On 31.10.2016 19:08, Joerg Sonnenberger wrote: > The duplicated check by itself is harmless. It isn't clear whether the > long long non-support glue works properly or even where it is needed at > all. As such, it is unclear on whether the second instance is necessary > or not. There was a discussion about "long long" in the ./build.sh tools (like a year ago) and - if I remember correctly Minix people gave up on supporting lack of longlong. Personally I don't see any reason to support systems (compilers) without it in pkgsrc and C99 might be official bare minimum. (I was told that lack of C99 in BSD/OS already halts possibility to bootstrap there). If there is an agreement on it, I will drop these harmless checks and so make clean room for strtoi(3) addition in libnbcompat, based on strtoll(3) with longlong support assumption (of course via intmax_t etc).
Attachment:
signature.asc
Description: OpenPGP digital signature