[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Scriptometer: measuring the ease of SOP (Script Oriented Programming) of programming languages



Daniel Weinreb <DLWeinreb@attbi.com> writes:

> Pixel wrote:
> 
> >Ken Anderson <kanderson@bbn.com> writes:
> >
> >>I like your approach, but i think it is weighted too much toward the
> >>empty expression, and "hello world" examples which aren't representative of
> >>scripts people write.
> >>
> >
> >the "empty" is merely for calibrating the scoring,
> >
> As I said in offline mail to you, the "empty" is really a cheap shot and
> doesn't demonstrate anything useful.  It hurts the credibility of the
> exercise and makes it look like we're trying to diss Java just for the hell of
> it.
> This is not needed.  You can make your points much more substantively.
> Don't give the critics cause to make you look like you're biased.

on the contrary, i'm using the "empty" (now renamed "smallest", is
that better?) to calibrate a little the other examples. The idea is to
remove the initial overhead of a program in java:

the java "hello_world" would be 107. Since i remove half the size of
"smallest" (which is 68), it gives 107-68/2 = 73 instead. I used to remove the
whole "smallest" size, giving 107-68 = 39, but i thought it did not
favour C# and C enough.

removing "smallest" and this callibration, it gives:

http://people.mandrakesoft.com/~prigaux/index_no_smallest.html

where java scores even worse.

WYT?

[...]

> >>How about something a little bigger, like writing something that could compile
> >>everything that needs compiling in a directory and it subdirectories.
> >>
> That sounds great.

ok, i try to add it (i may need some help though, it gets hard in some
languages, esp. for things http://pleac.sf.net/ doesn't help!)