@RUTGERS.ARPA:JEFFREY@OFFICE-2.ARPA (02/14/85)
From: JDS5.TYM@OFFICE-2.ARPA My particular programming history began in the days of batch (1960) and progressed through time-sharing (mostly Tenex) to PCs. In all my experiences on large and small projects there always seem to be two kinds of programmers: analytical and synthetic (my poor choice of terms). The analytical people are debuggers at heart - detectives. They are the people who want to start coding in DDT or the like (some of them claim to start in assembly language or even C, but that's just a rouse). The synthetic people start in some higher level language (or asm laden with macros) and probably never get down to DDT - mostly they are afraid of bugs (that's me). I am biased toward bug avoidance since to me, a bug is like a real pain in the neck (how in the world would you ever find out the cause for a bug?). My opinion is that the bug avoiders do better and spend less time in development to produce better looking code that does more. BUT, that does not say that batch is a better environment than a PC or Tenex etc. No sir, please do not stick me on one of those one-run-a-day projects. Gawd - that's painful. I prefer to avoid my bugs with good response time. Since I still don't avoid them all, I also like to debug interactively. I wouldn't look at a dump if you paid me (unless I was the first developer on a machine and there was no choice). Now there may be a class of young programmers who missed the healthy experience of developing with discipline for batch. That was certainlky good training and I wonder if we all don't need some of it. After all, if you gorw up with an interactive debugger, how long will it take you to learn the discipline required to produce well organized, maintainable code? Maybe some batchless programmers can comment on that question. Jeffrey Stone, Menlo Park, CA