Thank you for checking.
Gary
On Wed, Nov 11, 2015 at 8:45 AM, Remko Popma wrote:
> Gary, looks like the code in ReflectionUtil already has a comment with a
> reference to the log4j-perf benchmark. (Nice, Matt!)
> The duplication in this case is just "new Throwable()", so I think we're
> good.
>
Gary, looks like the code in ReflectionUtil already has a comment with a
reference to the log4j-perf benchmark. (Nice, Matt!)
The duplication in this case is just "new Throwable()", so I think we're
good.
On Thu, Nov 12, 2015 at 1:16 AM, Matt Sicker wrote:
> I know I used that in ReflectionUtil
I know I used that in ReflectionUtil because of the log4j-perf test that
demonstrates the difference between Throwable and Thread.
On 11 November 2015 at 10:11, Gary Gregory wrote:
> I suggest you add a comment there too or preferably refactor the
> duplication.
>
> Gary
> On Nov 11, 2015 8:05 A
I suggest you add a comment there too or preferably refactor the
duplication.
Gary
On Nov 11, 2015 8:05 AM, "Remko Popma" wrote:
> I agree I should have added a ref to the JIRA to document the reason for
> doing it this way. I added a comment just now.
>
> About moving this to a util class, I'd
I agree I should have added a ref to the JIRA to document the reason for
doing it this way. I added a comment just now.
About moving this to a util class, I'd be fine with that.
FYI, the other place that uses {{new Throwable().getStackTrace()}} is
ReflectionUtil#getEquivalentStackTraceElement.
O
I think this needs a code comment to avoid the code being changed in the
future to undo the improvement.
IMO all perf changes like this need good docs with a ref to the Jira. Who
knows how this kind of stuff will play out on top on Java 8 and 9.
If we do this kind of call in different places, we
