Re:The myth of using Scala as a better Java
by grzm on Thursday March 02, 2017 @12:58PM
(#13776414)
Not knowing C++, what are the types of changes in practices? Which are the most burdensome? I can imagine some accompanying changes in the language (as often best practices are there to cover infelicities in the language that are corrected in future versions). Are these of a scale that require large rewrites? Like, past best practices are now actual flaws (as opposed to just not the peak optimizations)?
[ Parent
| on HN ]
Re:The myth of using Scala as a better Java
by mcguire on Thursday March 02, 2017 @08:35PM
(#13780078)
I first used C++ around 1990, when it didn't have much of a standard library. Then came the Annotated Reference Manual with its crazy talk of templates and namespaces and what not, which weren't implemented by any compiler I had. Somewhere around here is a copy of PJ Plauger's book on the standard library that doesn't mention templates at all. The standards committee pulled a fast one on Plauger: between the publication of his book and the finalization of the draft, the STL took over the world. Then came C++11, with move constructors, anonymous functions, and a couple of new smart pointers (which was good). C++14 introduced type deduction. And there's Boost's smart pointers (and everything else).
Nothing requires a rewrite, but five year old code looks nothing like what people say is "best practices" now.