C0urante commented on PR #10566:
URL: https://github.com/apache/kafka/pull/10566#issuecomment-1478578470
Opened https://github.com/apache/kafka/pull/13433 to address this (and
cyclic schema support)
--
This is an automated message from the Apache Git Service.
To respond to the message, pl
C0urante commented on PR #10566:
URL: https://github.com/apache/kafka/pull/10566#issuecomment-1454040953
Ah thanks @urbandan, that helps. It's worth noting that the latest example
test case is probably missing calls to actually set the default value for the
schema builder instances? (Things
C0urante commented on PR #10566:
URL: https://github.com/apache/kafka/pull/10566#issuecomment-1452296446
@urbandan I've given the "tweak" fix (i.e., altering `ConnectSchema::equals`
to use interface methods and accept any `Schema` instance during equality
checking) a shot locally and, altho
C0urante commented on PR #10566:
URL: https://github.com/apache/kafka/pull/10566#issuecomment-1446878483
Hi @urbandan, thanks for clarifying the point about using `Schema` methods.
Agreed that this would be helpful.
I think we should try to keep validation of optionality if possible,
C0urante commented on PR #10566:
URL: https://github.com/apache/kafka/pull/10566#issuecomment-1432056618
@urbandan I don't think I follow what you mean by "Use the Schema methods";
it might be easier to illustrate with a code snippet?
I'm not really in favor of removing the validation
C0urante commented on pull request #10566:
URL: https://github.com/apache/kafka/pull/10566#issuecomment-914640232
Hey Gunnar! Sorry for the delay.
> Initializing a Struct with a builder rather than a Schema seems to be an
odd thing to do, no?
Indeed it is rather odd, but it's
C0urante commented on pull request #10566:
URL: https://github.com/apache/kafka/pull/10566#issuecomment-824292010
Hmm... I'm wondering if this might break existing setups. Since the
`SchemaBuilder` class does implement the `Schema` interface, it's currently
possible to do something like th