CurtHagenlocher commented on PR #73: URL: https://github.com/apache/arrow-dotnet/pull/73#issuecomment-3324876691
> From a performance point of view renting small buffers is slower than allocating a new array on the heap. https://adamsitnik.com/Array-Pool/ We already use the shared array pool for this purpose, including the reverse of this scenario where we're trying to read a serialized int32 value from a `Stream` (see line 208 of ArrowStreamReaderImplementation.cs). The only thing that distinguishes this case in the current code from those others is that for some reason we're creating a new pool instead of reusing the shared one. I'm inclined to stick with the patterns already in the code rather than creating a new one -- unless we have specific evidence which suggests sufficient benefit for doing it differently. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
