From extended Pascals to Ada 95 guide

From extended Pascals to Ada 95 guide

[gdemont]

Hi all. Here is new version of my "From extended Pascals to Ada 95"
guide started a while ago. There might be unaccuracies and surely
plenty of english mistakes. Don't hesitate to protest!

URL:  http://members.xoom.com/_XMCM/gdemont/pascada.htm

Have fun...

Gautier


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: From extended Pascals to Ada 95 guide

[James Smith]

Too bad the US gov didn't save some taxpayer dollars by just adopting Modula
2. Of course that would have made too much sense.

James

<gdemont@my-deja.com> wrote in message news:8o3s2a$9ph$1@nnrp1.deja.com...
> Hi all. Here is new version of my "From extended Pascals to Ada 95"
> guide started a while ago. There might be unaccuracies and surely
> plenty of english mistakes. Don't hesitate to protest!
>
> URL:  http://members.xoom.com/_XMCM/gdemont/pascada.htm
>
> Have fun...
>
> Gautier
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.

Re: From extended Pascals to Ada 95 guide

[Gautier]

James Smith:

> Too bad the US gov didn't save some taxpayer dollars by just adopting Modula
> 2. Of course that would have made too much sense.

To save dollars they also could have adopted C...
With Modula-2, they should have spent perhaps more than Ada to make it usable...

The main problem with Modula-2 is that it keeps the rigid types
of "classic" Pascal. No problem for teaching or write a
"Pascal in Pascal" compiler, but for the real world you need
a flexible typing like "array(integer range <>) of..."

The second problem is the modularity: you have to open manually
the visibility for *all* identifiers you need! It made things
easier for writing a compiler, but it rapidily took longer to
write and maintain the "FROM...IMPORT..."s, with correct casing,
than to write your programs. Just extrapolate the "Hello World"
to have an idea...

MODULE PrintHelloWorld;

FROM InOut IMPORT WriteString, WriteLn;

BEGIN
        WriteString('Hello world!');
        WriteLn;
END PrintHelloWorld.

Another big problem was the library: there were vague recommendations
for text I/O, and iirc, nothing more. As a result, even "Hello World"
was non portable: some compilers wanted "WriteString('Hello world!');",
others wanted "WRITESTRING('Hello world!');" I let you guess what the
mess was with Math libraries...

Finally I doubt that the average US programmer would have been
patient enough to stand more than 5 minutes before Modula-2...
Already in Europe many keyboards were severely damaged!

______________________________________________________
Gautier  --  http://members.xoom.com/gdemont/gsoft.htm

Re: From extended Pascals to Ada 95 guide

[Tarjei T. Jensen]

James Smith wrote in message <8o4bfq$v0h$1@slb7.atl.mindspring.net>...
>Too bad the US gov didn't save some taxpayer dollars by just adopting Modula
>2. Of course that would have made too much sense.


Don't agree. As far as I'm concerned, after Pascal it was downhill for Wirth.

Ada is a lot better language. The standard library leaves a lot to be desired,
but the language itself is as good as it gets. You would probably agree if you
have used it.

Reading about it on the web and swearing at the use of with at the top and use
of capials in the code does not constitute use.


Gretings,

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

In article <39A655BE.18E89020@maths.unine.ch>, Gautier wrote:
>James Smith:
>
>> Too bad the US gov didn't save some taxpayer dollars by just adopting Modula
>> 2. Of course that would have made too much sense.
>
>To save dollars they also could have adopted C...

Then the American subs would lie on the bottom of the ocean too I think:-)

C is simply known to introduce more bugs.

>With Modula-2, they should have spent perhaps more than Ada to make it usable...

Can't judge that, don't know ADA enough.

>The main problem with Modula-2 is that it keeps the rigid types
>of "classic" Pascal. No problem for teaching or write a
>"Pascal in Pascal" compiler, but for the real world you need
>a flexible typing like "array(integer range <>) of..."

That type safety avoids bugs. In the real world where you whack a GUI for
some users that is less of a problem. 
In subs, you don't want that.

>The second problem is the modularity: you have to open manually
>the visibility for *all* identifiers you need!

Or use them qualified. Again a very good principle, which I miss dearly from
pascal, introduced again to avoid bugs.

> It made things
>easier for writing a compiler, but it rapidily took longer to
>write and maintain the "FROM...IMPORT..."s, with correct casing,
>than to write your programs. Just extrapolate the "Hello World"
>to have an idea...

>MODULE PrintHelloWorld;
>
>FROM InOut IMPORT WriteString, WriteLn;
>
>BEGIN
>        WriteString('Hello world!');
>        WriteLn;
>END PrintHelloWorld.

I don't see any problem?

>Another big problem was the library: there were vague recommendations
>for text I/O, and iirc, nothing more. As a result, even "Hello World"
>was non portable: some compilers wanted "WriteString('Hello world!');",
>others wanted "WRITESTRING('Hello world!');" I let you guess what the
>mess was with Math libraries...

Also not a problem when an uniform system througout the military is used.

>Finally I doubt that the average US programmer would have been
>patient enough to stand more than 5 minutes before Modula-2...

If they get confused by this,  then they shouldn't be programming critical 
applications at all.

Re: From extended Pascals to Ada 95 guide

[Gautier]

> >The main problem with Modula-2 is that it keeps the rigid types
> >of "classic" Pascal. No problem for teaching or write a
> >"Pascal in Pascal" compiler, but for the real world you need
> >a flexible typing like "array(integer range <>) of..."

Marco:

> That type safety avoids bugs. In the real world where you whack a GUI for
> some users that is less of a problem. In subs, you don't want that.

Do you assume that Ada's "type vector is array(integer range <>) of float"
is not type safe ? Then this is a misunderstanding! Maybe you think it
is a sort of *float � la C ? Not at all: simply you declare variables
v: vector(1..100) and so on; you can pass them into "+"(v1,v2: vector) functions
and you have the range with attributes v1'first, v1'last, v1'range,...
To be even more reassured, use subtypes: "subtype matrix33 is matrix(1..3,1..3);"
It's that simple: type safety of Pascal, flexibility of C/Fortran!

As for submarines, you're right. Since C/C++ is said to replace Ada in US embedded
systems.. glups...

______________________________________________________
Gautier  --  http://members.xoom.com/gdemont/gsoft.htm

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

>Marco:
>
>> That type safety avoids bugs. In the real world where you whack a GUI for
>> some users that is less of a problem. In subs, you don't want that.
>
>Do you assume that Ada's "type vector is array(integer range <>) of float"
>is not type safe ?

I probably misread then. I assymed you were comparing with C. (instead of
Modula or Ada)

 Then this is a misunderstanding! Maybe you think it
>is a sort of *float � la C ? Not at all: simply you declare variables
>v: vector(1..100) and so on; you can pass them into "+"(v1,v2: vector) functions
>and you have the range with attributes v1'first, v1'last, v1'range,...
>To be even more reassured, use subtypes: "subtype matrix33 is matrix(1..3,1..3);"
>It's that simple: type safety of Pascal, flexibility of C/Fortran!

Sounds like Delphi style dynamic arrays. 

>As for submarines, you're right. Since C/C++ is said to replace Ada in US embedded
>systems.. glups...

Re: From extended Pascals to Ada 95 guide

[Pat Rogers]

"Preben Randhol" <randhol+abuse@pvv.org> wrote in message
news:slrn8qd0gn.bjs.randhol+abuse@kiuk0156.chembio.ntnu.no...
> I don't think a big project written in C would be very cost
efficient
> compared to Ada. Thus perhaps saving money in the short run (by not
> developing Ada), but not in the long run at all.

Some baggage never gets lost...

This idea that development in Ada is more expensive than in other
languages must be challenged whenever we come across it.  The tool
costs can be very reasonable and in my experience (and others' as
well) programmer productivity can be extremely high indeed.

All other things being equal, in a contest between highly-skilled Ada
programmers and highly-skilled C programmers, I'll bet on the Ada
people to produce the final code faster.

pat
---
Patrick Rogers                      Consulting and Training in:
http://www.classwide.com      Deadline Schedulability Analysis
progers@classwide.com        Software Fault Tolerance
(281)648-3165                       Real-Time/OO Languages

Adam ... does not deserve all the credit; much is due to Eve, the
first woman, and Satan, the first consultant.
Mark Twain

Re: From extended Pascals to Ada 95 guide

[Charles Hixson]

[-- Attachment #1: Type: text/plain, Size: 926 bytes --]

Marco van de Voort wrote:

> ... Then this is a misunderstanding! Maybe you think it
> >is a sort of *float � la C ? Not at all: simply you declare variables
> >v: vector(1..100) and so on; you can pass them into "+"(v1,v2: vector) functions
> >and you have the range with attributes v1'first, v1'last, v1'range,...
> >To be even more reassured, use subtypes: "subtype matrix33 is matrix(1..3,1..3);"
> >It's that simple: type safety of Pascal, flexibility of C/Fortran!
>
> Sounds like Delphi style dynamic arrays.
>
> >As for submarines, you're right. Since C/C++ is said to replace Ada in US embedded
> >systems.. glups...

I don't know Delphi, but that's a static (as in not changing characteristics at run
time) array.  There are also various more dynamic forms (e.g., unbounded strings), but
that one was static.

-- (c) Charles Hixson
--  Addition of advertisements or hyperlinks to products specifically prohibited


[-- Attachment #2: Card for Charles Hixson --]
[-- Type: text/x-vcard, Size: 145 bytes --]

begin:vcard 
n:Hixson;Charles
x-mozilla-html:FALSE
adr:;;;;;;
version:2.1
email;internet:charleshixson@earthling.net
fn:Charles Hixson
end:vcard

Re: From extended Pascals to Ada 95 guide

[Pat Rogers]

"Marin D. Condic" <mcondic-nospam@acm.org> wrote in message
news:39A6B3FF.73538A0E@acm.org...
> Pat Rogers wrote:
> > Some baggage never gets lost...
> >
> > This idea that development in Ada is more expensive than in other
> > languages must be challenged whenever we come across it.  The tool
> > costs can be very reasonable and in my experience (and others' as
> > well) programmer productivity can be extremely high indeed.
> >
> I would agree, but with a qualification. In some domains with some
> development environments, you get lots of prepackaged, well
integrated
> services. The language itself (Ada) is going to be
faster/better/cheaper
> to develop in than (say) C++ or some other popular languages because
of
> ease of understanding, extensive checking to avoid bugs, easier
> debugging, easier configuration management, etc. However, its hard
to
> compete with something like Microsoft Visual C++ for PC app
development
> simply because of the body of code leveraged through the MFC and the
> really spiffy, well integrated IDE. While similar tools are
available
> with Ada to some extent, you don't get the whole thing in one nice
kit,
> so you'll lose time in pulling the tools together, integrating them,
> figuring out how to use them, etc. For some domains you may not have
> these tools at all.

Yes, that's the quentissential example of "All other things being
equal" that I had in mind.

> Granted, this is not a "language" issue, but more of a "development
> environment" issue. Some other language may be faster to develop in
> simply because of the availability of the whole environment - not
> because of the language itself.
>
> > All other things being equal, in a contest between highly-skilled
Ada
> > programmers and highly-skilled C programmers, I'll bet on the Ada
> > people to produce the final code faster.
> >
> I'd bet the same way. There is a strong body of evidence to support
> this. But the "All other things being equal" qualification is a big
> issue.

True, but I work in embedded/real-time systems for the most part (as
do you, if memory serves), and my passions run accordingly.  One could
argue that today's "desktop language" might become tomorrow's
"embedded systems language", and then I should care a great deal more!
That's happening with C++ now, and will happen with Java eventually.
Neither are technologically supportable substitutions for Ada 95,
IMHO, but technology doesn't drive the decisions much.

We're thinking along the same lines.

pat

---
Patrick Rogers                      Consulting and Training in:
http://www.classwide.com      Deadline Schedulability Analysis
progers@classwide.com        Software Fault Tolerance
(281)648-3165                       Real-Time/OO Languages

Adam ... does not deserve all the credit; much is due to Eve, the
first woman, and Satan, the first consultant.
Mark Twain

Re: From extended Pascals to Ada 95 guide

[Larry Elmore]

"Marin D. Condic" <mcondic-nospam@acm.org> wrote in message
news:39A6B3FF.73538A0E@acm.org...
> >
> I would agree, but with a qualification. In some domains with some
> development environments, you get lots of prepackaged, well integrated
> services. The language itself (Ada) is going to be faster/better/cheaper
> to develop in than (say) C++ or some other popular languages because of
> ease of understanding, extensive checking to avoid bugs, easier
> debugging, easier configuration management, etc. However, its hard to
> compete with something like Microsoft Visual C++ for PC app development
> simply because of the body of code leveraged through the MFC and the
> really spiffy, well integrated IDE. While similar tools are available
> with Ada to some extent, you don't get the whole thing in one nice kit,
> so you'll lose time in pulling the tools together, integrating them,
> figuring out how to use them, etc. For some domains you may not have
> these tools at all.
>
> Granted, this is not a "language" issue, but more of a "development
> environment" issue. Some other language may be faster to develop in
> simply because of the availability of the whole environment - not
> because of the language itself.

I haven't used MS Visual C++ much (just worked my way through the tutorial
apps), and I haven't used Delphi since ver.2, but I thought Delphi was a
much better environment to work in than MFC, and I liked Object Pascal a lot
better than I've ever liked C++. That said, I just wish Delphi used Ada 95
instead of Object Pascal. _That_ would've been the best of all possible
worlds and I'd almost certainly still be using it today.

Larry

Re: From extended Pascals to Ada 95 guide

[Robert Deininger]

On Thu, Aug 24, 2000 7:51 PM, James Smith <jksmithiii@mindspring.com>
wrote:
>Too bad the US gov didn't save some taxpayer dollars by just adopting
Modula
>2. Of course that would have made too much sense.
>
>James

Were you ever a Reverend?  You sort of remind me of one I used to know.

---------------------------
Robert Deininger
rdeininger@mindspring.com

Re: From extended Pascals to Ada 95 guide

[Preben Randhol]

On Fri, 25 Aug 2000 13:17:19 +0200, Gautier wrote:
>James Smith:
>
>> Too bad the US gov didn't save some taxpayer dollars by just adopting Modula
>> 2. Of course that would have made too much sense.
>
>To save dollars they also could have adopted C...

I don't think a big project written in C would be very cost efficient
compared to Ada. Thus perhaps saving money in the short run (by not
developing Ada), but not in the long run at all.

-- 
Preben Randhol - Ph.D student - http://www.pvv.org/~randhol/
"i too once thought that when proved wrong that i lost somehow"
                               - i was hoping, alanis morisette

Re: From extended Pascals to Ada 95 guide

[Marin D. Condic]

Pat Rogers wrote:
> Some baggage never gets lost...
> 
> This idea that development in Ada is more expensive than in other
> languages must be challenged whenever we come across it.  The tool
> costs can be very reasonable and in my experience (and others' as
> well) programmer productivity can be extremely high indeed.
> 
I would agree, but with a qualification. In some domains with some
development environments, you get lots of prepackaged, well integrated
services. The language itself (Ada) is going to be faster/better/cheaper
to develop in than (say) C++ or some other popular languages because of
ease of understanding, extensive checking to avoid bugs, easier
debugging, easier configuration management, etc. However, its hard to
compete with something like Microsoft Visual C++ for PC app development
simply because of the body of code leveraged through the MFC and the
really spiffy, well integrated IDE. While similar tools are available
with Ada to some extent, you don't get the whole thing in one nice kit,
so you'll lose time in pulling the tools together, integrating them,
figuring out how to use them, etc. For some domains you may not have
these tools at all.

Granted, this is not a "language" issue, but more of a "development
environment" issue. Some other language may be faster to develop in
simply because of the availability of the whole environment - not
because of the language itself.


> All other things being equal, in a contest between highly-skilled Ada
> programmers and highly-skilled C programmers, I'll bet on the Ada
> people to produce the final code faster.
> 
I'd bet the same way. There is a strong body of evidence to support
this. But the "All other things being equal" qualification is a big
issue.

MDC
-- 
======================================================================
Marin David Condic - Quadrus Corporation - http://www.quadruscorp.com/
Send Replies To: m c o n d i c @ q u a d r u s c o r p . c o m
Visit my web site at:  http://www.mcondic.com/

    "Take away the punchbowl just when the party gets going" 

        --  William McChesney Martin, Former Fed chairman, explaining 
            what a sound central bank must always do. 
======================================================================

Re: From extended Pascals to Ada 95 guide

[steve]

In article <39A6ABD9.634308AA@earthlink.net>, Charles says...
 
>I don't know Delphi,
 
Delphi is Object pascal. It is much more modern language than
classical pascal, it is actually closer to Ada. It has
threads, it is single inheritance(sp?), multiple interfaces
(like Java), has 'class' construct, constructor, private, public,
units (packages?) and many more things. I am not sure if it is
an ISO or ANSI standard language though.

Borland are porting Delphi to Linux now, and it will be released
this year, they call it 'Kylix' on Linux. The nice thing about it,
is that all those current Windows delphi application can now very easily
be ported to Linux by using Kylix to compile the window delphi source 
code with Linux/Kylix. This is expected to increase the amount of good
commerical application on Linux in short time, and will help make Linux
more popular as a desktop platform as well.

steve

 

Re: From extended Pascals to Ada 95 guide

[Dimmy Timchenko]

"Larry Elmore" <lj.elmore@gte.net> wrote in message
news:2IBp5.1573$OE.204952@paloalto-snr1.gtei.net...

> I just wish Delphi used Ada 95  instead of Object Pascal. _That_ would've
been the
> best of all possible  worlds and I'd almost certainly still be using it
today.

It is my dream also. :)  But Ada is hard to implement, and Borland/Inprise
already have their
sector of market.  But, using Object Pascal, I often think: "but how nice it
could be made
in Ada!" :)   In fact, OP is rather weak and not as logical and strict, as
Ada.  Compared
with Ada it reminds C/C++. :)

Dimmy.

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

>classical pascal, it is actually closer to Ada. It has
>threads, it is single inheritance(sp?), multiple interfaces
>(like Java), has 'class' construct, constructor, private, public,
>units (packages?) and many more things. I am not sure if it is
>an ISO or ANSI standard language though.

No it isn't. That is probably the *reason* it is successfull:-) Somehow
standards never caught on under the bulk of Pascal users.
Most use a kind of UCSD dialect (specially the Borland dialects)

>Borland are porting Delphi to Linux now, and it will be released
>this year, they call it 'Kylix' on Linux.

I also thought this, but I read an article about this (from Jeroen Pluimers
last night), and he seemed to suggest that "Kylix" is the codename for the 
linux-port project, and not necessarily for the release product.
But that could be me of course.

>The nice thing about it,
>is that all those current Windows delphi application can now very easily
>be ported to Linux by using Kylix to compile the window delphi source 
>code with Linux/Kylix.

Yup. Only BDE seems to be a prob.

 This is expected to increase the amount of good
>commerical application on Linux in short time, and will help make Linux
>more popular as a desktop platform as well.

Why desktop? Personally I see Linux as a server OS, so the primary target
for Kylix to me seems to be configuration tools of server apps, and
the server apps themselves.

Re: From extended Pascals to Ada 95 guide

[Marin D. Condic]

Pat Rogers wrote:
> True, but I work in embedded/real-time systems for the most part (as
> do you, if memory serves), and my passions run accordingly.  One could
> argue that today's "desktop language" might become tomorrow's
> "embedded systems language", and then I should care a great deal more!
> That's happening with C++ now, and will happen with Java eventually.
> Neither are technologically supportable substitutions for Ada 95,
> IMHO, but technology doesn't drive the decisions much.
> 
> We're thinking along the same lines.
> 
Great minds do think alike, don't they? :-)

Yes, I have done considerable embedded/realtime work where you typically
don't find development kits with tons of pre-packaged software to do
most of your job. When you're developing most/all of the code going into
some app from bottom-dead-center, Ada is going to win out over a number
of other languages.

Of course in the embedded/realtime arena, there can still be issues
concerning development environments. The first being can you get a
compiler for the target machine at all? Sometimes Ada loses out in this
domain because of lack of a compiler or, if a compiler exists, there is
a lack of the surrounding tools needed for embedded programming. Often C
wins for lack of a competitor!

MDC
-- 
======================================================================
Marin David Condic - Quadrus Corporation - http://www.quadruscorp.com/
Send Replies To: m c o n d i c @ q u a d r u s c o r p . c o m
Visit my web site at:  http://www.mcondic.com/

    "Take away the punchbowl just when the party gets going" 

        --  William McChesney Martin, Former Fed chairman, explaining 
            what a sound central bank must always do. 
======================================================================

Re: From extended Pascals to Ada 95 guide

[Marin D. Condic]

Larry Elmore wrote:
> 
> I haven't used MS Visual C++ much (just worked my way through the tutorial
> apps), and I haven't used Delphi since ver.2, but I thought Delphi was a
> much better environment to work in than MFC, and I liked Object Pascal a lot
> better than I've ever liked C++. That said, I just wish Delphi used Ada 95
> instead of Object Pascal. _That_ would've been the best of all possible
> worlds and I'd almost certainly still be using it today.
> 
I'm afraid I've never used Delphi, so I can't compare. I have been
building up experience with MSVC++ for some time now and, while I still
find the Win32api (and hence the MFC built on top of it) to be
"Organically Grown" rather than designed, I can find reasons to like it.
If you know the MFC, you get a lot of work done for you right there.
Using the MSVC++ development environment wizards to get the core of an
app built is also a lot of leverage. I like the organization of the
various packages into "Classes" where you can easily see the pieces.
(I'd like to see an editor for Ada that let you do the same - without
imposing this as mandatory, mind you!) The debugging facilities are very
impressive and very dynamic - allowing code to be modified on the fly,
etc. Naturally, because its C++, you really *need* the debugger to be
great! :-)

Maybe Delphi is better, but I wouldn't know. What features/capabilities
does it have that you think makes it superior?

I'll say this: If there was a fully integrated development kit similar
to MSVC++ for PC development using Ada, it would be one heck of a
competitor. Perhaps something like Claw merged with Adagide and spiffed
up to be more feature-rich. I know that the only reason we're using C++
for this development is because any attempt to use Ada would have meant
a huge time-to-market lag. All the downside costs of C++ become
irrelevant in comparison.

MDC
-- 
======================================================================
Marin David Condic - Quadrus Corporation - http://www.quadruscorp.com/
Send Replies To: m c o n d i c @ q u a d r u s c o r p . c o m
Visit my web site at:  http://www.mcondic.com/

    "Take away the punchbowl just when the party gets going" 

        --  William McChesney Martin, Former Fed chairman, explaining 
            what a sound central bank must always do. 
======================================================================

RE: From extended Pascals to Ada 95 guide

[Robert C. Leif, Ph.D.]

From: Bob Leif
To: Gauter, James Smith et al.

I agree on Modula. However, Wirth et al. were able to create a very elegant
operating system with Oberon. The sources in Project Oberon appear to be
more Ada like. However, I am not a language expert. I believe that persons
who propose to develop an operating system in Ada should look at the Oberon
operating system. Wirth was correct in building an operating system with
useful tools. In retrospect that is what should have been done with Ada.

-----Original Message-----
From: comp.lang.ada-admin@ada.eu.org
[mailto:comp.lang.ada-admin@ada.eu.org]On Behalf Of Gautier
Sent: Friday, August 25, 2000 4:17 AM
To: comp.lang.ada@ada.eu.org
Subject: Re: From extended Pascals to Ada 95 guide


James Smith:

> Too bad the US gov didn't save some taxpayer dollars by just adopting
Modula
> 2. Of course that would have made too much sense.

To save dollars they also could have adopted C...
With Modula-2, they should have spent perhaps more than Ada to make it
usable...

The main problem with Modula-2 is that it keeps the rigid types
of "classic" Pascal. No problem for teaching or write a
"Pascal in Pascal" compiler, but for the real world you need
a flexible typing like "array(integer range <>) of..."

The second problem is the modularity: you have to open manually
the visibility for *all* identifiers you need! It made things
easier for writing a compiler, but it rapidily took longer to
write and maintain the "FROM...IMPORT..."s, with correct casing,
than to write your programs. Just extrapolate the "Hello World"
to have an idea...

MODULE PrintHelloWorld;

FROM InOut IMPORT WriteString, WriteLn;

BEGIN
        WriteString('Hello world!');
        WriteLn;
END PrintHelloWorld.

Another big problem was the library: there were vague recommendations
for text I/O, and iirc, nothing more. As a result, even "Hello World"
was non portable: some compilers wanted "WriteString('Hello world!');",
others wanted "WRITESTRING('Hello world!');" I let you guess what the
mess was with Math libraries...

Finally I doubt that the average US programmer would have been
patient enough to stand more than 5 minutes before Modula-2...
Already in Europe many keyboards were severely damaged!

______________________________________________________
Gautier  --  http://members.xoom.com/gdemont/gsoft.htm

RE: From extended Pascals to Ada 95 guide

[Robert C. Leif, Ph.D.]

From: Bob Leif
To: Marin D. Condic et al.

You were correct concerning, "However, its hard to compete with something
like Microsoft Visual C++ for PC app development simply because of the body
of code leveraged through the MFC and the really spiffy, well integrated
IDE."

One simple solution is to use XML with the next version of Microsoft Visual
Studio and interface it with Ada. In fact the ARA should ask Microsoft for
permission to collaborate on creation of an interface for Ada to Visual
Studio. I suspect that since Microsoft makes money selling to the DoD,
Microsoft would accommodate the Ada community.

-----Original Message-----
From: comp.lang.ada-admin@ada.eu.org
[mailto:comp.lang.ada-admin@ada.eu.org]On Behalf Of Marin D. Condic
Sent: Friday, August 25, 2000 10:59 AM
To: comp.lang.ada@ada.eu.org
Subject: Re: From extended Pascals to Ada 95 guide


Pat Rogers wrote:
> Some baggage never gets lost...
>
> This idea that development in Ada is more expensive than in other
> languages must be challenged whenever we come across it.  The tool
> costs can be very reasonable and in my experience (and others' as
> well) programmer productivity can be extremely high indeed.
>
I would agree, but with a qualification. In some domains with some
development environments, you get lots of prepackaged, well integrated
services. The language itself (Ada) is going to be faster/better/cheaper
to develop in than (say) C++ or some other popular languages because of
ease of understanding, extensive checking to avoid bugs, easier
debugging, easier configuration management, etc. However, its hard to
compete with something like Microsoft Visual C++ for PC app development
simply because of the body of code leveraged through the MFC and the
really spiffy, well integrated IDE. While similar tools are available
with Ada to some extent, you don't get the whole thing in one nice kit,
so you'll lose time in pulling the tools together, integrating them,
figuring out how to use them, etc. For some domains you may not have
these tools at all.

Granted, this is not a "language" issue, but more of a "development
environment" issue. Some other language may be faster to develop in
simply because of the availability of the whole environment - not
because of the language itself.


> All other things being equal, in a contest between highly-skilled Ada
> programmers and highly-skilled C programmers, I'll bet on the Ada
> people to produce the final code faster.
>
I'd bet the same way. There is a strong body of evidence to support
this. But the "All other things being equal" qualification is a big
issue.

MDC
--
======================================================================
Marin David Condic - Quadrus Corporation - http://www.quadruscorp.com/
Send Replies To: m c o n d i c @ q u a d r u s c o r p . c o m
Visit my web site at:  http://www.mcondic.com/

    "Take away the punchbowl just when the party gets going"

        --  William McChesney Martin, Former Fed chairman, explaining
            what a sound central bank must always do.
======================================================================

Re: From extended Pascals to Ada 95 guide

[Ronald Cole]

"James Smith" <jksmithiii@mindspring.com> writes:
> Too bad the US gov didn't save some taxpayer dollars by just adopting Modula
> 2. Of course that would have made too much sense.

I don't think the GNU Modula-2 compiler was ever finished/released.
Last time I did Modula-2 was with Volition Systems' compiler for Apple
Pascal on a //e fifteen years ago.  Nowadays, Oberon looks like the
successful, minimalist, follow-up to Modula-2; but if you like
Wirth-style languages, then my money's on Modula-3 as an Ada-level
contender.

-- 
Forte International, P.O. Box 1412, Ridgecrest, CA  93556-1412
Ronald Cole <ronald@forte-intl.com>      Phone: (760) 499-9142
President, CEO                             Fax: (760) 499-9152
My GPG fingerprint: C3AF 4BE9 BEA6 F1C2 B084  4A88 8851 E6C8 69E3 B00B

Re: From extended Pascals to Ada 95 guide

[tmoran]

>I suspect that if you put the full version (and lic.) of CLAW, GLIDE, GNAT,
>and GNATCOM on a single CD with a common install you may be able to make a
>buck or two doing so.
   Actually, I think I paid something like $100 for my first version of
Gnat, which was on a CD.  But I notice the company I bought it from seems
to have disappeared now that folks' download speeds have increased.

>I wouldn't buy a copy since CLAW is not Open Source.
  The source code is of course included (it pretty much has to be since
it runs with multiple different compilers), but of course you mean you
would insist on the right to sell or give away a copy to Marin.  So I
could sell Marin, but not you, a copy, or, if it was "Open Source (tm)",
I could sell you a copy, but Marin probably wouldn't buy one from me
since he could get it for less from you.  Doesn't sound like such a good
idea after all.

Re: From extended Pascals to Ada 95 guide

[David Starner]

On 27 Aug 2000 13:23:29 -0700, Ronald Cole <ronald@forte-intl.com> wrote:
>I don't think the GNU Modula-2 compiler was ever finished/released.

Not if you mean the front end to gcc. There is a project currently being
worked on to make one that's pretty close to producing result though. There
are several free Modula-2 compilers out there, though.

>Pascal on a //e fifteen years ago.  Nowadays, Oberon looks like the
>successful, minimalist, follow-up to Modula-2; but if you like
>Wirth-style languages, then my money's on Modula-3 as an Ada-level
>contender.

Okay, why isn't Ada a "Wirth-style language", if Modula-3 is? They 
share more in common with each other than either does with Wirth's 
Pascal, for instance.

-- 
David Starner - dstarner98@aasaa.ofe.org
http/ftp: dvdeug.net.dhis.org
It was starting to rain on the night that they cried forever,
It was blinding with snow on the night that they screamed goodbye.
	- Dio, "Rock and Roll Children"

Re: From extended Pascals to Ada 95 guide

[Richard Kenner]

In article <m3g0nqlagu.fsf@yakisoba.forte-intl.com> Ronald Cole <ronald@forte-intl.com> writes:
>I don't think the GNU Modula-2 compiler was ever finished/released.

Interesting coincidence!  Just a day or two ago there was a posting from
the folks on that project saying they are trying to get it finished.

Re: From extended Pascals to Ada 95 guide

[Al Christians]

David Starner wrote:
> 
> Okay, why isn't Ada a "Wirth-style language", if Modula-3 is? They
> share more in common with each other than either does with Wirth's
> Pascal, for instance.
> 

The principal criterium is size of the language definition/reference 
manual.  Modula-3's is under 60 pages, IIRC, much smaller than Ada's.
And Wirth was some kind of advisor to the M3 design team. 

M2 and M3 are both very nice languages if you want something simpler
than Ada and don't need what they leave out.  But neither one shows
much sign of life.  If you wish you had more mainstream/affordable
commercial support for Ada, the situation is about ten times worse
working with either M2 or M3.  

All three of these languages, M2, M3, and Ada, now support generics, 
and that gives them a big advantage over Delphi for coding without doing 
aribtrary conversions between data types.  In Delphi, the generic type 
stored by the standard VCL collection classes is the Pointer, and the 
program must cast it to whatever type it really represents.  One big 
reason Borland's Pascal became popular was because it let you go
around Pascal's strict type checking, but that was years ago, and it's 
regrettable to have to routinely write code that will be checked 
for correctness at run time when there are less worrisome alternatives.

I don't believe that Wirth's versions of M2 or Pascal allowed any way
to go around the type system, and in M3, the modules that might do
so must be marked as unsafe.  Ada does allow unchecked conversions.   

Al

Re: From extended Pascals to Ada 95 guide

[David Botton]

<steve@nospam.ada.eu.org> wrote in message
news:8o7top$4sb@drn.newsguy.com...
> I am not sure if it is
> an ISO or ANSI standard language though.

It is not and its syntax and symantics change frequently between versions
(many times to the surprise of developers :-)

> The nice thing about it,
> is that all those current Windows delphi application can now very easily
> be ported to Linux by using Kylix to compile the window delphi source
> code with Linux/Kylix.

Provided the application doesn't use any OS specific bindings, ActiveX
controls, or any third party units. Also, given statements by Borland to not
expect 100% source code compatability don't expect the language to be 100%
compatable or even the GUI units.

On the other hand, an application developed using Ada 95 and GtkAda is 100%
source code compatable between Linux and Windows versions. Using the POSIX
package for NT and Linux and you have cross platform on most of the OS
specific stuff. Of course you can use ActiveX controls only on the Windows
side (Using my thick Win32 framework, GWindows, you can easily add ActiveX
controls to GTKAda on Win32, see my example on the GWindows page at
http://www.adapower.com/gwindows)

David Botton

Re: From extended Pascals to Ada 95 guide

[David Botton]

You should have contacted ACT. GLIDE combined with CLAW and GNATCOM at the
time would have given you the IDE (GLIDE), the GUI (CLAW) and the leverage
of all of MS's latest technologies (GNATCOM), and would have offered you
just as much (BTW, as a _very_ long term developer of MFC apps, you quickly
have to throw out the generated code to get most things done on large
projects :-). There are additional solutions available (and up and coming)
now, so on your next project, I wouldn't back down from Ada :-)

David Botton


"Marin D. Condic" <mcondic-nospam@acm.org> wrote in message
news:39A7C4B7.E6D655C1@acm.org...
> I'll say this: If there was a fully integrated development kit similar
> to MSVC++ for PC development using Ada, it would be one heck of a
> competitor. Perhaps something like Claw merged with Adagide and spiffed
> up to be more feature-rich. I know that the only reason we're using C++
> for this development is because any attempt to use Ada would have meant
> a huge time-to-market lag. All the downside costs of C++ become
> irrelevant in comparison.

Re: From extended Pascals to Ada 95 guide

[David Botton]

"Marin D. Condic" <mcondic-nospam@acm.org> wrote in message
news:39A93469.7C1237F3@acm.org...
> I still find it a bit of a roadblock that there isn't a shrink-wrap kit
> available that has everything bundled together with suitable
> documentation.

You can get all the components from one vendor with documentation and all.
You don't have to hunt and peck for pieces. GNAT, GtkAda, GNATCOM, and GLIDE
are all available from a single source. They are all integrated and work
from the same IDE, no roll your own here.

> With MSVC++, a lot of the problems go away. I insert a disk and install
> the environment. I fire up a single program and every tool is available
> from there. I have a question about anything? I highlight it and punch
> F1 and the documentation is right there. There are books available that
> can teach you how to use it.

Same goes for the above set of tools. Documentation available in the IDE,
etc.

> While I'm no big proponent of C++ and I'd *much* rather use Ada than
> anything else, I've got to consider the ramp-up time and time to market
> issues. This is a business which needs to make money and the "best" way
> of doing something is not always the most profitable. Because I've got
> future projects that will probably need to cross platforms, I've got
> good reasons to look at things like GtkAda and GNAT as potential tools.
> But it sure would be nice if I could get something in a single box that
> did basically what MSVC++ does. (It could be a good business opportunity
> in itself to integrate what is out there into a kit and see if it could
> sell. :-)

See above, but Delphi is a far better packaged solution then MSVC++ (or even
VB) would ever dream of being. Using MSVC++ (or VB) is only needed to be
politicaly correct, never to get ahead of the game.

BTW, you may not have been working yet with MSVC++ long enough yet, but as
your project grows you will find that you need to dump much of what comes
packaged with MSVC++ for less bug prone components. You will need to build a
build enviorment that uses a different make then, MS's. You will need to get
a different dependancy generating tool. You will need to get a _correct_
standard library implementation, MS's is broken and poorly optimized. You
will need configuration management that works.. etc. etc. I and other long
time MSVC developers know that long term profession development with MSVC
means throwing away have of what MS delivers and then hunting down the tools
that work.

Delphi and Ada are both much better solututions on every level including
packaging :-)

David Botton

Re: From extended Pascals to Ada 95 guide

[tmoran]

>> You should have contacted ACT. GLIDE combined with CLAW and GNATCOM at the
>> time would have given you the IDE (GLIDE), the GUI (CLAW) and the leverage
>> of all of MS's latest technologies (GNATCOM), and would have offered you
>...
>I still find it a bit of a roadblock that there isn't a shrink-wrap kit
>available that has everything bundled together with suitable documentation.
  Hey, do you mean that if I burn those things onto one CD you'd buy it?

Re: From extended Pascals to Ada 95 guide

[Marin D. Condic]

David Botton wrote:
> 
> You should have contacted ACT. GLIDE combined with CLAW and GNATCOM at the
> time would have given you the IDE (GLIDE), the GUI (CLAW) and the leverage
> of all of MS's latest technologies (GNATCOM), and would have offered you
> just as much (BTW, as a _very_ long term developer of MFC apps, you quickly
> have to throw out the generated code to get most things done on large
> projects :-). There are additional solutions available (and up and coming)
> now, so on your next project, I wouldn't back down from Ada :-)
> 
I've got two projects that are in the discussion stages right now where
I'd like to consider using Ada. The big advantage would be one of
portability across Sun/PC platforms. I may end up trying to cobble
together all the pieces and construct a development environment that
will get the job done.

I still find it a bit of a roadblock that there isn't a shrink-wrap kit
available that has everything bundled together with suitable
documentation. People here will often respond with "Oh, well, you can
download this from here and then go get that from there and then get
this other piece from here and you can modify the source for ABC to make
it do XYZ and yada yada yada". Sure, you can do that. You'll end up with
something that maybe works, but you are spending time pulling together
the environment and customizing/integrating it, instead of building the
software you get paid to build. Documentation will be spotty and
inconsistent. The guy that comes after you on the project is going to
get stuck trying to figure out how to make it work. There are a lot of
costs associated with doing a "roll your own" environment.

With MSVC++, a lot of the problems go away. I insert a disk and install
the environment. I fire up a single program and every tool is available
from there. I have a question about anything? I highlight it and punch
F1 and the documentation is right there. There are books available that
can teach you how to use it. 

While I'm no big proponent of C++ and I'd *much* rather use Ada than
anything else, I've got to consider the ramp-up time and time to market
issues. This is a business which needs to make money and the "best" way
of doing something is not always the most profitable. Because I've got
future projects that will probably need to cross platforms, I've got
good reasons to look at things like GtkAda and GNAT as potential tools.
But it sure would be nice if I could get something in a single box that
did basically what MSVC++ does. (It could be a good business opportunity
in itself to integrate what is out there into a kit and see if it could
sell. :-)

MDC
-- 
======================================================================
Marin David Condic - Quadrus Corporation - http://www.quadruscorp.com/
Send Replies To: m c o n d i c @ q u a d r u s c o r p . c o m
Visit my web site at:  http://www.mcondic.com/

    "Take away the punchbowl just when the party gets going" 

        --  William McChesney Martin, Former Fed chairman, explaining 
            what a sound central bank must always do. 
======================================================================

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

In article <39A6ABD9.634308AA@earthlink.net>, Charles Hixson wrote:

>> >v: vector(1..100) and so on; you can pass them into "+"(v1,v2: vector) functions
>> >and you have the range with attributes v1'first, v1'last, v1'range,...
>> >To be even more reassured, use subtypes: "subtype matrix33 is matrix(1..3,1..3);"
>> >It's that simple: type safety of Pascal, flexibility of C/Fortran!
>>
>> Sounds like Delphi style dynamic arrays.
>>
>> >As for submarines, you're right. Since C/C++ is said to replace Ada in US embedded
>> >systems.. glups...
>
>I don't know Delphi, but that's a static (as in not changing characteristics at run
>time) array.  There are also various more dynamic forms (e.g., unbounded strings), but
>that one was static.

then I went a step to far, and it is plain and simple a simple type.

type matrix33=array[1..3,1..3] of float;


You seem not to specify the element type in Lady Byron's language. Is that
the special part?

But that is only possible using 
- Dynamic vars
- interpreters
- Only types with the same size are allowed (e.g. only integers etc)


So what am I missing here?

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

In article <8obv01$7hu1@news.cis.okstate.edu>, David Starner wrote:
>On 27 Aug 2000 13:23:29 -0700, Ronald Cole <ronald@forte-intl.com> wrote:
>>I don't think the GNU Modula-2 compiler was ever finished/released.

There are several stirrings in the GNU M2 field. Afaik they were working
hard the last year.

>Not if you mean the front end to gcc. There is a project currently being
>worked on to make one that's pretty close to producing result though. There
>are several free Modula-2 compilers out there, though.

IIRC they were first a p2c spinoff, but now they use gcc.



I myself checked most M2 compiler out, before making the cross to pascal
(and FPC), when I didn't found a good free compiler. The free was added as
requirement as a result of the stopped development of TopSpeed which left me
with some trauma's :-)

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

In article <39A991F3.A8D8BED7@easystreet.com>, Al Christians wrote:
>David Starner wrote:
>> 
>> Okay, why isn't Ada a "Wirth-style language", if Modula-3 is? They
>> share more in common with each other than either does with Wirth's
>> Pascal, for instance.
>> 
>
>The principal criterium is size of the language definition/reference 
>manual.  Modula-3's is under 60 pages, IIRC, much smaller than Ada's.
>And Wirth was some kind of advisor to the M3 design team. 

Why is that the principal criterium?

>M2 and M3 are both very nice languages if you want something simpler
>than Ada and don't need what they leave out.  But neither one shows
>much sign of life. 

That's true. This is the reason I started with Pascal again. Commercially
there are still two fairly decent M2 compilers (XDS and StonyBrook), but I
didn't want a commercial compiler.

>All three of these languages, M2, M3, and Ada, now support generics, 
>and that gives them a big advantage over Delphi for coding without doing 
>aribtrary conversions between data types.  

Could you elaborate on this?

>program must cast it to whatever type it really represents.  One big 
>reason Borland's Pascal became popular was because it let you go
>around Pascal's strict type checking, but that was years ago, 

BP was usable at a time that very strict principals (in either libraries,
typing etc) were simply to slow. IIRC Borland didn't only strip some typing, 
but a lot of other features that stood in the way of fast programs.
(e.g. interprocedural goto's?)

> and it's 
>regrettable to have to routinely write code that will be checked 
>for correctness at run time when there are less worrisome alternatives.

Of course in a standard Delphi program those stress tested VCL classes are
used mostly.

>I don't believe that Wirth's versions of M2 or Pascal allowed any way
>to go around the type system, and in M3, the modules that might do
>so must be marked as unsafe.  Ada does allow unchecked conversions.   

Re: From extended Pascals to Ada 95 guide

[Gautier]

Marco van de Voort:

> then I went a step to far, and it is plain and simple a simple type.
> 
> type matrix33=array[1..3,1..3] of float;
> 
> You seem not to specify the element type in Lady Byron's language.

Yes, you specify it...

  type matrix is array(integer range <>, integer range <>) of my_float;
  subtype matrix33 is matrix(1..3,1..3);

An expression of type matrix33 can be used as parameter for a function
like ` "*"(A:matrix; v:vector) return vector; '.

You could also define

  type matrix33 is array(1..3,1..3) of my_float;

But then you lose the generality of the "matrix" type.
That's the big problem of "classic" Pascal, or Modula-2: you can only
program things with fixed sizes, except for variant records.

______________________________________________________
Gautier  --  http://members.xoom.com/gdemont/gsoft.htm

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

>An expression of type matrix33 can be used as parameter for a function
>like ` "*"(A:matrix; v:vector) return vector; '.
>
>You could also define
>
>  type matrix33 is array(1..3,1..3) of my_float;
>
>But then you lose the generality of the "matrix" type.
>That's the big problem of "classic" Pascal, or Modula-2: you can only
>program things with fixed sizes, except for variant records.

Pretty much yes. But because the size of "matrix" is fixed, and not
dynamical, it is too limited IMHO.

When I had to design a new language, I would add it I think, because it
seems nice. But to change to another language (... ADA..) for it?

Re: From extended Pascals to Ada 95 guide

[Gautier]

Marco:

> Pretty much yes. But because the size of "matrix" is fixed, and not
> dynamical, it is too limited IMHO.

No no... size of "matrix" is NOT fixed; only subtypes of it or objects of this
type are - that's the whole difference :-) ! If you want resizable things, it
is also doable in Ada. E.g.

 type matrix( max_m,max_n: positive ) is record
   a: array( 1..max_m, 1..max_n ) of my_float;
   m: positive:= max_m;
   n: positive:= max_n;
 end record;

Or without maximum size, you could use OO techniques with controlled types, maybe -
but a bit expensive for efficient matrix calculations...

> When I had to design a new language, I would add it I think, because it
> seems nice. But to change to another language (... ADA..) for it?

Just for it ? It depends - on needs and personal tastes...
My opinion is that the global advantage of the change
is worth. Also for most details - e.g. for this matrix affair:
in "classic" Pascal there is no solution! Anyway the jump to
Ada is not a frightening change - rather an upgrade...

______________________________________________________
Gautier  --  http://members.xoom.com/gdemont/gsoft.htm

Re: From extended Pascals to Ada 95 guide

[Larry Kilgallen]

Whereas I only use things I can buy on CD (or other tangible media).

Of all the definitions of Open Source I have heard, none has been
sufficiently attractive for me to adopt it as a purchase criterion.

In article <017801c0105d$06e88ac0$cf18b70a@db2000>, "David Botton" <David@Botton.com> writes:
> I suspect that if you put the full version (and lic.) of CLAW, GLIDE, GNAT,
> and GNATCOM on a single CD with a common install you may be able to make a
> buck or two doing so.
> 
> I wouldn't buy a copy since CLAW is not Open Source.
> 
> David Botton
> 
> 
> ----- Original Message -----
> From: <tmoran@bix.com>
> 
> 
>  Hey, do you mean that if I burn those things onto one CD you'd buy it?
> 
> 
> 

Re: From extended Pascals to Ada 95 guide

[David Botton]

FLORIST on UNIX and Pascal's on NT.

David Botton

"Ole-Hjalmar Kristensen" <ohk@clustra.com> wrote in message
news:umqvgwlj3ak.fsf@gong2.clustra.com...

> Interesting. But which POSIX interface implementation are you
> referring to?

Re: From extended Pascals to Ada 95 guide

[Gautier]

> >> Okay, why isn't Ada a "Wirth-style language", if Modula-3 is? They
> >> share more in common with each other than either does with Wirth's
> >> Pascal, for instance.

> >The principal criterium is size of the language definition/reference
> >manual.  Modula-3's is under 60 pages, IIRC, much smaller than Ada's.
> >And Wirth was some kind of advisor to the M3 design team.

Marco:

> Why is that the principal criterium?

Wirth's conception - it appears - is to have a minimalist language definition,
say a BNF codification, some explanations, and bye bye. It's to maintain
"simplicity". It's true, it makes books lighter.
As a result, you get a myriad of proprietary dialects, libraries,
I/O routines etc. IMHO The summit was M2 with different, uncompatible casings
for WriteLn !

______________________________________________________
Gautier  --  http://members.xoom.com/gdemont/gsoft.htm

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

In article <39AA5EC0.A9B8AB6A@maths.unine.ch>, Gautier wrote:
>
>> >The principal criterium is size of the language definition/reference
>> >manual.  Modula-3's is under 60 pages, IIRC, much smaller than Ada's.
>> >And Wirth was some kind of advisor to the M3 design team.
>
>Marco:
>
>> Why is that the principal criterium?
>
>Wirth's conception - it appears - is to have a minimalist language definition,
>say a BNF codification, some explanations, and bye bye. It's to maintain
>"simplicity". It's true, it makes books lighter.

Sure, back then :-) But I'm not using standard pascal, but other dialects
with more functionality (and probably thicker specs). Like Object Pascal
(Delphi's language)

>As a result, you get a myriad of proprietary dialects, libraries,
>I/O routines etc. IMHO The summit was M2 with different, uncompatible casings
>for WriteLn !

I think a language specification and a library specification are two
separate things.

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

In article <39AA6526.A33A4330@acm.org>, Marin D. Condic wrote:
>Marco van de Voort wrote:
>> 
>> When I had to design a new language, I would add it I think, because it
>> seems nice. But to change to another language (... ADA..) for it?
>
>Well, not just for that one feature. Just remember that if you like
>Pascal you'd likely feel very much at home with Ada. 

I do. But to program apps, not to create academical perfect programs.
I also think portability issues are often severely overrated.

>Ada is kind of like
>"Pascal++" in that the syntax is very much like Pascal's, but it cures
>certain ills that existed there and then added a *lot* more stuff to
>make the language much more powerful.

Over standard pascal, or over Delphi?

Re: From extended Pascals to Ada 95 guide

[Gautier]

> >> When I had to design a new language, I would add it I think, because it
> >> seems nice. But to change to another language (... ADA..) for it?

Marin:

> >Well, not just for that one feature. Just remember that if you like
> >Pascal you'd likely feel very much at home with Ada.

Marco:

> I do. But to program apps, not to create academical perfect programs.
> I also think portability issues are often severely overrated.

It depends on your needs. When you have to maintain a large commercial
program that must run on NT for some customers, on Linux for others
it's perfectly feasible in Ada (real case). Try just to manage accurately
file I/O in a portable way in Pascal, even details like 'what happens when
such file is not there ?'. Not to make the hot-line explode...

Another important thing: you should have the choice between compilers, even
when your program already has more than 100'000 lines. E.g. switch from
a compiler firm A (which suddenly closed down, stopped development,
begins to procude a unusable bloatware, has too expensive support or
doesn't provide optimization or a Linux version) to firm B or C.

> >Ada is kind of like
> >"Pascal++" in that the syntax is very much like Pascal's, but it cures
> >certain ills that existed there and then added a *lot* more stuff to
> >make the language much more powerful.

> Over standard pascal, or over Delphi?

Anyway, a lot over BP7. I find even Ada 83 (current is 95) with its
generics (templates), polymorphism, overloadable operators, subtypes
a quantum leap for programming numerics (here on an Alpha server) compared
to BP7...

I would be glad to incorporate some details about Delphi in my comparison...

______________________________________________________
Gautier  --  http://members.xoom.com/gdemont/gsoft.htm

Re: From extended Pascals to Ada 95 guide

[Al Christians]

Marco van de Voort wrote:
> 
> In article <39A991F3.A8D8BED7@easystreet.com>, Al Christians wrote:
> >
> >The principal criterium is size of the language definition/reference
> >manual.  Modula-3's is under 60 pages, IIRC, much smaller than Ada's.
> >And Wirth was some kind of advisor to the M3 design team.
> 
> Why is that the principal criterium?
> 

Because a concise language definition has been a design goal of the 
Wirth-type languages.  The last, Oberon-2, was considered a successful
design because it got this down to about 15 pages.


> >All three of these languages, M2, M3, and Ada, now support generics,
> >and that gives them a big advantage over Delphi for coding without doing
> >aribtrary conversions between data types.
> 
> Could you elaborate on this?
> 
> 
> Of course in a standard Delphi program those stress tested VCL 
> classes are used mostly.
> 


If I define my own type in a Pascal program and want to save it in
one of the VCL containers, the code would look like this:

    { To Insert }      AContainer.Objects[i] := Pointer(MyObject);
    { To Retrieve }    MyObject := TMyObject(AContainer.Objects[i]);

The compiler can't check if the conversion of type from Pointer back
to TMyObject is applicable.  Some typographical or thinkographical
error causing an object of the wrong type to be converted to a Pointer 
when inserting is not detected until retrieval at run time.  Delphi is
far from the worst offender in this regard.  Look at any book on COM
programming in C++, and you see about three programmer specified 
conversions in each four lines of code.  It's almost like programming
with no user-defined types at all.

In Ada, this would be done with generic containers, and no type 
conversions would be required.   By defining languages with strict 
rules for type conversions, Wirth made progress against certain 
kinds of errors.  By giving us ways around the rules, Borland made
Wirth's languages practical for day-to-day use.  Generics in Ada,
M3, and now M2, combine good features from each approach.


Al

Re: From extended Pascals to Ada 95 guide

[Ray Blaak]

"Marin D. Condic" <mcondic-nospam@acm.org> writes:
> Maybe Delphi is better, but I wouldn't know. What features/capabilities
> does it have that you think makes it superior?

Besides the language, which aside from the lack of destructors, is more
amenable to an Ada programmer, the interface to the OS, and all the support
libraries are much better designed.

With C++ one gets the nitty gritty code in your face immediately -- not
because of language problems, but because that's the way VC does it (e.g. try
setting up some automation objects and note all of the manual reference
increments and decrements that must be done, even with the OLE dispatch
classes).

With VB nothing is in your face, but as soon as you need good control of
things (which is all too often) the programmer has to do some extreme
contortions to get things done.

With Delphi, about 80% of the time the presupplied classes are good
enough. When you need to extend/change things, you can do so in a reasonable,
well designed way.

-- 
Cheers,                                        The Rhythm is around me,
                                               The Rhythm has control.
Ray Blaak                                      The Rhythm is inside me,
blaak@infomatch.com                            The Rhythm has my soul.

Re: From extended Pascals to Ada 95 guide

[Charles Hixson]

[-- Attachment #1: Type: text/plain, Size: 1161 bytes --]

Gautier wrote:

> Marco:
>
> > Pretty much yes. But because the size of "matrix" is fixed, and not
> > dynamical, it is too limited IMHO.
>
> No no... size of "matrix" is NOT fixed; only subtypes of it or objects of this
> type are - that's the whole difference :-) ! If you want resizable things, it
> is also doable in Ada. E.g.
>
>  type matrix( max_m,max_n: positive ) is record
>    a: array( 1..max_m, 1..max_n ) of my_float;
>    m: positive:= max_m;
>    n: positive:= max_n;
>  end record;

The dimensionality of matrix is, however, fixed.  I'm not certain that this matters,
since I usually know the dimensions of the type before I start, but perhaps that is
because I've always worked with languages that specify things that way.  OTOH, what
operations could one define without knowing the dimensionality ahead of time?

Well, there's projection, magnitude, scalar addition, scalar subtraction, application
(apply this operation to every element of the matrix), application2 (apply every
element of this matrix to this thing), store, recall ...

-- (c) Charles Hixson
--  Addition of advertisements or hyperlinks to products specifically prohibited


[-- Attachment #2: Card for Charles Hixson --]
[-- Type: text/x-vcard, Size: 145 bytes --]

begin:vcard 
n:Hixson;Charles
x-mozilla-html:FALSE
adr:;;;;;;
version:2.1
email;internet:charleshixson@earthling.net
fn:Charles Hixson
end:vcard

Re: From extended Pascals to Ada 95 guide

[Ronald Cole]

kenner@lab.ultra.nyu.edu (Richard Kenner) writes:

> In article <m3g0nqlagu.fsf@yakisoba.forte-intl.com> Ronald Cole <ronald@forte-intl.com> writes:
> >I don't think the GNU Modula-2 compiler was ever finished/released.
> 
> Interesting coincidence!  Just a day or two ago there was a posting from
> the folks on that project saying they are trying to get it finished.

Synchronicity!!

-- 
Forte International, P.O. Box 1412, Ridgecrest, CA  93556-1412
Ronald Cole <ronald@forte-intl.com>      Phone: (760) 499-9142
President, CEO                             Fax: (760) 499-9152
My GPG fingerprint: C3AF 4BE9 BEA6 F1C2 B084  4A88 8851 E6C8 69E3 B00B

Re: From extended Pascals to Ada 95 guide

[Ray Blaak]

Al Christians <achrist@easystreet.com> writes:
> If I define my own type in a Pascal program and want to save it in
> one of the VCL containers, the code would look like this:
> 
>     { To Insert }      AContainer.Objects[i] := Pointer(MyObject);
>     { To Retrieve }    MyObject := TMyObject(AContainer.Objects[i]);
> 

With Delphi, I have always found it worthwhile to make my own container
classes that do the appropriate checking:

type MyContainer = class
public
   procedure Insert(atIndex : Integer; obj : MyObject);
   function GetAt(index : Integer) : MyObject;
private
   flist : TList;
end;

where we have:

procedure MyContainer.Insert(atIndex : Integer; obj : MyObject);
begin
  flist.Items[i] := Pointer(obj);
end;

function MyContainer.GetAt(index : Integer) : MyObject;
begin
  result := MyObject(flist.Items[i]);
end;

and then I can do:
   
   AContainer.Insert(i, MyObject);
   MyObject := AContainer.GetAt(i);

Now I have the benefit of good typechecking. The one disadvantage of this
approach is that Delphi's lack of generics can mean constructing these wrapper
classes is tedious. However, the benefit outweighs the tedium, in my opinion.

-- 
Cheers,                                        The Rhythm is around me,
                                               The Rhythm has control.
Ray Blaak                                      The Rhythm is inside me,
blaak@infomatch.com                            The Rhythm has my soul.

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

About the Matrix type:

The subtype system, to me, looks like a cross between and attempt to
implement multidimensional openarrays, and GPC pascal's schemata types.

If I were you, I wouldn't be too sad that it doesn't 

> Marco:
 
> > I do. But to program apps, not to create academical perfect programs.
> > I also think portability issues are often severely overrated.
> 
> It depends on your needs.

Definitely. But portability demands always need extra time. Except if
you allow the applications limits  to be set by the portability
requirements.

> When you have to maintain a large commercial
> program that must run on NT for some customers, on Linux for others
> it's perfectly feasible in Ada (real case). 

That is the same processor, and NT is POSIX compliant. If you use some
IBCS format, you might even be able to use the same binary :-) (in
theory of course).

Well, I can only put the FPC compiler itself. 100000 lines, and
increasing daily. FreeBSD will be the fifth i386 OS.

> Try just to manage accurately
> file I/O in a portable way in Pascal, even details like 'what happens when
> such file is not there ?'. Not to make the hot-line explode...

First:
What does ADA in such case? The program always has to react to what
happens and doesn't happen in such cases. An non existant file (that
should exist) always triggers some special treatment. That treatment can
be as simple as halt-program, or reinitialise, but it is a condition set
by the programmer, not by the language.

Second: What pascal does in this case depends on the runtime library I
use, not on pascal. This is a critical difference, since in free
compilers with access to the compiler source, you can finally
remove/replace nearly
all the RTL. TS M2 was pretty much the Borland under the Modula2
compilers. (and even related to BP IIRC)

> Another important thing: you should have the choice between compilers, even
> when your program already has more than 100'000 lines. E.g. switch from
> a compiler firm A (which suddenly closed down, stopped development,
> begins to procude a unusable bloatware, has too expensive support or
> doesn't provide optimization or a Linux version) to firm B or C.

That happened to me. Not with a large codebase, but with a bunch of
small utils, non of which rocketscience, that kept a few BBSes running.
At a certain point people started using longfilenames, and I was stuck
with the TopSpeed RTL.

That's why I moved to a free compiler, and for private, or even medium
use, I'll never consider a commercial compiler again.

Not only because in problem cases you can fix problems yourself, but
compilerprojects are relatively large projects, and probably will last
for a decade in some form. (iow several generations of developpers)
 
> > >Ada is kind of like
> > >"Pascal++" in that the syntax is very much like Pascal's, but it cures
> > >certain ills that existed there and then added a *lot* more stuff to
> > >make the language much more powerful.
> 
> > Over standard pascal, or over Delphi?
> 
> Anyway, a lot over BP7. 

Yeah, but I dumped BP7 in 1993 for TopSpeed :-) (Modula-2)

I don't know much about ADA. I liked TopSpeed M2, but didn't like
ISO-M2, M3 or Oberon. 
 
> I would be glad to incorporate some details about Delphi in my comparison...

I think we should call it Object Pascal :-) Delphi is more the entire
project (including IDE, libraries etc)

Check the manuals or specs:
- Inprise (Borland) is quite generous with the helpfiles, and they are
downloadable from ftp.inprise.com/pub/delphi/techpubs (or something like
that)
- JEDI, the Delphi open source project, has a documentation project (and
tool), it is called Dolphin, and contains some nice texts.
- For the FPC implementation: download the reference manual from
	the FPC site (www.freepascal.org)

-- 

Marco van de Voort (Marcov@stack.nl or marco@freepascal.org)

Re: From extended Pascals to Ada 95 guide

[Larry Elmore]

"Marco van de Voort" <marcov@toad.stack.nl> wrote in message
news:slrn8qksee.oft.marcov@toad.stack.nl...
> In article <39AA6526.A33A4330@acm.org>, Marin D. Condic wrote:
>
> >Ada is kind of like
> >"Pascal++" in that the syntax is very much like Pascal's, but it cures
> >certain ills that existed there and then added a *lot* more stuff to
> >make the language much more powerful.
>
> Over standard pascal, or over Delphi?

Having used both, I'd say Ada is a _serious_ improvement over Delphi's
Object Pascal, and of course a much greater improvement over standard
Pascal.

Larry

Re: From extended Pascals to Ada 95 guide

[nabbasi]

In article <39A991F3.A8D8BED7@easystreet.com>, Al says...
 
>All three of these languages, M2, M3, and Ada, now support generics, 
>and that gives them a big advantage over Delphi for coding without doing 
>aribtrary conversions between data types.  In Delphi, the generic type 
>stored by the standard VCL collection classes is the Pointer, and the 
>program must cast it to whatever type it really represents.

Well, generics are nice, but Java do not have them, and it seems
that half the world now program in Java. So, I do not think that Borland
delphi not having generics is such a big deal for most people.
 
Nasser
 

Re: From extended Pascals to Ada 95 guide

[Marin D. Condic]

Marco van de Voort wrote:
> 
> When I had to design a new language, I would add it I think, because it
> seems nice. But to change to another language (... ADA..) for it?

Well, not just for that one feature. Just remember that if you like
Pascal you'd likely feel very much at home with Ada. Ada is kind of like
"Pascal++" in that the syntax is very much like Pascal's, but it cures
certain ills that existed there and then added a *lot* more stuff to
make the language much more powerful.

MDC
-- 
======================================================================
Marin David Condic - Quadrus Corporation - http://www.quadruscorp.com/
Send Replies To: m c o n d i c @ q u a d r u s c o r p . c o m
Visit my web site at:  http://www.mcondic.com/

    "Take away the punchbowl just when the party gets going" 

        --  William McChesney Martin, Former Fed chairman, explaining 
            what a sound central bank must always do. 
======================================================================

Re: From extended Pascals to Ada 95 guide

[Marin D. Condic]

David Botton wrote:
> You can get all the components from one vendor with documentation and all.
> You don't have to hunt and peck for pieces. GNAT, GtkAda, GNATCOM, and GLIDE
> are all available from a single source. They are all integrated and work
> from the same IDE, no roll your own here.
> 
Well, I've downloaded the various versions of GNAT for the PC and the
other tools don't seem to come with it. Maybe they can be had, but its
not all in one self-extracting file or even in the same directory. I've
found GtkAda on the web as a separate entity and got the impression that
the Windows version was not quite mature yet - maybe still usable. If
all that stuff is integrated nicely, maybe it ought to be in a single
install file somewhere?

If you have to go to ACT to get it as a supported customer, then the
situation changes. I can currently go to the local computer store and
plunk down a few hundred dollars and get a shrink-wrap kit from
Micro$oft that includes documentation in printed and CD-ROM format and a
subscription for quarterly updates. (I do not delude myself into
believing I get any real support, but mostly in this context, I don't
really need it.) The last time I talked to ACT about being a supported
customer (different context), the price tag was a bit higher than that -
maybe more than 30x. Bzzzzzzt! Wrong answer! Can't afford that kind of
price tag for this level of development and don't need that much
support.



> BTW, you may not have been working yet with MSVC++ long enough yet, but as
> your project grows you will find that you need to dump much of what comes
> packaged with MSVC++ for less bug prone components. You will need to build a
> build enviorment that uses a different make then, MS's. You will need to get
> a different dependancy generating tool. You will need to get a _correct_
> standard library implementation, MS's is broken and poorly optimized. You
> will need configuration management that works.. etc. etc. I and other long
> time MSVC developers know that long term profession development with MSVC
> means throwing away have of what MS delivers and then hunting down the tools
> that work.
> 
I have no illusions about the overall quality of MSVC++. I have already
run into a number of compiler bugs and related problems. Its a complex
piece of software that is (probably) written in C++. (I rest my case!
:-) Yes, many features have problems and the language itself makes
matters worse (dependency, for example) so I know it is far from
perfect. And remember, I'm a big fan of Ada so I'm not trying to make
excuses for MSVC++ just because I don't want to use Ada. I'm just saying
I have to give MSVC++ some credit for having put together a spiffy
environment that gets you to market very fast and appears to be more
advanced (or at least all in one place) than what I've seen of Ada
programming environments.

Being an "old timer" I have absolutely no problems with a command line
compiler and in many ways, I'd prefer to work that way. I've seen some
IDEs that I thought spent more time getting in your way than anything
else. (I don't generally like all the "help" that IDEs try to give you
in terms of project management.) But I have to conceed that there are
advantages to what you get from a full-up toolset like MSVC++ . 

> Delphi and Ada are both much better solututions on every level including
> packaging :-)
> 
Don't know Delphi. Ada is a better language - no argument. I'd like to
see a kit that uses Ada and is feature-wise/price-wise competitive with
MSVC++. If its out there, I've just not bumped into it. Point me at a
URL. Thanks.

MDC
-- 
======================================================================
Marin David Condic - Quadrus Corporation - http://www.quadruscorp.com/
Send Replies To: m c o n d i c @ q u a d r u s c o r p . c o m
Visit my web site at:  http://www.mcondic.com/

    "Take away the punchbowl just when the party gets going" 

        --  William McChesney Martin, Former Fed chairman, explaining 
            what a sound central bank must always do. 
======================================================================

Re: From extended Pascals to Ada 95 guide

[Marin D. Condic]

tmoran@bix.com wrote:
> 
> >> You should have contacted ACT. GLIDE combined with CLAW and GNATCOM at the
> >> time would have given you the IDE (GLIDE), the GUI (CLAW) and the leverage
> >> of all of MS's latest technologies (GNATCOM), and would have offered you
> >...
> >I still find it a bit of a roadblock that there isn't a shrink-wrap kit
> >available that has everything bundled together with suitable documentation.
>   Hey, do you mean that if I burn those things onto one CD you'd buy it?

Well, lets take a look at it and see. Just having a CD with several
megabytes of programs on it is not good enough. If its all pulled
together nicely, I think it would be marketable. People lay down money
for ObjectAda from Aonix, don't they? (Have not looked at their stuff in
over a year - maybe its got more features now.)

There's more to it than burning it on a CD. But I'm sure you understand
that.

MDC
-- 
======================================================================
Marin David Condic - Quadrus Corporation - http://www.quadruscorp.com/
Send Replies To: m c o n d i c @ q u a d r u s c o r p . c o m
Visit my web site at:  http://www.mcondic.com/

    "Take away the punchbowl just when the party gets going" 

        --  William McChesney Martin, Former Fed chairman, explaining 
            what a sound central bank must always do. 
======================================================================

Re: From extended Pascals to Ada 95 guide

[Marin D. Condic]

tmoran@bix.com wrote:
>   The source code is of course included (it pretty much has to be since
> it runs with multiple different compilers), but of course you mean you
> would insist on the right to sell or give away a copy to Marin.  So I
> could sell Marin, but not you, a copy, or, if it was "Open Source (tm)",
> I could sell you a copy, but Marin probably wouldn't buy one from me
> since he could get it for less from you.  Doesn't sound like such a good
> idea after all.

I'm not particularly hung up on "Open Source (tm)". For my needs, all I
really care about is that I can build a useful product from it and am
not going to have enormous problems giving the object code to my
customer. If the customer wants source, they'll pay us extra for that
and then all I need is the ability to transfer the required source for
the development tool (CLAW, or whatever libraries are needed to compile)
to them.

But I suppose that is going to depend on the job one is trying to do. If
I wanted to put the end product out on the net and Open Source my stuff,
then I guess it means I need Open Source for anything I use.

MDC
-- 
======================================================================
Marin David Condic - Quadrus Corporation - http://www.quadruscorp.com/
Send Replies To: m c o n d i c @ q u a d r u s c o r p . c o m
Visit my web site at:  http://www.mcondic.com/

    "Take away the punchbowl just when the party gets going" 

        --  William McChesney Martin, Former Fed chairman, explaining 
            what a sound central bank must always do. 
======================================================================

Re: From extended Pascals to Ada 95 guide

[Richard Riehle]

Marco van de Voort wrote:

> >As for submarines, you're right. Since C/C++ is said to replace Ada in US embedded
> >systems.. glups...

Can you document this?  I know of some cases where irresponsible or ignorant systems
engineers
have specified C++ instead of Ada, but there continues to be a lot of Ada in many of
the systems I know about.

Anyone who would select C++ instead of Ada for a safety-critical military weapon
systems either
1) does not understand Ada, 2) does not understand C++, 3) does not understand either
one, or
4) has some agenda that supercedes reliability and technical excellence.   I frequently
encounter
all four situations in my software practice.

There may be some alternative emerging within the C family of languages:  C#.  I
recently attended
a briefing on that language.   It may improve on the reliability issues so rampant in
C++ and rectify
some of the concerns relative to Java.   Of course, it is still a new language offering
and has a
job of persuading real software developers to adopt it.

So far, though, no programming language quite measures up to Ada when reliability and
safety
concerns are the key issues.   I personally like Smalltalk and Eiffel, but would not
choose either
when developing something as important as a military system, a communications
satellite, a
cardiac control unit, flight control systems, or the many other kinds of software that,
if there is
an accident, can result in the death or maiming of unintended humans.

Modula-2 and Modula-3 fall short of the requirement.  C++ falls far short of the
requirement.
Oberon is a step in the right direction.   Eiffel, as a language, has potential for the
future but the
implementations are not at the right level of maturity.   Java is an interesting
offering, but more
of a research topic in many respects.  It is also a good teaching tool. Perhaps it will
improve with time.

When one studies this issue really carefully, with the goal of reliability and safety
in mind, the decision in favor
of Ada naturally falls out as the right one.  It is an engineering decision, not a
programming language decision.
Very few software developers, and even fewer software development managers, are trained
to consider
engineering issues in constructing software.  If one is designing for one of the
popular desktop operating
systems, many of which are inherently defect-infested themselves, there may be minimal
benefit in
selecting a language focused on error-free software.   In that case, C++, Java,
Smalltalk and other popular
languages are fun to use.   I wonder how many C++ programmers would trust their lives
to flying in
an airplane programmed in that language.

Well, this is just Richard preaching to choir again.   Sorry for the rant.   It is the
end of a long day and I just
reviewed this thread, discovering it failed to focus on what is and is not important --
at least to my mind.

Richard Riehle

Re: From extended Pascals to Ada 95 guide

[tmoran]

>Perhaps because you have not yet found a way to be succesful with that sort
>of lic. arangement.
  True, true.  But let's not get into another "open source" economics
debate.  I think neither of us can understand how the other could
possibly believe what he apparently does believe.  The bottom line is:

>>I suspect that if you put the full version (and lic.) of CLAW, GLIDE, GNAT,
>>and GNATCOM on a single CD with a common install you may be able to make a
>>buck or two doing so.
> Doesn't sound like such a good idea after all.

Re: From extended Pascals to Ada 95 guide

[Ole-Hjalmar Kristensen]

"David Botton" <David@Botton.com> writes:

> <steve@nospam.ada.eu.org> wrote in message
> news:8o7top$4sb@drn.newsguy.com...
> > I am not sure if it is
> > an ISO or ANSI standard language though.
> 
> It is not and its syntax and symantics change frequently between versions
> (many times to the surprise of developers :-)
> 
> > The nice thing about it,
> > is that all those current Windows delphi application can now very easily
> > be ported to Linux by using Kylix to compile the window delphi source
> > code with Linux/Kylix.
> 
> Provided the application doesn't use any OS specific bindings, ActiveX
> controls, or any third party units. Also, given statements by Borland to not
> expect 100% source code compatability don't expect the language to be 100%
> compatable or even the GUI units.
> 
> On the other hand, an application developed using Ada 95 and GtkAda is 100%
> source code compatable between Linux and Windows versions. Using the POSIX
> package for NT and Linux and you have cross platform on most of the OS
> specific stuff. Of course you can use ActiveX controls only on the Windows
> side (Using my thick Win32 framework, GWindows, you can easily add ActiveX
> controls to GTKAda on Win32, see my example on the GWindows page at
> http://www.adapower.com/gwindows)
> 
> David Botton
> 
> 

Interesting. But which POSIX interface implementation are you
referring to?

--
"Plus I remember being impressed with Ada because you could write an
 infinite loop without a faked up condition.  The idea being that in Ada
 the typical infinite loop would normally be terminated by detonation."
                                             -Larry Wall

Ole-Hj. Kristensen

Re: From extended Pascals to Ada 95 guide

[David Starner]

On 28 Aug 2000 22:20:19 -0700, <nabbasi@pacbell.net.NOSPAM> wrote:
>Well, generics are nice, but Java do not have them, and it seems
>that half the world now program in Java. So, I do not think that Borland
>delphi not having generics is such a big deal for most people.

How I hate Java. I can't believe how all the programmers who would 
slam Pascal and Ada for "bondage and discipline" turn around and
use this language that doesn't do procedural programming, doesn't
do generic programming, and doesn't have goto's. The fact that Java
doesn't have generics says more about Java then the value of generics.

-- 
David Starner - dstarner98@aasaa.ofe.org
http/ftp: dvdeug.net.dhis.org
It was starting to rain on the night that they cried forever,
It was blinding with snow on the night that they screamed goodbye.
	- Dio, "Rock and Roll Children"

Re: From extended Pascals to Ada 95 guide

[Larry Kilgallen]

In article <8ofh6j$2anb@drn.newsguy.com>, nabbasi@pacbell.net.NOSPAM writes:

> Well, generics are nice, but Java do not have them, and it seems
> that half the world now program in Java.

I do not believe that half the (programming) world programs in Java.
Perhaps half of the loudest programmers.

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

In article <F2pc75UUKWbP@eisner.decus.org>, Larry Kilgallen wrote:
>In article <8ofh6j$2anb@drn.newsguy.com>, nabbasi@pacbell.net.NOSPAM writes:
>
>> Well, generics are nice, but Java do not have them, and it seems
>> that half the world now program in Java.
>
>I do not believe that half the (programming) world programs in Java.
>Perhaps half of the loudest programmers.

I think 50% is exaggerated, but I wouldn't bet anything serious on it.
 Everywhere they are abandoning almost any programming language for Java.
Even for teaching programming!!!

(I was yesterday totally flabbergasted when I heard that the Math+IT faculty
will teach the next generation programming with Java as introductionary
course!)

Re: From extended Pascals to Ada 95 guide

[Gautier]

Larry:

> >I do not believe that half the (programming) world programs in Java.
> >Perhaps half of the loudest programmers.

Marco:

> I think 50% is exaggerated, but I wouldn't bet anything serious on it.
> Everywhere they are abandoning almost any programming language for Java.
> Even for teaching programming!!!

> (I was yesterday totally flabbergasted when I heard that the Math+IT faculty
> will teach the next generation programming with Java as introductionary
> course!)

Precisely: my impression is that the Java(tm) marketing is far more efficient
towards university professors than for the "real" computing world...

Maybe they are attracted by pedantic, unefficient things - and the Pascal
avatars aren't pedantic enough ?...

G.

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

In article <39ABC2C2.1AEC92E0@maths.unine.ch>, Gautier wrote:
>Larry:
>
>> course!)
>
>Precisely: my impression is that the Java(tm) marketing is far more efficient
>towards university professors than for the "real" computing world...

That is not my impression. I have a feeling that specially the small application
programmers have switched to Java en masse.
Big solid apps are still coded in C, C++.  

>Maybe they are attracted by pedantic, unefficient things - and the Pascal
>avatars aren't pedantic enough ?...

Childish remark. 

But to add something constructive: Anything with some structure, and 
reasonably strong typing is good to teach programming I think. Maybe Modula2 
would be best due to its simplicity, and being less fragmented than Pascal,
while even being stronger typed.

Re: From extended Pascals to Ada 95 guide

[Gautier]

Marin:

> Well, lets take a look at it and see. Just having a CD with several
> megabytes of programs on it is not good enough. If its all pulled
> together nicely, I think it would be marketable. People lay down money
> for ObjectAda from Aonix, don't they? (Have not looked at their stuff in
> over a year - maybe its got more features now.)

There is a new 7.2 demo version @

  ftp://ftp.aonix.com/pub/ada/public/pal/windows-zip-small/

- compare the sizes between 7.1 and 7.2...

You may also fill their Web forms.

A somewhat Pavlovian description is there:

  http://www.aonix.com/content/news/pr_05.25.00.html 

______________________________________________________
Gautier  --  http://members.xoom.com/gdemont/gsoft.htm

Re: From extended Pascals to Ada 95 guide

[Charles Hixson]

[-- Attachment #1: Type: text/plain, Size: 1576 bytes --]

David Starner wrote:

> On 28 Aug 2000 22:20:19 -0700, <nabbasi@pacbell.net.NOSPAM> wrote:
> >Well, generics are nice, but Java do not have them, and it seems
> >that half the world now program in Java. So, I do not think that Borland
> >delphi not having generics is such a big deal for most people.
>
> How I hate Java. I can't believe how all the programmers who would
> slam Pascal and Ada for "bondage and discipline" turn around and
> use this language that doesn't do procedural programming, doesn't
> do generic programming, and doesn't have goto's. The fact that Java
> doesn't have generics says more about Java then the value of generics.

But what it does have is:
a) Garbage Collection
b) Standardized Screen Drawing library  --  portable!
c) Standardized Print formatting library -- portable!
If it weren't for jGnat, then Ada would be impossible for many applications.
And that's still quite new.  I'm really hoping that it will develop into
something really smooth, but so far I haven't given it a thorough test.
(It's nice to see that the new edition of Gide supports it though.)  And it
(supposedly, I haven't tested!) allows seamless integration between Ada
native and jvm programs.  Even without that, at least one wouldn't need to
pay the overhead for using handles just because the abstract type wasn't
primitive.  But again, the public release version is 1.0, so if you are doing
commercial work, you had probably best purchase a support contract.

-- (c) Charles Hixson
--  Addition of advertisements or hyperlinks to products specifically
prohibited


[-- Attachment #2: Card for Charles Hixson --]
[-- Type: text/x-vcard, Size: 145 bytes --]

begin:vcard 
n:Hixson;Charles
x-mozilla-html:FALSE
adr:;;;;;;
version:2.1
email;internet:charleshixson@earthling.net
fn:Charles Hixson
end:vcard

Re: From extended Pascals to Ada 95 guide

[Jonas Maebe]

In article <slrn8qndkq.2v02.marcov@toad.stack.nl>, Marco van de Voort
<marcov@toad.stack.nl> wrote:

> (I was yesterday totally flabbergasted when I heard that the Math+IT faculty
> will teach the next generation programming with Java as introductionary
> course!)

Last year they also changed from Pascal to Java at our university for
all introductory courses to programming (in all faculties). Java seems
to be seen here as the second coming in terms of programming languages.


Jonas

Re: From extended Pascals to Ada 95 guide

[Brian Rogoff]

On 28 Aug 2000 nabbasi@pacbell.net.NOSPAM wrote:
> Well, generics are nice, but Java do not have them, and it seems
> that half the world now program in Java. So, I do not think that Borland
> delphi not having generics is such a big deal for most people.

I hope you're not being serious, because there are a flaws with both the 
facts (half the world, or even "most programmers" using Java) and the 
inference that if this were the case that it isn't a big deal. 

Generics are one of the most requested features in Java, and it looks like
some variant of Generic Java (http://www.cs.bell-labs.com/who/wadler/pizza/gj/)
will influence Java 0X. 

It's also a bit sad that this kind of argument (most people are using X 
so there is no need for anything better than X) is ever used in
programming language discussions. 

-- Brian

Re: From extended Pascals to Ada 95 guide

[Ehud Lamm]

On 28 Aug 2000 nabbasi@pacbell.net.NOSPAM wrote:

|In article <39A991F3.A8D8BED7@easystreet.com>, Al says...
|
|>All three of these languages, M2, M3, and Ada, now support generics, 
|
|Well, generics are nice, but Java do not have them, and it seems
|that half the world nowprogram in Java. So, I do not think that Borland
|delphi not having generics is such a big deal for most people.
|

Notice that Ada generics combine nicely with inheritance, and can be used
to achieve some of the goals of multiple inheritance. Java has interfaces
to deal with some of the limitations of single inheritance.

This is without really discussing the mertis of generic coding. I leave
that for another time...

Ehud Lamm mslamm@mscc.huji.ac.il
http://purl.oclc.org/NET/ehudlamm <== My home on the web 
Check it out and subscribe to the E-List- for interesting essays and more!

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

In article <39AAED55.6EA4E810@ix.netcom.com>, Richard Riehle wrote:
>
>Marco van de Voort wrote:
>
>> >As for submarines, you're right. Since C/C++ is said to replace Ada in US embedded
>> >systems.. glups...
>
>Can you document this? 

No. One of the original msgs stated such I think.

Re: From extended Pascals to Ada 95 guide

[Gary Scott]

My company is now developing several projects using Java.  The end
product typically is not up to snuff from a usability standpoint.  It is
extremely slow and sluggish and the quality of the graphic rendering is
typically below the native OS applications in visual quality.

$0.02

Charles Hixson wrote:
> 
> David Starner wrote:
> 
> > On 28 Aug 2000 22:20:19 -0700, <nabbasi@pacbell.net.NOSPAM> wrote:
> > >Well, generics are nice, but Java do not have them, and it seems
> > >that half the world now program in Java. So, I do not think that Borland
> > >delphi not having generics is such a big deal for most people.
> >
> > How I hate Java. I can't believe how all the programmers who would
> > slam Pascal and Ada for "bondage and discipline" turn around and
> > use this language that doesn't do procedural programming, doesn't
> > do generic programming, and doesn't have goto's. The fact that Java
> > doesn't have generics says more about Java then the value of generics.
> 
> But what it does have is:
> a) Garbage Collection
> b) Standardized Screen Drawing library  --  portable!
> c) Standardized Print formatting library -- portable!
> If it weren't for jGnat, then Ada would be impossible for many applications.
> And that's still quite new.  I'm really hoping that it will develop into
> something really smooth, but so far I haven't given it a thorough test.
> (It's nice to see that the new edition of Gide supports it though.)  And it
> (supposedly, I haven't tested!) allows seamless integration between Ada
> native and jvm programs.  Even without that, at least one wouldn't need to
> pay the overhead for using handles just because the abstract type wasn't
> primitive.  But again, the public release version is 1.0, so if you are doing
> commercial work, you had probably best purchase a support contract.
> 
> -- (c) Charles Hixson
> --  Addition of advertisements or hyperlinks to products specifically
> prohibited

Re: From extended Pascals to Ada 95 guide

[Charles Hixson]

[-- Attachment #1: Type: text/plain, Size: 1099 bytes --]

Gary Scott wrote:

> My company is now developing several projects using Java.  The end
> product typically is not up to snuff from a usability standpoint.  It is
> extremely slow and sluggish and the quality of the graphic rendering is
> typically below the native OS applications in visual quality.
> $0.02
> ...

But it does do the graphics and the printing.  What I was thinking of was using
java for the screen and printing, and jGnat as a bridge to Gnat where the
computation, IO, etc. would be done.  This said, I do understand that the Java
event model has heavy compute penalties.  The quality may be below native OS
applications (seems that way to me, also), but all I need are dialog boxes (large
ones, 'tis true, and several of them).  I don't really see ANY other answer.
Outside of do everything in Java, which while confining would work.  If most of the
Java overhead is in the screen handling, then that might actually be the best
choice.  I DON'T want to tie myself to one OS.

-- (c) Charles Hixson
--  Addition of advertisements or hyperlinks to products specifically prohibited


[-- Attachment #2: Card for Charles Hixson --]
[-- Type: text/x-vcard, Size: 145 bytes --]

begin:vcard 
n:Hixson;Charles
x-mozilla-html:FALSE
adr:;;;;;;
version:2.1
email;internet:charleshixson@earthling.net
fn:Charles Hixson
end:vcard

Re: From extended Pascals to Ada 95 guide

[Gary Scott]

Hi,

Charles Hixson wrote:
> 
> I DON'T want to tie myself to one OS.

That's why I use a cross-platform GUI builder like GINO.  Near native
performance, native look and feel, very minimal portability problems. 
Of course, it requires multiple sets of development tools and production
of multiple executable modules and associated fauna and flora, but not
very difficult at all.

> 
> -- (c) Charles Hixson
> --  Addition of advertisements or hyperlinks to products specifically prohibited

Re: From extended Pascals to Ada 95 guide

[John English]

"Marin D. Condic" wrote:
> I still find it a bit of a roadblock that there isn't a shrink-wrap kit
> available that has everything bundled together with suitable
> documentation...

Hmm. I feel an advert break coming on... :-)

BURKS 5 is now out (a 3 CD set, now it's reached its 5th year) and has
GNAT 3.13, GNATCOM, ADAGIDE, GRASP, and a bunch of other stuff (LRM95,
the Rationale, the Style Guide, the Lovelace tutorial, etc). Will this
fit the bill?

The reason it takes up 3 CDs is that it also includes Red Hat 6.2,
compilers and tutorials for another couple of dozen languages, a
dictionary of computing with about 13,000 entries, hyperlinked copies
of all the online RFCs, and a shedload of other useful odds and sods.
It costs �6.00 (about $10.00 US?) for about 2 gig of software and
documentation. It's fully indexed, and the 3 CDs act together as a
single co-ordinated website (there's a bundled copy of Netscape or
you can use your own browser, and you get prompted to change disks
when necessary).

The whole lot is available online at http://burks.brighton.ac.uk/ --
the Ada section is at http://burks.brighton.ac.uk/burks/language/ada/
and
the Pascal section is at
http://burks.brighton.ac.uk/burks/language/pascal/.
Might be worth a look, IMHO.

-----------------------------------------------------------------
 John English              | mailto:je@brighton.ac.uk
 Senior Lecturer           | http://www.comp.it.bton.ac.uk/je
 Dept. of Computing        | ** NON-PROFIT CD FOR CS STUDENTS **
 University of Brighton    |    -- see http://burks.bton.ac.uk
-----------------------------------------------------------------

Re: From extended Pascals to Ada 95 guide

[John English]

nabbasi@pacbell.net.NOSPAM wrote:
> 
> In article <39A991F3.A8D8BED7@easystreet.com>, Al says...
> 
> >All three of these languages, M2, M3, and Ada, now support generics,
> >and that gives them a big advantage over Delphi for coding without doing
> >aribtrary conversions between data types.  In Delphi, the generic type
> >stored by the standard VCL collection classes is the Pointer, and the
> >program must cast it to whatever type it really represents.
> 
> Well, generics are nice, but Java do not have them, and it seems
> that half the world now program in Java. So, I do not think that Borland
> delphi not having generics is such a big deal for most people.

Generics are at the top of the Java wishlist, and you can get a Java
extension (Generic Java) that implements generics in Java from
http://www.cis.unisa.edu.au/~pizza/gj/ (or from
http://burks.bton.ac.uk/burks/language/java/).

-----------------------------------------------------------------
 John English              | mailto:je@brighton.ac.uk
 Senior Lecturer           | http://www.comp.it.bton.ac.uk/je
 Dept. of Computing        | ** NON-PROFIT CD FOR CS STUDENTS **
 University of Brighton    |    -- see http://burks.bton.ac.uk
-----------------------------------------------------------------

Re: From extended Pascals to Ada 95 guide

[John English]

Marco van de Voort wrote:
> I was yesterday totally flabbergasted when I heard that the Math+IT faculty
> will teach the next generation programming with Java as introductionary
> course!

Expect some interesting times... examples might include (from a 2nd
year group I taught, having taught Ada as a first language in their
1st year):

1) Is it spelt RuntimeException or RunTimeException? Case sensitivity
   is fine as long as you *know* whether the API designers considered
   "run time" to be one word or two...
2) Is it "myDate.getMinutes()" or "myDate.getMinute()"? Can I really
   be bothered to spend my whole life with my nose in an API reference?
3) Null statements are fun:
     while (someCondition);
       doSomething();
   Shown this on a poor quality hardcopy produced by the student, the
   faint smudge at the end of the condition might not be recognised as
   a semicolon at first glance... but the compiler is happy to compile
   your code without warnings either way...
4) Things like "i = j;" or "if (i == j) ..." have a nasty habit of
   surprising students when i and j aren't simple types like ints...
5) All the compilers I use (JDK, Jikes, GJ) have very poor error
reporting
   compared to GNAT. The messages can point to anywhere within
half-a-dozen
   lines of the real error and tell you something completely misleading
   about it.
6) I'd better stop now, hadn't I? :-)

-----------------------------------------------------------------
 John English              | mailto:je@brighton.ac.uk
 Senior Lecturer           | http://www.comp.it.bton.ac.uk/je
 Dept. of Computing        | ** NON-PROFIT CD FOR CS STUDENTS **
 University of Brighton    |    -- see http://burks.bton.ac.uk
-----------------------------------------------------------------

Re: From extended Pascals to Ada 95 guide

[Charles Hixson]

[-- Attachment #1: Type: text/plain, Size: 2086 bytes --]

Generics ... it depends on how they are implemented.  I'd much prefer functions
that could tell the kind of value they were supposed to return, and selected
the correct function from a full match of the parameters.  So one could, e.g.,
distinguish between:
StringBuffer s = new ("this is a StringBuffer");
    and
String s = new ("this is a String");

Forwarding of dependencies (was that Pizza also?  Or was it Jamie?) would also
be a very nice feature, so that one could have a sort of sideways inheritance.
Very useful in systems that don't permit multiple inheritance.

Then, perhaps, generics.

Speeding things up wouldn't hurt either.

John English wrote:

> nabbasi@pacbell.net.NOSPAM wrote:
> >
> > In article <39A991F3.A8D8BED7@easystreet.com>, Al says...
> >
> > >All three of these languages, M2, M3, and Ada, now support generics,
> > >and that gives them a big advantage over Delphi for coding without doing
> > >aribtrary conversions between data types.  In Delphi, the generic type
> > >stored by the standard VCL collection classes is the Pointer, and the
> > >program must cast it to whatever type it really represents.
> >
> > Well, generics are nice, but Java do not have them, and it seems
> > that half the world now program in Java. So, I do not think that Borland
> > delphi not having generics is such a big deal for most people.
>
> Generics are at the top of the Java wishlist, and you can get a Java
> extension (Generic Java) that implements generics in Java from
> http://www.cis.unisa.edu.au/~pizza/gj/ (or from
> http://burks.bton.ac.uk/burks/language/java/).
>
> -----------------------------------------------------------------
>  John English              | mailto:je@brighton.ac.uk
>  Senior Lecturer           | http://www.comp.it.bton.ac.uk/je
>  Dept. of Computing        | ** NON-PROFIT CD FOR CS STUDENTS **
>  University of Brighton    |    -- see http://burks.bton.ac.uk
> -----------------------------------------------------------------

-- (c) Charles Hixson
--  Addition of advertisements or hyperlinks to products specifically
prohibited


[-- Attachment #2: Card for Charles Hixson --]
[-- Type: text/x-vcard, Size: 145 bytes --]

begin:vcard 
n:Hixson;Charles
x-mozilla-html:FALSE
adr:;;;;;;
version:2.1
email;internet:charleshixson@earthling.net
fn:Charles Hixson
end:vcard

Re: From extended Pascals to Ada 95 guide

[Stefan Skoglund]

John English wrote:
> 
> Marco van de Voort wrote:
> > I was yesterday totally flabbergasted when I heard that the Math+IT faculty
> > will teach the next generation programming with Java as introductionary
> > course!
> 
> Expect some interesting times... examples might include (from a 2nd
> year group I taught, having taught Ada as a first language in their
> 1st year):

note also that you need to imprint real early on a small taste
of OOP.

Understanding a bit about objects is necessary if you want
something like Hello World.

Early on it is good if the student can try ideas (even stupid ones) and
get immediate results ie like a language like ML or LISP.

The student shouldn't need to handle such technicalities like files
or units and so on.

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

Stefan Skoglund wrote:
> 
> John English wrote:
> >
> > Marco van de Voort wrote:
> > > I was yesterday totally flabbergasted when I heard that the Math+IT faculty
> > > will teach the next generation programming with Java as introductionary
> > > course!
> >
> > Expect some interesting times... examples might include (from a 2nd
> > year group I taught, having taught Ada as a first language in their
> > 1st year):
> 
> note also that you need to imprint real early on a small taste
> of OOP.

Why?

> Understanding a bit about objects is necessary if you want
> something like Hello World.

No, you can make a perfectly good "Hello World" without objects.
 
> Early on it is good if the student can try ideas (even stupid ones) and
> get immediate results ie like a language like ML or LISP.

Again, why?

> The student shouldn't need to handle such technicalities like files
> or units and so on.

A file is an important generic concept used by most languages (including
OOP ones).  The same for units, or modules.

Why do you get the idea that they are useless technicalities?

-- 

Marco van de Voort (Marcov@stack.nl or marco@freepascal.org)

Re: From extended Pascals to Ada 95 guide

[Stefan Skoglund]

Marco van de Voort wrote:
> > Understanding a bit about objects is necessary if you want
> > something like Hello World.
> 
> No, you can make a perfectly good "Hello World" without objects.
Dont you need to create a subclass of some obscure Java
class to even generate a program with a null body ??

> > The student shouldn't need to handle such technicalities like files
> > or units and so on.
> 
> A file is an important generic concept used by most languages (including
> OOP ones).  The same for units, or modules.
> 
> Why do you get the idea that they are useless technicalities?

Because in the beginning of CS1 you don�t really need persistent storage
of your work. You as a teacher would prefer not getting questions like
this:

Pupil: I'm working with a small a little bit of trigonometry.
      Why is the system bitching about no cos ?

Teacher: you need to add something like this '-lm'

What I'm saying is this: a language which is likened by programmer
professionals isn't implicitly a good way of expression for students
which
has started programming for the first time.

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

> > > Understanding a bit about objects is necessary if you want
> > > something like Hello World.
> >
> > No, you can make a perfectly good "Hello World" without objects.
> Dont you need to create a subclass of some obscure Java
> class to even generate a program with a null body ??

I was talking about programming in general.
If you need that for Java, I consider that just another reason to NOT
use Java.
 
> > > The student shouldn't need to handle such technicalities like files
> > > or units and so on.
> >
> > A file is an important generic concept used by most languages (including
> > OOP ones).  The same for units, or modules.
> >
> > Why do you get the idea that they are useless technicalities?
> 
> Because in the beginning of CS1 you don�t really need persistent storage
> of your work. You as a teacher would prefer not getting questions like
> this:

You also don't need to with e.g. Pascal. So that is no argument for
Java.

> Pupil: I'm working with a small a little bit of trigonometry.
>       Why is the system bitching about no cos ?
 
> Teacher: you need to add something like this '-lm'

That is than the problem of one language  (and probably not even all
compilers). Not a reason why all languages except Java are invalid.
 
> What I'm saying is this: a language which is likened by programmer
> professionals isn't implicitly a good way of expression for students
> which
> has started programming for the first time.

Not implicitly no. But you don't have to use all features directly in
the first week. I assume you don't do that for Java and its libraries
either.

-- 

Marco van de Voort (Marcov@stack.nl or marco@freepascal.org)

Re: From extended Pascals to Ada 95 guide

[Ken Garlington]

"Marco van de Voort" <marcov@stack.nl> wrote in message
news:39BBB97E.FD03B359@stack.nl...
> > > > Understanding a bit about objects is necessary if you want
> > > > something like Hello World.
> > >
> > > No, you can make a perfectly good "Hello World" without objects.
> > Dont you need to create a subclass of some obscure Java
> > class to even generate a program with a null body ??
>
> I was talking about programming in general.
> If you need that for Java, I consider that just another reason to NOT
> use Java.

// example complete Java "Hello World" program
class HelloWorld {
  public static void main (String args[]) {
    System.out.println("Hello World!");
  }
}


// alternative complete Java "Hello World" program, as a Web-page applet
import java.awt.Graphics;
public class HelloWorldApplet extends java.applet.Applet {
  public void paint(Graphics g) {
    g.drawString("Hello World!", 5, 25);
  }
}

Re: From extended Pascals to Ada 95 guide

[John English]

Stefan Skoglund wrote:
> John English wrote:
> > Marco van de Voort wrote:
> > > I was yesterday totally flabbergasted when I heard that the Math+IT faculty
> > > will teach the next generation programming with Java as introductionary
> > > course!
> >
> > Expect some interesting times... examples might include (from a 2nd
> > year group I taught, having taught Ada as a first language in their
> > 1st year):
> 
> note also that you need to imprint real early on a small taste
> of OOP.

Indeed. I'm a fan of Lynn Andrea Stein's "Rethinking CS1" approach,
myself (http://www-cs101.ai.mit.edu/ipij/).

-----------------------------------------------------------------
 John English              | mailto:je@brighton.ac.uk
 Senior Lecturer           | http://www.comp.it.bton.ac.uk/je
 Dept. of Computing        | ** NON-PROFIT CD FOR CS STUDENTS **
 University of Brighton    |    -- see http://burks.bton.ac.uk
-----------------------------------------------------------------

Re: From extended Pascals to Ada 95 guide

[Ehud Lamm]

On Mon, 11 Sep 2000, John English wrote:

|Stefan Skoglund wrote:
|> 
|> note also that you need to imprint real early on a small taste
|> of OOP.
|
|Indeed.

I am really not sure about this.
Sometimes good intentions lead to bad results... Indeed, exposure to OO
concepts is important, however in my experiense those who learned Java
first have huge problems with simple imperative notions like loops and
structured programing (i.e., breaking code into routines).
I find that, when taught well, most students with Pascal/C background
learn the OOP approach rather easily (Note: I am not talking about
becoming experts in patterns, frameworks or the more inticate uses of
inheritance). 
However, it is quite uncommon to tell students who learned Java as a first
language, that they have to take a course on "structured programming" -
only so that they relearn about loops and such,.
Since this is not done, they usually stay, almost, programming
illiterate. 


Who knows? Maybe the best approach will be to teach a functinal/object
language first... :-)

Ehud Lamm mslamm@mscc.huji.ac.il
http://purl.oclc.org/NET/ehudlamm <== My home on the web 
Check it out and subscribe to the E-List- for interesting essays and more!

Re: From extended Pascals to Ada 95 guide

[Marco van de Voort]

>|Indeed.
>
>I am really not sure about this.
>Sometimes good intentions lead to bad results... Indeed, exposure to OO
>concepts is important, however in my experiense those who learned Java
>first have huge problems with simple imperative notions like loops and
>structured programing (i.e., breaking code into routines).

That is my experience too. 

Re: From extended Pascals to Ada 95 guide

[John English]

Ehud Lamm wrote:
> On Mon, 11 Sep 2000, John English wrote:
> |Stefan Skoglund wrote:
> |>
> |> note also that you need to imprint real early on a small taste
> |> of OOP.
> |
> |Indeed.
> 
> I am really not sure about this.
> Sometimes good intentions lead to bad results... Indeed, exposure to OO
> concepts is important, however in my experiense those who learned Java
> first have huge problems with simple imperative notions like loops and
> structured programing (i.e., breaking code into routines).

At Brighton, we teach Ada in the first year and Java in the second
year for exactly this reason... ;-)

-----------------------------------------------------------------
 John English              | mailto:je@brighton.ac.uk
 Senior Lecturer           | http://www.comp.it.bton.ac.uk/je
 Dept. of Computing        | ** NON-PROFIT CD FOR CS STUDENTS **
 University of Brighton    |    -- see http://burks.bton.ac.uk
-----------------------------------------------------------------