[This note has been sent to the following InterNet address(es):
[log in to unmask]]
 
 
I have an application where RoadRunner is executed on a remote system. It uses
an input file which is on the host system and makes use of file equations to
point to the host input file. E.g. The file equation on the host is :
 
REMOTE FILE MYSTDIN;DEV=#
 
Then the program is run  REMOTE RUN RR.PUB.TYM;PARM=1;STDIN=*MYSTDIN
 
This works fine (and still works) if the host and remote systems have the same
operating system (e.g. both on MPE/iX 4.0 or both on 5.0) but fails with a
nasty message if the host is on MPEX/iX 4.0 and the remote is on 5.0.
 
The display is :
**** OPEN OF STDIN FAILED (CERR 14)
 
ABORT: RR.PUB.TYM
       PC=a.00a1f35c dbg_stackdump_cmds+$28
NM* 0) SP=41837c60 RP=a.004d4e10 trap_handler.handle_setdump+$f4
NM  1) SP=41837be0 RP=a.004d62a4 trap_handler+$3a4
NM  2) SP=41837b60 RP=a.0021038c subsystem_abort_handler+$28
NM  3) SP=41837520 RP=a.00210330 ?subsystem_abort_handler+$8
         export stub: 6e0.000c1ce0 _csu_error+$e4
NM  4) SP=418374e0 RP=6e0.000a99a0 _open_std_file+$70
NM  5) SP=41837360 RP=6e0.000a991c ?_open_std_file+$8
         export stub: dac.0008da70 _start+$20
NM  6) SP=41837320 RP=dac.0003403c $START$+$1c
NM  7) SP=418361c0 RP=dac.00000000
     (end of NM stack)
 
As I read somewhere about the internal file numbers being changed in 5.0 to
bring it into compliance with the POSIX file system where STDIN is file #0
rather than #1 for 4.0, I am assuming this is a sympton of this change.
 
Anyone want to comment on this ?
 
TIA,
John Dunlop ([log in to unmask])