From: Kevin Wheatley (hxpro_at_[hidden])
Date: 2005-10-12 10:38:20
David Abrahams wrote:
> "troy d. straszheim" <troy_at_[hidden]> writes:
> > 326.17s user 153.82s system 92% cpu 8:40.03 total
> > the first ~30 seconds of which is bjam calculating the dependencies
> > for each test module.
> Unless you have used bjam's profiling feature to break the execution
> time down carefully, you shouldn't draw any conclusions about how much
> time it spends calculating dependencies -- they're usually wrong.
as an example, in a few cases recently it has been filesystem
operations e.g. pwd, that has caused Boost.Build to sit there spinning
its thumbs, the CVS version has some big improvements over the
bjam/Boost.Build in 1.33. However reducing the dependancies/number of
compile/link operations will no doubt also help, the proportions may
now be different.
-- | Kevin Wheatley, Cinesite (Europe) Ltd | Nobody thinks this | | Senior Technology | My employer for certain | | And Network Systems Architect | Not even myself |
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk