Review request: JMC-6127: Live Objects Page do not forward selections correctly

Henrik Dafgård hdafgard at
Tue Nov 27 15:47:09 UTC 2018


The leaves in the aggregated reference trees are not necessarily leaks,
they're just known to be alive when the recording is dumped and the
documentation should reflect that. I think referring to the leaf nodes as
closer to OldObjectSamples or something similar would be better since it
differentiates them more clearly from the leak candidates.

There also seems to be a change in MemoryLeakPageUI constructor where the
initial state of the table is set to display model.getLeakObjects() instead
of model.getRootObjects(), which should probably be configurable. But there
is also the case where selecting a time range will show
model.getLeakObjects(IRange) and then unselecting a time range will revert
to model.getRootObjects() even in this change.

* A helper method to calculate number of leaked object with in specified time

Within is one word, and since this is a sentence should be followed with a

Other than that it looks good.

// Henrik Dafgård

On Tue, 27 Nov 2018 at 15:00, Marcus Hirt <marcus.hirt at> wrote:

> Hi Guru,
> Looks good, but documentation could be a bit improved e.g. on the
> ReferenceTree related classes. It should not say "Object which is leaked"
> rather "referenced object" or something similar. Also, it should
> preferably
> explain what the method does, and not only list the parameter. For example:
>           * @param leafNode
>           *            Object which is leaked
>           */
> Kind regards,
> Marcus
> From: Guru <guru.hb at>
> Date: Tuesday, 27 November 2018 at 07:59
> To: <jmc-dev at>, Marcus Hirt <marcus.hirt at>
> Subject: Review request: JMC-6127: Live Objects Page do not forward
> selections correctly
> Hi All,
> Please review the fix for
> JBS :
> Webrev :
> Root cause and its solution is updated in JBS.
> Thanks
> Guru

More information about the jmc-dev mailing list