Add Pthreads (and possibly other) flags to the pkg-config file

[wferi]

No description provided.

[jnpkrn]

Just gathering the context, is this to serve building libqb
(it's users?) statically or does just Libs.privates receive
a special treatment in Debian or something like that?

[wferi]

Libs.private enables linking libqb statically, yes. (Not that I know anybody doing that, that part is mostly for completeness.) Cflags, on the other hand, is emitted all the time and (under GCC) propagates the -pthread option to all users of libqb (even when compiling modules or linking everything dynamically).

[jnpkrn]

Thanks, was just curious since I vaguely remember Libs.private
references seldom scattered in the Fedora devel ML and casting
some opinionated ideas about its proper treatment
+ there's at least a single crucial difference regarding libtool
already as discussed elsewhere under this project, for instance.

Can you add this "static compiling" + "completeness" justifications
to the commit message, please? There's some bad experience with
not entirely obvious changes that charge an extra time to understand
in retrospect if need be, so better to prevent that systemically

Also would mention that gethostbyname/libnsl and libglib counterparts
don't qualify since they are only used in the examples (if I'm not
missing anything), not in the library proper -- in case one would
wonder in the plain cross-check.

Thanks

[wferi]

I rebased the commit keeping the content but with extended message.
If you want real visibility, wouldn't comments be more effective? Those can get stale, though...

[chrissie-c]

I think the commit message is fine as documentation. As you say comments can get stale and it seems to be routine to check commit messages these days when checking when/why things happen.

Thanks Feri & Poki,

[wferi]

Thanks for the merge.
Why/how was my full name replaced in the commit? I don't mind, just curious...

[chrissie-c]

I'm not sure. github being weird I suspect! I should maybe do manual commits if it's going to start happening regularly.

which removes half the reason for having GH features in the first place. sigh

Sorry

[jnpkrn]

Happens easily when single-button-laziness prevails, and I've already
pointed this out in the past (over a year ago if not earlier):

#251 (comment)

So yes, please finally do retain a full control over the commits to hit
the tree, everything else carries these spoiling risks (and that GitHub
will continue to be the primary forge is far from being set in stone,
especially when none of us has a bit of a power over the hosting).

That would also avoid the silliness of undesired authorship handling:
44a9379

Hopefully that's enough of a justification.