jackye1995 commented on pull request #3326: URL: https://github.com/apache/iceberg/pull/3326#issuecomment-961308952
Sorry I am a bit late to the whole conversation. @rymurr @rdblue I remember we went through the exact same lines of thoughts when we introduced Glue and Nessie catalogs, and there was an attempt to do exactly this, but ended up not merged because: 1. difference between Flink and Spark, with the confusing term of `type`. Even with a unification, Trino itself has it's own `CatalogType` https://github.com/trinodb/trino/blob/master/plugin/trino-iceberg/src/main/java/io/trino/plugin/iceberg/CatalogType.java (PrestoDB as well) and we will never unify everything 2. by enabling catalog type, we are forcing everyone to register the catalog as a type which is just more work without too much benefit compared to using `impl` for all new catalogs, and only keep `type` for backwards compatibility. There will always be private catalog implementations that you cannot express as an open source type and has to be loaded using `impl` anyway. Are we trying to revisit the decision made back then? if so, what are some new reasons to add types now? -- 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]
