Duplicate thanks for signing the CLA

[wilkinsona]

spring-projects/spring-boot#7239

The users was thanked three times, and, barring a coincidence in timing, they appear to have been triggered by me interacting with the issue rather than the user signing the CLA.

I see two problems:

1. Has the CLA actually been signed, or did my interaction trigger the thanks comment incorrectly?
2. The comment was posted three times

[snicoll]

See also #140

[wilkinsona]

Three payloads were delivered to the web hook for spring-projects/spring-boot#7239 at 2016-10-28 09:51:02. One unlabelled action and two labelled actions. I can't see why any of these alone should have triggered a comment that the CLA had been signed.

[mp911de]

Reading data from Github is eventually consistent... When an object was updated and that data is immediately read, then consistency issues may occur.

@rwinch we should reconsider "thanking for signing the CLA". In this case, the user signed the CLA without updating the PR (which is likely to happen if CLA was signed just jumping to cla.pivotal.io and not using the link in the PR). We had several scenarios where short-time persistence of "active" PR's would be beneficial. In this case, we would compensate for a slow Github.