Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!rcj
From: rcj@clyde.UUCP
Newsgroups: comp.society.futures
Subject: Whither Unix Gurus
Message-ID: <15669@clyde.ATT.COM>
Date: Sat, 24-Oct-87 20:03:01 EST
Article-I.D.: clyde.15669
Posted: Sat Oct 24 20:03:01 1987
Date-Received: Wed, 28-Oct-87 23:02:12 EST
References: <4692@elroy.Jpl.Nasa.Gov>
Sender: nuucp@clyde.ATT.COM
Reply-To: rcj@moss.UUCP (Curtis Jackson)
Organization: AT&T Bell Laboratories, Whippany NJ
Lines: 38

In article <4692@elroy.Jpl.Nasa.Gov> earle@jplopto.JPL.NASA.GOV (Greg Earle) writes:
}	each can take up all of one's time.  In the face of this, how can one
[ refers to the information explosion in Unix-related areas ^^^^  - rcj]

}	enter the fray at this late stage and hope to become one of
}	the all-knowing Guru Elite?  It would appear impossible at this stage,
}	and only those who (through early entrance to this netherworld) have
}	already accumulated the knowledge will be able to keep up.  When they
}	evolve out of this, this knowledge will be lost.  Looking for all
}	arguments pro and con ...

It is a known fact in most research areas that you are required to document
your findings.  This is *not* true in the type of work that hackers rightfully
call "hacking" -- it ain't establishment-style research; it's just playing
around.  Consequently, my solution to the information explosion and the
resultant loss of valuable discoveries when persons move from one area to
another is:

Modify the hacker ethic to include documentation (not fancy, just complete)
as a Feature rather than an Annoyance (and good luck doing it!).  I decided
to change jobs about a year ago and in my interviewing saga I was appalled
at the absolute crap I saw people working on -- little or no documentation,
garbage thrown together at midnight even though there was no great deadline
to meet, etc. etc.  And most of this was at big companies that have
guidelines for code development, documentation, testing, and the like.
I shudder to think about some of the incomprehensible, untested, bug-ridden
software that must be put out by smaller places with no controls whatsoever.

Of course, perhaps it is best that the knowledge of the work put out by these
so-called "Gurus" be lost -- how many people do you know who want detailed
design documentation on the Ford Pinto?

Yours for fast, compact, innovative but *readable* and *maintainable* software,

The MAD Programmer -- 201-386-6409 (Cornet 232)
			      ^^^^ new extension
alias: Curtis Jackson	...![ ihnp4 ulysses cbosgd allegra ]!moss!rcj
			...![ ihnp4 cbosgd akgua watmath  ]!clyde!rcj