6 ms·
It's been five years (and two standard revisions). Can we stop calling C++11 a new standard?
by maxlybbert 10y ago
It's been five years (and two standard revisions). Can we stop calling C++11 a new standard?
- pjmlp 10y agoTell that to embedded system guys still having to deal with C++98, not even C++03.
- Strangework 10y agoReally depends on where you go. In our office, we use C++11 functionality in our AVR chips.
- 2bitencryption 10y agoStandards are not adopted overnight. C++11 is still, effectively, a "new" standard.
- maxlybbert 10y agoI realize that many projects haven't made the switch, and in some cases they have good reasons to be cautious. And I realize that many programmers aren't familiar with the various changes, even after five (or more) years to catch up. But that doesn't change the age of the standard, or the fact that C++14 was published, and the Committee is finishing up C++17. At some point it's silly to call C++11 "new." "Unfamiliar," sure, but not "new." It reminds me of Larry Wall's talk ( http://www.wall.org/~larry/pm.html http://www.wall.org/~larry/pm.html ) "Do you know when New College was founded. Any guesses? New College was new in 1379."
- marssaxman 10y agoThe industry is still in the process of adopting it. My workplace only switched over six months ago. (That had been annoying, as I picked it up in roughly 2013, and it was hard to unlearn the new habits when I started working here.) My recollection is that C99 took roughly ten years to go from "the new thing" to "normal".