About the pausing = Garbage Collection:
"Jean-Yves Avenard" <me@privacy.net> wrote in message news:BB76747A.12B42%me@privacy.net...<VPN snipped>
*<< 1 500 FOR I I NEXT 500 ->LIST LIST-> MEM >>: 7.98s (HP49G: 45.81s) (This
test the GARBAGE collector and SysRPL kernel)
====================================================
TO measure just the CG, I changed this to << MEM DROP # 0d # 999d FOR k k NEXT
1000 ->LIST LIST-> >> EVAL {MEM} HEAD TEVAL
and it takes only around 15% of the "old" time to do CG now.
"Jean-Yves Avenard" <me@privacy.net> wrote in message news:20030915085021446+1000@news.cis.dfn.de...<VPN snipped>
> Must be a fair bit of native code in there otherwise you wouldn't see
> such a speed gain.
The CG routine could be ARM native to gain ~ 6*speed increase or time decrease.
Just buy the HP 49G+ !!! - VPN