Dear All,
I've just read the abovementioned report. URL pointer:
ftp://ftp.dfrc.nasa.gov/pub/x3j3/ncsa/doc/meeting/140post/97-106.txt.gz
Personally, I'm a little bit in horror. I'm wondering why I don't
see any alarming messages about just one simple and visible issue:
EACH OBJECT INSTANCE is supposed to hold all pointers to methods.
That means huge overheads for fine grain OO programs (say two or
three simple data objects, e.g., REAL numbers, plus a similar number
of pointers). With a large array of such objects you can forget any
dreams of efficiency.
Thus a half of OO uses are thrown out of the window. You can use
OO only when you have only very large objects with relatively few
methods, or when you have very few instances of the objects.
What was so wrong with, e.g., virtual tables that THIS is better?
Still terrfied but recovering,
Artur Swietanowski.
----------------------------------------------------------------------
Artur Swietanowski mailto:[log in to unmask]
Institut fuer Statistik, Operations Research und Computerverfahren,
Universitaet Wien, Universitaetsstr. 5, A-1010 Wien, Austria
tel. +43 (1) 407 63 55 - 120 fax +43 (1) 406 41 59
----------------------------------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|