I submitted this PR years ago using ConcurrentWeakKeyHashMap from Doug Lea and the Netty team but Gary had concerns from legal about being able to use it.
PR is here: https://github.com/apache/commons-beanutils/pull/56 On Sat, Sep 14, 2024 at 7:22 AM Xeno Amess <xenoam...@gmail.com> wrote: > I thought old d-lea bro have a concurrent weak hashmap implement > somewhere...why not just ask him > after all nobody here more expert than him in this area > ________________________________ > From: Niall Pemberton <niall.pember...@gmail.com> > Sent: Saturday, September 14, 2024 1:46:32 PM > To: Commons Developers List <dev@commons.apache.org> > Subject: Re: [beanutils] For 2.0, WeakFastHashMap vs ConcurrentHashMap > > On Thu, 12 Sep 2024 at 19:59, Gary D. Gregory <ggreg...@apache.org> wrote: > > > Hi All, > > > > For the upcoming 2.0.0-M1, I plan on replacing the custom WeakFastHashMap > > with the stock ConcurrentHashMap. > > > > If you think this is a bad idea, please tell us why. > > > It’s a good idea for the “fast” part, but the “weak” aspect also needs to > be retained - so in principle yes, but the implementation detail matters. > > BeanUtils caused memory issues in multiple ClassLoader environments (e.g. > WebApp containers) because of strong references to classes prevented them > from being garbage collected if the app was reloaded/restarted. So this > weak map was introduced to resolve that issue. > > Niall > > > > > > Gary > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > > For additional commands, e-mail: dev-h...@commons.apache.org > > > > > -- ============================== Melloware melloware...@gmail.com http://melloware.com ==============================
