GH-2930: Add issue templates #2931
Conversation
| attributes: | ||
| label: Component(s) | ||
| multiple: true | ||
| options: |
There was a problem hiding this comment.
Do we actually need these options? Or should it be consistent with modules like parquet-avro, parquet-cli, etc.? Same question for other yamls.
There was a problem hiding this comment.
Also please note I introduced as many new labels as I could during migration. We now might want to reduce the current set by deleting them here (delete is irreversible). Same goes for milestones.
There was a problem hiding this comment.
- I would not add the modules as is. It does not make sense to differentiate e.g.
parquet-common,parquet-column, orparquet-hadoop(etc.). These are all parts of the "core". - It might make sense to mention the "bindings" separately (
parquet-arrow,parquet-avro,parquet-pig,parquet-protobuf,parquet-scala,parquet-thrift). - If we want to list components, Java is too wide.
- I think it makes more sense to somehow sum up Release+CI+Packaging.
- What is Documentation? Adding javadocs? I would add these changes to the related module instead.
- I would add parquet-cli as a separate item
There was a problem hiding this comment.
Thanks for the feedback @gszadovszky. I've change the list like so:
options:
- - Java
- Release
- - Documentation
- Continuous Integration
- - Packaging
+ - Arrow
+ - Avro
+ - Pig
+ - Protobuf
+ - Scala
+ - Thrift
+ - CLI
- Benchmarking
- OtherI think it makes more sense to somehow sum up Release+CI+Packaging.
Agreed, but I don't have a good suggestion how. For now I just removed packaging.
There was a problem hiding this comment.
Thank you, @rok!
It would be nice to see somehow how many issues we have related to CI and Release. Maybe Other is enough but not sure. I am fine as is, though.
One thing is clearly missing. This is the core functionality that most PRs have updates in. These are the modules parquet-column, parquet-common, parquet-encoding, parquet-format-structures, parquet-generator, parquet-hadoop. I am fine calling it Core if there are no better suggestion.
parquet-hadoop is definitely part of core currently, even though we want to get rid of the hadoop dependencies. After this work would be done, the separate component Hadoop would make sense, but currently I don't think we need it.
There was a problem hiding this comment.
Thanks @rok for working on this 🙌 I would be inclined to remove Other and make it optional. WDYT?
There was a problem hiding this comment.
Thanks for the feedback! :D I've pushed a change and here's the current list:
- Core
- Release
- Continuous Integration
- Arrow
- Avro
- Pig
- Protobuf
- Scala
- Thrift
- CLI
- BenchmarkingI'm glad to defer to your suggestions as I'm quite unfamiliar with parquet-java (except for the migration).
There was a problem hiding this comment.
One more thing, I was looking at Iceberg regarding @gszadovszky's comment.
I think it makes more sense to somehow sum up Release+CI+Packaging.
At Iceberg we've packaged these into a Build label.
There was a problem hiding this comment.
Seems like a good idea. Will do.
Co-authored-by: Gang Wu <[email protected]>
|
I will merge this after next 24 hours if no objection. |
This introduces Arrow-like issue templates to structure and categorize issues somewhat as they are reported.
See #2930