rdblue commented on code in PR #16425: URL: https://github.com/apache/iceberg/pull/16425#discussion_r4648382753
########## format/spec.md: ########## @@ -530,6 +531,8 @@ When an existing row is moved to a different data file for any reason, writers s Engines may model operations as deleting/inserting rows or as modifications to rows that preserve row ids. +When column files are added to a data file, `_last_updated_sequence_number` is handled as described in [Writing Column Files](#writing-column-files). Review Comment: The referenced section is huge so this statement makes it much harder to understand the requirements that are covered by _this_ section. The reader has to go read a huge section and search for relevant content. I think the content should be inlined so that this section can be easily read and understood. It's okay to add a reference for more information, but the core ideas should be stated here. Isn't the important part that when a column file is added, it will typically contain `_last_updated_sequence_number` because the rows have changed? (I don't think that's a hard requirement because you could add multiple column files in one commit and only one would have `_last_updated_sequence_number`.) Also, do we allow `_row_id` to be present in column files? It should not change. ```suggestion Note that when updating the rows of a data file with column files, `_last_updated_sequence_number` must also be updated to reflect the changes by writing `null` for any row that is modified by the column file. ``` -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
