From: Vladimir Prus (ghost_at_[hidden])
Date: 2008-05-05 12:03:51
On Friday 02 May 2008 16:31:37 John Maddock wrote:
> John Maddock wrote:
> > siliconman wrote:
> >> In article <fv23l0$o52$1_at_[hidden]>,
> >> "Robert Ramey" <ramey_at_[hidden]> wrote:
> >>> In V2 - I tried to update my Jamfile to handle this but have
> >>> been unsuccessful. I did ask for help on this a couple of
> >>> times on the build list - but the response I got suggested
> >>> what I did should work - so here we are.
> >>> I would love to get this fixed so I could test this
> >>> compiler on my own system.
> >> So what needs done here? As it stands I am unable to do regression
> >> testing at all. I'm very much willing to put time into this, but I
> >> need at least a starting point.
> > I think I *may* have a fix: it seems that if a target has both a
> > <build>no and a <dependency>some-target, then you get a Boost.Build
> > error from "some-target". Commenting out the <build>no option in
> > libs/serialization/build/Jamfile.v2 fixes things temporarily, but
> > isn't much of a fix: I think I have a better one, but it needs
> > tidying up and testing!
Hmm, I'll look into why it happens.
> The best I have is below - it requires a few rules to be duplicated in
> several places which is not good at all - but it retains the <build>no
> optimisation for older compilers.
> I tried to use the import rule to define the new rules in one place only and
> then import it where required, but I couldn't get it to work for some
> reason. If anyone can see a way of tidying this up that would be great:
> otherwise it may be better to simply remove the <build>no optimisation, but
> then I guess Borland users will get an alarming number of error messages at
> install time?
> Anyway I'm out of time on this now: can one of David or Robert take a look
> at this and see if it can be improved? Or maybe a BBv2 expert can help out?
What is util/test.jam? If this is supposed to be some helper module, why can't
it have the only definition of those rules?
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