Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Our app powered car rental lost cell service and now I'm stuck here (twitter.com/kari_paul)
36 points by scrollaway on Feb 18, 2020 | hide | past | favorite | 6 comments


I don't run a car-sharing company but I don't think this is the way to handle the situation:

> after @GIGCarShare told us to sleep in our car on the side of the road and try again in the morning we called a tow truck on our own and made it back to our Airbnb. TBD on whether I’ll be refunded for this.

> also we were able to turn the car back on somehow but now we are afraid to turn it off because it may not start again and Gig told us we used our “allotted restarts” of the car so we are on a literal endless road trip through California now


This incident is fantastic. I worked at Zipcar on the firmware deployed in the cars, and these kinds of issues are _hard_ to get right, though this failure looks to be spectacularly naive. I had to deal with cars parked in garages underground, as well as off the grid (and moved between countries, and with dead batteries, etc).

There is an inescapable tension between default-permit (the customer _must_ be able to get in the car and drive it) and default-deny (it should be impossible to steal a car) that results in all kinds of trade-offs with UX implications. These are, at their core, sociological problems that the technology can guide but not solve.


Seattle had a problem with Lime Cars.

People would put their Lime Car on a ferry and head west. As soon as the car lost reception from its range of allowance, it essentially was bricked. The car would have to wait for the ferry to come back to Seattle and be towed off.

This was shared by a friend over Labor Day.


I saw this on Internet of Shit: https://twitter.com/internetofshit


[flagged]


I’m a “sheltered young urbanite” but when I bring up the need for testing under a wide variety of real-life scenarios, I frequently get “ain’t nobody got time for that” from the PM (or management says yes, but does not provide enough resources).

There’s no need to make this an attack on the developers, I’m pretty sure they’re running into software development economic factors that transcend individual personalities here.


You're assuming this was a "didn't know" failure mode, not a "didn't care" failure mode.

I'd wager it was a "didn't care". They knew about the limitations, figured the product would work often enough to make some money and labeled everyone pointing out the problem as haters or technophobes.

Someone probably even pitched a keypad password based redundancy and got drown out in valley buzzwords about MVPs and shipping fast until they gave up.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: