S0C4 after OPEN failed - we kept going anyway

Another one for the archives. OPEN failed (bad status), no check, later WRITE, S0C4. Dev swore the DISPLAY after OPEN proved the file was fine because an earlier test showed 35 and RC 0.

How do you get juniors to treat file-status like SQLCODE?

Shop standard: after every OPEN/READ/WRITE/START, evaluate status, non-ok goes to a common abend paragraph that dumps status + filename.

Code review rejects bare OPEN. Same energy as missing SQLCODE checks.

We added an excerpt from the file-status table to the onboarding doc. 35/37/39/96 show up in week one otherwise.