On 2/7/18 2:26 PM, Adam Petcher wrote:
On 2/7/2018 12:33 PM, Sean Mullan wrote:
This looks fine although I don't really think you need the othervm on
line 36; that seems like something that is done separately as part of
testing all of the tests on different locales.
I included that line as a regression test for this bug. I would prefer
to keep it, because it uses a locale that is known to cause problems for
(earlier versions of) this test. I don't know how common these
problematic locales are, and how often the test suite is run under them.
If there is a regression here, it would be nice to know about it before
release testing.
Do you expect this @run line to cause any problems (other than being
unnecessary)?
I guess it's fine; it is just that I have seen default locale test bugs
filed in other places (do a search on the i18n label), so it seems like
there is some overall broader test coverage on this already. Maybe ask
the submitter for more info as to how this came about. Either way, it
was a minor comment, so I am ok if you want to keep it.
--Sean
