My parents were raccoon sympathizers.
Back when I was young and tiny, and before we got the big living room TVs, my parents’ favorite evening entertainment was loading up a tin plate with scraps from dinner and feeding the raccoons. Just before sundown, my mother would place the plate on the cement patio by their bedroom, and they would watch from bed behind the sliding glass door, and wait for the families of raccoons to come.
Around 10pm, the raccoons would come, occasionally a family. There was Tripod, a three-legged mother, who brought her kids, and let them feast. There were nights with twenty or so, and then when the plate was done, they’d wrap up and look in the window to see if anything else was forthcoming. Eventually, there would be some other animals: opossums, including the notably loud one dubbed Henry Hissinger, and a skunk that came for several nights, sadly standing off at the edge of the lights, waiting her turn only to find an empty tray, until one night when she figured out that if she turned as if to spray the crowd and then walked backwards over the plate, she’d be able to eat everything for herself.
By the time I was five, and we had cats that wanted to be outdoors, the era of the raccoon food bank was over; to be fair, they had broadened their diet out to what was in the birdfeeders, and what was in the metal garbage cans (mostly birdseed). And by the time I was ten, I had an impression of raccoons as fantastically clever critters, who managed to work their way around every security system my father could place between them and food, short of the electronics of the garage door.
The reason I’m starting with that description today is I ended up listening to this podcast on the way home from Virginia:
https://www.pushkin.fm/podcasts/revisionist-history/rat-vs-raccoon
While the podcast goes down the path of why don’t we use raccoons for testing generally, I’m going down the path of “why didn’t they use raccoons for testing the particular problem they’re fighting here?” In the podcast they make the case that rats are kept as a test subject for psychological experiments because they take the test so well, and the raccoon is occupied with escape and other things so a raccoon’s not producing reproducible results.
Where I think the raccoons were wronged here is later in the podcast. The scientist who then tested the raccoons claims they tested very poorly and concludes the raccoon is merely destructive, and its destructive impulse is not sufficient demonstration of intelligence.
No. They may or may not be intelligent, but the conclusion doesn’t get them there. And as a tester myself, I think I see the problem.
Rats take the test. Their performance is a reflection only of how they do things.
Raccoons test the design of the experiment. And they don’t necessarily need to be smart to do so, because part of testing the experiment includes brute forcing the experiment. And their performance grades the test designer as much as themselves.
When one person tries to defeat their local raccoon, they’re coming up with their own solution. And that solution usually is not successful the first time, you wake up to a noise in the night, or find a knocked over can in the morning. But you the human adapt, and apply more solutions, putting more barriers in place. And the barriers you put in place are not the barriers your neighbor has put before this same raccoon. If the barrier is sufficient, the raccoon may consider another target.
Aside from it being a over-promise destined to be under-delivered, the decision to introduce a raccoon-proof garbage can was probably flawed from the point of standardizing the problem for the raccoon. As an equivalent, say corporate IT installed a secure login on everyone’s work computer, but with the same password on each machine. If that password is on a list of passwords that someone could try in a brute force attack, every computer can be broken into. Now consider that the people who designed the lock missed a flaw, but it’s something that only the 1% strongest raccoon could exploit by flipping the can over, or with the 1% most sensitive paws could discern where to push to slip the lock. Even at that limitation, every garbage can is vulnerable, and that vulnerability is across the board, and that 1% can open up the cans for their less fortunate brethren (I don’t mean this to imply intelligence, organization, or charity, just that an open and spilled garbage can typically contains more garbage than a single raccoon could eat in a night.)
From a testing standpoint, there’s two ways to approach a problem: regression testing or interactive testing. Regression is the repeated application of the same test to the same subject, observing how they perform over time. It’s assumed they’ll improve, or they’ll not get any worse. In my day job, that means if a regression test starts failing, something has been introduced to cause the error. Regression is a job for rats, they are the invariant application of input against a target. If there’s a change in the outcome, there’s a change in the target, and we have to address it. Interactive testing doesn’t constrain the input to a particular path or amount or value. It’s unpredictable what the results could be, because the input varies from tester to tester, and from attempt to attempt. Interactive testing is a raccoon’s job.
The part that annoys me about this is they have the pieces here. A mechanical lock is just the sort of attractive nuisance that would attract a raccoon’s attention, especially once they detect the presence of food beyond it. You know they use their paws as sense organs (the podcast mentioned why they ‘wash’ things is because their paws are more sensitive underwater,) so a puzzle like a lock is appealing to them. You’ve gained their attention and interest in two separate ways, they’ve been motivated, and this has ceased to be psychology experiment, now it’s a security penetration experiment. And as I’ve seen, the raccoon will not stop on the first try. No, they will take many attempts at the problem, until they achieve an outcome: either they solve it, are shoed away, or daylight comes out and they hear activity. Essentially they will interactively test that raccoon-proof garbage can hundreds of times over a short time.
This comparison put me in the mind of a quote from the pilot of Columbo, where he’s probing the killer to see where he’s overconfident.
https://www.imdb.com/title/tt0061496/quotes/
Lt. Columbo: Tell me, Doctor, how do you catch a man like that?
Dr. Ray Flemming: You don't.
Lt. Columbo: You're probably right. He sounds just too clever for us. What I mean is, you know, cops, we're not the brightest guys in the world. Of course, we got one thing going for us: we're professionals. I mean, you take our friend here, the murderer. He's very smart, but he's an amateur. I mean, he's got just one time to learn. Just one. And with us, well, with us, it's - it's a business. You see, we do this a hundred times a year. I'll tell ya, Doc. That's a lot of practice.
By the estimate in the story, Toronto had 10,000 raccoons (an estimate that seems way light), and they standardized the garbage can that those 10,000 raccoons could try to access. And in doing so the designers had one time to learn, and put their raccoon-proof garbage can into production. And they were very certain of their design, and very smart, but they were up raccoons for whom breaking into a food source was a business, and against whom they’ve had a lot of practice.
Columbo may have been three raccoons in a raincoat, it’s not that hard to imagine.
[I tried to tie this to quiz bowl for this week but failed, best I could do was to note that due to the prohibition against direct repeats, writers have to function like raccoons, but desperately want the reaction of players to be like rats. But that wasn’t fair to either players or writers.]
Tomorrow is the day. I’ve packed the office up, and readied it to move back into the other building. Tomorrow the moving team will box up the two desktop machines on my desk and drop them and the four boxes of books and papers into my new cube. They’ve only planned to move and give permanent cubes to those who work here on a regular basis, if you’re working from home now and having to come in, you’ll be in the pool of twelve cubes on the other side of the aisle from me. On first glance it’s smaller, darker, and a more avant-garde space (being hexagonal and my two monitors will now be at 60 degrees from each other.) I suspect it won’t be any louder over there, but I’m going to have a period of adjustment, just getting used to the notion of other people being around, and possibly not having all work interactions take place over a Teams call.
I’m going to be only two rows over from where I left the building in 2018. But it’s a very different environment from what I left when I got sick. The one time lounge area has been commandeered by a locked-up IT room. I couldn’t find the kitchen we started with, and I’m seriously doubting I’ll be given one of the lockers I was removed from last year. It’s almost like I’m going into a new job piecemeal. Being over here for seven years, I’m not really sure who’s going to be around, who’s gone to work from home completely, and who’s just gone. I have to assume there’s a bunch of people over there who assume I’ve been gone all this time, and they’re kind of right. HQ2 has been almost a secret hideaway the rest of the company barely noticed was there, and now it’s going to be gone.