<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
Tim Rowledge wrote:<br>
<blockquote cite="midfb132c084d.rowledge@Gravious.shaw.ca" type="cite">
<pre wrap="">Rob Lally <a class="moz-txt-link-rfc2396E" href="mailto:smalltalk@roblally.plus.com"><smalltalk@roblally.plus.com></a> wrote:</pre>
<blockquote type="cite">
<pre wrap="">f course this is a highly contrived example, purpose designed to make
java look bad by using contrived non-idiomatic code.
</pre>
</blockquote>
<pre wrap=""><!---->No, it's trying to compare vaguely alike code. The original algorithm
is clearly a nonsensical piece of code with no plausible value for the
real world. There aren't many places where performance of a trivial
loop doing trivial arithmetic on guaranteed same-type value is actually
important. (except perhaps in signal processing type work)
But if one has to use such pointless code at least it should be
producing realistic semantics and since Smalltalk treats numbers as
proper objects and not some dememented 'primitive type' with restricted
capabilities then the java code must be written to do the same.
</pre>
</blockquote>
By that logic we could write a small java app to loop over some
primitive types and add them together and then declare that since there
is no Smalltalk analog that Smalltalk is less useful, capable and
powerful than Java. Something that is obviously nonsense.<br>
<blockquote cite="midfb132c084d.rowledge@Gravious.shaw.ca" type="cite">
<pre wrap="">For a more meaningful comparison one must use bigger programs with plausible
behaviour. Not that I've actually looked, but I can't say I've ever
actually seen a java program that does anything useful. Please, please
don't anybody send me a list though - I really don't care.</pre>
</blockquote>
<br>
Don't care? I have to say I reckon you do.<br>
<br>
The organization where I work has about 5000 (yup that's thousand) java
applications in production, most of which do something useful, at least
to them; so a list would be rather protracted. <br>
<br>
Since these applications were written by developers much like any
other, you could probably guess the names of most of them - take any
Greek or roman god, mountain, sea, forest, jungle, animal, sport,
robot, sci-fi character or philosopher and you'll probably hit one. For
the remaining a random TLA/ETLA generator would be required.<br>
<br>
My company also has a number of Smalltalk applications deployed, 10's
of applications I believe, rather than thousands, but they are in
general well thought of and gaining mind-share.<br>
<br>
I don't see why Java has to be bad for Smalltalk and Squeak to be good.
Both are excellent tools with strengths and weaknesses.<br>
<br>
I'm sure that most IT managers have a heavy preference for Java, even
in cases where Smalltalk would be better, which is unfortunate; for
them, for their organization and for developers who would like to work
in Smalltalk (although we have Smalltalk vacancies we just can't seem
to fill). <br>
<br>
I'm sure that the Smalltalk community feels that these are attitudes
they have to battle against, and I wish them luck. But battle against
it with positive information, by demonstrating the power and
flexibility of smalltalk, by demonstrating the creativity and
innovation that has placed Smalltalk developers at the forefront of the
OO, patterns and agile movements, by demonstrating the ability to
invent the future. Smalltalkers have been in the vanguard of almost
everything good in software over the past two decades, use that and
build on it. Don't do it by trying to bash the opposition, you're
better than that.<br>
<br>
Here endeth the rant.<br>
<br>
Sorry about that, it all just sort of ... welled up in me.<br>
<br>
<br>
Rob.<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
</body>
</html>