BetterC Is Not Ready

[Adam Ruppe]

On Friday, 3 June 2022 at 00:15:06 UTC, Jack Stouffer wrote:
> I didn't know this was possible sans compiling a custom version 
> of DMD yourself.

oh no, you just make a file called `object.d` and add it to your 
build and you'll override the normal druntime, and/or you can 
then just not link in the default one (`dmd -defaultlib= 
yourfile.d`). It is easiest to keep the default one around even 
if you don't link it in since some things will be auto-generated 
as needed.

I've written a little about this in the past: see my "D Cookbook" 
if you wanna buy one (though no-runtime D is a moving target and 
that is abit old now), or some of my blog posts:

http://dpldocs.info/this-week-in-d/Blog.Posted_2019_02_25.html#no-runtime-d
http://dpldocs.info/this-week-in-d/Blog.Posted_2020_07_27.html
http://dpldocs.info/this-week-in-d/Blog.Posted_2020_08_10.html#webassembly-in-d

There is a project called "lightweight D runtime" based on my 
webassembly code but retargeted to some arm devices, though the 
author had something come up in real life that derailed it. Some 
info here:
https://forum.dlang.org/post/zlnplbmdlcazkokccbxx@forum.dlang.org

> Based on the way people talked about it I had assumed the only 
> way to get rid of druntime was to use BetterC.

This is part of why I hate betterC so much. It is a major 
distraction from better approaches to the point of both 
denigrating D itself and obscuring facts about real druntime.