akp@isrnix.UUCP (12/15/83)
#N:isrnix:17500001:000:258 isrnix!akp Dec 14 04:35:00 1983 I got a strange error message from Eunice today (I don't remember the circumstances); can anybody enlighten me? The message was: %NONAME-E-NOMSG, Message number 00000002 I wonder what that could mean... -- Allan Pratt ...decvax!ihnp4!iuvax!isrnix!akp
jla@houxt.UUCP (12/15/83)
This problem is caused by interaction between UNIX and VMS program return values. UNIX treats 0 as success, and non-zero as an error. VMS has a more complicated scheme, using the bottom three bits to hold the severity of the error. Eunice does a very half-hearted job of mapping UNIX onto VMS status. It maps return of 0 to a VMS return of 1 (Success) and passes any non-zero UNIX return through as it stands. Most UNIX utilities return error codes of 1, which comes through looking exactly the same as a successful return. This makes writing VMS scripts using UNIX utilities very difficult. The DCL script to run the CC command is a case in point. Has anybody out there done any fixes in this area? Really: Brendan Lynch
stevesu@azure.UUCP (Steve Summit) (12/17/83)
Funny messages from VMS aren't unusual when using Eunice. The normal exit status for processes in VMS is 1, as opposed to Unix' 0. Of course, under Unix the exit status is rarely checked. If the last command you executed in a Eunice shell bombed out with a strange error code, and you dropped into VMS, or if you were running one of your programs directly under VMS, VMS tends to notice that the exit status isn't 1, but isn't an error code it recognizes. Thus the "%NONAME-E-NOMSG, Message number 00000002." I tend to get "%NONAME-W-NOMSG, Message number 00000000." As I recall, the 1 bit is success, the 2 bit is error, and neither is a warning. Steve Summit tektronix!tekmdp!stevesu
sdc@uvacs.UUCP (12/22/83)
References: pur-ee.1243 That means that only VMS is dumb enough to have a message that says "I have no message". That reminds me of a funny story... We used to use "Pascal MT+" on our z80 systems. One of its command line options was "quiet mode", which was supposed to turn off all the "user friendly" feedback, like when it said parsing... and so on. Well, it turns out that all the compiler did when you selected "quiet mode" was to output an extra message: **** QUIET MODE SELECTED **** on top of all the other garbage. Neat, huh? But seriously, I think that message number 2 means that some program exited with a "vms status" of 2, and that vms did not recognize the value 2 in its repertoire of exit conditions. So rather than just ignoring the message, vms decided to say something, however meaningless. I think it's safe to say you can ignore it. Scott Carson ...mcnc!ncsu!uvacs!sdc at Univ. of Virginia ...mcnc!ncsu!uvacs!iedl02!sdc at General Electric