
to using Windows? I have no explanation for such behavior.
My experience (not long enough to weight in that discussion) however,
was the opposite of yours: I do not really have the time and patience
to debug Forth programs. In Bigloo I get very robust prgrams which are
easy to change and predestined for re-using legacy code.
Regards,
Förster vom Silberwald
PS: Not to everyones pleasure in clf. I am writing my PhD in physics.
My supervisor of my PhD assigned me a new Master student of him. I will
have to supervise the Master student in the next couple of months or
years. However, his programming experience is limited. I said to him he
may use what he wants (even he is that lunactic and wants to use
Fortran 95) I do not mind. But I would make one exception: he should
really refrain from using Forth. That will not put him far because he
will have to manage big data sets from the field of atmopsheric
sciences.
F�rster vom Silberwald wrote:
...
> I really wonder what your programs are about and which kinds of
> problems you try to solve. I by all honesty cannot understand why you
> once upon a time felt into that Forth trap. I mean you were using Lisp
> - right? Sorry I cannot understand why people chose a lesser
> programming language. Have you ever wondered why people resort themself
> to using Windows? I have no explanation for such behavior.
>
> My experience (not long enough to weight in that discussion) however,
> was the opposite of yours: I do not really have the time and patience
> to debug Forth programs. In Bigloo I get very robust prgrams which are
> easy to change and predestined for re-using legacy code.
>
>
> Regards,
> F�rster vom Silberwald
> PS: Not to everyones pleasure in clf. I am writing my PhD in physics.
> My supervisor of my PhD assigned me a new Master student of him. I will
> have to supervise the Master student in the next couple of months or
> years. However, his programming experience is limited. I said to him he
> may use what he wants (even he is that lunactic and wants to use
> Fortran 95) I do not mind. But I would make one exception: he should
> really refrain from using Forth. That will not put him far because he
> will have to manage big data sets from the field of atmopsheric
> sciences.
Sadly, you are misinformed. With broader experience, you would know
better. Too bad!
Jerry
--
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������
Förster vom Silberwald wrote:
>
> I am not saying that Forth is useless. But honestly speaking: I like
to
> go down-hill by a full suspension mountainbike as opposed to a bike
> which is just pure steel.
You're approaching Forth from the wrong paradigm. You build the bike
you want, then you ride it. You don't ride the bike provided to you
unless you like it the way it is. You don't have to take it or leave
it, as is the case with most other languages. Years of Fortran
experience may have ingrained habits that don't fit Forth.
As you may have noticed, you can't win a technical argument against
Forth in c.l.f. Maybe you can force a draw, but it's very hard.
It's possible to argue about the decline of Forth, since that topic is
much less understood than Forth itself. The psychology behind the use
of Forth and the cultural bias against it would be an interesting topic
to write about for a humanities course.
Brad
F�rster vom Silberwald wrote:
> PS: Not to everyones pleasure in clf. I am writing my PhD in physics.
> My supervisor of my PhD assigned me a new Master student of him. I
> will have to supervise the Master student in the next couple of
> months or years. However, his programming experience is limited. I
> said to him he may use what he wants (even he is that lunactic and
> wants to use Fortran 95) I do not mind. But I would make one
> exception: he should really refrain from using Forth. That will not
> put him far because he will have to manage big data sets from the
> field of atmopsheric sciences.
This doesn't make sense.
First, I can't think of any rational reason why Forth (or nearly any
other language I can think of) wouldn't be suitable for "big data sets."
What *specific* quality of those data sets makes Forth unsuitable?
Second, I would imagine the goal here is to analyze the "big data set."
If the student is productive in language X and will get results in
language X, on what rational reason why would you reject language X?
Replace "X" with "Forth" if you're having a hard time understanding the
point.
F�rster vom Silberwald wrote:
> Once you start: managing huge data sets, reading and writing to files
> and at the same time steering external legacy C or Fortran programs
> then you will need a better programming language.
Or a better Forth than the one you know (though I doubt there is one you
know).
> The problem in Forth I gather are not so much related to writing a spot
> on program which runs from the very first start on. I believe you when
> you say so. But the hell will open if you try to change and adopt that
> particular program in Forth.
I've written and maintained a number of bigger Forth programs. It's a
non-issue. While (standard) Forth does not *contain* abstractions like
lists, objects, or arrays, it allows you to *build* them in a few lines
(and most systems indeed have things like that as part of their library).
I've written programs where I changed the underlying abstraction how to
store data. Thinking Forth (http://thinking-forth.sourceforge.net/)
contains a small example where you extend a variable to be redirect to
different classes without changing the rest of the code.
There are differnet "reasons" (or rather sound-bites) why you can't write
big programs in Forth. Let's examine them all together:
* you can't write big programs, because Forth's philosophy is "keep it
simple", and you'll just write a small simple program that does all the
functionality of what would be a large program in another language.
- Ok, we accept that "sound-bite"; we like to keep programs simple ;-).
* you can't write big programs, because Forth is 16 bit.
- Get up and leave the cave where you have been hiding. There are 64 bit
Forths available today.
* you can't write big programs, because Forth source has to fit into one 1k
screen.
- Even more funny, as it never was true ;-)
* you can't write big programs, because Forth is write-only, and therefore,
you can't change what you have written.
- Forth is as "unreadable" as Chinese, Arabic, Russian, or whatever language
you don't understand. Now go and try to write a single correct sentence in
Chinese, and come back when you manage to do that without being able to
read or understand Chinese. Once you've learned Forth and got used to a
style with short definitions, useful stack comments, and good names,
there's no problem. People have written and maintained Forth programs with
up to almost a million lines of code. If you stick your head into the
ground, cover your ears with your hands, and sing "lalala" to avoid hearing
tales like that, all we can do is kick your ass ;-).
Forth and Lisp share a number of features, like interactivity and
meta-programming. They are different on other features, like Lisp having
named, and Forth having unnamed parameters, but allowing named locals. Or
data typing (Lisp: dynamic, the data knows its type, Forth: the
user/programmer names the type). You can combine things by having a reverse
polish lisp (RPL, like on the HP calculators, or PostScript), or an object
oriented Forth extension, which gives you typed data, too. I haven't seen a
Lisp-like syntax with a Forth-like semantics, though.
--
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/
Bernd Paysan wrote:
> Or a better Forth than the one you know (though I doubt there is one
you
> know).
Hello:
I have to be late with my presponse due to a meeting in Zurich where I
have been participating last week.
However, one should keep discussions in a fair manner. I mean one could
easily infer from your reply that PowerMops is not well intented for
means of Forth programming.
Surely, I say Forth is crap. But PowerMops as it is is really good and
it is a pity that it doesn't run on other architectures as well.
I once showed Scheme to my colleagues and they think now I am on drugs.
They cannot understand that parentheses never comes into ones way if
used properly.
It is indeed a miracle why people get quickly upset if one tells them
that they are using a crapy language (e.g. Forth). People have told me
the same for Scheme. So what? I do not mind.
Where is the problem when people do not like things or behaviors? I
cannot understand, sorry. I smoke cigars and prefer the ones from
Brazil. I would ridicule myself if I am going now to say that every
other man should smoke cigars from Brasil as well.
Förster vom Silberwald
F�rster vom Silberwald wrote:
> Surely, I say Forth is crap.
You said the exact same thing about CL a while ago. In view of this, I
wonder if I should take the above as a recommendation of Forth?
alex goldman wrote:
> Förster vom Silberwald wrote:
>
> > Surely, I say Forth is crap.
>
> You said the exact same thing about CL a while ago. In view of this,
I
> wonder if I should take the above as a recommendation of Forth?
I cannot remember. Maybe in the context of comparing it to Scheme.
However, if the following makes your day: Scheme is crap.
What makes me behaving not gentlemen like is the fact that Forther
believe that they are the better programmers. That is just lunatic.
Okay, the same could be said for Schemers or Lispers who thinks they
are upper class programmers.
I wonder really about: what are the amount of the people who are
promoting Forth and using it at the same time in their daily work. I
for one can say that I am using Scheme every day.
The problem what I often see is the following: if you define your life
or working perspective solely on a programming job then you have lost
because you will rarely find a programming job which will suit you and
your beloved programming language.
Maybe I am in a better position in academia because I may use the tools
what will do my job as long as I deliver the expected results to my
boss. In all the years working on my PhD degree in physics my boss has
never asked me which tools I used.
Förster vom Silberwald
In comp.lang.forth F�rster vom Silberwald <··········@hotmail.com> wrote:
> alex goldman wrote:
>> F�rster vom Silberwald wrote:
>>
>> > Surely, I say Forth is crap.
>>
>> You said the exact same thing about CL a while ago. In view of this,
> I
>> wonder if I should take the above as a recommendation of Forth?
> I cannot remember. Maybe in the context of comparing it to Scheme.
> However, if the following makes your day: Scheme is crap.
> What makes me behaving not gentlemen like is the fact that Forther
> believe that they are the better programmers. That is just lunatic.
Don't you have a bridge to hide under?
Andrew.
········@littlepinkcloud.invalid wrote:
>
> Don't you have a bridge to hide under?
--
Cheers, www.indiegamedesign.com
Brandon Van Every Seattle, WA
"Trollhunter" - (n.) A person who habitually accuses
people of being Trolls.
In article <························@o13g2000cwo.googlegroups.com>,
=?iso-8859-1?q?F=F6rster_vom_Silberwald?= <··········@hotmail.com> wrote:
>alex goldman wrote:
<SNIP>
>What makes me behaving not gentlemen like is the fact that Forther
>believe that they are the better programmers. That is just lunatic.
>Okay, the same could be said for Schemers or Lispers who thinks they
>are upper class programmers.
They probably are. Come on. If you are able to select a language
you want to promote, you belong to an elite.
<SNIP>
>F=F6rster vom Silberwald
Groetjes Albert
--
--
Albert van der Horst,Oranjestr 8,3511 RA UTRECHT,THE NETHERLANDS
One man-hour to invent,
One man-week to implement,
One lawyer-year to patent.
F�rster vom Silberwald wrote:
> Surely, I say Forth is crap.
That's your opinion.
> I once showed Scheme to my colleagues and they think now I am on drugs.
> They cannot understand that parentheses never comes into ones way if
> used properly.
And I'm sure some people thought Einstein was on crack. Not everyone is
unprejudiced enough to actually try out a totally different culture.
Even if you can't live like them, can you tolerate them without insult?
> It is indeed a miracle why people get quickly upset if one tells them
> that they are using a crapy language (e.g. Forth). People have told me
> the same for Scheme. So what? I do not mind.
But why do you want to be someone who thinks or says other languages are
crap? Just because you don't like them (languages)?
I think C dialects are crap, but I've seen more than those. Most people
who think Scheme/Forth etc. are crap have never tried them. Everyone I
know who likes Java knows not much else (and even those often don't like
it).
> Where is the problem when people do not like things or behaviors? I
> cannot understand, sorry. I smoke cigars and prefer the ones from
> Brazil. I would ridicule myself if I am going now to say that every
> other man should smoke cigars from Brasil as well.
Then why do you say cigarettes are crap (in the Forth context)?
On 7 Mar 2005 04:17:51 -0800,
"=?iso-8859-1?q?F=F6rster_vom_Silberwald?=" <··········@hotmail.com>
wrote:
>But help you God once you start developing larger applications in
>Forth. I mean with larger application not numerical recipies functions.
>
>Once you start: managing huge data sets, reading and writing to files
>and at the same time steering external legacy C or Fortran programs
>then you will need a better programming language.
See:
http://www.ccssa.com
This is a construction planning package. It planned the Hong
Kong airport and metro. It consists of 850,000+ lines of Forth
source code. It has been ported across at least four platforms
in the last 15 years.
The primary issue for maintaining and developing large applications
with a team of is management.
Stephen
--
Stephen Pelc, ··········@INVALID.mpeltd.demon.co.uk
MicroProcessor Engineering Ltd - More Real, Less Time
133 Hill Lane, Southampton SO15 5AF, England
tel: +44 (0)23 8063 1441, fax: +44 (0)23 8033 9691
web: http://www.mpeltd.demon.co.uk - free VFX Forth downloads