COBOL woes

The language won't die, but that doesn't mean the programmers won't!

Funny quote:

'"Just because a language is 50 years old, doesn't mean that it isn't good," said Donna Dillenberger, an IBM Fellow.'

Right Donna: it's the fact that COBOL sucks that means it isn't good, not the fact that it is old.


  1. Do you think that COBOL sucked even 50 years ago, or that it sucks now because it is outdated ?

    1. At the time of invention, it did its job... but even then, a Fortran program would typically be 10x shorter. By 1970, far better languages in terms of expressiveness, structure, etc. were available.

  2. I had a friend who made huge money during the Y2K scare fixing software written in COBOL. Struck me as the equivalent of spending as much money on baling wire and chewing gum to hold an old car together as they would have spent on a new car. But maybe there was something I was missing.

    1. There was: the need for testing. Testing is slow and expensive, and unpredictable. Hard to commit to new programs with a precise to the millisecond deadline when there's a simple alternative. That's one reason that cobol programs still run much of the world. It's the reason old C programs -- a vastly beter language -- will continue to run much of the rest of it for years and years.
      It's not just a software thing. Imagine relocating a heavily used airport like La Guardia.


Post a Comment

Popular posts from this blog

Central Planning Works!

The biggest intellectual nothing burger of the last century?