On Friday, 11 December 2015 at 07:40:55 UTC, Jacob Carlborg wrote:
I'm not sure how related rdmd is to the above mentioned
features. If one would use rdmd for the above, it would require
to compile the code as a dynamic library and the load that. I
guess that could be possible.
I was really trying to get a handle on what their point was.
rdmd provides an immediacy that is similar to using some
scripting languages. For me, rdmd is better to use when
prototyping something than C++, but I'm still more productive
prototyping something with R or Matlab.
Nevertheless, while I think there is value in an REPL-like
environment for D, I would also give it a low, low priority.
Some people have said things like D is an AOT compiled language.
Fine. But imagine you had a scripting language with the exact
same syntax and semantics as D, but this language can be used
with an REPL. Maybe there would be a few differences, but for the
most part a program written in this language could also be
compiled with dmd.
Consider the relationship between C and Ch. It provides an REPL
interactive shell for C along with some other changes. While
there are some differences, you're still basically using an
interpreted version of C.
Let's suppose there's a Dh that is to D as Ch is to C. Would some
people find value in Dh? I think yes. Would there be some overlap
between implementing this hypothetical language and dmd/rdmd? I
would suspect quite a bit (though I don't know enough of the
technical details). Would it be possible to use a JIT in the
implementation? I don't see why not. Should smart people work on
creating Dh? I'm guessing other priorities are more important.
