Some kind of magic

Charles Pergiel sent this over from a Chromebook forum, and the truth of it hurts just a little:

I remember my early days with computers, asking a question of the developers of a fairly large (for 1980) piece of IBM mainframe software, and their response was “*We don’t know*.” At the time, I found this unbelievable — *you guys wrote it; why don’t you know how it works!?*”

Things have gotten more complicated by orders of magnitude since then, and sometimes the only reasonable answer is “We don’t know.”

I was working on an IBM mainframe in those days, and I can assure you that this wasn’t at all a unique situation: a lot of legacy stuff outlived its original developer, and subsequent developers opted, quite reasonably, to leave well enough alone.





2 comments

  1. Jack Baruth »

    1 January 2015 · 9:51 pm

    The biggest problem now is that everything’s done on top of some stupid-ass framework so you have two sets of liabilities: the idiocy of the devs and the idiocy of the framework. IT’s why so few systems last for more than a few years now, too; the frameworks effectively expire.

  2. CGHill »

    1 January 2015 · 10:04 pm

    In some ways, I should be grateful I got into midranges: IBM has had no qualms about adapting existing packages into System i. Still, that means more frameworks down the road.

RSS feed for comments on this post