From: Vladimir Prus (ghost_at_[hidden])
Date: 2005-07-04 09:25:08
On Friday 01 July 2005 07:18, David Abrahams wrote:
> > a. It seems to be missing, at least from my CVS version of the docs.
> > And I have trouble getting it built.
> > b. Ask Volodya ;-))
> Well, I've been expecting him to weigh in here and I'm wondering why
> he hasn't. Volodya?
Initially I though you'll two will figure out everything ;-)
> > Well, I'm not sure this was an "explicit" design choice.
> Well, now that's a different story than you were telling me before.
> > It _can_ be a mere oversight when implementing "alias" since its a
> > newer feature as usage-requirements.
> Well, I'd be surprised if this had anything to do with "alias."
> Doesn't it work the same way for built library targets?
> > I believe there might be use-cases when this "surprising" behaviuor
> > might be wanted.
> I don't. Or if there are such cases, they should be handled via a
> more explicit and direct mechanism.
> > On the other hand, "use-project" always meant "this projects Jamfile
> > and its contents" to me. So, I'm not surprised at all ;-))
> But that's crazy! The fact that you need library X from project A
> shouldn't cause all of the usage-requirements from libraries Y and Z
> in project A to be applied to everything in your project.
I think the situation is that Jurgen has <use>/boost somewhere. That, in turns
builds every single target in Boost, and propagates usages requirements back
to the dependent. Including that <async-exceptions> thing.
The right solution is a combination of:
1. <use>-in only the needed libraries
2. Add "alias headers ; " to top-level Boost Jamfile.v2, and
<use>/boost//headers. This way, you'll get <include> usage requirements from
top-level Jamfile.v2, and nothing else.
-- Vladimir Prus http://vladimir_prus.blogspot.com Boost.Build V2: http://boost.org/boost-build2
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