"BCS" <n...@anon.com> wrote in message news:a6268ff154f88ccde5b99a63...@news.digitalmars.com... > Hello Walter, > >> BCS wrote: >> >>> Maybe someone should make a DMD-EE ("embedded edition") with a truly >>> minimal runtime and library along with a compiler patched to support >>> it (e.g. no GC, no AA's, no hidden memory allocations, no new/delete, >>> D-ified versions of the c std lib). The DMD patch shouldn't be to >>> hard, the runtime would mostly amount to dropping stuff and the lib >>> would mostly just copy some stuff from phobos and wrapping, porting >>> or binding (depending on what's more reasonable) stuff from C. >>> >>> I'd offer to do it but I have almost no experience in any of those. >>> >> There's not a whole lot of point to do this, other than checking a >> box, because embedded systems developers have no problem doing this >> themselves (create a custom runtime library). >> > > It would lower the barrier to entry. While not to important once someone > has chosen D, I would think it would be much more important when they are > just considering it. >
Exactly. D is enormously better than C for many such things, but C/C++ have so much inertia in all the fields C/C++ are used that regardless of how much better D is, barrier to entry and overcoming C/C++'s inertia is still going to be a significant issue for D. We want it to be as easy as possible for potential D users in any field appropriate for D to try it out. Then they can customize it further if they really need to, but the real issue is getting them to decide to use it in the first place, and I'd be very surprised if a lot of them would be willing to do *all* of the necessary adjustments merely to test drive it. Plus, there's embedded homebrew too, and the homebrewers aren't likely to have a lot of time to do the necessary adjustments.