On 10/31/06, Jim Weirich <jim.weirich@xxxxxxxxx> wrote:
On 10/30/06, Edward Sumerfield <esumerfd@xxxxxxxxxxxxxx> wrote:
> I am not suggesting that Republicans like strongly typed languages,
> just that there are lots of common characteristics between those that
> do and those that don't.
Oh wow! If I didn't like the strong typing/checked exception
generalization, you can imangine my reaction to the above. MUST RESIST!
If you want to see the canonical work on assigning politcal views based
upon OO design beliefs, check out comp.object circa 1998 when Eliot declared
anyone who didn't create comprehensive, fully ellaborated, graphical ( e.g.
UML) designs before starting coding must be represive right-wingers.
I new that analogy would get me in trouble. I didn't resist though.
My intention was analogy, not comparison. A is to B as C is to D but A
and B are not comparable to C or D. Two exclusive sets, complete unto
themselves but unrelated.
I wonder if you have heard the argument that says that two opposing
views tend to diverge over time because of the human tendency to
overly express their perspectives to achieve convincing arguments.
This idea came out of a historical look at the political divide and
how it has changed. Anyway, thats why the analogy struck me as
appropriate.
> My thought was that the problem exists within the Java community
> alone, and then is also addressed in conversations about whether we
> should use alternative languages.
I agree that the problem exists in Java alone. That's why I objected to the
broad generalization of the statement. The original was:
You will tend to find that people that like strongly
typed languages also like checked exceptions, where as those who
prefer dynamically typed languages tend toward the unchecked variety.
I agree with your disagreement, I wonder if my perception more tightly
correlates in the Java community and is less relevant elsewhere.
Perhaps we could increased the correlation by limiting our set to
those Java programmers that have not experienced other languages. Of
coarse, this line of thought leads me to the unavoidable conclusion
that my ideal set is those few of people that I have met and agree
with the statement :)
First it is a false dicotomy. Dynamically typed is not the opposite of
strongly typed, but of statically typed. Second, you will find broad
disagreement amoung those who favor statically typed languages on value of
checked exceptions.
Having been lucky enough to hear your presentation on typing
techniques a few times, I continue to have to admit to not having the
best grasp of each of the facets. I was originally going to use the
"strong" and "loose" terms but you have scolded me on those terms in
past discussions. I learn, but not well.
One interesting result from this thread is that we have no posts that
strongly argue for the use of checked exceptions everywhere. Mark
mentioned that a checked exception should be used when "calling code
can make a recovery from the failure" which I interpret as a
semi-convinced statement.
Not sure what our list membership is now (over 1000?), obviously we
have a lot of shy lurkers out there but I was expecting a more
categorical divide on this point.
Ed "the inciter" Sumerfield
|