Sebastian Lövdahl

Anyone managed to work around this?

This seems to be caused by the `_updateHeaders` method not being called before `_resize` is invoked. I worked around it locally by inserting a second call to `_resize` in `refresh`...

I have used double-sided tape between the housing and the meter, and duct tape on it because the double-sided tape was not strong enough to keep it in place 😁...

@klaasnicolaas Yeah, that makes sense. 👍 I'll try to restructure it a little bit.

Done, what do you think? 🙂

Just posting this here for reference. The issue I described in https://github.com/ben-manes/caffeine/issues/323#issuecomment-924620934 and https://github.com/ben-manes/caffeine/issues/323#issuecomment-924664002 would most likely not have occurred at all with bulk refresh, because we implement `loadAll` for...

FWIW, the same happens if there would be cyclical dependencies, the method logs and returns silently.

I now tried to override `AbstractPluginManager.loadPlugins()` and unload all plugins that have missing dependencies, but that did not work out of the box, because `AbstractPluginManager.stopPlugin(String pluginId, boolean stopDependents)` isn't able...

> Should unloadPlugin check the plugin state before calling stopPlugin? I opened #460 with a suggested fix for this.

This turned into a rabbit hole.. :grin: Should `AbstractPluginManager.loadPlugin(Path pluginPath)` also unload the plugin under the hood if `resolvePlugins()` fails? ```java @Override public String loadPlugin(Path pluginPath) { if ((pluginPath ==...