On Sunday, 11 January 2015 at 14:10:56 UTC, ponce wrote:
On Sunday, 11 January 2015 at 13:37:33 UTC, Ola Fosheim Grøstad
wrote:
On Sunday, 11 January 2015 at 12:57:17 UTC, MattCoder wrote:
Since I'm relative new here, I want know from you agree with
this statement:
"
[–]clay_davis_sheeit 4 points 17 hours ago*
get real. D is more dead now than it was a year ago. if you
won't accept repo counts, look at how many people attended D
con vs Gophercon
"
"more dead" is a very subjective term.
It is "more dead" in the sense that you got @nogc and there
was a sense of movement towards getting to a workable memory
model, but since then nothing has happend. One step forward,
then stagnation.
The Rust team have announced that they are moving towards a
non-breaking stability situation within 6 weeks. And they have
a working memory model.
Andrei and Walter need to stop focusing on details for a
moment and focus more on presenting "a great plan" within 2
months. Meaning stating goals and plans which gives D a
direction that developers want and can believe in.
If no clear statements on where D is heading appears in the
near future... Well, then I am pretty sure that many of those
who prefer D will give Rust a spin when Rust hits 1.0, out of
boredom.
Rust is not complete feature wise, but a working memory model
and stability is more important than having single inheritance
and other convenience features...
So D is not dead, but is currently in a position where it can
be hit by both Go and Rust. The space between Rust (system
programming) and Go (server programming) is very tiny.
The problem with Rust and Go is that they only deliver in
theory, while D kicks some asses in practice. How?
Eg: at this very moment, D is more stable than Rust, ground
truth.
D has backends for GCC/LLVM/custom, Go has backends for GCC /
Plan9, Rust only for LLVM. None of Rust+Go can link with
binaries produced by eg. the Microsoft compiler. None of them
has Visual Studio integration with debugging support and that
is pretty important for native and enterprise programmers.
Go is actively behind the times by preventing shared libraries
and discouraging exceptions, let alone generics. None of the
C++ programmers I know give Go any credit, cause it would make
their work more difficult, and it's already pretty difficult.
Despite efforts, Rust don't get syntax right. They will enjoy
huge amount of complaining as soon as people actually use the
language, only to discover it is not fun enough and fun is more
important than "memory safety without GC". Looks like it
inherited its boringness from Ocaml.
I don't buy in the Rust team stability guarantees, you can't go
from pondering about removing "box" this very week (the syntax
is from this year) then promising stability forever starting
next month. But for some reason everything they say has a ring
of truth, because it's Mozilla they only do Good Things right?
They will come to the same model as D, minimizing code breakage
but do it anyway, because it's way more practical. And as soon
as Servo is interrupted because of internal politics at Mozilla
or rebudgeting (ie. very high probability), Rust will be halted
in a heartbeat since loosing its purpose. Ever noticed the Rust
original designer jumped off ship long ago?
That won't happen with D, whatever the ratio of github projects
in the fashion industry.
I am known to complain about some of the Go missing features.
However I am closer to use Go than D at work alongside .NET and
JVM.
Why? Because of Docker. Now with big names adopting Docker, our
enterprise
customers are looking into it for cloud deployments.
D really needs a some kind of killer application/framework.
--
Paulo
