Compatibility with older C++-Standards under C++11

[Anonymous]

Hi all,

Parsing functionallity (Example ORED) is implemented using modern C++-Standard. Is there any special reason for that?

Well, I understand, that modern C++ is more pythonic and is better for data conversion tasks.

Is there any plans for earlier C++ versions?

Alex

[Anonymous]

Hi Alex,

Most of ORE is written in C++11 (but not C++14 or 17, so it’s modern, but not cutting edge), as it says in the FAQ this is for convenience, features like autoboxing, foreach loops, etc. There are a few lambdas used, but nothing that would be difficult to back port to C++98/03 (e.g. no use of std::thread or anything like that)

That said, there are no plans to support earlier versions of C++, if you are a windows user, you need MSVC 2013 or higher, this was released over 3 years ago. For GCC it is version 4.8.1 which again is over 3 years ago, so it should be accessible. Restricting ourselves to C++98/03 would be a retrograde step and it’s hard to maintain for windows users (MSVC does not have an equivalent of gcc’s –std= flag).

Regards,
Niall.

[Anonymous]

Thanks Niall. It is a great work you’ve done.

I will talk to our IT and try to move to MSVC 2015, but I don’t think it will happen soon. You are right, it is still possible to backport it to MSVC 2010.

Looking at the examples, looping is the most performance killer. Do you have any idea, what would be the best way to parallelize it? Schould we open a separate thread for it?

[Anonymous]

Hi Alex,

The binaries in the release should still work for you, even if you can’t build it with MSVC2010.

By all means open a new thread on parallelisation, there is a small comment in the FAQ already.

Regards,
Niall.

[Author]

Posts