On Thursday, 3 June 2021 at 23:47:07 UTC, zjh wrote:

Yes, for Phobos v2 one of the primary goals should be to not being forced to rely on GC. Phobos should only rely on malloc/free. Phobos may be using reference counting internally as it also only relies on malloc/free.

On Thursday, 3 June 2021 at 23:48:16 UTC, zjh wrote:

As a small language, if you want to succeed.There is no way out except to be the best.
Otherwise, why don't I use C++?

rust,go,zim.I dislike them.
Rust: slow compilation, weak template function.
Go: there is no template function, and the abstraction is not strong,error process,time question.spaghetti....
Zim: the grammar is ugly.
D's advantage is template.But now the advantage over C++20 is smaller.
So D should have a sense of urgency.
D's position is too vague. Don't want to gain everything,Everything is nothing.
what D need to do is enhancing advantages, discarding disadvantages, so simple!
Template metagramming is a big advantage, we should enhance. GC is disadvantage, we should discard it!
discard GC,attract C++er ,pythoner/scripter/learner will all come.
otherwise,who help you write the lib?

On Friday, 4 June 2021 at 00:14:11 UTC, zjh wrote:

Zim? Is that what they speak in Zimbabwe?

On Friday, 4 June 2021 at 00:39:41 UTC, IGotD- wrote:

Zig.

Yet another GC vs NoGC thread :sigh:

On Friday, 4 June 2021 at 11:41:49 UTC, sighoya wrote:

Routine determines success or failure.

On Friday, 4 June 2021 at 00:14:11 UTC, zjh wrote:

GC won't go away tho. What might happen is more flexibility. The GC-phobia is irrational.

On Friday, 4 June 2021 at 12:44:07 UTC, Imperatorn wrote:

The topic doesn't fit in this thread, but it isn't irrational.

You have to wait for all participating threads to be ready to collect, so it isn't only about collection speed. In essence you end up with some of the same issues as with cooperative multitasking.

And it is also obvious that collection speed will drop as your application grows and you start working with larger datasets.

So, you might initially think it is fine, but end up rewriting your codebase because it only worked well with the simple prototype you started with.

That's not a good strategy.

(but ok for batch programs)

On Friday, 4 June 2021 at 13:32:37 UTC, Ola Fosheim Grøstad wrote:

The most irrational issue here is that the language itself prevents precise collection, and there is no willpower to change it. If you combine task-local GC with fully precise compiler-guided scanning, then you'd have something that would work.