Oh, yes. A hundred times... and usually said dev then leaves the
company and nobody can maintain this completely bespoke thing they
left behind with no documentation
In the development shop I used to work in, we had a mainframe dev team
and a "distributed" dev team (i.e. their stuff ran on servers and PCs).
The problem you describe usually happened in the latter group, but was
usually triggered by a manager who decided that X-newish language was the
way to go, i.e. java or whatever M$ was pushing at the time.
I have probably mentioned this before but a lot of their devs were the "point-and-click" type that had difficulty maintaining code that required
any "manual" coding that the "framework interface" would not provide them.
We did sometimes have similar issues on the mainframe side, but it was a
lot less of someone using newish things and a lot more of a case of some
of the devs not being familiar with common utilities (like syncsort) that
could really help then get simple tasks taken care of more quickly than
writing their own COBOL program to do so.
> As one of my university lecturers used to say - there's a difference
> between being clever and being smart...
Indeed. We had a few COBOL devs that liked to make things much more
"clever " (read "difficult to maintain") than they needed to. I think it
was called job security. ;)
--- Talisman v0.53-dev (Linux/armv7l)
* Origin: possumso.fsxnet.nz * telnet:2123/ssh:2122/ftelnet:80 (21:4/134)