30 Jan
2009
30 Jan
'09
9:49 a.m.
On 30 Jan 2009 at 12:28, Tim McNerney wrote:
P.S. Back in the 70's version control systems were
rare.
Personal Software's backups were done on floppies.
Rare only in the microcmputer world. The mainframe people knew about
the importance of source-code control for a long time. At CDC, we
used a program called UPDATE that was old in 1970, and by no means
the first of its kind, even at CDC.
ONe of the things that got lost in the transition away from punched
cards was the use of columns 73-80 as a non-parsed "sequence" field
in source code. With utilities like Update, one could look at a
listing and immediately see which correction set a statement was part
of. The people in Integration assigned correction set IDs according
to PSR numbers, so you even knew what problem the code was supposed
to fix.
It surprised me that the micro types took so long to figure basic
things like SCCS and regression testing out.
Cheers,
Chuck