Learning Squeak
Alan Kay
Alan.Kay at squeakland.org
Sat Jan 19 23:22:09 UTC 2002
Les --
I think it would be great for some of the experienced Smalltalkers on
the list to take a crack at making a "filtered Browser" for a core
Squeak for beginners. How about ~ 100 classes to do most things? My
personal preference would be for a set of abstractions that include
Morphic instead of MVC, but any nice filtering would be a great
start. To do this really nicely, it might require a few new classes
to be made that are the higher-level abstraction for what is now
"over subclassing" in the current system.
Cheers,
Alan
At 1:46 PM -0800 1/19/02, Les Tyrrell wrote:
>( In Noel's reply, Lex wrote: )
>> > The hard parts of Smalltalk are things like OO design, the code browser,
>> > the way collections and streams work, and so on -- not the syntax.
>
>Definitely. When you've stripped away all the other ways to make a
>programming environment complex, this is all that is left. There is nothing
>in the programming environment to stand between you and the core concepts in
>the object-oriented approach to programming. While these should be natural
>enough to understand, it seems that for many people these things are still
>hard to grasp at first, irregardless of the language used to express them.
>But I would still definitely prefer to reccomend Smalltalk as the language of
>choice for this.
>
>> Which gets us full circle back to the original point:
>>
>> > [The] lack of really good tutorials geared to take existing
>> > C(++)/Pascal/Java programmers (or any programmer for that matter) from
>> > newbie to mastery is one of the most significant failings.
>>
>> You made too much out of the "odd syntax", which was in quotes in the first
>> place to represent it as a somewhat tongue-in-cheek comment. Odd? LISP
>> (which I loath) and APL (which I love) are odd, not Smalltalk.
>>
>> > It's taken me years to learn Smalltalk, too
>>
>> It shouldn't, but look how many times that's been said recently.
>
>Keep in mind that many of us have said it in the context of Smalltalk not
>being available to mere mortals shortly after being publicly announced. It
>took me years to learn because first there was no way to learn except to read
>about it, and then because I could not afford to buy a computer to run it on
>when commercial versions did become available. Once I had both of those
>things, it went relatively smoothly. Perhaps it is better to mention that,
>based solely on what I had read in 1981, I was willing to put the effort into
>learning as much as I could long before it was even possible for me to use it.
>I bought Digitalk's Smalltak/V for DOS and a Logitech mouse years before I
>ever owned a computer myself- the same is also true of V/286 and V/PM. In
>the early years I could only play with these versions for a few minutes at a
>time on a computer owned by a friend, or in computer stores, or ( much
>later ), in campus computer labs. So it isn't exactly a fair assessment to
>take these anecdotes as being indicative of the expected learning curve!
>
>However, I do sometimes feel that in other ways I had an advantage over
>today's newcomers- V/DOS only had about 90 classes visible to the programmer.
>V/286 ( I believe ) had fewer than 300. It was *much* easier to get a grasp
>of the important, useful classes in such a smaller system. I would not want
>to learn by wading through the thousands of classes that are more typical now,
>wondering which of those many hundreds of classes are *really* essential to
>getting things done for the problems I want to solve. In that sense, I could
>believe that learning Smalltalk has become much harder. But, you can only
>learn something for the first time once in your life. So, I'd have to hear
>more from the "newbies" about what they encountered in discovering and
>learning Smalltalk to pass any further judgement on that.
>
>- les
--
More information about the Squeak-dev
mailing list
|