Subject: Re: [Boost-build] Locating libraries
From: Brian Ravnsgaard Riis (brian_at_[hidden])
Date: 2012-01-06 17:13:59
-----BEGIN PGP SIGNED MESSAGE-----
On 06-01-2012 11:50, Mateusz Loskot wrote:
> On 6 January 2012 00:35, Brian Ravnsgaard Riis
> <brian_at_[hidden]> wrote:
>> But what about pre-built libs?
> I'm sure Volodya has already solve it at least conceptually ;-) It
> is only due his time constraints we haven't seen it implemented.
> (I'm amazed how I am so unable to grasp the Boost.Build, it is
> easier for me to learn C++ or any new langauge than the Jam'fu.
> Quite unfortunate for myself.)
I've had quite the time of it myself, but I'm finally starting to wrap
my head around it, so to speak. I think a main difficulty is the
declarative approach; I keep trying to grok it as an imperative thing,
and it simply doesn't work that way (for very good reasons!).
>> A given piece of software may be quite able to compile and run
>> without, say, freetype or pango, but take advantage of these
>> libraries if they are present. If b2 could determine whether they
>> are available it could configure that build correspondingly.
> Yes, it could.
>> Even better, it might be possible to have b2 locate headers and
>> libraries (or even complete software packages) given some
>> standard locations (and some supplied non-standard?) to search
>> for the files in. This is what CMake's FIND_PATH and FIND_LIBRARY
>> commands do.
> Indeed, the difference is clear.
>> Just thinking out loud here... [...]
> It is interesting reading, but I'm hardly able to comment it with
> anything relevant or new.
Hey, thanks for reading anyway. :-)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.12 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
-----END PGP SIGNATURE-----
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk