Re: Programming skills required by mathematicians?

On Jun 27, 7:46 am, Phil Hobbs
<pcdhSpamMeSensel...@[EMAIL PROTECTED]
> wrote:
> Marshall wrote:
> >>>> Or did I miss something ultra-im****tant?
> >>>
> >>> You did, actually: the existence of other languages that are better
> >>> at those things.
> >>
> >> So in other words, "No."
>
> > If you are happy with where you are, and are not concerned
> > about learning anything beyond what you already know,
> > then by all means continue as you have been.
>
> > If you are going to advise other people, then doing so from
> > a place of ignorance is being less than fully responsible
>
> Ooooohhhh.  I'm put in my place, for sure.

Not really. You've been exposed publicly as an ignorant, but
to put you in your place requires you to realize it, and for that
to happen you'd first have to overcome your difficulties with
reasoning and reading comprehension.


> How much physics do you know? How many actual working things that are
> not programs have you built? How many theorems have you published?

Here I see that you specifically exclude experience with programs
from a question about programming, and attack the person and
not the message. And apparently since the above is your list of
non-programming-related things, it's clear you think theorems are
not related, either. How interesting.

Please explain how my level of understanding of physics will
help readers interpret my comments on the topic of what
programming languages a math major should pay attention to.


> If the OP wants to be a language lawyer, or primarily a computer
> scientist, then of course he doesn't want to be limited to C++. On the
> other hand, some of us need to build things that just work.

The implication that people who need to build things that just work
should be limited to C++ is flat-out hilarious.


> I don't claim that C++ is the only way to do this.

No; just the only way you're interested in.


> On the other hand, your
> defending Java as a numerics language is, well, parochial,

How amusing that you think I defended Java for numerics.
Clearly you have some difficulty taking in written information.
And the one advocating for a single language approach is
calling the one advocating for a broader dialog on programming
languages "parochial"! It's like rain on your wedding day.


> and your defense of JIT compilers as producing *cough* more
> efficient code than static compilers

I described Java's performance as "almost at the level of C++".
Somehow you read that as "more efficient"? I didn't even mention
"JIT compilers". I did contrast static compilation with a HotSpot
VM approach, in terms of cl***** of optimizations possible.
I pointed out that a compiler that only has access to static
information cannot perform as many optimizations as one that
in addition has access to some runtime data. Do you disagree?
Do you see this as a language-dependent issue?


> suggests that your interests don't overlap mine by very much.

At last a true statement! I am interested in programming languages
and intellectual honesty.


> I care about getting stuff working, and C++ works for me and for
> a lot of other people. You don't like it, but I can't help that.

Actually I explicitly said "by all means continue as you have been."
Somehow you get from this that I'm not happy.


> When JIT compilers vectorize loops well or sup****t MPI, I'd be quite
> happy to use Java, if it offered something more than "See me! I can do
> that too!"

Whatever. My post mentioned C++, Java, Fortran, OCaml, SML,
Haskell, Prolog, Mercury, SQL, Erlang, Ruby, Python, and Perl.
Yet somehow you've reached the conclusion that I'm a Java
lover. I guess that as a rabid single-language advocate, you
can't imagine someone might have a more nuanced position than
"my language is better than yours."


Marshall