Heh, eleventh-hour change. When this version first showed its face, a few months ago, the UA string was
LinkedInBot/1.0 (compatible; Mozilla/5.0; Jakarta Commons-HttpClient/3.1 +http://www.linkedin.com)/1.0 (LinkedInBot; https://www.linkedin.com/; wkrupa@linkedin.com)
This month they changed to omni-crawler. Possibly W. Krupa was getting too much irate mail.
I used to deny "Jakarta Commons" comprehensively, but it's now subsumed under a generic header-based lockout. (Haven't bothered to check, but LinkedInBot has got straight 403s to date so they must be missing
something.) I've yet to see if they obey robots.txt; I only denied them after the most recent visit, last week I think. They've never happened to request anything in a roboted-out directory, so an unconditional Disallow is the only test. We Shall See.
UA strings that contain the same name more than once annoy me because a quickie global search without fancy RegEx turns up twice as many hits as there should be. Yes, yes, robot, I heard you the first time.
