June 2017 - vcf-midatlantic

COBOL
by Evan Koblentz 01 Jun '17

01 Jun '17

Relax: I don't intend to learn COBOL. :) Lots of tech reporters are writing stories about why COBOL is dead/obsolete. They're wrong. I am writing a story about COBOL's merits, roadmap, etc. -- already working on setting up some interviews with people in the ISO standards community, IBM, and so on. Here's a specific question for you guys. The page at https://en.wikipedia.org/wiki/List_of_compilers#COBOL_compilers lists lists nine COBOL compilers. Other than IBM and Gnu, which are the most common ones used in industry? (banks, federal agencies, etc.) ________________________________ Evan Koblentz, director Vintage Computer Federation a 501(c)(3) educational non-profit evan(a)vcfed.org (646) 546-9999 www.vcfed.org facebook.com/vcfederation twitter.com/vcfederation instagram.com/vcfederation

6 9

Re: [vcf-midatlantic] COBOL: vintage vs. legacy code
by VAXman＠tmesis.org 01 Jun '17

01 Jun '17

Bob Applegate <bob(a)corshamtech.com> writes: > On May 31, 2017, at 2:36 PM, Joseph S. Barrera III via vcf-midatlantic = <vcf-midatlantic(a)lists.vintagecomputerfederation.org> wrote: >=20 > On Wed, May 31, 2017 at 11:00 AM, Brian Schenkenberger via = vcf-midatlantic < > vcf-midatlantic(a)lists.vintagecomputerfederation.org> wrote: >=20 >> I've been writing assembly for 40+ years and fully expect to be = writing it >> for another 10+. ;) >>=20 >=20 > Are you asserting that, say, VAX assembly and 6502 assembly are the = *same > language*? :-) Somehow I missed this comment. Funny you should mention VAX; it shows how out of touch you are with the VMS world. I have written my fair share and more of VAX assembly but I've written volumes of Alpha and Itanium too. I used to play on PDPs and the micro-computers (CP/M) in the physics lab in college writting assembly. One of my first year EE classes involved 6502 processors too. >Same concepts, even if the target processor is different. Pretty much. >FWIW, I was in a code review this week for a few hundred bytes of = >assembly code. The number of people fluent in assembly at the company = >is very small, but this product sells several hundred thousand units a = >year and is very important. The code was in C using a highly optimizing = >compiler but it couldn=E2=80=99t quite squeeze the code enough once a = >new feature was added, and going to a slightly larger processor with = >more RAM (32 bytes) was about four cents more per chip, so switching to = >assembly was the best approach. > >Not all computers have 2 TB disk and 32 GB RAM. When you need to save a = >few bytes, assembly is your friend. Yeah, like the specialized processors incorporated in to satellites which I coded for in my Astrospace days.

1 0

Re: [vcf-midatlantic] COBOL: vintage vs. legacy code
by Jeffrey Jonas 31 May '17

31 May '17

My dad's happily retired for about 25 years now. Some of his COBOL IBM mainframe code is STILL IN PRODUCTION. Why? Because of the investment of untold man-YEARS of effort, debugging and diligence to yield nearly bullet-proof code. Because the code implements algorithms and logic that's no longer documented or understood by anyone else. Porting code introduces new bugs, thus IBM's captive audience with all the new mainframes ALWAYS being "backwards compatible". > It's not obsolete as long as it's *not* too much trouble to maintain. There's a whole wild and wacky world of backwards compatible retrofits to "legacy systems". Screen scraping interfaces new apps to programs that still think they're talking to 3270 "green screen" terminals. Dave McGuire: > Organizations are using it for a reason: > It's the right tool for their job, > based on whatever criteria they've set forth. FORTH? Did somebody say FORTH :-) Most of my dad's career was writing COBOL programs. I'll never have that stability in today's "language of the month" club: Perl, Python, Go, Ruby on Rails, Java and all its variants and environments, ADA (before it was a fruit), FORTRAN/RATFOR/WATFIV, ... -- Jeff Jonas

5 6