Varun Patil

One user reported this on his Droid Maxx with Adreno 320: [screenshot_2017-06-17-06-46-31](https://user-images.githubusercontent.com/10709794/27258073-bfa1267e-540d-11e7-9cbf-6110c1c83636.png) Seems totally new, but I wouldn't really bother much about the 3XX and 2XX series' any longer, since...

For the record, I can see no obvious delay with pathCache enabled now.

This can be fixed with https://github.com/radialapps/mkxp/commit/84226422b54174ac324a3ef53057335449cc8d29. For some reason, just changing the search path without reducing the size of `buffer` was producing weird errors on my device.

This is out now ... seems like a Unity remake

@Ancurio, quoted [here](http://www.androidauthority.com/to-the-moon-android-release-date-761875/) and [here](http://toucharcade.com/2017/04/06/to-the-moon-mobile-launching-in-may/) for a start.

@Ancurio, will do. Just pushed a commit https://github.com/xperia64/android-mkxp/commit/e0acfca42e69283e644f473ae1d2c3fcdc3552a6 to my fork that fixes this by killing the app altogether on destroy. I have put the method right at the top...

I would still prefer to kill it at the end either way, since a running process which is no longer need is, after all, a waste.

Actually none of the other nextcloud packages (at least the ones I used) depend on node polyfills, which allowed me to get rid of `NodePolyfillPlugin` altogether. But then I had...

Also noteworthy: `path` depends on `process` so this also adds completely redundant code to the output bundle. The impact may be negligible though.

Hmm, the docs of `path-browserify` claim webpack injects it by default. I'm a bit confused now. EDIT: I guess this was old documentation. We can still import from `path-browserify` directly...