14 Jul
2023
14 Jul
'23
8:14 a.m.
On Thu, Jul 13, 2023 at 6:36 PM segaloco via TUHS <tuhs(a)tuhs.org> wrote:
The conclusion I'm coming to from what has been
said thus far is that people who were moving from COBOL and the mainframe world to UNIX
didn't have as much of a need for COBOL. Since that transition often involved change
in enough other aspects of an operation, moving to UNIX with the same COBOL applications
just wasn't the path to success for most folks, as opposed to folks deeply invested
in FORTRAN. Would that be a fair characterization?
Echoing what others have said, that seems reasonable to me. Don't tell
the mainframe infosec folks about this; they get really touchy about
defending COBOL's honor for some reason I don't quite understand.
Thanks for the feedback by the way, one of the matters
I'm trying to suss out is what a typical COBOL environment on UNIX would've
looked like back when, and what it sounds like is a COBOL environment on UNIX was anything
but typical.
A way to think about COBOL is that it's like a DSL for business
transaction processing, but is itself a small part of the overall
offering. In the mainframe world, it's often intimately tied to things
like CICS, ISAM, VTAM and 3270 access methods, SNA, and so on, and in
that sense the language itself is a rather small part of the
ecosystem. Transferring everything into a new environment (e.g., on
Unix) raises a lot of questions about the surrounding technologies and
their non-availability, and ultimately just having the language by
itself isn't terribly useful if you don't have all the other stuff as
well.
Also, COBOL is just a terrible language.
- Dan C.