I actually started building ROC before redis-objects was released. But I stuck with using and developing ROC for a couple of reasons:
1. TransientStore is really useful for testing and for re-using redis code in scenarios that don't require persistence
2. As well as implementing the Redis commands, the ROC objects also perfectly mimic the APIs of their Ruby-core equivalents (except for a few destructive methods that are not possible to mimic).
In addition, there's now a third reason:
3. I recently added eval/lua support, which I don't think redis-objects has yet.
1. TransientStore is really useful for testing and for re-using redis code in scenarios that don't require persistence
2. As well as implementing the Redis commands, the ROC objects also perfectly mimic the APIs of their Ruby-core equivalents (except for a few destructive methods that are not possible to mimic).
In addition, there's now a third reason:
3. I recently added eval/lua support, which I don't think redis-objects has yet.