Rogers and BitTorrent: another datapoint

[Christopher Browne]

On 10/28/06, JoeHill <joehill-rieW9WUcm8FFJ04o6PK0Fg at public.gmane.org> wrote:
> On Sat, 28 Oct 2006 13:23:28 -0400
> Michael MacLeod got an infinite number of monkeys to type out:
>
> > I'm not being a Java apologist here, I'm just glad that I understand both
> > the OO and the procedural paradigms, and that I know both C and Java (in
> > addition to others) equally well. I think that these skills and the ability
> > to know when to apply which ones will make me much more valuable to any
> > potential employers than a programmer hellbent on sticking to the one
> > language he/she knows.
>
> That's a whole other argu...er, discussion, though.
>
> But you've got two different debates goin' on here:
>
> 1. Java is just annoying. It's bloated, it's really unnecessary (same tasks
> accomplished with less overhead), and, correct me if I'm wrong, but is it not
> easily used as a tool for creating malware? (note I said 'easily', as in
> 'script kiddie' easy)
>
> Agree, disagree, whatever.

It's bloated and it's not...

You need to build a fair bit of infrastructure, whether that be:
a) J2EE-like stuff
b) Other libraries you're dependent on
c) Some bunch of homegrown dependancies
in order to get Java to a point of usefulness.

Almost certainly, by the time you get to 'a point of usefulness,'
you're invoking Greenspun's Tenth Rule.
<http://en.wikipedia.org/wiki/Greenspun's_Tenth_Rule>  That is, Any
sufficiently complicated C or Fortran program contains an ad hoc,
informally-specified, bug-ridden, slow implementation of half of
Common Lisp...

As far as "script kiddie misuse" is concerned, I'm not sure Java
installations are standardized enough for that to work out
meaningfully.
- There is a designed-in security model which is of
somewhat-greater-than-zero value;
- Nonstandardization of "where Java goes" (our installs at work are
definitely not in any 'official standard' location) means it's liable
to be harder to hijack than, say, C...
- Unlike C/C++, Java has somewhat stronger typing and tries rather
hard to avoid pointer referencing errors, which should make Java apps
somewhat harder to hijack via (say) buffer overruns.

The really irritating thing that has happened over the last 10-15
years is the growth of unthinking espousal of "abject object
orientation."  It seems to me like the computer science equivalent to
Lysenkoism.  <http://en.wikipedia.org/wiki/Lysenkoism>    Only in some
of the more unreadable areas is a "theory of objects" emerging (such
as Pierce's work <http://en.wikipedia.org/wiki/Benjamin_C._Pierce> on
type theory).

OO is, to my mind, merely another programming technique to be applied
when useful, NOT fitting of being the end-all of how systems ought to
be "shaped."

To my mind, the common focus on "what are the object classes?" or
"show me the UML!" are mistaken replacements for analysis of the
problem that is to be solved.
-- 
http://www3.sympatico.ca/cbbrowne/linux.html
Oddly enough, this is completely standard behaviour for shells. This
is a roundabout way of saying `don't use combined chains of `&&'s and
`||'s unless you think Gödel's theorem is for sissies'.
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists