Conversation
Signed-off-by: David Radley <david_radley@uk.ibm.com>
Signed-off-by: david_radley@uk.ibm.com <david_radley@uk.ibm.com>
|
@rmoff fyi |
rmoff
left a comment
rmoff
left a comment
There was a problem hiding this comment.
Thanks for this @davidradl!
I'd suggest reordering the sections into perhaps a dev -> user -> governance flow. At the moment it feels like it jumps around a bit.
Signed-off-by: david_radley@uk.ibm.com <david_radley@uk.ibm.com>
|
|
||
| Previous Blog: [Flink community update February 2026](https://flink.apache.org/2026/02/01/flink-community-update-february26/) | ||
|
|
||
| A strong February; there was a lot of commit activity with 320 open PRs and 4 new releases. |
There was a problem hiding this comment.
do we have 320 PRs opened within one months?
is it a number for just Flink repo, I guess releases are counted here not only for the Flink main repo
There was a problem hiding this comment.
total in the core repo. I will clarify that
There was a problem hiding this comment.
total in the core repo.
then we need to have the same approach, like why commits are mentioned only for main repo and releases across all projects ?
There was a problem hiding this comment.
I understand your concern - this is inconsistent:
1 I can say something like "A strong February; there was a lot of commit activity with 320 open PRs in the main repository (600+ across all Flink projects) and 4 new releases."
2 I realise we are only really interested in active repositories, i.e. ones where something has happened in February. I can give those numbers - but then I will be asked to evidence base them. I could share a shell script in the repo.
3 the total commits - is useful really for the repos with stale bot on. Otherwise the total commits could be very old.
I am tempted for 1 to keep it simple, but am open to suggestions. WDYT?
There was a problem hiding this comment.
my point is that if we are talking about what's changed for the month, then we shouldn't operate things like total, rather how many PRs were submitted and how many were merged within this month
| ### Community Announcements | ||
|
|
||
| No new committers or PMC members this month. | ||
|
|
There was a problem hiding this comment.
do we need to put anything if nothing changed?
There was a problem hiding this comment.
I was following the style of the previous blogs - will remove
| ## Apache Flink Dev List activity | ||
|
|
||
| You can view the dev list archives [online](https://lists.apache.org/list.html?dev@flink.apache.org) or [subscribe](https://flink.apache.org/community.html#mailing-lists) to receive emails. | ||
|
|
There was a problem hiding this comment.
is there a reason to keep activity in other repos separate?
I'm still not convinced we need to put all the commits, may be rather having kind of summary with some highlights
davidradl
force-pushed
the
communityBlogMar26
branch
from
ac4f14a to
268764a
Compare
Signed-off-by: david_radley@uk.ibm.com <david_radley@uk.ibm.com>
davidradl
force-pushed
the
communityBlogMar26
branch
from
268764a to
5ed4601
Compare
| - [[hotfix] Log watermark alignment duration (and all other stages)](https://github.com/apache/flink/commit/f6fffd7d33f617acc69f405deecf050ef710d439) | ||
| - [[hotfix] Log unexpected non-terminal task state](https://github.com/apache/flink/commit/79e79fcd26eeaabb956198ce2c6293cc221a9f5e) |
There was a problem hiding this comment.
do we really need all these hotfix commits description here?
| * Flink | ||
| * All CI builds were failing; this was fixed in a [PR](https://github.com/apache/flink/pull/27619) that upgraded Kubernetes client libraries. | ||
| * Connector parent | ||
| * Added [license checking](https://github.com/apache/flink-connector-shared-utils/commit/6e1af169012db40306a56ba8bcf31bb8a309e6e6) to Parent Connector. |
There was a problem hiding this comment.
I tend to think nobody is interested in license checking except for people who are directly involved in development/release process wdyt?
| 5. Table planner enhancements | ||
| - [[FLINK-37924] Introduce Built-in Function to Access field or element in the Variant](https://issues.apache.org/jira/browse/FLINK-37924) | ||
| - [[FLINK-35661] Fix MiniBatchGroupAggFunction silently dropping records](https://issues.apache.org/jira/browse/FLINK-35661) | ||
| - [[FLINK-38913] ArrayIndexOutOfBoundsException while unparsing ExtendedSqlRowTypeNameSpec](https://issues.apache.org/jira/browse/FLINK-38913) |
There was a problem hiding this comment.
to be more precise this is not a planner change, rather parser
| - [[hotfix] Try to get last checkpoint on recovery regardless of checkpointing interval](https://github.com/apache/flink/commit/c0ada39aafa23e58e6c80883bdb740a89d02c1ff) | ||
| - [[hotfix] Move checkpointing configuration code to CheckpointCoordinatorConfiguration](https://github.com/apache/flink/commit/b2b044dc66814498f0cdfb4e9249f446e2c6fce9) | ||
| - [[FLINK-38939] Pause Sources until the first checkpoint barrier is received](https://issues.apache.org/jira/browse/FLINK-38939) | ||
| 4. Enhancements around splits |
There was a problem hiding this comment.
this is very vague definition.
For instance a question that might pop up after first reading: we have code-splitter module, does it have anything in common?
3 participants
No description provided.