autoconf-patches
[Top][All Lists]
Advanced

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

Re: TODO update


From: Ralf Wildenhues
Subject: Re: TODO update
Date: Fri, 28 Mar 2008 16:11:42 +0100
User-agent: Mutt/1.5.17 (2008-03-09)

Hi Eric,

* Eric Blake wrote on Fri, Mar 28, 2008 at 03:48:57PM CET:
> Does anyone have a preference of whether I should release M4 1.4.11
> first, and update autoconf to recommend it?

If otherwise it doesn't matter to you, yes, please do M4 first.  Thanks.

> We currently recommend M4 1.4.8 or later.  I
> can't in good conscience recommend 1.4.10 since it is broken out-of-the-box on
> BSD-style stdio systems when dealing with any configure.ac that would result 
> in
> a configure larger than a half megabyte.

No, recommending 1.4.10 would be a bad idea.

> --- a/TODO
> +++ b/TODO
> @@ -13,18 +13,6 @@ these suggestions... their presence here doesn't imply my 
> endorsement.
>  and the like, don't have a consistent way to handle multi-line
>  arguments.  Fix, test, and document.
>  
> -** AC_PROG_INSTALL
> -This test should be extended to check that install supports the GNU
> -Install syntax: install FILES... DIR.  This will relieve everybody
> -form having to use mkinstalldirs to create the directories, as install
> -does it itself.  install-sh is already handling this case.  This also
> -makes it simple not to create the directories where nothing will be
> -installed because of configuration options, which is next to
> -impossible using the current setting.
> -
> -In other words: everything is ready (install-sh and Automake), we just
> -need a good reimplementation of AC_PROG_INSTALL.

FWIW.  This is not exactly what we do.  Even GNU coreutils install does
not do what the TODO item claimed (namely, creating DIR; it doesn't even
with -D), nor does install-sh.  So I wonder whether the TODO item should
be left in but reformulated to match whatever it really desired.  Or
maybe that which was meant was what we did now?

> -For AC_TYPE_SIGNAL signal handlers, provide a way for code to know
> -whether to do "return 0" or "return" (int vs void) to avoid compiler
> -warnings.  (Roland McGrath)

In what way is that fixed?

Thanks,
Ralf




reply via email to

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