http://www.xs4all.nl/~jmvdveer/algol.html

I would like to thank all persons who have helped over the past years.
Algol68G would not be what it is without their help.

To satisfy my curiosity about this "Improved interpreter efficiency" I
did a crude performance test.

In the (truely crude) tradition of the bogo_mips program I wrote some
"gips" snippets in various languages to see what the difference was.

$ cat ResultsDec06-O5.txt
c-O5_gips:      0.10841284     -64%     slower then c
ada-O5_gips:    0.06760411     -43%     slower then c
f77-O5_gips:    0.05355613     -28%     slower then c
a68r-O5_gips:   0.04530627     -14%     slower then c
c_gips:         0.03880782       0%     slower then c
ada_gips:       0.03556188       9%     slower then c
a68r_gips:      0.03387534      15%     slower then c
f77_gips:       0.02894859      34%     slower then c
Interpretors:
a68g_gips.a68:  0.00045913     84 x     slower then c
py_gips.py:     0.00039505     97 x     slower then c
pl_gips.pl:     0.00029557    130 x     slower then c
gawk_gips.awk:  0.00028388    136 x     slower then c
ruby_gips.ruby: 0.00023380    165 x     slower then c
bc_gips.bc:     0.00010461    370 x     slower then c
csh_gips.csh:   0.00000220  17639 x     slower then c
bash_gips.sh:   0.00000182  21322 x     slower then c
nawk_gips.awk:       not installed.
sh_gips.sh:          not installed.

Conculsion: A68G... NOT BAD!!
My thanx to Marcel

Happy New Year all.
NevilleDNZ

gips psuedo code:
#!/usr/bin/env a68g
INT double:=1024;
INT max=double*double;
FOR i TO max DO
  IF i >= double THEN
    print (("i=",i," "));
    double*:=2
  FI
OD;
print ((newline,"DONE",max,newline))