Re: Question about languages
Robbert Haarman
What functionality do you expect out of an IDE?
As for a language recommendation, I assume you're going to be working
with email and parsing text. My recommendation, then, would be a
language for which libraries/modules/whatever that do this have already
been developed and used extensively.
Languages that come to mind are Perl and Ruby (also Python and Java, but
my knowledge about the text processing libraries for these is very much
out of date). Of the two, Perl seems the more obvious choice, because it
is the language usually chosen for this kind of task, and it is
procedural (albeit with objects bolted on).
On the other hand, Ruby's designer claims that one of the goals for Ruby
was to be "a better Perl than Perl", and I think it has achieved that
goal. Where Perl feels to me like features where thrown in without any
regard to overal consistency or elegance, Ruby feels like it has a nice,
clean design with a few minor mistakes. At the very least, in Ruby you
don't have to worry about scalar vs. list context.
Now, getting back to IDEs, I've never used any IDE for Perl, but I would
be surprised if there weren't any. I know for sure that Emacs supports
Perl. For Ruby, you would probably want to use one of the Rails IDEs,
even if you're not going to use Rails. I know there's one for Emacs and
one for Eclipse. There's also FreeRIDE, which I think is a pure Ruby
IDE. I haven't used that one, though.
Good luck with your decision, and, more importantly, good luck with your
project!
Regards,
Bob
--
"Common sense is not so common."
Marco van de Voort
> That's my previous experience. What I'm looking for is suggestions on
> programming languages. To start, I'd like a language that I can u
> se on
> both Linux and Windows XP, with a _free_ IDE/development kit/etc. Most
> important of all, I'd like it to be an effective language for a
> procedural programmer.
http://www.freepascal.org (IDE http://lazarus.freepascal.org ) with the
additional benefit that books can be found (Turbo Pascal or Delphi)
Free Pascal allows both procedural as OOP programming.
James Harris
...
> That's my previous experience. What I'm looking for is suggestions on
> programming languages. To start, I'd like a language that I can use on
> both Linux and Windows XP, with a _free_ IDE/development kit/etc. Most
> important of all, I'd like it to be an effective language for a
> procedural programmer. I still don't grokk OO programming and I'm not
> really looking to start right now. I'd just like to make my life
> easier and work with my presently known style of programming. The only
> other quirk I can think of is that it's going to have to read
> information from CSV spreadsheet files, although that seems to be
> pretty commonly supported.
>
> Java isn't proving very friendly for procedural programming, otherwise
> I'd use it. Any suggestions?
Given what you say, try Python. It is mainstream, has a clear syntax
and can be procedural or OO. Just remember to keep indentations
consistent using spaces OR tabs, not both. It is free for either Linux
or XP. Perl is too but is much more cryptic.
Robbert Haarman
> On Mon, 18 Jun 2007 13:59:06 +0200, Robbert Haarman
> <ingl...@inglorion.net> wrote:
>
> >Jenny,
>
> To start, thank you to all the people who responded, I read all of the
> answers.
>
> >What functionality do you expect out of an IDE?
>
> development kit to be free. I meant whatever I had to install on my
> computer to end up being able to create programs in that language
> needed to be free.
Ah, ok. *Sighs in relief* That would hold for any language I could
recommend, because I have only used languages for which open source
implementations exist.
> I can write programs in a text editor with syntax highlighting. The
> part that makes them "run" has to be free :).
Yes, so the word you were looking for is "implementation", not "IDE".
> OTOH, Ruby seems to be very much focused on being Object Oriented (the
> topics in the Ruby newsgroup are overwhelmingly OO).
Yes, Ruby is definitely a language that supports and encourages
object-oriented programming.
On the other hand, I program predominantly in functional style, using
objects here and there where it makes sense (e.g. for grouping together
values that belong together). I can assure you that Ruby is an excellent
multi-paradigm language, supporting object-oriented, but also functional
and procedural styles.
Regards,
Bob
--
The chief cause of problems is solutions.
-- Eric Sevareid
Torben Ægidius Mogensen
If you mostly program in functional style but want a bit of OO from
time to time, you should try O'Caml, F#, SML or Haskell.
O'Caml and F# have traditional object models, SML has a module system
and Haskell has type classes, both of which are akin to static
classes, but more powerful. SML modules have somewhat more syntactic
overhead than Haskell type classes, though.
You can see a comparison at
http://faculty.cs.tamu.edu/jarvi/papers/cmp_gp.pdf
or, if you prefer slides,
http://faculty.cs.tamu.edu/jarvi/papers/comp_gp_slides.pdf
Torben
Robbert Haarman
> Robbert Haarman <comp.la...@inglorion.net> writes:
> >
> > On the other hand, I program predominantly in functional style, using
> > objects here and there where it makes sense (e.g. for grouping together
> > values that belong together). I can assure you that Ruby is an excellent
> > multi-paradigm language, supporting object-oriented, but also functional
> > and procedural styles.
>
> If you mostly program in functional style but want a bit of OO from
> time to time, you should try O'Caml, F#, SML or Haskell.
Absolutely. These are great languages. However, I think, for the task
that Jenny described, Ruby would be easier. Also note that the F#
implementation has a pretty nasty license...
Regards,
Bob
--
For a list of the ways which technology has failed to improve our
quality of life, press 3.
Torben Ægidius Mogensen
> On Tue, Jun 19, 2007 at 09:41:11AM +0200, Torben Ægidius Mogensen wrote:
>> Robbert Haarman <comp.la...@inglorion.net> writes:
>>>
>>> On the other hand, I program predominantly in functional style, using
>>> objects here and there where it makes sense
>>
>> If you mostly program in functional style but want a bit of OO from
>> time to time, you should try O'Caml, F#, SML or Haskell.
>
> Absolutely. These are great languages. However, I think, for the task
> that Jenny described, Ruby would be easier.
Probably. More due to Rails than Ruby, I suspect. Anyway, my
suggestion was mainly aimed at you, since you mentioned using
functional style.
> Also note that the F# implementation has a pretty nasty license...
True. But since F# is basically O'Caml.NET, you can use O'Caml if
that bothers you. You wouldn't be using .NET anyway if you are
bothered about licenses.
Torben
Robbert Haarman
> Robbert Haarman <comp.la...@inglorion.net> writes:
>
> >> If you mostly program in functional style but want a bit of OO from
> >> time to time, you should try O'Caml, F#, SML or Haskell.
> >
> > Absolutely. These are great languages. However, I think, for the task
> > that Jenny described, Ruby would be easier.
>
> Probably. More due to Rails than Ruby, I suspect.
Err, no. I wouldn't be dragging rails into that project. Why would you
build on top of a database abstraction layer and web application
framework if all you need to do is reading and sending mail and a bit of
game mechanics?
> Anyway, my suggestion was mainly aimed at you, since you mentioned
> using functional style.
Ah, I see. Thanks for the suggestion.
> You wouldn't be using .NET anyway if you are bothered about licenses.
Perhaps. I haven't read the license for Microsoft's .NET implementation.
However, many companies are using it, so I am inclined to assume the
license isn't too bad. In either case, there's always Portable.NET and
Mono.
Regards,
Bob
--
An astronaut in space in 1970 was asked by a reporter, "How do you feel?"
"How would you feel," the astronout replied, "if you were stuck here, on
top of 20,000 parts each one supplied by the lowest engineering bidder?"
Waldek Hebisch
> I'm posting here because the comp.lang group seems to be effectively
> dead, and I'm looking for information on languages.
>
> Hopefully short explanation: I'm involved in a fantasy sport league
> that's PBEM. "Cards" are sent from players to judges, points are
> assigned and random dice are rolled, and then results are sent back to
> the community mailing list.
>
> What I'm looking for is a language that will allow me to write a
> program that will automate many of these steps, so that I can focus on
> getting results out to competitors, faster.
>
<snip>
> That's my previous experience. What I'm looking for is suggestions on
> programming languages. To start, I'd like a language that I can use on
> both Linux and Windows XP, with a _free_ IDE/development kit/etc. Most
> important of all, I'd like it to be an effective language for a
> procedural programmer. I still don't grokk OO programming and I'm not
> really looking to start right now. I'd just like to make my life
> easier and work with my presently known style of programming. The only
> other quirk I can think of is that it's going to have to read
> information from CSV spreadsheet files, although that seems to be
> pretty commonly supported.
>
I would suggest that you look deeper at the things that you want
to do. If you want to manipulate text: search, replace patterns
etc Perl is hard to beat. Also Perl has large libraries so
that is good chance that you will find alredy written code that
you can just call. OTOH if you need more compicated data
structures you can do them in Perl, but then Perl looses much
of its advantages.
It seems that you care very much about extras coming with
implementation, but just in case you are more intresed in
the language proper let me mention Pop11. Pop11 has
Pascal like syntax, but semantically is closer to Lisp.
If you wish you can do sophisticated OO or functional programming
in Pop11, but you can as well program in procedural way. You
can get more information about Pop11 (and download the
implementation) at:
http://www.cs.bham.ac.uk/research/poplog/freepoplog.html
If you want to see how various languages feel, more precisely
compare solutions to a bunch of simple programming tasks
in many languages, look at:
http://www.rosettacode.org/wiki
--
Waldek Hebisch
heb...@math.uni.wroc.pl
Rod Pemberton
"Jenny Purcell" <donts...@spam.com> wrote in message
news:vvfb73lfe7n6o4in5...@4ax.com...
> Hopefully short explanation: I'm involved in a fantasy sport league
> that's PBEM. "Cards" are sent from players to judges, points are
> assigned and random dice are rolled, and then results are sent back to
> the community mailing list.
>
Well, I've no familiarity with PBEM. But, I just did a number of quick
searches. From that, the various tasks related to PBEM: email servers,
parsing email, programming gameplay, programming databases, etc., appears to
require a broad set of skills to program from scratch. Just programming a
simple database can be quite a challenge - which is why MySQL is so popular:
Anyway, a few more quick Yahoo searches pulled up two PBEM engines that may
be useful to you:
http://www.pbemengine.nl
http://digilander.libero.it/zak965/thoth/
The second seems to be just for deck of card games.
The first doesn't seem to be game specific, i.e., you can write your own
ruleset. It's coded for Java and SQL. But, it was originally written for
some game called Ovoron's Legacy, i.e., there could be some feature you need
which isn't implemented.
I'm not sure if I agree with their interpretation of the GPL and how it
applies to your ruleset: see "Licensing." They state that if you release
your ruleset it must be GPL'd. Data and grammars aren't usually required to
be GPL'd since they're your property, any compiled results derived from them
are your legal property too. Bison has to deal with this issue. They add a
special exception because GPL'd and non-GPL'd sourcecode is mixed. I doubt
that your ruleset and GPL code is being mixed. The problem is probably the
binding requirement of the GPL: non-GPL compiled code bound with GPL'd code
is supposed to be released GPL'd. Or, non-GPL compiled code can be bound
with LGPL'd code without being released. Of course, that doesn't really
explain what happens to code compiled with GCC and GLIBC. GCC is GPL and
GLIBC is LGPL. So, non-GPL code can link with GLIBC without problems, but
non-GPL code can't legally link to code GCC emits (since some of it is part
of GCC - e.g., internal functions)... So, why is GCC GPL without an
exception? Where are the lawyers? And why does GNU believe this when it
clearly isn't true due to the license they applied?
From the Bison FAQ:
"...The other GNU programming tools, such as the GNU C compiler, have never
had such a requirement. They could always be used for nonfree software. The
reason Bison was different was not due to a special policy decision; it
resulted from applying the usual General Public License to all of the Bison
source code."
Rod Pemberton
ernobe
> So, my program will not be parsing e-mails.
It seems to me that you want a good general string-processing
language, not one that has specialized libraries for processing mail.
Icon, http://www.cs.arizona.edu/icon and Unicon, http://unicon.sourceforge.net
which is the current version of Icon undergoing development, are very
well documented ( see the 'Books about Icon' section at the website )
and are general purpose languages that will do everything you're
asking about.
Robbert Haarman
Just to get you started, I have written a little bit of Ruby code that
should give you an idea of how what you want can be implemented in Ruby.
If you give me your email address, I will send it to you.
Regards,
Bob
--
You can't cross a large chasm in two small jumps.
James Harris
wrote:
...
> Just to get you started, I have written a little bit of Ruby code that
> should give you an idea of how what you want can be implemented in Ruby.
> If you give me your email address, I will send it to you.
Bob, if not too long could you share it here? I followed a suggestion
you made a while ago of looking into Ruby and must say I didn't like
much what I saw. It may be what the only book the library had was a
poor introduction. There seemed too many ways to express everything
and a fascination with technical issues and syntax. If you have some
code that shows a better side to the language............
--
James
Rod Pemberton
> I'm leaning towards Ruby right now, but, I want to be sure that it's
> going to be right for what I want to do.
>
I wouldn't know for sure. C is my current strength. I do like the simple
Ruby OOP examples I've seen. They appear to be much simpler than the C++
OOP examples I've seen.
But, I do know that most dynamicly typed languages are written in C (at
least today they are...), i.e., Ruby, Python, Java, Lisp. And, many
staticly typed languages can be completely converted to C, e.g., Fortran,
Pascal, C++, etc. So, another language may be easier to program than C, but
they have no more functionality than C does. The major remaining languages,
which aren't written in C or convertable to C, are (effectively) dead.
There are some exceptions like GNU Prolog or MS products like Visual Basic.
Also, C derived languages (C,C++,C#,D,Java) cover the majority, i.e., 40%,
of popular language usage (according to the TIOBE Programming Community
index).
For example, GCC supports these (and maybe more) C frontends (most by GNU,
some by others...): Walter Bright's (Digital Mars) D, GNU C++, GNU Common
Lisp, Ada (GNAT), Objective-C, Java (GCJ), Fortran, Pascal (GPC), Cobol (in
development), Modula-2, PL/1, GNU FORTH, Mercury, VHDL (GHDL), Unified
Parallel C (UPC), LLVM, VMGEN (by Anton Ertl, part of GNU FORTH), BASIC (in
development: LibertyBASIC, FreeBASIC).
1) i.e., if I didn't list it, don't use it.
2) i.e., if it can't be done in C, you won't be able to do it.
Rod Pemberton
Ole Nielsby
> But, I do know that most dynamicly typed languages are written in C (at
> least today they are...), i.e., Ruby, Python, Java, Lisp. And, many
> staticly typed languages can be completely converted to C, e.g., Fortran,
> Pascal, C++, etc. So, another language may be easier to program than C,
> but
> they have no more functionality than C does.
This would depend on your definition of "functionality".
I'd say languages implemented on top of C can certainly have functionality
not found in C, like a brick house (even if built from bricks and nothing
else)
can have functionality not found in the bricks.
One example of functionality that can exist "on top of C" even if not in C
is the ability to generate and execute code on-the-fly, found in Javascript,
Lisp and others. Others are reflection, partial/symbolic evaluation, pattern
matching, type inference...
Marcin 'Qrczak' Kowalczyk
> But, I do know that most dynamicly typed languages are written in C (at
> least today they are...), i.e., Ruby, Python, Java, Lisp.
A language is not written in a language.
A particular language *implementation* (compiler or interpreter) is
written in a language.
There are several implementations of Python, Java and Lisp. One
implementation of Python is implemented in C, another is implemented
in Java, and another is implemented in C#. Many Lisps are implemented
in Lisp. Et caetera.
So this is quite wrong.
> And, many staticly typed languages can be completely converted to C,
And C can be completely converted to assembler. But what does it imply?
> So, another language may be easier to program than C, but
> they have no more functionality than C does.
This is false. For example you can embed a JavaScript program on a web
page, such that it's executable by any modern web browser. You can't do
that with a C program.
Moreover, since in most fundamental aspects nearly all languages are
equivalent, choosing languages is usually based on convenience, not
functionality.
--
__("< Marcin Kowalczyk
\__/ qrc...@knm.org.pl
^^ http://qrnik.knm.org.pl/~qrczak/
Marcin 'Qrczak' Kowalczyk
> But, I do know that most dynamicly typed languages are written in C (at
> least today they are...), i.e., Ruby, Python, Java, Lisp. And, many
> staticly typed languages can be completely converted to C, e.g., Fortran,
> Pascal, C++, etc.
I haven't heard of a compiler of Java or C# which translates Java or C#
programs to C. So in what sense "they have no more functionality than C
does"?
Robbert Haarman
before, but it appears not to have made it through.
On Fri, Jun 22, 2007 at 01:38:15PM -0700, James Harris wrote:
>
> Bob, if not too long could you share it here?
Alright. The code is below. It assumes the existence of a file named
horses.csv, which contains lines, each containing the horse name,
owner's initials, post, and pre-points, separated by comma's.
Some of the code might be a bit arcane. I hope the comments explain what
is happening. If any clarifications are needed, I would be happy to
provide them.
-----BEGIN RUBY CODE-----
#! /usr/bin/env ruby
# For each line in the CSV file...
File.open('horses.csv').each_line do |line|
# Parse the line.
# The thing being passed to match is a regular expression
# that matches four elements, separated by commas, each
# consisting of any number of characters that aren't commas.
# This is a crude CSV parser, but probably a good first start.
match = line.match(/([^,]*),([^,]*),([^,]*),([^,]*)/)
# Now we assign the appropriate parts of the match to
# variables named after the fields in the CSV file.
# Note that we need to call .to_i on the 3rd and 4rd matches
# to convert them from strings to integers.
horse_name = match[1]
owners_initials = match[2]
post = match[3].to_i
pre_points = match[4].to_i
# Roll the dice.
# This generates an array containing four die rolls,
# each die roll ranging from 1 to 10
dice = (1..4).map { rand(10) + 1 }
# Calculate score.
# The score is the sum of post, pre_points, and each of the die rolls
score = post + pre_points
dice.each { |n| score = score + n }
# At this point, you would translate the score to a race time, but
# I don't know the specifics of that, so I'll leave it up to you
# Display this horse's name, owner, score, and pre_points
puts "#{horse_name},#{owners_initials},#{score},#{pre_points}"
end
-----END RUBY CODE-----
Given a file horses.csv that looks like this:
Horse A,J. R.,2,9
Horse B,J. D.,8,8
Horse C,D. D.,7,8
an example run of the above code (saved into the file horses and made
executable with chmod +x) might look as follows:
$ ./horses
Horse A,J. R.,31,9
Horse B,J. D.,35,8
Horse C,D. D.,43,8
Hope that helps,
Bob
--
"The nice thing about standards is that you have so many to choose from."
-- Andrew S. Tanenbaum
Robbert Haarman
shipped with Ruby. So, instead of using a regular expression to
implement a crude parser, we can use the csv module to do the work for
us. An updated program is below. The usage is the same as the previous
version.
-----BEGIN RUBY CODE-----
#! /usr/bin/env ruby
require 'csv'
# For each record in the CSV file...
CSV::Reader.parse(File.open('horses.csv')) do |record|
# Now we assign the appropriate fields of the record to
# variables named after the fields in the CSV file.
# Note that we need to call .to_i on the 3rd and 4rd fields
# to convert them from strings to integers.
horse_name = record[0]
owners_initials = record[1]
post = record[2].to_i
pre_points = record[3].to_i
# Roll the dice.
# This generates an array containing four die rolls,
# each die roll ranging from 1 to 10
dice = (1..4).map { rand(10) + 1 }
# Calculate score.
# The score is the sum of post, pre_points, and each of the die rolls
score = post + pre_points
dice.each { |n| score = score + n }
# At this point, you would translate the score to a race time, but
# I don't know the specifics of that, so I'll leave it up to you
# Display this horse's name, owner, score, and pre_points
puts CSV.generate_line([horse_name, owners_initials, score, pre_points])
end
-----END RUBY CODE-----
thomas...@gmx.at
wrote:
> Dnia 23-06-2007, So o godzinie 04:30 -0400, Rod Pemberton napisa³(a):
>
> > But, I do know that most dynamicly typed languages are written in C (at
> > least today they are...), i.e., Ruby, Python, Java, Lisp. And, many
> > staticly typed languages can be completely converted to C, e.g., Fortran,
> > Pascal, C++, etc.
>
> I haven't heard of a compiler of Java or C# which translates Java or C#
> programs to C. So in what sense "they have no more functionality than C
> does"?
Maybe this refers to things like
- closures (in C a closure is delimited with braces, eg.:
{ i++; print("hello there\n"); }
) as objects and parameters.
- local functions (functions defined localy in other functions)
- there are more (left as exercise)
which are not supported in C.
A programming language implementation which generates C needs to
code around this shortcomings.
For example: The Seed7 compiler (which generates C) uses
macro expansion to emulate closure parameters.
Greetings Thomas Mertes
Seed7 Homepage: http://seed7.sourceforge.net
Project page: http://sourceforge.net/projects/seed7
Steven Nichols
> I'm posting here because the comp.lang group seems to be effectively
> dead, and I'm looking for information on languages.
> That's my previous experience. What I'm looking for is suggestions on
> programming languages. To start, I'd like a language that I can use on
> both Linux and Windows XP, with a _free_ IDE/development kit/etc. Most
> important of all, I'd like it to be an effective language for a
> procedural programmer. I still don't grokk OO programming and I'm not
> really looking to start right now. I'd just like to make my life
> easier and work with my presently known style of programming. The only
> other quirk I can think of is that it's going to have to read
> information from CSV spreadsheet files, although that seems to be
> pretty commonly supported.
>
and Linux, ant it has an IDE.
Steve
James Harris
...
> > Bob, if not too long could you share it here?
>
> Alright. The code is below.
Thanks. It looks fairly easy to follow. The single line array
operations seem to be useful, like this one: