Thanks for updating the KIP.
I think you can start a vote.
-Matthias
On 4/13/25 10:54 PM, Herman K. Jakobsen wrote:
Thank you for your inputs!
It looks like my previous answer messed up the mailing thread due to the
use of the hashtag symbol when creating the following list. I've tried to
re
Thank you for your inputs!
It looks like my previous answer messed up the mailing thread due to the
use of the hashtag symbol when creating the following list. I've tried to
reformat the list:
(1) I agree and I have changed the `startTime` parameter to be of type
`Instant`.
(2) I have added a se
Thank you for your inputs!
#1, I agree and I have changed the `startTime` parameter to be of type
`Instant`.
#2, I have added a section for the unintuitive cases and semantics that you and
Sophie have mentioned. In short, I’m proposing to skip forward to the next
trigger time.
#3, It was a ty
Thanks for the KIP!
I pretty much echo what Matthias has said so far regarding the API.
Regarding #4-5, assuming we would just be leaving out stream-time in the
initial implementation for time/scope reasons and might want to add this in
the future, I think it's best to just throw an exception if
Herman,
thanks for the KIP, and sorry for late response. Overall the KIP makes
sense to me, and the propose API change is neat and contained, so I
don't have any concerns about it.
Couple of questions/comments.
(1) I think the propose `startTime` parameter should not be a `long` but
in `Ins
