Re: make: new modifyer :Y

[David Holland <dholland-tech%netbsd.org@localhost>]

On Mon, Nov 07, 2011 at 01:52:44PM +0300, Aleksey Cheusov wrote:
 > I'm surprised that 30 lines of code can cause objections. 

It's not that it's 30 or 300 or 300,000 or even 3; it's that the
primary problem with make, both now and historically, has that people
have gone and added special-case hacks to it, time and again, without
worrying about generalizing what they're doing, being systematic, or
pursuing any kind of coherent high-level design.

It is true that I haven't had time to work on make in the past year to
any significant extent. That's how it goes. But it's not like I'm the
only person who can work on make's input language, or who can
reorganize the code to a level of sanity where significant semantic
improvements are feasible.

Although for the record I think ${FOO:${ISTRUE}} is a fine approach.

(As for the suggestion of having make understand '.if yes', that's not
necessarily a bad idea but someone needs to sit down and figure out if
adding that automatic coercion makes the eval rules for expressions
inconsistent. Some such conversions can.)

-- 
David A. Holland
dholland%netbsd.org@localhost