Ross,
just to be on the safe side. Which version are you on?
CMS bREXX 1.0.1 Jul ?5 2022 ? ?
Q CPLEVEL ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ??
SYSTEM 4381-A ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ??
VM/370 Community Edition Version ?1 Release ?1.2 07/19/22 13:20:46 ?
IPL at 13:22:39 GMT WEDNESDAY 12/13/23 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
PEAK LOAD= 006 USERS ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ??
These are, I think, the current levels of VM/CE and BREXX.
I have not been able to reproduce, using some large TEXT files.
best regards,
¸é±ð²Ô¨¦.
toggle quoted message
Show quoted text
On 16 Mar 2024, at 06:57, rvjansen@... wrote:
Hi Ross,
I'd suggest??where Bob Bolch will see it and can look with Adrian Sutherland (who is the person that got the BREXX and associated C runtime in usable state). I don't know yet if the TEXT deck contents is relevant to the bug but I would suggest to attach that to the issue.
best regards,
¸é±ð²Ô¨¦.
On 15 Mar 2024, at 23:19, Ross Patterson <ross.patterson@...> wrote:
I've seen this randomly a couple of times, but today I built a repeatable test case.? Using a simple Rexx iterative loop, BREXX occasionally goes quite literally crazy, generating either floating point numbers for the loop index, or randomly resetting the index to 0.
I thinned down the Rexx program I was running until it couldn't get any plainer.? The below files BOOM11 EXEC and CONSOLE demonstrate both behaviours.? BOOM11 EXEC reads a TEXT deck file via EXECIO DISKR * (STEM, and then simply loops over it, with tracing active, writing the loop index to the console. At line 144 in BOOM11 CONSOLE, you can see it resetting from 7 to 0, and later at 248, the "8" is shown as "8E+20".
I apologize for cluttering everyone's mailbox with the files in-line, but this list doesn't allow attachments.
Anybody know what's going on here?? And if not, which of several "BREXX" places on the web should I report the bug to?
