Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent

[Ola Fosheim Grøstad via Digitalmars-d]

On Saturday, 21 November 2015 at 13:28:20 UTC, Joakim wrote:
> an inroads into.  They simply optimize programmer convenience 
> over efficiency and that's an acceptable tradeoff in certain 
> niches.  However, even in that market, there are badly-designed 
> languages that do unreasonably well.

That's true. The computer market is heavily dominated by critical 
mass. That means any poorly thought out technology will survive 
on a long death trail when it reaches critical mass. Php was 
really only for adding a little bit of scripting within HTML, and 
was shipped as easy installs for web servers. Then Php got 
installed by default by providers on the cheap offerings and 
people wanted apps that they could just install and configure 
with minimal effort. Which is why we have seriously bad code like 
Wordpress all over the web...

Anything can be made to scale... just reach critical mass and it 
scales since people stick to their investments. And knowhow is a 
big investment...

> Not interested in an argument here, so I'll end with a 
> constructive question: what do you believe D "should learn from 
> [PHP/JS] rather than dismissing it?"

 From Php:

- Focus on ONE domain and make it very easy to get started with 
in that domain.

- Gain critical mass in that domain before spreading out.

 From JS:

- A modern language that is aiming for high level convenience has 
to transpile to javascript/asm.js. That's a very important target 
if you want to increase the amount of portable code written for 
your language.

- One big-big advantage Node.js has is the ability to make 
backwards compatible views of a web-app by running client code on 
the server for older browsers.

- When designing a language around a garbage collector, make sure 
it is top notch before extending the language.