[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: ANN: ncurses-6.2-20211016
From: |
Milan P . Stanić |
Subject: |
Re: ANN: ncurses-6.2-20211016 |
Date: |
Mon, 25 Oct 2021 23:39:23 +0200 |
On Mon, 2021-10-25 at 16:26, Thomas Dickey wrote:
> On Mon, Oct 25, 2021 at 03:28:38PM -0400, Thomas Dickey wrote:
> ...
> > hmm - the drawback that I knew of was that it would pick the first directory
> > scanned (which isn't necessarily the one most often used). Standardization
> > across pkg-config's always been very poor, so I've machines with all of the
> > choices being used by one package or another.
> >
> > To recap, the current issue was begun by a report that the existing
> > (Debian-biased) predefined list of directories wasn't suitable.
> >
> > The manual page mentions
> >
> > --variable=VARIABLENAME
> > This returns the value of a variable defined in a package's
> > .pc
> > file. Most packages define the variable "prefix", for
> > example,
> > so you can say:
> > $ pkg-config --variable=prefix glib-2.0
> > /usr/
> >
> > and I overlooked its usefulness for the task (because the manual page
> > doesn't
> > point out the relationship between "pc_path", PKG_CONFIG_PATH and
> > PKG_CONFIG_LIBDIR (actually reading the source code is needed to see the
> > distinction between the latter two):
> >
> > pc_path
> > The default search path used by pkg-config when searching
> > for
> > .pc files. This can be used in a query for the pkg-config
> > module
> > itself itself:
> > $ pkg-config --variable pc_path pkg-config
> >
> > Using "pkg-config --variable=pc_path" gives me this:
> >
> > /usr/local/lib/x86_64-linux-gnu/pkgconfig:/usr/local/lib/pkgconfig:/usr/local/share/pkgconfig:/usr/lib/x86_64-linux-gnu/pkgconfig:/usr/lib/pkgconfig:/usr/share/pkgconfig
> >
> > which needs some work to make it usable when developing a package.
> >
> > (I did read about "pkgconf" while working on the macro -- because I tested
> > it
> > with the BSDs which generally don't use pkg-config -- and the point was
> > made in _its_ documentation that one should not actually count on it being
> > compatible in fine details).
> >
> > fwiw, I test with machines older than 15 years as well...
>
> for instance, this doesn't work on my Solaris10 machine (0.15.0),
> so my configure script will have to check for availability of the feature.
>
> Checking for when this feature was available in this form,
> pkg-config's git log says it was added here:
>
> commit e170895bc6241e933c6c6e85dc800fef5ed44615
>
> Author: Tollef Fog Heen <tfheen@err.no>
>
> Date: Mon Mar 30 21:17:43 2009 +0200
>
>
>
> 2009-03-30 Tollef Fog Heen <tfheen@err.no>
>
>
>
> * pkg.c (add_virtual_pkgconfig_package): Add pc_path as a
> variable
> which you can use to get at the compiled-in PKG_CONFIG_PC_PATH.
>
> which (for the upstream) makes it a little over 12 years.
> For real use, I'd have to subtract a couple of years.
>
> For pkgconf, it's a different story:
>
> commit 7d89f659ffa45d381c3c4ee03c55f74635042829
>
> Author: Baptiste Daroussin <bapt@FreeBSD.org>
>
> Date: Sun Jan 22 10:44:09 2017 +0100
>
>
>
> Fix regression introduced in 7b39c38
>
>
>
> now pkgconf --variable=pc_path pkg-config works
>
> The feature itself was added here (9 years ago):
>
> commit 111046f100a1a63b125494406351ce72a1a486bf
>
> Author: Ariadne Conill <ariadne@dereferenced.org>
>
> Date: Sun Jul 22 23:16:39 2012 -0500
>
>
>
> pkg: add support for $(pc_path) variable for extracting PKG_DEFAULT_PATH
>
> But this isn't the pkgconf bug reporting list; further discussion
> of that aspect is off-topic.
Thomas, I understand this and I also don't like to discuss what is
happened years ago and why.
I only asked if you have fix for installing .pc files in
$libdir/usr/lib/pkgconfig instead of root install dir.
--
Kind regards
- Re: ANN: ncurses-6.2-20211016, (continued)
- Re: ANN: ncurses-6.2-20211016, Thomas Dickey, 2021/10/24
- Re: ANN: ncurses-6.2-20211016, Milan P . Stanić, 2021/10/24
- Re: ANN: ncurses-6.2-20211016, Thomas Dickey, 2021/10/24
- Re: ANN: ncurses-6.2-20211016, Milan P . Stanić, 2021/10/24
- Re: ANN: ncurses-6.2-20211016, Thomas Dickey, 2021/10/24
- Re: ANN: ncurses-6.2-20211016, Milan P . Stanić, 2021/10/24
- Re: ANN: ncurses-6.2-20211016, Thomas Dickey, 2021/10/24
- Re: ANN: ncurses-6.2-20211016, Milan P . Stanić, 2021/10/25
- Re: ANN: ncurses-6.2-20211016, Thomas Dickey, 2021/10/25
- Re: ANN: ncurses-6.2-20211016, Thomas Dickey, 2021/10/25
- Re: ANN: ncurses-6.2-20211016,
Milan P . Stanić <=
- Re: ANN: ncurses-6.2-20211016, Thomas Dickey, 2021/10/25
- Re: ANN: ncurses-6.2-20211016, Milan P . Stanić, 2021/10/28
- Re: ANN: ncurses-6.2-20211016, Thomas Dickey, 2021/10/28
- Re: ANN: ncurses-6.2-20211016, Thomas Dickey, 2021/10/24