Alexey Mayshev

> JCache is an alternative standardized Java api To be honest, it looks overloaded. Actually, I quite like the alternative with options, but I'm still not completely sure about this...

Oh, I can hardly imagine cases in which the ttl is not known in advance. It seems that in such cases, network requests are not needed and such a function...

Okay, it seems that such an api should be good enough. So I can start developing.)

Haha, it really works. I tried to allocate on the stack and pass a pointer, but it leaked onto the heap and I didn't try to fix the escape analysis...

# Thoughts about Otter v2 **Disclaimer**: I'm sorry for such a long message. I have noted questionable (relatively) points with ⚠️. Comments will be very helpful. ## Basic concepts and...

Unfortunately, I didn't really understand how this will help you determine what data needs to be saved in the disk cache.

It looks relatively useful, but it seems that you don't need exactly hit count. Checking for one-hit wonder is more than enough. I think I'll try to add this after...

> can run concurrently, when Delete run into this [line](https://github.com/maypok86/otter/blob/main/internal/hashtable/map.go#L379) and Set run into this [loop](https://github.com/maypok86/otter/blob/main/internal/hashtable/map.go#L352), there is chance that Set node to an invalid(shrinked) table. But this one has...

> So there is a chance that delete creates a brand new shrinked table for hashmap, and set will add or update the node in the old table. If these...

> So if m.resizeInProgress() and if m.newerTableExists(t) are soft limit in set, right? What do you mean by "soft limit"? > how about remove the code below from set, does...