Re: [boost] [1.40.0] Beta 1 release candidate available

Subject: Re: [boost] [1.40.0] Beta 1 release candidate available
From: KTC (ktc_at_[hidden])
Date: 2009-08-12 05:33:58

Vladimir Prus wrote:
> Václav Haisman wrote:
> 
>> Beman Dawes wrote, On 10.8.2009 17:50:
>>> 1.40.0 beta 1 release candidates are available at http://boost.cowic.de/rc/
>>>
>>> Before pushing the beta out to SourceForge, I'd appreciate it if
>>> several people would try the release candidate and report success or
>>> problems.
>> It builds on Cygwin 1.5 with GCC 4.3 if I just choose release variant.
>> --build-type=complete fails:
> 
> Why do you want --build-type=complete?

It doesn't and shouldn't matter what the answer to that question is. 
Whether it's because someone is not sure what they want yet so want all 
variant, or they generally have a need for all variant. It's a common 
and basic option, if you want to build all supported version pass 
--build-type=complete. I think it actually say something along that line 
in the docs. The user shouldn't have to worry about the details. That's 
the whole point of having a catch all option instead of making users 
specify each and every variant separately.

> 
>> amber2::wilx:~/tmp/boost_1_40_0_beta1> nice -n 20 ./bjam -d+2
>> -sEXPAT_INCLUDE=/usr/include -sEXPAT_LIBPATH=/usr/lib toolset=gcc-4.3
>> --without-mpi stage --layout=versioned --build-type=complete
>> Note: Building Boost.Regex with Unicode/ICU support enabled
>>     Using ICU in  /usr/include
>> warning: Graph library does not contain MPI-based parallel components.
>> note: to enable them, add "using mpi ;" to your user_config.jam
>> error: link=shared together with runtime-link=static is not allowed
>> error: such property combination is either impossible
>> error: or too dangerious to be of any use
> 
> Hmm, ICU's acting up again.
> 
> John,
> this seems to become a recurring issue. Can you suggest some alternative
> behaviour I should implement?

Well, it's the same issue as before.

Either, the static combination with ICU that causes the conflict should 
not be built, or the static combination should be built without linking 
to ICU. With or without a warning message to the user and documentated 
in docs.

KTC

-- 
Only two things are infinite,
  the Universe and Stupidity.
And I'm not quite sure about the former.
         - Albert Einstein