867140 - Notify inactive committers prior to disabling accounts

Status: RESOLVED INCOMPLETE

(Infrastructure & Operations :: Infrastructure: Other, task)

Description

[Ashish Vijayaram [:ashish]]

Committers who haven't had a commit in 6 months are considered inactive and their accounts disabled. They then have to file a bug to get their accounts re-enabled. It would be useful to notify them a few days in advance so that they could maybe make a no-op commit to avoid a longer process.

Comment 1

[Shyam Mani [:fox2mike]]

I like the idea, but I'm not sure if this means people will "dummy" commit to keep privileges alive. I don't like the latter. CC'ing Gerv for his input too.

Comment 2

[Gervase Markham [:gerv]]

Do we have data or stats on how many "please reactivate my account" requests we have received? Is this proving to be a burden for the IT team?

If someone hasn't used their account in 6 months, it's statistically unlikely that they were "just about to use it". I don't think we should notify before (although we might want to notify after, so people know what's happened and aren't surprised when and if they try and use it again, and can file the message with the procedure somewhere safe). 

As Shyam says, it'll lead to dummy commits. My online banking has a timeout window saying "I'm about to log you off". I keep saying "No", long after I've actually finished banking, "just in case". The same thing might happen here. And having to make the no-op commit will be effort that will cause resentment.

So I think we shouldn't do this.

Gerv

Comment 3

[Justin Dow [:jabba]]

So, wontfix, or e-mail notification *after* expiration?

Comment 4

[Shyam Mani [:fox2mike]]

Gerv,

Notify after has the same effect as right before. So I'm not sure that's a good idea....

Comment 5

[Gervase Markham [:gerv]]

(In reply to Shyam Mani [:fox2mike] from comment #4)
> Notify after has the same effect as right before. So I'm not sure that's a
> good idea....

I don't think it's the same, because it prevents the "make a bogus checkin to keep my account, just in case" behaviour, while having the benefits of people understanding what has happened and not being surprised later.

But anyway, does anyone have any data on how many reactivation requests we've seen?

Gerv

Comment 6

[Justin Dow [:jabba]]

Being resource constrained atm, I'm going to close this bug for now. The status quo is cleanly outlined in the policy on the wiki and notifying after still makes them file a bug, notifying before promotes bogus checkins, so there isn't much overall benefit to us investing a lot of time on this.