Date: Thu, 13 Apr 2023 15:58:22 UTC
On Thu, Apr 13, 2023 at 07:03:52AM -0700, David Wolfskill wrote: > ... > So the "U_NOEXCEPT" lines are in the (older) icu-72.1,1. > > But on machines that have been updated this morning, I see: > > freebeast(12.4-S) pkg info -o icu\* > icu-73.1,1 devel/icu > freebeast(12.4-S) grep -n U_NOEXCEPT > /usr/local/include/unicode/locid.h > freebeast(12.4-S) > > [Note that the above is in the machine's "native" environment -- not in > a poudriere jail.] > > I am not at all clear on how poudriere could be getting an old version > of the file. In the mean time, I have started a biuld of www/chromium > on my laptop (using portmaster) -- laptop already updated devel/icu & > all ports that depend on it except for www/chromium. > .... After my local (laptop) build also failed -- for the same reason -- I was able to determine that the copy (of locid.h) that is being used in the failure cases is one from the distfiles used to build www/chromium: localhost(12.4-S) pwd /common/ports/www/chromium/work localhost(12.4-S) grep -n U_NOEXCEPT `find . -name locid.h` ./chromium-112.0.5615.49/third_party/icu/source/common/unicode/locid.h:297: Locale(Locale&& other) U_NOEXCEPT; ./chromium-112.0.5615.49/third_party/icu/source/common/unicode/locid.h:323: Locale& operator=(Locale&& other) U_NOEXCEPT; localhost(12.4-S) Maybe (upstream) www/chromium needs a fix to work with icu-73.1 (instead of icu-72.1)? Peace, david -- David H. Wolfskill email@example.com FD&C Red #3: OK to eat, but banned in cosmetics since 1990. -- FDA See https://www.catwhisker.org/~david/publickey.gpg for my public key.