[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Nmh-workers] obsolete optional slocal argument format
|
From: |
Valdis . Kletnieks |
|
Subject: |
Re: [Nmh-workers] obsolete optional slocal argument format |
|
Date: |
Mon, 26 Nov 2012 16:49:26 -0500 |
On Wed, 21 Nov 2012 23:13:46 -0600, David Levine said:
> Sometime between MH 6.8.5 and entry into CVS in April 1999,
> the argument processing was changed to use getarguments().
> But that broke the handling of [address info sender]. The
> relevant code is:
So you other dinosaurs out there - that bug was there for almost 14 years.
What's the record for longest time between when a bug was introduced in a code
tree and somebody noticed?
Even longer-lived was a bug in OS/360, where the JCL code checked for a
"skipped" jobstep by checking for a non-zero execution time, as even IEFBR14
took at least 0.01 seconds to run once you counted the initiator CPU time and
all that stuff. Then the 303x showed up and the CPU time for an IEFBR14 could
register as 0.00. Bug must have been in the original MFT/MVT code in
1964-ish, and I remember applying the OS/VS1 patch in 1984 or so....
pgp5ujDfqcJWe.pgp
Description: PGP signature