Closed Luisav1 closed 1 year ago
On-hold because of CT issue #129.
https://github.com/phetsims/build-a-nucleus/issues/129 is closed. let's do it!
As part of the code review in https://github.com/phetsims/build-a-nucleus/issues/165 I ran a memory test. There is a screenshot of the results below. The bottom line is that there may be a slow leak, but nothing egregious, and what we do see could be explainable by pooling in Scenery.
Everything was done in a Chrome incognito window using a freshly built version of build-a-nucleus_all_phet_debug.html
.
Explanation:
fuzz
parameter added and one minute of fuzzingphet.chipper.queryParameter.fuzz = false
, reset both screens, went to the home screenWhen I create 10 of each nucleon and then reset, the comparison memory tool doesn't show any obvious memory leaks. Yay!
We found one problem with how we were recreating a Text, PatternStringProperty, AND a DerivedProperty each time the mass number changed. Oops. Fixed here.
After fuzzing for 45 minutes and capturing snapshots intermittently, I don't see problems anymore. Woo! Closing. Thanks @Luisav1 for all your help.
Arg! Found another nested listener adding:
?listenerLimit=1000
has been helpful for this nesting issue
After fuzzing for a while I saw that the maxListenerCount was 783, then when I restarted the sim with ?listenerLimit=782, it failed on startup. No further work to be done here.
From self code review https://github.com/phetsims/build-a-nucleus/issues/112.