Merge lp:~jontai/openvista-gtm-integration/bug429699 into lp:openvista-gtm-integration

[Jon Tai]

[Jon Tai (jontai)]

I think it would be good if Joel or one of the other MUMPS guys weighed in here to say if they felt this change was necessary or not. The bug didn't include the circumstances under which the problem occurred, so I wonder if it was during a debugging session or if ZRULE can really be undefined when EXIT is called under "normal" use.

Or is our policy to go ahead and use $G whenever we can?

[JSHER (joel-sher)]

We already have
ZRULE=$G(ZRULE) in the routine

I like Jon am wondering how we got here without ZRULE defined.

Although the change will not hurt anything and the code would fix the bug as described.

review: Approve

[Jon Tai (jontai)]

> We already have
> ZRULE=$G(ZRULE) in the routine
>
> I like Jon am wondering how we got here without ZRULE defined.
>
> Although the change will not hurt anything and the code would fix the bug as
> described.

I discussed this with Jeff and we decided to drop this patch. Under normal circumstances, ZRULE should be set to $G(ZRULE) already -- we must have been calling EXIT directly during debugging when we ran into this. We can re-open the bug if we run into a real-world situation where EXIT is called and ZRULE is undefined.

Unmerged revisions

95. By Jon Tai on 2009-10-22

make change as suggested by Jeff

=== modified file 'mumps/ZISTCPS.m'
--- mumps/ZISTCPS.m	2009-06-05 20:09:12 +0000
+++ mumps/ZISTCPS.m	2009-10-22 23:40:24 +0000
@@ -1,4 +1,4 @@
-%ZISTCPS ;ISF/RWF MSC/JDA - DEVICE HANDLER TCP/IP SERVER CALLS ;22APR2009
+%ZISTCPS ;ISF/RWF MSC/JDA - DEVICE HANDLER TCP/IP SERVER CALLS ;22OCT2009
  ;;8.0;KERNEL;**78,118,127,225,275,388,MSC**;Jul 10, 1995
  Q
  ;
@@ -55,7 +55,7 @@
 EXIT() ;See if time to exit
  I $$S^%ZTLOAD Q 1
  N ZISQUIT S ZISQUIT=0
- I $L(ZRULE) X ZRULE I $G(ZISQUIT) Q 1
+ I $L($G(ZRULE)) X ZRULE I $G(ZISQUIT) Q 1
  Q 0
  ;
 LGTM ;GT.M multi thread server