ICFP2004 results are in

ICFP2004 results are in

[Martin Dowie]

Sad to see not one Ada entry... :-(

http://www.cis.upenn.edu/proj/plclub/contest//results.php

I've stuck a 'reminder' in my desktop calender to keep an ear out for next 
years competition - anyone else up for it? It's obviously just for fun but 
it would be at least some publicity!

Cheers

-- Martin

Re: ICFP2004 results are in

[Cesar Rabak]

Martin Dowie escreveu:
 > Sad to see not one Ada entry... :-(
 >
 > http://www.cis.upenn.edu/proj/plclub/contest//results.php
 >
 > I've stuck a 'reminder' in my desktop calender to keep an ear out for
 > next years competition - anyone else up for it? It's obviously just
 > for fun but it would be at least some publicity!
 >
Martin,

This contest is geared towards functional languages (meaning dynamic�)
and rapid prototyping.

It would be interesting you spent a time trying to solve _this_ year 
problem with Ada and see if it would be worth the effort!

--
Cesar Rabak

[1] even untyped languages.

Re: ICFP2004 results are in

[Hyman Rosen]

Cesar Rabak wrote:
> This contest is geared towards functional languages (meaning dynamic�)
> and rapid prototyping.

Nevertheless, it attracted more entries in C++ than in any other language.
The problems tend to be ones that do not have clearly optimal solutions or
approaches. For the past two years, they have designed problems where the
entries compete against each other to accomplish a goal. This makes for a
rather objective competition. They also award subjective Judges' prizes.

Re: ICFP2004 results are in

[Jacob Sparre Andersen]

Martin Dowie wrote:

> Sad to see not one Ada entry... :-(

Yes.  A friend and I worked on something, but good weather, nice
beaches, and beautiful girls _are_ more important. :-)

The task was very interesting, and maybe I will play a bit more with
the problem just for the fun of it.  But not until the autumn comes to
Sardinia.

> I've stuck a 'reminder' in my desktop calender to keep an ear out
> for next years competition - anyone else up for it? It's obviously
> just for fun but it would be at least some publicity!

We are definitely two who are going to take a look at it again next
year.  My advice - based on our experience from the previous years -
is to make it a social event with everybody on the team collected in
the same location.

Jacob
-- 
"If you think Tuck has said something that is completely
 wrong, you are almost certainly missing something :-)"

Re: ICFP2004 results are in

[Martin Dowie]

Jacob Sparre Andersen wrote:
> Martin Dowie wrote:
>
>> Sad to see not one Ada entry... :-(
>
> Yes.  A friend and I worked on something, but good weather, nice
> beaches, and beautiful girls _are_ more important. :-)

indeed... :-)

...and you forgot 'beer'! :-)))

Re: ICFP2004 results are in

[Marius Amado Alves]

The *problems* are interesting, but IMO they have nothing to do with 
programming language. It is almost invariably an artificial intelligence 
problem. The success depends on the artificial intelligence method 
devised, not on the choice of programming language. (For the 2004 ants 
problem I think I would have bet on a genetic programming solution.)

Unfortunately the problems are the only interesting thing in these 
contests. The prize money is too low (specially after division by 5 or 
10 team members), the choice of programming language is a false target 
as I said, and the winners are called bad names ("discriminating 
hacker") go figure!

Re: ICFP2004 results are in

[Hyman Rosen]

Marius Amado Alves wrote:
> The *problems* are interesting, but IMO they have nothing to do with 
> programming language.

Why should they? Problems are problems. Anyway, Ada is intended to be
a general purpose language, and it would gain some notoriety if it were
used to create a successful entry.

One of the rigorous standards of the contest is that entries which crash
are immediately eliminated, so this should gave Ada a leg up, no?

 > It is almost invariably an artificial intelligence problem.

This is good, because it allows a broad range of implementation strategies.

Re: ICFP2004 results are in

[Martin Dowie]

Marius Amado Alves wrote:
> The *problems* are interesting, but IMO they have nothing to do with
> programming language. It is almost invariably an artificial
> intelligence problem. The success depends on the artificial
> intelligence method devised, not on the choice of programming
> language. (For the 2004 ants problem I think I would have bet on a
> genetic programming solution.)
>
> Unfortunately the problems are the only interesting thing in these
> contests. The prize money is too low (specially after division by 5 or
> 10 team members), the choice of programming language is a false target
> as I said, and the winners are called bad names ("discriminating
> hacker") go figure!

But - there's no such thing as bad publicity...

Re: ICFP2004 results are in

[stephane richard]

"Martin Dowie" <martin.dowie@baesystems.com> wrote in message 
news:415187d0_1@baen1673807.greenlnk.net...
>
> But - there's no such thing as bad publicity...
>
>

Indeed, as the old saying goes: "Talk about me, in good or in bad, but TALK 
ABOUT ME!" ;-) 

Re: ICFP2004 results are in

[Cesar Rabak]

Hyman Rosen escreveu:
> Cesar Rabak wrote:
> 
>> This contest is geared towards functional languages (meaning dynamic�)
>> and rapid prototyping.
> 
> 
> Nevertheless, it attracted more entries in C++ than in any other language.

Yes this is a fact: C++ and Java, but see below.

> The problems tend to be ones that do not have clearly optimal solutions or
> approaches. 

Yes, as a lot of interesting problems in AI are.

> For the past two years, they have designed problems where the
> entries compete against each other to accomplish a goal. This makes for a
> rather objective competition. They also award subjective Judges' prizes.

I agree with the 'objective competition' and the "subjective Judges'".

This just rises my point as a question: Does anyone has pointers or 
experience on the use of Ada in lieu of 'Rapid Prototyping' languages 
and/or extreme Programming?

--
Cesar Rabak

Re: ICFP2004 results are in

[Ed Falis]

On Wed, 22 Sep 2004 12:08:44 -0300, Cesar Rabak <crabak@acm.org> wrote:

> This just rises my point as a question: Does anyone has pointers or  
> experience on the use of Ada in lieu of 'Rapid Prototyping' languages  
> and/or extreme Programming?
>

The use of Ada is not in conflict with the Extreme Programming development  
approach.

- Ed

Re: ICFP2004 results are in

[Samuel Tardieu]

>>>>> "Ed" == Ed Falis <falis@verizon.net> writes:

Ed> The use of Ada is not in conflict with the Extreme Programming
Ed> development approach.

Well, compared to languages such as SmallTalk, Ada lacks a good
true refactoring editor for example, which is a must for agile
development and extreme programming.

  Sam
-- 
Samuel Tardieu -- sam@rfc1149.net -- http://www.rfc1149.net/sam

Re: ICFP2004 results are in

[Ed Falis]

On 22 Sep 2004 19:30:27 +0200, Samuel Tardieu <sam@rfc1149.net> wrote:

> Well, compared to languages such as SmallTalk, Ada lacks a good
> true refactoring editor for example, which is a must for agile
> development and extreme programming.

I disagree.  A refactoring editor is nice to have, but definitely not a  
must to use agile methods.  Remember, they didn't exist when XP was being  
developed.

- Ed

Re: ICFP2004 results are in

[Samuel Tardieu]

>>>>> "Ed" == Ed Falis <falis@verizon.net> writes:

Ed> A refactoring editor is nice to have, but definitely not a must to
Ed> use agile methods.  Remember, they didn't exist when XP was being
Ed> developed.

Wrong. The Smalltalk refactoring browser was created in mid 1994,
while extreme programming has been developped circa 1999 if I am not
mistaken (at the beginning of 2000, few books existed on the subject,
but the famous "White Book"). Moreover, Ward Cunningham and Kent Beck,
the fathers of XP, were Smalltalk programmers and used refactoring a
lot. So I'd really say a refactoring editor is really a must. Sure,
you can do without it, but the productivity will be highly
decreased.

I hope we'll see a refactoring editor in Ada one day, perhaps in GPS,
but this is not an easy thing to do.

  Sam
-- 
Samuel Tardieu -- sam@rfc1149.net -- http://www.rfc1149.net/sam

Re: ICFP2004 results are in

[Ed Falis]

On 22 Sep 2004 21:25:08 +0200, Samuel Tardieu <sam@rfc1149.net> wrote:

> Moreover, Ward Cunningham and Kent Beck,
> the fathers of XP, were Smalltalk programmers and used refactoring a
> lot.

Using refactoring and using a refactoring browser are two different  
things.  I never saw a requirement for a refactoring browser in any of  
Beck's early books on XP.  Having used many of the XP basic practices, I  
still don't think it's a requirement (and how many other languages had  
such browsers when people started applying XP and agile methods)?

Yes, it would be nice to have a refactoring browser for Ada, but it XP and  
agile methods can still be applied without such a magic bullet.

- Ed

Re: ICFP2004 results are in

[Jacob Sparre Andersen]

Martin Dowie wrote:
> Jacob Sparre Andersen wrote:
> > Martin Dowie wrote:

> >> Sad to see not one Ada entry... :-(
> >
> > Yes.  A friend and I worked on something, but good weather, nice
> > beaches, and beautiful girls _are_ more important. :-)
> 
> indeed... :-)
> 
> ...and you forgot 'beer'! :-)))

That's because I prefer wine. :-)

But a good stout (or two :-) is okay now and then.

Jacob
-- 
People in cars cause accidents. Accidents in cars cause people.

Re: ICFP2004 results are in

[Jacob Sparre Andersen]

Hyman Rosen wrote:
> Marius Amado Alves wrote:

> > The *problems* are interesting, but IMO they have nothing to do
> > with programming language.
> 
> Why should they?

It is implicit (or maybe actually explicit) in the contest that it
favours functional programming languages.

> Problems are problems. Anyway, Ada is intended to be a general
> purpose language, and it would gain some notoriety if it were used
> to create a successful entry.

I don't think it matters that much which language was used for the
winning entries.

> One of the rigorous standards of the contest is that entries which
> crash are immediately eliminated, so this should gave Ada a leg up,
> no?

Yes.

>  > It is almost invariably an artificial intelligence problem.
> 
> This is good, because it allows a broad range of implementation
> strategies.

Yes.  I wouldn't quite call the problems AI-like, but they are
definitely interesting and possible to solve in many different ways.

Jacob
-- 
"I don't want to gain immortality in my works.
 I want to gain it by not dying."

Re: ICFP2004 results are in

[Hyman Rosen]

Jacob Sparre Andersen wrote:
> It is implicit (or maybe actually explicit) in the contest that it
> favours functional programming languages.

Not unlike this n.g., I think they believe that *all* contests
favor functional programming languages :-) In the last few years
we've had a ray-tracing problem, a robots wars problem, an auto
race simulation, and now battling ant colonies. I don't see any
a priori reason that these favor FP.

> I don't think it matters that much which language was used for the
> winning entries.

I know I'm happy when C++ does well.

Re: ICFP2004 results are in

[Stephen Leake]

Samuel Tardieu <sam@rfc1149.net> writes:

> >>>>> "Ed" == Ed Falis <falis@verizon.net> writes:
> 
> Ed> The use of Ada is not in conflict with the Extreme Programming
> Ed> development approach.
> 
> Well, compared to languages such as SmallTalk, Ada lacks a good
> true refactoring editor for example, which is a must for agile
> development and extreme programming.

What are the features of a "refactoring editor" that Emacs + GNAT
lack? 

Is there a good summary/intro to a "refactoring editor" somewhere?

-- 
-- Stephe

Re: ICFP2004 results are in

[Rod Chapman]

"Ed Falis" <falis@verizon.net> wrote in message news:
> I disagree.  A refactoring editor is nice to have, but definitely not a  
> must to use agile methods.  Remember, they didn't exist when XP was being  
> developed.

FWIW, I'm here in Pittsburgh taking the SEI's PSP I course, which
has daily programming assignments that must be completed.  Pretty much
everyone is using jazzy "modern" environments like Java, VB.Net, and
VC++.  I'm using SPARK, EMACS/GLIDE, SPARK Examiner, and GNAT Pro.

Needless to say, I'm holding my own so far... :-)
 - Rod

Re: ICFP2004 results are in

[Ed Falis]

On 22 Sep 2004 19:13:34 -0400, Stephen Leake <stephen_leake@acm.org> wrote:

> What are the features of a "refactoring editor" that Emacs + GNAT
> lack?
>
> Is there a good summary/intro to a "refactoring editor" somewhere?
>


The idea behind a refactoring editor is that it automates ways you might  
want to reorganize code to make it cleaner, or more flexible.  This would  
include moving declarations, combining common code fragments, and similar  
transformations on your code base.  It's really a great kind of tool in  
that it assists developers in common tasks while helping to avoid missing  
changes that need to be made for such an operation to be safe.

A good reference on refactoring is Martin Fowler's "Refactoring"  ;-)

It goes into the philosophy behind it, enumerates common kinds of  
refactorings, and discusses tool support, as well as providing a number of  
good references.

Like a like of the components of "agile methods", the subject material is  
about something all of us do after some experience developing software,  
but the book reframes its importance and provides a lot of guidance about  
how to do it effectively.


- Ed

Re: ICFP2004 results are in

[Ed Falis]

Rod,

I'm sure you know I have nothing against modern, jazzy environments (I  
still think VisualAge was about the best I've ever seen).  My difference  
with Sam was just one of necessity for being able to use agile methods vs  
desirability.

- Ed

Re: ICFP2004 results are in

[Martin Krischik]

Ed Falis wrote:

>  (I
> still think VisualAge was about the best I've ever seen)

Strange that none ever made it to implement something just as cool.

BTW: Did you mean VisualAge for Smalltalk, VisualAge for C++ or VisualAge
for Java. Actually they where all cool. AFAIK only one of them is still
supported (not considering special contracts).

With Regards

Martin

-- 
mailto://krischik@users.sourceforge.net
http://www.ada.krischik.com

Re: ICFP2004 results are in

[Ed Falis]

On Thu, 23 Sep 2004 10:35:19 +0200, Martin Krischik  
<krischik@users.sourceforge.net> wrote:

> BTW: Did you mean VisualAge for Smalltalk, VisualAge for C++ or VisualAge
> for Java. Actually they where all cool. AFAIK only one of them is still
> supported (not considering special contracts).

Most of my exposure was to VA for Java, though I played a bit with the  
Smalltalk version on eval.  I've heard that VAJ got rolled into Web Sphere.

- Ed

Re: ICFP2004 results are in

[Samuel Tardieu]

>>>>> "Stephen" == Stephen Leake <stephen_leake@acm.org> writes:

Stephen> What are the features of a "refactoring editor" that Emacs +
Stephen> GNAT lack?

Ed made a very interesting answer, so I won't repeat it. Here are two
concrete examples of things I often do in Smalltalk:

  - select a piece of code in a method, choose "make new method" --
    the piece of code is extracted, moved into a new method with the
    appropriate signature, temporary variables are put in the
    appropriate place, original location is replaced by a call,
    similar piece of codes in other methods are replaced (after user
    validation) by a call

  - move a method up or down the hierarchy -- if it is up, remove
    unnecessary redundant implementation in subclasses

 Sam
-- 
Samuel Tardieu -- sam@rfc1149.net -- http://www.rfc1149.net/sam

Re: ICFP2004 results are in

[Stephen Leake]

"Ed Falis" <falis@verizon.net> writes:

> On 22 Sep 2004 19:13:34 -0400, Stephen Leake <stephen_leake@acm.org> wrote:
> 
> > What are the features of a "refactoring editor" that Emacs + GNAT
> > lack?
> >
> > Is there a good summary/intro to a "refactoring editor" somewhere?
> >
> 
> 
> The idea behind a refactoring editor is that it automates ways you
> might  want to reorganize code to make it cleaner, or more flexible.
> This would  include moving declarations, combining common code
> fragments, and similar  transformations on your code base.  It's
> really a great kind of tool in  that it assists developers in common
> tasks while helping to avoid missing  changes that need to be made for
> such an operation to be safe.
> 
> A good reference on refactoring is Martin Fowler's "Refactoring"  ;-)

Ok. I'm familiar with the concept of refactoring, but I had not
considered how a tool could help do it.

> It goes into the philosophy behind it, enumerates common kinds of
> refactorings, and discusses tool support, as well as providing a
> number of good references.
> 
> Like a like of the components of "agile methods", the subject material
> is  about something all of us do after some experience developing
> software,  but the book reframes its importance and provides a lot of
> guidance about  how to do it effectively.

Right. I'll put it on my list of things to read.

-- 
-- Stephe

Re: ICFP2004 results are in

[Stephen Leake]

Samuel Tardieu <sam@rfc1149.net> writes:

> >>>>> "Stephen" == Stephen Leake <stephen_leake@acm.org> writes:
> 
> Stephen> What are the features of a "refactoring editor" that Emacs +
> Stephen> GNAT lack?
> 
> Ed made a very interesting answer, so I won't repeat it. Here are two
> concrete examples of things I often do in Smalltalk:
> 
>   - select a piece of code in a method, choose "make new method" --
>     the piece of code is extracted, moved into a new method with the
>     appropriate signature, temporary variables are put in the
>     appropriate place, original location is replaced by a call,
>     similar piece of codes in other methods are replaced (after user
>     validation) by a call

Interesting. The editor clearly has to know the syntax of the
language, and some of the semantics.

I can see using ASIS to get there with Ada, but it would have to run
fast to be acceptably interactive.

>   - move a method up or down the hierarchy -- if it is up, remove
>   unnecessary redundant implementation in subclasses

I do this on occasion, and GNAT doesn't provide much help in
identifying what needs to be removed. I think the "require override"
feature in Ada 05 will help here.

-- 
-- Stephe

Re: ICFP2004 results are in

[Ed Falis]

The whole idea of a refactoring browser for Ada is exciting.  Hopefully we  
can do something along these lines in the not too distant future.

- Ed

ICFP2004 results are in

[Andrew]

> ------------------------------
> Message: 7
> Date: Tue, 21 Sep 2004 20:19:41 +0000 (UTC)
> From: "Martin Dowie" <martin.dowie@btopenworld.com>
> Subject: ICFP2004 results are in
>
> Sad to see not one Ada entry... :-(
>
> http://www.cis.upenn.edu/proj/plclub/contest//results.php
>
> I've stuck a 'reminder' in my desktop calender to keep an ear out for
next
> years competition - anyone else up for it? It's obviously just for fun
but
> it would be at least some publicity!
>
> Cheers
>
> -- Martin
>

I am interested.

Andrew Carroll
Carroll-Tech
720-273-6814
andrew@carroll-tech.net