This is exactly the problem with unlimited PTO. It makes employees to feel guilty about taking time off and allows employers to avoid paying out unused time off when they leave.
I've been running a PC Engines APU2 (https://www.pcengines.ch/apu2.htm) for many years now. It has enough power to route over my gigabit fiber connection. It has mini-PCIE slots for wifi cards. If you want 2.4GHz and 5GHz bands, you'll need two cards and run two instances of hostapd.
Traffic and accidents might push you into different routes, especially if you have major decision points like which route you take around some geographic feature. When I drive in Los Angeles decades back even people who’d been driving there for ages wanted traffic reports to pick between roughly equivalent freeways because almost inevitably one of them would have much worse congestion due to an accident and you could save 20+ minutes.
Distances in LA aren't measured in miles, they're measured in time -- and the distance between two points changes based on time of day, direction, and events like sportsball or presidential visits.
Knowing where to switch between freeways and surface streets, and what route to take in what conditions, can cut half off the driving time.
Some examples of varying conditions from my personal history, even with optimal routing:
Commute to work: 22 minutes. Commute home: 1h 16min.
Go visit a friend: 45 minutes. Come back home at night: 16 minutes.
> Knowing where to switch between freeways and surface streets, and what route to take in what conditions, can cut half off the driving time.
Yes - I remember as a kid learning how to read maps and noticing my dad's face when I was like "These cities are only 10 miles apart. Why does it take an hour to get to grandfather's?".
The surface streets are an especially interesting blindspot a lot of drivers have. There was a period around the turn of the century where I was commuting between Santa Ana and Garden Grove and freeway traffic was generally manageable early in the morning but on the trip home it was often better to cruise down the surface streets where the lights were timed around 20mph than to sit in stop-and-go traffic on 405 or 55/5.
Driving in the London suburbs, I know how to get most places I want to go. In fact I usually know several different routes that are roughly similar in travel time. Google maps is excellent for telling me which one of those routes is least congested right now. So I'm not really using it for navigation, but for congestion information. It very rarely suggests a route I don't know, but I use it all the time anyway.
There is a gate right behind me that is closed with chains.
It isn't possible to drive through there, though you can walk in and out from the smaller gate.
Google still always tells me to drive through the gate.
I don't think Google Maps auto-learns.
Maybe it needs a bigger sample size so people don't abuse it?
If you use the reporting interface they seem to update promptly. There’s a major road near us where Google used to tell people to make a illegal left turn across 4 lanes of traffic, which was common enough that I used to see it almost every time I went by for years. Once I found the well-hidden reporting UI, that stopped two weeks later.
There is a no-left-turn street near my work, and Google maps keeps telling me to turn left there. I've been reporting it every now and then since at least 2018 and it does nothing.
It learns other patterns, too (which is kinda creepy). On Thursdays, I like to swing past a local taco shop on my way home after work and when I plug my phone into CarPlay, it prompts me navigation to the taco shop; it doesn't do this on any other day.
Kinda a neat feature, but it makes you wonder what else it is learning.
The Tesla maps do it too, when you tap on the navigation address entry there's a list of 'Suggested' which is learned based on time of day / day of week.
Intent matters. The maintainer very clearly intended to do harm. They abused end user trust which is a common attack vector for many pieces of malware.
I'm curious if you think the same applies to a developer that writes any kind of ransomware when an end user downloads and installs it knowingly. End user trust is a common attack vector for malware and the developer here took advantage of that just like any other malware developer.
