Anastasia Braginsky

Hi @fujian-zfj , As we already discussed in issue #192 keys are not released for now in OakMap. Specifically SeqExpandMemoryManager does not support deletion of the key, unless the key...

> I don't quite understand why ctx.key.getSlice() is not released when OakMap.remove() is called, but only ctx.value.getSlice() is released. By the way, keys in OakMap are used for navigation and...

Hi @fujian-zfj , Indeed we are still working on the issue of releasing the memory, because we want to make it correct and without performance implications. If you want you...

Hi @dynaxis , Thanks for your comment. Here is the comment/documentation about `OakUnsafeDirectBuffer`'s `getByteBuffer()` ``` public interface OakUnsafeDirectBuffer { /** * Return the underlying memory address wrapped as a ByteBuffer....

Hi @dynaxis , Thank you for your reply. Your suggestion makes sense. Although I would stress that this buffers brings you no protection from possible concurrent update of the same...

Thank you very much @dynaxis , do as you think is the best way! Very much appreciated!

@dynaxis , thank you so much! Review done. Code looks good, just one comment left!

Hi @fujian-zfj , Thanks for your interest in Oak and sorry for late reply. The issue that you see is because currently keys in the OakMap are not released and...

Hi @liuxinpeister , With how many threads do you run? Is the failure re-creatable? Generally, there is assert for infinite loop there, but most likely you are not running in...

Indeed those optional APIs are not yet implemented. They are on the TODO list. Meanwhile instead of `Iterator.remove()` it is possible to use an iterator and invoke `remove()` on relevant...