Hacker News Reader

Comment by chuckadams | Hacker News Reader

They'll sit there and meow for a bit, and if you're not around to let them in, they'll wander off and try again a little while later. I'd be more concerned about complicated gadgets like this breaking down completely.

View on HN

Yeah - I recently replaced a SureFlap with a mechanical flap, because it failed and let my cat out the day she had come home from an operation and needed to be kept in. 'Failure with no warning' is more likely with complex products.

And if you’re away on vacation?

You shouldn't be leaving an animal without supervision for extended periods like that! If nothing else, what if they fill up the litterbox?

Who said anything about extended? Do you think there might be a period of time a person might be away such that they would not want their cat locked out the whole time, but their cat also has no chance of filling up their waste receptacles?

Vacation generally means an extended time, so you did.

If the false positive rate is not 100%, their next try is likely to succeed. Unless, as I mentioned, the contraption just breaks entirely, which is why I'd be reluctant to have one of them in the first place.

We have no evidence that the failures are randomly distributed. What about when the failure mode is such that any time there’s not a massive blizzard outside, it works great. But when there is… icecat.

Sounds like you and I are actually in agreement here. It's not regularly-distributed but still occasional glitches I'd be worried about, it's the modes of total failure. Or as I like to put it: "gadgets break".

> If the false positive rate is not 100%, their next try is likely to succeed

Assuming they’re randomly distributed. If it’s 20% of cats getting rejected 50% of the time, that could be a problem.

Reply on news.ycombinator.com