Re: Clearsgts

[cld@xxxxxxxx (Carl Distefano)]

Reply to note from "Robert Holmgren"  Tue, 30
Nov 2004 01:41:05 -0500

Robert:

> However, let's suppose, for the sake of a good argument, that
> it really is necessary to clear these 44 S/Gs, but you don't
> want to take the time to let CLEARSGTS do it. And you also
> don't want to have the code bloat of
> "...". ... Here's what I'd do:
>
> +",>"> +1>

Yup, works like a charm. Nulls out S/Gs 555-598 -- and fast, too
(100 ITERations in .28 secs. on my new machine). What's also
interesting, as a theoretical matter, is that the analogous code to
_uninitialize_ the same sequence of S/Gs, using our forced-error
procedure, is only about twice as slow (ITER 100 = .61 secs.) -- in
other words, still very fast:

;*;
BX es 1Q2 >>>;*;
BX d wa=0,mb=0,eb=0,0Q2 ;*;
+",+",>>+">
">+1>;*;
BX d wa=,mb=,eb=Q2 

Here's the encoded version:

XPLeNCODE v2.0
b-gin [UNTITLED]
{<}SV01,555{>}{<}SX02,44{>};*;[cr|lf][BX_]es 1[Q2_]{<}SX03,{<
}VAWA{>}{>}{<}SX04,{<}VAMB{>}{>}{<}SX05,{<}VAEB{>}{>};*;[cr|l
f][BX_]d wa=0,mb=0,eb=0,0[Q2_];*;[cr|lf]{<}CUA,02{>}{<}SX06,"
{<}SV"+{<}IS01{>}+",{<}SV"+{<}IS01{>}+",{>}{>}{<}PV"+{<}IS01{
>}+"{>}{<}PR {>}"{>}{<}PV06{>}{<}SX01,{<}PV01{>}+1{>}{<}LBA{>
};*;[cr|lf][BX_]d wa={<}PV03{>},mb={<}PV04{>},eb={<}PV05{>}[Q
2_]{<}EX{>}[cr|lf]
-nd
XPLeNCODE

So if Harry wanted to uninitialize all of those permanent S/Gs
rather than just empty them, he could do it without much additional
overhead, using your template. Whereas 100 ITERations of JM 2.clearsgtsQ2 takes a whoppping 3.90 seconds.

But Harry still has a lot of explaining to do. Forty-four permanent
S/Gs!

--
Carl Distefano
cld@xxxxxxxx