Martin Kinkelin
Martin Kinkelin
https://dlang.org/changelog/2.100.0.html#command-variables
Have you tried a dub build instead of meson, seeing that the repo has a regular dub.json? Dub's all-at-once builds (per dub package) are waaay less likely to hit all...
Well fighting with missing symbols is unfortunately a constant pain with D, especially with separate compilation. Most of the time, `-allinst` now suffices to fix those. These are (old and...
FWIW, I can confirm that this happens with DMD v2.100.1 too, and also with `dub build --build-mode=singleFile`, but `dub build` (compiling all modules in one compiler invocation) works. - I've...
I can't say anything about GDC (other than the jump from v2.076 to v2.100 obviously being huge), but we haven't had any problems with dustmite bundled with LDC (FWIW, from...
> `powIntegralImpl` is marked [pragma(inline)](https://github.com/dlang/phobos/blob/1d93fa32ebed51f6de51c2cfd244678043738dea/std/math/algebraic.d#L956), so should never be pruned by the front-end. The culling doesn't and AFAIK has never taken that into account; it's the glue layer that needs...
> but it's dependency powIntegralImpl isn't. Clearly is for LDC, it's inlined into `nextPow2` (even at `-O0`), but doesn't make it to the object file itself.
> I am having issues with throwing exceptions in constructors. The testcase obviously works just fine without ASan, just for clarification.
Wow, sounds extremely weird! :D - A factory returning a variant seems much more reasonable. If you think the original code should really have worked (**edit**: or reported a better...
I think we should actually be in better shape wrt. this than DMD nowadays - class TypeInfos are emitted once in their owning CU only, struct TypeInfos into each referencing...