You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Commit Message Bot (anonymous) said:
(extempore in r25452) Attacked classpaths to get "." off of it when it's not actually
specified. The commit makes me nervous, but there's no invisible way to
fix something like this.
** Attention, this commit changes classpath handling **
We desperately need some way of testing that the classpath has
certain qualities and does not have others; partest is not that way.
Closes #4857, no review.
@SethTisue said:
Ow, this bit me today. Is there a workaround? (And if there's a list somewhere of candidates for inclusion in 2.9.2, I hope this is on it.)
@SethTisue said:
In addition to being a correctness problem, this is also a performance problem in that it can potentially take a long time for the compiler to recursively descend the entire subtree of the current directory, looking for any and all class files.
@SethTisue said:
fwiw, in the case of Scaladoc I was able to work around this by patching my local copy of the scaladoc shell script to pass -Dscala.usejavacp=false instead of -Dscala.usejavacp=true (it isn't enough to override it in JAVA_OPTS since that comes earlier in the command line)
The text was updated successfully, but these errors were encountered: