I see a fair few people here getting pissed at the pet owners for using a “smart” feeder, and I kind of get some of the sentiment, but the issue here is not purely with their use of a device to do a task. A washing machine has a timer, a dishwasher has a timer, and a smart feeder has a timer. These are all valid things to have timers on and even to schedule operation to make them automatic. Automated feeding ensures that your pets get food even if you have a bad day and end up forgetting. I have severe ADHD and the only reason my cat ends up fed and I end up taking my meds is using systems.

The issue here is the structure of the system. If the smart feeder had a local copy of the instructions and its own clock and then it dispensed food according to that schedule unless the plan was updated we would never had heard about this issue and nobody would be judging anyone. But what we have here is a system that puts all the control in the hands of a company over a device in your home. If you could update the feeding schedule via a Bluetooth connection or use Home Assistant to manage it then it would just be another modern convenience. The problem is the closed source application which actually has control over the feeder and nothing you do can control it directly.

Also, from an engineering perspective, come on, this is a simple failure state. Internet connection go down. DNS fails. Servers go down. If something like that happens the feeder should continue without intervention. To do otherwise is intentionally bad design. If you don’t code for your failure states you are guaranteeing their consequences. I would rather have a machine that can connect to an app for config but can also be configured by a simple web portal or via Bluetooth.