Alberto Islas

Sure, every 1000 items seems fine that's the number we're currently using on `cl_index_parent_and_child_docs`. I'll start working on this then.

Some ideas of improvements: - It would be nice to display the `date_created` in the `ProcessingQueueAdmin` and `PacerFetchQueueAdmin` list_display, making it easier to find the desired instance when filtering and...

Yeah the root issue here is that in Solr search, indexing was based on `RECAPDocuments` instead of `Dockets` and `RECAPDocuments`, as we do now in ES. Previously, in Solr, it...

Yes, definitely. It makes sense to do this now to avoid adding additional content that is not from RECAP. I see this as part of the [sweep indexer ](https://github.com/freelawproject/courtlistener/pull/3862) so...

Got it. Yes, the second step of removing all the cases that don't belong to RECAP makes sense for creating a separate script. The first step of avoiding adding content...

Sure. The sweep indexer is ready with the tweak to avoid indexing Dockets that don't belong to RECAP, and I've opened an issue to remind us to apply the same...

Maybe something like: `f"recap.pdf.processing:{rd_pk}"` makes sense?

Got it. yeah, using the SHA in the key will allow us to process a following file if it's different, while avoiding processing the same file. Sounds good!

Upgraded django-ses to the latest version 4.1.1, tested email sending using a SES sandbox AWS account, as well as SNS webhook notifications for bounces and complaints, and is working properly....

> This will always be limited to a few different opinions, so I'd say that both the API and the front end should always show all of them. If you...