You are viewing a plain text version of this content. The canonical link for it is here.

Posted to github@arrow.apache.org by GitBox <gi...@apache.org> on 2022/06/16 21:27:57 UTC

[GitHub] [arrow] westonpace commented on pull request #13364: ARROW-16756: [C++] Introduce non-owning ArraySpan, ExecSpan data structures and refactor ScalarKernels to use them

westonpace commented on PR #13364:
URL: https://github.com/apache/arrow/pull/13364#issuecomment-1158146289

   > Who would be the canonical owner of a datatype?
   
   Not a detailed thought but could the canonical owner be `arrow::Schema`?  Outside of Acero, every record batch has a schema (might have to do some work so that collections of record batches, like `arrow::Table`, share a single schema instance).  Within Acero the canonical owner could be the ExecPlan (each node has an output schema and any intermediate types needed could be managed by the nodes).  That being said...
   
   > I think we should first run benchmarks after the ExecSpan changes have been propagated to the compute engine, to see if it's still worth doing something on the DataType front.
   
   +1


-- 
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: github-unsubscribe@arrow.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org