Subject: Re: some packages don't work well with MANZ unset...
To: NetBSD Packages Technical Discussion List <firstname.lastname@example.org>
From: Greg A. Woods <email@example.com>
Date: 12/23/1998 19:58:08
[ On Wed, December 23, 1998 at 15:32:52 (-0800), I presume I need no introduction. wrote: ]
> Subject: Re: some packages don't work well with MANZ unset...
> The really misleading one happens to be MANZ, since _I_ understood
> "undefined" to mean "not set to anything in particular" and not "never
> even mentioned", especially when you read the comment.
Now that I understand that these are boolean flags, using .ifdef(VAR), I
see how this works, and what the alternatives "defined, not defined"
mean, and at least the *intent* makes sense.
I realize too that what's most misleading in mk.conf.example is that the
examples which "set" the flag put something in the variable, while the
examples which don't "set" it, both comment out the setting (implying
that the default will take effect), *and* leave the value empty
(implying that an empty value is the same as an unset variable).
> This is _really_ shoddy, gang. "Undefined" and "empty" should mean the
> same thing, unless you can give a working, concrete example as to why
> they should not. This means that we should be using ".if !empty(VAR)"
> and not ".ifdef(VAR)".
The "gang" in this case isn't likely to be anywhere in sight (unless
they're peering over this particular fence), being that this stuff was
all "derived" from the FreeBSD "ports" stuff. ;-)
> .undef (MANZ)
hmm.... (RTFM in earnest) ....that makes sense.
> System V any flavor: just say NO!
Tell it to my 3B2! ;-)
(Once upon a time the correct cry was "free unix source!" and nobody
really cared what denomination the free one turned out to be....)
Greg A. Woods
+1 416 218-0098 VE3TCP <firstname.lastname@example.org> <robohack!woods>
Planix, Inc. <email@example.com>; Secrets of the Weird <firstname.lastname@example.org>