Adam Pallozzi
Adam Pallozzi
@existentialmutt that looks good to me if you were able to get around the `send_keys` issue. Without that, it would have been really difficult!
Thanks, that makes sense re the difference between TTL and Timeout. Could you explain how TTL relates to the time a job takes to complete? If I set a TTL...
Thanks, I'll take a look at v7 👍
@gazayas yep that's a tricky one that we haven't been able to fix in ActiveHash yet. However, it's mainly an issue with the test suite. I've been running that version...
@dansingerman @andrewculver would love any feedback from either of you on this change. In my local testing, I was saving up to 100ms per request so there's definite performance improvements...
> I think this is a good change. I like having the `cache` option, as I can imagine there may be cases where it's harder to know when to invalidate...
Makes sense. Unless @andrewculver has any other thoughts at this stage, I'll keep this PR moving forward.
I like that. However, won't we still need to pass the cache key all the way through so the `parent_ids_for` method knows if it should check the cache?
Yeah I think given that we're making the cache happen "under the hood" as long as when they set `cache: false` the result is calculated live each time it's ok...
@dansingerman after merging #21, using `cache: true` as the key here feels confusing. Now that we've introduced caching at two different levels, I think we need to be explicit about...