From: Robert Ramey (ramey_at_[hidden])
Date: 2022-12-06 20:37:11
On 12/6/22 9:52 AM, John Maddock via Boost wrote:
> On 06/12/2022 16:52, Peter Dimov via Boost wrote:
>> Rainer Deyke wrote:
>>> At a first glance, the following libraries would have to justify
>>> their continued
>>> existence as "core" Boost libraries:
>>>
>>> Â Â Â Any (superseded by std::any)
>>> Â Â Â Array (superseded by std::array)
>>> Â Â Â Assign (superseded by std::initializer_list)
>>> Â Â Â Atomic (C++11 feature emulation library)
>>> Â Â Â Bind (superseded by lambda expressions)
>>> Â Â Â Chrono (superseded by std::chrono)
>> ...
>>
>> This gets us back to our never-ending discussion of Boost dropping
>> C++03 support.
>>
>> I'm sure this time it will be more productive than our last three
>> attempts.
>
> Perhaps the question should be this: are there libraries still using
> C++03 Boost versions of these libraries, that are still actually
> supporting C++03 themselves? Or are these dependencies just cruft left
> over from previous C++03 support which is no longer there in practice?
>
> Just wondering ;)
>
> John.
FYI - serialization uses C++03 libraries (e.g. MPL) and builds and
passes tests on C++03. I don't know how many users might use that. My
motivation for keeping this is:
a) the library is for the most part correct. (though it took a while).
b) so I disinclined to mess with it.
c) it costs nothing to leave it as it is
d) it maximizes the number of users/applications it might support.
e) There would be no effective benefit to anyone to alter it to use more
recent C++ features.
So whatever boost "decides", I don't see any reason to change anything
Robert Ramey
>
>
> _______________________________________________
> Unsubscribe & other changes:
> http://listarchives.boost.org/mailman/listinfo.cgi/boost