Re: [boost] [Endian] Review

Subject: Re: [boost] [Endian] Review
From: Beman Dawes (bdawes_at_[hidden])
Date: 2011-09-10 08:05:48

On Fri, Sep 9, 2011 at 3:20 PM, Vicente J. Botet Escriba
<vicente.botet_at_[hidden]> wrote:
> Le 09/09/11 15:05, Beman Dawes a écrit :
>>
>> On Fri, Sep 9, 2011 at 2:32 AM, Vicente
>> Botet<vicente.botet_at_[hidden]>wrote:
>>
>>> Hi,
>>>
>>> I have separated the endian class into a endian_pack that is integer
>>> agnostic and an endian class that behaves as an integer (You could see
>>> the
>>> details in the Sandbox under endian_ext).
>>>
>>> Reference docs:
>>>
>>>
>>> http://svn.boost.org/svn/boost/sandbox/endian_ext/libs/integer/endian/doc/html/toward_boost_integer_endian/reference.html#toward_boost_integer_endian.reference.integer_endian_pack_hpp
>>>
>>>
>>>
>>> http://svn.boost.org/svn/boost/sandbox/endian_ext/libs/integer/endian/doc/html
>>>
>>> (Sorry but there is a bad link to the reference)
>>>
>>> This is not complex
>>
>> Complexity is in the eye of the beholder. To me, that approach was
>> markedly
>> more complex and confusing.
>>
> What do you find complex in separating the classes endian_pack and endian?

That particular approached seemed to markedly increase surface area.
It might be a good thing to provide the basic endian buffer facility
in addition to the full endian integer facility, but not at the cost
of doubling the surface area.

A policy, as Tomas mentioned, would be one way to do that. I'm about
to reply to him asking for his reaction to an inheritance based
approach.

--Beman