Subject: Re: [boost] Is there a macro to test ifstd::numerc_limits<>::max_digits10 is supported by a platform?
From: Paul A. Bristow (pbristow_at_[hidden])
Date: 2012-07-02 14:32:09
> -----Original Message-----
> From: boost-bounces_at_[hidden] [mailto:boost-bounces_at_[hidden]] On Behalf Of Jürgen
> Sent: Saturday, June 16, 2012 11:49 AM
> To: boost_at_[hidden]
> Subject: Re: [boost] Is there a macro to test ifstd::numerc_limits<>::max_digits10 is supported by
> Hi Paul,
> On Thursday, 14. June 2012 13:26:20 Paul A. Bristow wrote:
> > > -----Original Message-----
> > > From: boost-bounces_at_[hidden]
> > > [mailto:boost-bounces_at_[hidden]]
> > > On Behalf Of John Maddock
> > >
> > > Try BOOST_NO_NUMERIC_LIMITS_LOWEST as a proxy?
> > OK, but I'm on holiday for a while so I'll wait until I get back
> > before I create more trouble ;-)
> > Anyone trying multiprecision and wanting this feature can patch their
> > \boost-trunk\boost\test\tools\impl.hpp locally.
> I've tested BOOST_NO_NUMERIC_LIMITS_LOWEST with gcc-4.6.3 on Linux and this seems to work.
> Patch attached.
Thanks - looks good, but I think it might be wise to test how the macro works on a specific test not
actually using Boost.Test before causing a nuisance - as I have discovered to several peoples cost
The trouble is that Boost.Test is used by all platforms. So it will be disruptive to just patch
like this and hope for the best.
> I'd like to commit it and check the test results for breaks.
> Or is a special macro needed here?
Well perhaps but hopefully not - only a check on all platforms in the testing regime will make sure.
I'll check this out, perhaps using a temporary program attached to Boost.Math testing schedule and
then see what platforms work.
There may still be obscure platforms that are not routinely tested that don't work?
--- Paul A. Bristow, Prizet Farmhouse, Kendal LA8 8AB UK +44 1539 561830 07714330204 pbristow_at_[hidden]
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk