ZENOTME commented on issue #596:
URL: https://github.com/apache/iceberg-rust/issues/596#issuecomment-2371283315
> Sorry, I don't quite get the point, if updates are sent to rest catalog
server, why we need to update it in local first?
E.g. the user wants to batch multiple updates and
liurenjie1024 commented on issue #596:
URL: https://github.com/apache/iceberg-rust/issues/596#issuecomment-2370693889
> Why for rest catalog, it should be sent to rest catalog server.🤔
> According to API from pyiceberg, it seems possible to [create a
transaction without auto commit
](htt
liurenjie1024 commented on issue #596:
URL: https://github.com/apache/iceberg-rust/issues/596#issuecomment-2336534759
We have
[check](https://github.com/apache/iceberg-rust/blob/a1ec0fa98113fc02b2479d81759fccd9dab10378/crates/iceberg/src/transaction.rs#L50)
to avoid such duplicated case. Fo
liurenjie1024 commented on issue #596:
URL: https://github.com/apache/iceberg-rust/issues/596#issuecomment-2328031227
Hi, @ZENOTME Could you elaborate on this? I'm kind of confusing about the
proposal.
--
This is an automated message from the Apache Git Service.
To respond to the message,
ZENOTME commented on issue #596:
URL: https://github.com/apache/iceberg-rust/issues/596#issuecomment-2323395450
> Exactly. Internally you want to stack the changes together. For example,
within a single transaction, you add a new field and then write the data, then
the latest schema should
ZENOTME opened a new issue, #596:
URL: https://github.com/apache/iceberg-rust/issues/596
This will work for now but might get problematic later on. Just a heads up.
An important concept for Iceberg is to stack snapshots in a single commit.
For example, now with append being added in t
