Hi there,
On Sun, Mar 15, 2015 at 11:33 PM, MZMcBride z@mzmcbride.com wrote:
Keegan Peterzell wrote:
Single-user login[1] finalization will be taking place next month.[2]
[...]
All accounts that will be affected by this will be contacted on their talk page within the next couple of days.[4]
Hi.
Do you think a month of warning to affected users is sufficient? Ideally we want as many users as possible to self-rename before being forcibly renamed, I think. And if a lot of users take us up on the opportunity to self-rename (i.e., request a user rename, we still don't have true self-renames...), we want to make sure there aren't big backlogs. Sending out messages to 2.8 million accounts seems like it could create a flood.
MZMcBride
Great question, and it's one I've wrested with personally and in meetings.
Ultimately the timeline that is chosen is arbitrary. We can look at the data of the accounts like the number of edits v. age, potentially how much time does the average person with a conflicting account login based on user_touched, and several other metrics, but in the end there is no science to come up with a predictive timeframe. Another arbitrary number is how many of these accounts will show up to request a rename, no matter the timeframe. I can set this to take place six months from now but that's not predictive of behavior. As you can imagine, there's many more scenarios :)
So, given that, in my opinion the biggest tell is going to be what happens in the first two weeks. There are two potential bottlenecks: technical and in personnel. If the job queue or any other sort of technical function is regularly getting stuck, we'll revisit the month time frame. If the technical side is holding up but there are too many rename requests to fulfill before the deadline, we'll revisit the deadline. I'm willing to be flexible depending on the circumstance.
The bottom line is that this is something that we're only doing once, and it's something that I want to do right. I'd be willing to postpone if major technical and/or personnel blockers show up. However, I'm committed to the month deadline as much as I can be - at some point this has to be done, we clearly can come up with blockers for over a decade. 2.8 million accounts could create a flood, or a trickle. We're going to see, and very soon :)