In new 1.4.0 release, nimbase.h defies the macro NIM_STATIC_ASSERT, trying some possible implementations depending on C compiler version used. The last fallback option uses a NIM_STATIC_ASSERT_AUX trick (nimbase.h line 275), which does not work. This macro is used twice in the following source (lines 323 and 542), both lines typedef the same symbol NIM_STATIC_ASSERT_AUX, gcc blames on redefenition. We need some better fallback static assert (there are many possible solutions floating around, one is mentioned in the code).
For me, the strong point of Nim is, it is still usable for ancient systems (and some people are bound to support them). Let us keep it, well, if not forever, but a little more at least? (works for me after a dirty but trivial fix with my current gcc).
Sorry for a poor wording, I registered just to report this.
Please report bugs on Nim's issue tracker so we can track and let you know if the bug is fixed.
AFAICT support for older gcc is also a priority in Nim as we aim for embedded support (where there are odd versions of gcc everywhere).
thanks for reporting this, and please report on nim's github issue tracker as mentioned. > works for me after a dirty but trivial fix with my current gcc
writing a PR would also be appreciated