- Boost - lists.boost.org

[geometry] MSVC needs namespace even if it is hoisted
by Barend Gehrels 24 Dec '11

24 Dec '11

Hi, We got a strange MSVC compilation problem in a Unit Test for Boost.Geometry. Attached a miminal reproduction. The essention is that we have stated: using strategy::distance::services::returning_type; and somewhere after that: typedef typename returning_type<strategy_type>::type the_return_type; but this is not accepted, in this case, by MSVC. It is well accepted by gcc, clang and comeau. If the namespace is specified, it is accepted by MSVC. The error message is: reproduce_msvc_using_bug.cpp(257): error C2059: syntax error : '<' And that is the same message as if the returning_type would not be a template. If we state: std::cout << typeid(returning_type).name() << std::endl; It compiles in MSVC, but it should not compile, because it is a template. So somehow it picks up this "returning_type" as if it is not a template. But of course it is. I've tried another (much smaller) minimal reproduction to mimic this problem but that is compiled normally by MSVC. So attached this one with (reworked) pieces from Boost.Geometry and even Boost, but still all contained in one file. We've discussed this internally and the current thought is that it is a compiler problem. If so, is this issue known to anyone on this list, or Microsoft? Otherwise I will report it. If not so, what is the problem here? It is of course easy to work around: giving the full namespace is OK for any compiler. So this is more out of curiosity: what is wrong here? Regards, Barend

3 3

TRAC hit hard last night
by Marshall Clow 23 Dec '11

23 Dec '11

We got a major infusion of garbage; strictly speaking, it doesn't appear to be spam, because it's not advertising anything, and the text doesn't contain any links. However, it doesn't have anything to do with boost. I've deleted about 40 tickets this morning. Any suggestions on how to prevent this in the future? -- Marshall Marshall Clow Idio Software <mailto:mclow.lists@gmail.com> A.D. 1517: Martin Luther nails his 95 Theses to the church door and is promptly moderated down to (-1, Flamebait). -- Yu Suzuki

3 2

C++11 challenge: ASSERT_INVALID_EXPR
by Dave Abrahams 23 Dec '11

23 Dec '11

I don't know whether this is actually possible, but I was thinking today that there might be a way to use the new SFINAE to detect errors in a large class of invalid expressions without actually causing a hard error, and thereby build an assertion that fails compilation when passed a valid expression and passes for any invalid expression in the large class mentioned above. Any takers? -- Dave Abrahams BoostPro Computing http://www.boostpro.com

4 4

[concept_check] Some issues
by Robert Kawulak 23 Dec '11

23 Dec '11

Hi, I've spotted some problems when playing with BCCL: - The implementation of assignable_archetype defines private default and copy constructors, so for example assignable_archetype<copy_constructible_archetype<>> will not actually be copy constructible. The same for convertible_to_archetype and convertible_from_archetype, but additionally with copy assignment. Is this by purpose? - copy_constructible_archetype defines public default constructor - again, am I overlooking something or it should not be there? - Function and predicate archetypes also seem to provide too wide interface (they are copy constructible and assignable). Additionally, they don't allow for archetypes chaining since they lack the Base parameter (with the only exception of binary_predicate_archetype - why the inconsistency?) - Apart from the provided function object concepts and archetypes it seems it would be useful to have a general Callable concept and the corresponding callable_archetype, parameterised with the types of arguments and return value. Wouldn't it be valuable to add them? - There's boost/concept_check directory in the trunk - are the files in it actually used anywhere? I've found no reference to them in Boost headers, documentation doesn't mention them too. Best regards, Robert

3 10

boost and google search
by Edward Rankin 23 Dec '11

23 Dec '11

Hi, I am sure it is pretty common for users of boost to use google to search for documentation/tutorials on their favorite library in boost. I have found that the top hits are typically not the latest documents. For example I just searched for boost,foreach and a documentation from 1.35 was the first hit. It would be great if there is a way to tell google to update their links to the latest version. Anyone knows whether this is possible, and how to make it happen? This problem probably applies to any search engine but I can only comment on google search. Thanks, Edward

11 23

constness in barrier.hpp
by Gaetano Mendola 23 Dec '11

23 Dec '11

Hi all, simple patch to make const the private member m_threshold in class barrier Regards Gaetano --- barrier.hpp.original 2011-12-22 12:05:05.921720854 +0100 +++ barrier.hpp 2011-12-22 12:05:36.936205847 +0100 @@ -52,7 +52,7 @@ private: mutex m_mutex; condition_variable m_cond; - unsigned int m_threshold; + const unsigned int m_threshold; unsigned int m_count; unsigned int m_generation; };

3 3

[config] ok to merge forceinline changeset to release?
by Eric Niebler 23 Dec '11

23 Dec '11

I got a little overzealous and merged to release some code that uses BOOST_FORCEINLINE, only to discover that that macro doesn't exist on the release branch yet. The right fix is probably to get config in sync between the two branches. OK to merge [74917] to release? -- Eric Niebler BoostPro Computing http://www.boostpro.com

2 2

Re: [boost] RFC: Multiprecision arithmetic library
by John Maddock 22 Dec '11

22 Dec '11

>I really like the idea, and the API looks good. > > Have you considered an MPIR backend, in addition to GMP and others? Being > the > Windows port of GMP, the interfaces and usage are identical. Nod, the two are treated the same - in fact I use MPIR for testing under Visual Studio. John.

1 0

it would be nice if people logged into trac before commenting on tickets.
by Marshall Clow 22 Dec '11

22 Dec '11

That way we don't get things like this: > Changes (by anonymous): > > * status: new => closed > * resolution: => invalid Thanks, everyone! -- Marshall Marshall Clow Idio Software <mailto:mclow.lists@gmail.com> A.D. 1517: Martin Luther nails his 95 Theses to the church door and is promptly moderated down to (-1, Flamebait). -- Yu Suzuki

4 3

Re: [boost] ASIO Query
by Olaf van der Spek 21 Dec '11

21 Dec '11

On Wed, Dec 21, 2011 at 9:21 PM, james(a)jamesbb.co.uk <james(a)jamesbb.co.uk> wrote: > Blocking - I've thought about this, but the MySQL C/C++ libraries unfortunately > don't provide asynchronous queries. > > I'm thinking of moving database reads into a seperate thread(s). It would be > nice if I can integrate this with the io_service, but I'm unsure how to do so. > Is it possible to pass arbitrary work items to an io_service, in the same manner > as async_read/write? Does every request involve blocking calls? If so, I'd not go with a 'shared nothing' design. Olaf

2 1