I'll commit this. There are occasionally tests too expensive to run on a regular basis. For those, we should use a common mechanism (like test labelling) to ensure that expensive tests are run less frequently, but not never.
On Mon, Aug 18, 2014 at 10:43 AM, Naoto Sato <naoto.s...@oracle.com> wrote: > Hi Martin, > > Initial intention for not copying the JDK was to avoid timeout for some > cases in slow networked environment. But it still does not avoid all the > cases. So, yes I am fine with your fix. > > Naoto > > > On 8/16/14, 6:11 PM, Martin Buchholz wrote: > >> Hi Naoto, Alan, >> >> I'd like you to do a code review. >> >> http://cr.openjdk.java.net/~martin/webrevs/openjdk9/Currency-concurrency/ >> https://bugs.openjdk.java.net/browse/JDK-8055253 >> >> This will fix the transient failures in Currency ONCE AND FOR ALL! >> >
