Merge back upstream to LevelDB? about hyperleveldb

[rescrv]

We've replied to your question on the LevelDB mailing list.

We chose to fork because we need the ability to push fixes to HyperDex users without being dependent upon the LevelDB developers to cut a new release. They're great people, but also very busy, and have no relationship with the HyperDex project. It'd be quite rude to impose a release schedule onto them for our users.

For the same reason, we've not pushed any performance-related changes upstream.

If you check some of their bug reports, we have sent patches and other fixes directly upstream. HyperLevelDB is a friendly fork of LevelDB.

[jtolio]

It's not entirely clear to me why the release schedule affects pushing performance-related changes upstream, but thanks for your reply all the same. :)

[rescrv]

Our release schedule requires that we maintain a fork of some kind, even if it were in name only.

We haven't pushed the performance changes upstream because the upstream devs and the hyperleveldb devs are all busy people, and there would be a lot of wasted effort to push upstream for the sake of pushing upstream.