Re: [I] Write DataFusion paper for (SIGMOD / VLDB / ICDE) [arrow-datafusion]

["ozankabak (via GitHub)" <gi...@apache.org>]

ozankabak commented on issue #6782:
URL: https://github.com/apache/arrow-datafusion/issues/6782#issuecomment-1748871435

   I worked on the paper and made edits in Sections 1, 2 and 3. I made a lot of improvements so I will not list all of them individually here, but I'd say the single most important change is the stronger emphasis on composability and modularity in the last sentence of the abstract, and the reference to Voltron's composable codex.
   
   The only place I think we can do better in the parts I've covered so far is the following sentence from the introduction:
   
   > Section 6 partially enumerates the query processing optimizations built into DataFusion, illustrating the level of effort required, and why creating a new engine is a substantial undertaking.
   
   I think the emphasis should not be establishing/illustrating the fact that creating a new engine is hard (this should be common knowledge, or at least shouldn't be surprising), but the fact that we can mitigate/obviate the need to do so by being modular, extensible and composable (this is rather novel and our main contribution as the community). I couldn't find a succinct way to express this while staying in tune with the existing context just yet (getting tired), but shouldn't be hard.
   
   PTAL and I will continue with Section 4 if all looks good to you all.


-- 
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