an online class hierarchy?

Tom,

 > we're not all heathens.

No, but I start to wonder whether you might be a troll.  If you are not,
please note that we are a helpful bunch and will certainly try to help
you succeed at Smalltalk.  The first step in that is (often, and not
just for heathens) to get the newbie to recognize that the key is in the
browser/image, and that many things they think are necessary are simply
ways to compensate for hurdles their current languages and tools place
before them.

Sincerely,

Bill

--
Wilhelm K. Schwab, Ph.D.
[hidden email]

Tom,

> > Smalltalk's dynamic typing and use of blocks of code allow things
> > to do far more than you might envision.  You really have to dive in
> > in order to get your brain around it.
>
> No doubt but Smalltalk isn't the only typeless language as you know.
> I used to work with Clipper which is both typeless and supports code
> blocks.  Interestingly BCPL (the precursor to B and ultimately of the
> C language) was (and still is) typeless and supports anonymous
> procedures.

Well, Smalltalk *isn't* a typeless language; it is strongly typed
(unlike BCPL which I worked with some years ago; I'm not sure about
Clipper). The difference is that the types in Smalltalk are dynamically
checked at runtime and by the receiver of a message. In statically
typed languages the types are checked at compile time by the compiler
(obviously) and at every function call site, rather than by the called
function.

Both statically and dynamically typed languages are generally
"type-safe". But one can go further and split languages into strongly
and weakly typed. In a strongly typed language the type system cannot
be circumvented and broken. With this in mind you have the following:

Dynamically, Strongly typed = Smalltalk
Statically, Strongly typed = Haskell
Statically, Weakly typed = Java, C++ (by nature of their cast operators
which can break stuff)
Typeless = BCPL, K&R C.

Best regards


Andy Bower
Dolphin Support
www.object-arts.com

"Bill Schwab" <[hidden email]> wrote in message
news:ckmsaj$vbi$[hidden email]...
> No, but I start to wonder whether you might be a troll.

Uhhh... I don't have the time for you Professor Schwab.

> Well, Smalltalk *isn't* a typeless language; it is strongly typed
> (unlike BCPL which I worked with some years ago; I'm not sure about
> Clipper). The difference is that the types in Smalltalk are dynamically
> checked at runtime and by the receiver of a message. In statically
> typed languages the types are checked at compile time by the compiler
> (obviously) and at every function call site, rather than by the called
> function.

I get the feeling this has become something of a debate and to be perfectly
honest nothing could be less interesting to me.  You are no doubt correct.

I was able to locate the following in Ward Cunningham's Wiki... you'll see
they list numerous languages, they arrange the languages into a quadrant and
importantly that there is an honest ongoing discussion of the definitions.
http://c2.com/cgi/wiki?TypingQuadrant

Personally... I don't need to debate right and wrong when it comes to
programming languages.  I get paid (as many of us do) to solve business
problems.  Lots of decision points go into that equation right?

Don't get me wrong but that was asked somewhat rhetorically, you don't have
to answer.  I excused myself from comp.lang.smalltalk because I knew in time
it would come down to somebody calling me names.  It almost happened here,
caught me a little earlier than I thought but I wasn't careful enough in
hindsight.  I've got what I need now and thanks for your help.

Tom

Hi Andy...

> Statically, Weakly typed = Java, C++ (by nature of their cast operators
which can break stuff)

This is IMO not correct. In Java Casts are safe in that they issue a
ClassCastException when the cast is not possible due to incompatibility. It
is not possible like in e.g. C/C++ to cast "Apples" to "Pears" and happily
write into a completely incompatible binary representation. Java is as
strongly typed as Smalltalk, but as you said, checks types statically rather
than at runtime.

Java and C++ are REALLY different (I developed in both languages), sentences
stating Java/C++ in context of OO I therefore rarely found to be correct.

Ciao

...Jochen

Jochen,

Ok, you're right, of course.

Best regards

Andy Bower
Dolphin Support
www.object-arts.com

Tom Leylan wrote:
> "Bill Schwab" <[hidden email]> wrote in message
> news:ckmsaj$vbi$[hidden email]...
>
>>No, but I start to wonder whether you might be a troll.
>
>
> Uhhh... I don't have the time for you Professor Schwab.

Raw nerve??

Bill

--
Wilhelm K. Schwab, Ph.D.
[hidden email]

"Jochen Riekhof" <[hidden email]> wrote in message
news:416f7db4$[hidden email]...
Well no, since it is only really strongly typed at runtime. So you have all
the "weight" of static typing bearing down on you during development, and
yet that up front cost provides you no guarantee that you will not get
runtime type errors. Of course additional type information does have uses
for both programmer and tools, but one does have to question whether the
compromise is worth it in these hybrid languages.

Regards

Blair

"Bill Schwab" <[hidden email]> wrote in message
news:ckodd6$14m6$[hidden email]...
You're determined right?  Your turn!

Tom,
Smalltalk is NOT a programming language, is an ambient. There are only 5
reserved words, and in fact they could not exist and it would still be a
Smalltalk. You can change the sintax and it would still be a Smalltalk
too. So don't look at the class hierarchy or language capabilities, is not
important, don't think that you will find a programming language called
Smalltalk, is not like that. You don't program writting text, in Smalltalk
text is just one way to communicate with the already running objects. And
is not the only way to communicate with objects.
In Smalltalk there is no such thing as run-time or design-time or
compile-time, it's complety different to everything you compared, you
talked about (just) programming languages. Those programming languages are
very good ones, but are just that, declaratives ways of making software.
When you talk about reading papers in the park I don't understand, you
can't do Smalltalk in paper, you can do C or Clipper because they are
languages, they are text. Smalltalk does not run on paper because is an
ambient, Java does because is text. You also talk about something called
sun I don't understand either, the only light I see come from my monitor
:-).

Diego Coronel

DiegoC wrote:

[... all snipped ...]

Nice.

    -- chris

"DiegoC" <[hidden email]> wrote in message
news:[hidden email]...
Here! Here! (and various other contented and supportive murmurings from the
back benches)

Regards

Blair

(Apologies if this reference to the British system of representative
parliamentary democracy does not translate well)

> Here! Here! (and various other contented and supportive murmurings from
> the back benches)
>
> (Apologies if this reference to the British system of representative
> parliamentary democracy does not translate well)

It translated well for me.. then again, I'm Australian.

Australia, the social experiment the poms conducted when they sendt all the
scum of society to the other end of the earth....

Diego,

Blair McGlashan wrote:

> Here! Here! (and various other contented and supportive murmurings
> from the back benches)

In fact, can we place a quote from your message on the new OA website
(suitably attributed to you of course)?

Best regards,

Andy Bower
Dolphin Support
www.object-arts.com

Chris:
> DiegoC wrote:
> [... all snipped ...]
> Nice.

You'll, and anybody who is interested, can find more about this approach on
objects technology at www.smalltalking.net.

Best regards.

--
Esteban.

Andy,
If you mean my email, would be an honour to put on OA site, but please
correct my grammar, english is not my native language.

Diego Coronel

Blair,

> Here! Here! (and various other contented and supportive murmurings from the
> back benches)
>
> (Apologies if this reference to the British system of representative
> parliamentary democracy does not translate well)

In the US, it's more "There! There!" - the sound of passing the buck :)

Have a good one,

Bill


--
Wilhelm K. Schwab, Ph.D.
[hidden email]