autoconf-patches
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: list of --host options


From: Ralf Wildenhues
Subject: Re: list of --host options
Date: Mon, 1 Oct 2007 20:03:52 +0200
User-agent: Mutt/1.5.13 (2006-08-11)

* Benoit SIGOURE wrote on Mon, Oct 01, 2007 at 05:54:05PM CEST:
> On Oct 1, 2007, at 2:22 PM, Ralf Wildenhues wrote:
>>> On Oct 1, 2007, at 1:30 PM, Benoit SIGOURE wrote:
>>>
>>>> By the way, I have a question to the list, AC_CANONICAL_HOST implies
>>>> AC_CANONICAL_BUILD, maybe this should be mentionned in the 
>>>> documentation?
[...]
>> But also, why do we want to document this at all?
[...]
> Because you can be surprised to see:
>
> checking build system type... i386-apple-darwin8.10.1
> checking host system type... i386-apple-darwin8.10.1
> checking build system type... (cached) i386-apple-darwin8.10.1
>
> even though it doesn't harm.  But if you want to keep this as an internal 
> detail, fair enough.

Ah, ok.  Here's a different view to this: whenever you see cached checks
or for that matter repeated tests at all on a configure run without sub
configure scripts or a cache file, then either the macro ordering may be
improved, or there is a certain redundancy in the tests.  Sometimes the
redundancy may be eliminated by having macros that call other macros
instead AC_REQUIRE the latter ones.  In Autoconf code, this is not always
possible, for example due to semantic or backward compatibility reasons
(e.g., a macro that should be expandable inside a shell conditional).

So if you like, I would prefer that this principle be documented rather
than every macro dependency; that's less work, too.  ;-)

Cheers,
Ralf




reply via email to

[Prev in Thread] Current Thread [Next in Thread]