Boost logo

Boost :

From: John Phillips (phillips_at_[hidden])
Date: 2004-04-29 16:45:52

   Sorry I don't have time at the moment to give a more complete
description of the problem (and hopefully a fix, as well), but
big_integer does not yet compile under CW 9.2 on Mac OSX (10.2.8). The
error list is huge, but the important one seems to be :

Unimplemented C++ feature
big_integer.hpp line 153

The portion in question is

    template<class T>
         friend bool operator<(T lhs, const expression<Base>& rhs)
     { return rhs > lhs; }

which doesn't strike me as all that odd, but maybe I'm missing something.
   I'll try to look at it more, later and see if I can fix the problem.

   Also, I would love to see a big_float to go along with big_integer. I
haven't thought about it enough to know if they should be separate
libraries or different facets of the same.

                        John Phillips

Richard Peters wrote:
> I have uploaded a new version of the big_integer library which also compiles
> with GCC 3.4.0 (cvs version). This version now compiles on compilers which
> implement two-stage name lookup. The library is available at
>, the version is 0.1.
> best regards,
> Richard Peters
> "Giovanni Bajo" <giovannibajo_at_[hidden]> wrote in message
> news:c5mf2u$bt0$
>>Richard Peters wrote:
>>>A few months ago there was some discussion about an arbitrary
>>>magnitude integer library. I have now implemented such a library,
>>>which is located at
>>Compiling the code generate hundreds of failures because of two-stage name
>>lookup. You tested it under MSVC 71 and GCC 3.2 which both fail to
> implement
>>lookup rules correctly. I suggest you try again with GCC 3.4.0 (a
> pre-release
>>binary package should be available now) or EDG-based compilers.
>>Giovanni Bajo
> _______________________________________________
> Unsubscribe & other changes:

Boost list run by bdawes at, gregod at, cpdaniel at, john at