Boost logo

Boost :

From: Rene Rivera (grafik666_at_[hidden])
Date: 2003-02-05 18:35:02


[2003-02-05] Ronald Garcia wrote:

>
>
>On Wed, 5 Feb 2003, Rene Rivera wrote:
>
>> [2003-02-05] Ronald Garcia wrote:
>> >
>> >I would like to request the addition of the -ansi flag to the
>> >[ ... ]
>>
>> Some questions...
>>
>> Are the errors in Boost code, or your code?
>err...both (boost code that happens to be my code :) ).

Ah, got it ;-)

>> Are you using Boost.Build for Boost only, or for your builds also?
>I'm only using Boost.Build to run the regression tests -- specifically for
>Boost.Multiarray, a library I maintain. My rationale for requesting the
>addition is that at least multi_array passed all regression tests without
>it, though stricter (wrt the C++ Standard) compilers (that I do
>not have access to :) ) reject it. It seems to me that enabling
>stricter adherence to the Standard (at least as per the EDG front-end used
>in icc) by default would be a Good thing.

It would be a good thing, IMO. But not for those that don't want it ;-] For
example those people using Boost.Build for non-Boost building.

What this suggest is adding support in Boost.Build for specifying standards
conformance levels for the toolsets.

But for now I would suggest you change the multi_array tests to add that
flag, that is if you think it's important in this case. Taking one of your
tests as an example:

    [ run libs/multi_array/test/constructors.cpp
        <lib>../libs/test/build/boost_test_exec_monitor
        : : : <intel-linux><*><cflags>-ansi ]

-- grafik - Don't Assume Anything
-- rrivera_at_[hidden] - grafik_at_[hidden]
-- 102708583_at_icq


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk