[E-Lang] empirical and statistical comparison of languages

[Mark S. Miller]

At 09:40 AM Monday 3/5/01, zooko@zooko.com wrote:
>http://wwwipd.ira.uka.de/~prechelt/Biblio/jccpprt_computer2000.pdf

Looks like an interesting paper.  I wish he'd included some very different 
languages as well.  To take just two examples, compared to Prolog and APL, 
all the languages he compares are more similar to each other than different.

Nevertheless, for E's purposes he makes the right choice, as E is trying to 
appear to be a member of the same family.  Actors in C's clothing.


>I especially recommend to E folk the comments at the end about using an extant
>data structure (e.g. tree or hashtable) or writing a new one.

Quoting the comment:

>It is an interesting observation that despite the existence
>of hash table implementations in both the Java and
>the C++ class libraries none of the non-script programmers
>used them (but rather implemented a tree solution by hand),
>whereas for the script programmers the hash tables built
>into the language were the obvious choice.
>that need to handle fair amounts of computation and
>data. Their relative run time and memory consumption
>overhead will often be acceptable and they may
>offer significant advantages with respect to programmer
>productivity — at least for small programs like the

A very interesting observation.  Unfortunately, in a quick skim, I didn't 
see that he anywhere speculated on why.  Certainly if I were programming in 
Java, I expect I would have used the existing collection class library, and 
I find it strange that these Java programmers didn't.  OTOH, in C++ I would 
have written my own collection classes before touching the STL.

In any case, whatever these reasons are, I believe the E programmer will 
chose to use the existing collection classes, whether those classes are the 
current ones or Hydro-based ones.  If you see any reason to be less 
confident that we're "scripting-like" in this regard, please lemme know.


        Cheers,
        --MarkM