Posted by Schwab,Wilhelm K on Oct 13, 2004; 8:34pm
URL: https://forum.world.st/an-online-class-hierarchy-tp3371800p3371840.html

Tom,

> Well that was bad news but I note that better news has followed in other
> replies.  Primarily I am (probably) most interested in the MVP aspect so I'd
> like to see what the base model and view provide along with whatever they
> are subclassed from of course.  Ultimately what happens when browsing docs
> however is something strikes ones interest and that leads to something which
> then leads to something else.  I'll guess that I wouldn't wander too deeply
> into the representations of numerics, booleans. stacks, queues,
> linked-lists, et. al.  I can pretty much guess how they work.

It sounds as though you are already aware that you can see how they work
from the trial version.  I probably also do not need to tell you that
the code is often more informative than the method-level comments.


> Ultimately I would need to be able to produce a listview containing the
> names and id's of customers (for instance.)  Then choosing from the list
> fetch, edit and then save the properties of that customer.

You might want to look at Ted Bracht's book.  There are also some
tutorials lurking around; IIRC, a "challenge" here recently resulted in
something much like you probably want.

MVP starts out a little confusing, but after a while, you will begin to
wonder why it was so difficult.  I keep hoping we'll get better at
reducing the time for that to happen.  Trust me that it does happen.



 > The kinds of
> things that a typical business would expect out of typical software.  The
> companies are already doing these things in C++, VB6 and what have you.
> Some are considering a migration to C# or VB.Net.  I can't point at a
> single-user version of Etch-A-Sketch and suggest they consider using
> Smalltalk based upon that.  Well, I _can_ suggest it but they aren't likely
> to do it.

Does this make you a Smalltalker trying to convert a shop?  Are you an
employee or consultant/other?  I spotted Smalltalk long before I was
able to afford it, either financially or in time to learn it.  The price
dropped first thank to a Digitalk educational discount: $50 for ST/V
DOS.  I spent a while trying to do things with, setting _very_ short
deadlines for switching to things that I knew how to use.  One day I
actually beat the deadline, and I was hooked :)  By that time, ST/V Win
16 hit (also educationally discounted) $100.  That was starting to reach
its limits, or at least the limits of what I could do with it at the
time, when Dolphin came along.  Then I bet the farm on my ability to use
Smalltalk to deliver what is now my most productive cash cow.  It might
be more accurate to say that I bet the farm that I would not be able to
do it any other way, and I suspect I was correct.

Most C* developers are not going to see the benefits of Smalltalk, and
many don't even want to try to see it.  If you can get anything based on
Smalltalk in front of them, it seems best to first let them observe that
"Tim's thing doesn't crash".  Make some changes in a few minutes and let
them observe that it still doesn't crash.  Then get them to agree to
building something sufficiently complex that things will eventually go
wrong.  Now you hit them with crash dumps, logs, and reflectivity
(introspection at least) to get the program to at least tell you what it
was doing at the time of an error or (very rarely) a crash.

Just my 2 asCents on the subject.


> Similarly pointing out "IBM uses it" doesn't mean much to companies one
> one-millionth the size.  IBM uses everything somewhere.

The financial industry is a good source of success stories.  The broker
who can sell this AM's newest derivatives is the one making money.  I
haven't checked in a while, but JWARS and FedEx at least were projects
to mention.  I like to run these things down immediately before
inserting my foot in my mouth though =:0

In general, take a look at http://www.whysmalltalk.com.

Happy Smalltalking!

Bill

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