Once again, 4.8 is <i>not</i> yet out. It is "officially" released only when an announcement is sent to this mailing-list: <a href="http://gcc.gnu.org/ml/gcc-announce/2013/" rel="nofollow">http://gcc.gnu.org/ml/gcc-announce/2013/</a><p>Even the 4.8 tarballs have <i>not</i> been uploaded to FTP servers: <a href="http://ftp.gnu.org/gnu/gcc/" rel="nofollow">http://ftp.gnu.org/gnu/gcc/</a><p>Until then, let's wait.<p>EDIT: Looks like the tarballs are being mirrored, and 4.8 is indeed out.
This page shows the C++11 features that are added by each version of GCC, in case you're curious how C++11 support has changed from 4.7 to 4.8: <a href="http://gcc.gnu.org/projects/cxx0x.html" rel="nofollow">http://gcc.gnu.org/projects/cxx0x.html</a><p>EDIT: I see that some of the C++11 changes are also listed on the change log (<a href="http://gcc.gnu.org/gcc-4.8/changes.html" rel="nofollow">http://gcc.gnu.org/gcc-4.8/changes.html</a>), but that page seems to be missing the changes in concurrency support: bidirectional fences, memory model, `at_quick_exit`, and thread-local storage.
With all these folks excited by gcc releases (and really the C++11 features) it would seam like there's lots of folks still using C++. However, I have the hardest time in the world recruiting good C++ engineers (here in NYC at least).
Link to official announcement by Jakub: <a href="http://gcc.gnu.org/ml/gcc-announce/2013/msg00001.html" rel="nofollow">http://gcc.gnu.org/ml/gcc-announce/2013/msg00001.html</a>
Off Topic Question: As a compiler (C++ compiler esp) slides through its versions, it gets better optimising features and every newer version can produce faster binaries.<p>Can this cycle end? Or, will the cycle continue forever? (Like gcc 5 more efficient, gcc 6 even more efficient.....etc)