This has happened now on the Wikidata mailing list, and is continuing to happen regularly on the accounts-enwiki-l mailing list. This is starting to get a bit problematic. I don't think this is "call in someone from their holidays" level problematic, but I think it would be good (and I would certainly appreciate) to please get some additional prioritization of this issue.

Looks like this has just happened on Wikimedia-l. Here's a link to the archived message: https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/DL2A5HHYG5RQPWNH26IGFGSCWFIUW25M/. Unsure if the message was held for moderation since I'm just a regular subscriber to the list.

I have been increasingly having to deal with these as an admin for the accounts-enwiki-l mailing list, which is problematic because it means I have to dig into my folder of moderation notifications, find the message that was lost, and manually forward it to the mailing list. For accounts-enwiki-l specifically, this started on September 27th, then next happened on October 26th, and has happened a couple times per week since then.

@stwalkerster Patch looks good to me.

Bumping this task.

@Addshore I'm okay with giving the go-ahead for both site tests and the expected failure tests if you have no issues with them.

One last question, I promise (maybe) :P

What exactly do these load tests and write tests entail? What's being tested?

I definitely see both sides of the argument here. I agree that, in most cases, maintaining two versions of key parts of a single piece of software is quite costly. However, as a small wiki operator myself, I feel that at this point, restricting Flow to Parsoid-only would only hurt wide adoption, because at this point, Parsoid is still rather experimental (definitely at least "Beta" software).