Okay, we make mistakes in the ST Section too. In the December, 1985 review of 4xFORTH we
compared our October, 1985 Sieve of Eratosthenes test of 3.8 seconds for Digital Research
C with a onepass time of 1.1 seconds for 4xFORTH. Unfortunately, the C benchmark was for
10 passes, so the equivalent 4xFORTH time is actually 11 seconds. That's still very good.
68000 SUM
Also in the December, 1985 issue, Introducing 520ST Assembly Language states, "the
sum of the first N even natural numbers‹or 2 + 4 + . . . + N." This should be 2+4+...+2N.
LOGO TYPO
Finally, there is a typo on page 61 of ST Logo Exploration (December, 1985). The example,
TYPE :HCR should read, TYPE :HCHR. and line 2 of that example should read,
MAKE "HCHR PIECE :A + :A + 1 :A+ :A+2 :HEX1.
We'd like to thank eagle-eyed Bill Wilkinson of Optimized Systems Software for spotting
every single one of the above errors.
MORE FORTH
Mike Starling, from the Dragon Group also spotted the 4xFORTH error. (Probably because
he wrote the language.) With the Level 2 Accelerator, Mike tells us, 4xFORTH can
accomplish 10 passes in high level execution in 3.8 seconds.
There have also been a number of updates to 4xFORTH that resolve most problems
mentioned in our review. If you own the package, please remember to send your
registration card to 4xFoRTH. As Mike says, "It's tough to provide good support
if we don't have your address."
NEW TOS
Of course it had to happen. Antic had no sooner published a "definitive"
outline of TOS (January, 1986) when we received a new TOS version from Atari. It's a bit smaller
(198,126 bytes), but otherwise seems pretty much the same as the previous TOS.
Well, almost... You may recall we wrote last month about Desk Accessories needing the
filename, DESK1.ACC. It now seems that desk accesories may have any filename as long
as the extender is .ACC.
One other obvious change: The thermonuclear bombs generated by a system crash now are round,
black anarchist bombs with fuses.
ANTIC VOL. 4, NO. 10 / FEBRUARY 1986 / PAGE 9