Re: [boost] List of C++ 11 only Boost libraries and their status?
On Sat, Dec 13, 2014 at 08:37 AM, Rogerio dos Santos <rogerio.santos.main@gmail.com> wrote:
On Sat, Dec 13, 2014 at 8:27 AM, Nat Goodspeed <nat@lindenlab.com> wrote:
I'm just contradicting the assumption that those who must work with old compilers aren't doing any new development, or have no interest in new libraries.
To add to this point, I would like to mention that even new developments trends might require support to old or limited compilers. E.g.: IoT, real time projects, industrial HMIs and mobile applications.
Roger
Indeed. I work on embedded, real-time industrial control systems in a field where certain specifications (unfortunately, IMO) severely restrict the platform choices available. In many cases, the systems developed must be built using pre-C++98 compiler technology. While that is a problem I am attempting to address, I suspect it is going to take at least another year to do so, and I am confident that at least one of the platforms will probably never support multi-threading even if its vendor does end up improving or replacing its toolchain (and I have no iron-clad guarantee from the vendor that the toolchain will actually be improved or replaced, though I have a reasonable hope). None of the above constraints imply that the libraries developed by Boost are not of use for these systems, or that we are not interested in using Boost libraries to the extent that it is possible to do so. I think it is possible to structure compatibility code so that it doesn't egregiously interfere with future development. I always get nervous when Steve posts asking to hack away the support for a few more of the older toolchains. While I fully appreciate that his end-goal is well-intentioned, there's still important code out there relying upon some of those pre-C++98 workarounds that can still benefit from the fixes and new libraries introduced into new versions of Boost. Consequently, I'm much in favour of allowing Boost library maintainers who wish to retain support for older compilers the latitude to do so. Dave
Dave Gomboc wrote
Consequently, I'm much in favour of allowing Boost library maintainers who wish to retain support for older compilers the latitude to do so.
I don't think anyone can prohibit a library maintainer from maintaining compatibility with older platforms nor impose any requirements that old platforms not be supported. It's still a free country. I expect to maintain such support for the serialization library as long as it's practicable. In fact, it's easier to do this than it would be to convert the library to use C++11 features. This leaves aside the fact that such a conversion would diminish the usage of the library which would be contrary to my main goal. New libraries are an entirely different thing however. It's easier to make a new library using the C++11+ features. The number of extra users that would benefit from trying to make the library with C++03 would be very small in relation to the extra amount of work involved. So I would recommend to authors of new libraries that they target the most recent versions of the language that are widely used. Robert Ramey -- View this message in context: http://boost.2283326.n4.nabble.com/List-of-C-11-only-Boost-libraries-and-the... Sent from the Boost - Dev mailing list archive at Nabble.com.
participants (2)
-
Dave Gomboc -
Robert Ramey