Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!rpi!uupsi!cmcl2!adm!smoke!gwyn
From: gwyn@smoke.brl.mil (Doug Gwyn)
Newsgroups: comp.unix.questions
Subject: Re: random numbers in awk?
Message-ID: <15992@smoke.brl.mil>
Date: 29 Apr 91 02:41:49 GMT
References: <1991Apr24.041134.14519@athena.mit.edu> <1991Apr24.173502.9731@berlioz.nsc.com> <7453@auspex.auspex.com>
Organization: U.S. Army Ballistic Research Laboratory, APG, MD.
Lines: 18

In article <7453@auspex.auspex.com> guy@auspex.auspex.com (Guy Harris) writes:
>The theory was, I guess, that people with scripts that depended on the
>"old AWK" should change them to use "oawk", so that in the future "awk"
>could become the "new AWK".  However, according to the S5R4
>documentation, "awk" is still the old AWK....

We did this sort of thing with C compilers, too.  The idea is:
	PHASE 0 (existing practice):  All applications use "awk".
	PHASE 1:  New awk installed as "nawk", old awk remains as "awk",
		"oawk" linked to "awk".
	PHASE 2 (transition):  Applications that require the new
		behavior invoke "nawk".  Old applications are tested
		against "nawk" and if they really do require the old
		version, they are changed to invoke "oawk"; otherwise
		they continue to invoke "awk".
	PHASE 3: "awk" is changed to be a link to "nawk".
	PHASE 4 (catch-up):  Applications that rely on "oawk"-specific
		behavior are updated to work with regular (new) "awk".