On 5/3/23 19:57, Kenneth Knowles wrote:
My big picture hot take: this is useful, but the problem we really
need to solve is topology change, which will obsolete coder evolution.
I think Beam model has a role in this. It isn't just a runner-specific
thing. We need to ensure the model makes it p
My big picture hot take: this is useful, but the problem we really need to
solve is topology change, which will obsolete coder evolution.
I think Beam model has a role in this. It isn't just a runner-specific
thing. We need to ensure the model makes it possible/easy to replace one
pipeline with an
I think I'm not understanding the use case here? Are we talking about
encoding of data in motion (e.g. between stages of a streaming pipeline) or
data at rest? (e.g. input formats and output formats) Or maybe something
else?
On Wed, May 3, 2023 at 6:58 AM Jan Lukavský wrote:
> Hi,
>
> I'd like t
