Well last time I tried that it prompted me to run the shortcut.
It’s outside of Shortcuts, but when attempting to set up an automation through the Home app it displays a message that gives a little insight.
My attempt was to make my garage door automatically open as I arrive home. Unfortunately that isn’t an option. I can understand why, but also be bummed that I couldn’t make it happen. The garage door button in CarPlay is pretty dope tho.
“This automaton required confirmation to run because it may allow entry to your home”
So the scenario apple envisions is:
- The user creates a shortcut that opens his garage door when arriving home.
- Thief watches users create the shortcut.
- Thief steals iPhone and goes to users’ home and loots it.
First, it seems pretty unlikely, second, it is the person that owns the housing choice to make if they are so paranoid then put up a warning while creating the shortcut.
I’m on the current 13.1 public beta (iOS and iPadOS) and the situation is the same
Pop to neighbours for a nice cup of tea, neighbours is on edge of geo boundary. Door keeps unlocking with a click, courier delivering hears this and decides to have a look around.
Live in flats, stalker waits by front door. You reach the ground floor within geoboundary, door unlocks, stalker walks in and waits whilst your taking the lift.
If (non whitelisted human detected) {
unleash gamma cannon
wait 60
activate roomba
spray pine air freshener
}
2 Likes
The bluetooth lock market must be going through hell right now…ohh wait it isnt.
Point being, its not necessary for a bad actor to view your creating a Shortcut.
What’s bluetooth locks got to do with geofencing, when they use bluetooth to unlock?
Because Bluetooth is about as reliable as geofencing and people still use it to unlock doors. Anyway, all this is possible on Android and there has not been a wave of Android-related break-ins. If Apple is so paranoid then write a good disclaimer and as with location sharing ask the user. Otherwise, do not implement it, focus on something else.
Bluetooth can have a 10m/100m range. Just because bluetooth locks can unlock doesnt mean the accompanying app is allowing a 100m range geofence, they might be using 10m. Allowing 100m geofence could still be inadvisable, especially as location data isnt always accurate. Preventing people from doing it on 100m range seems very sensible.
You can have auto unlock doors for Homekit that unlock with a 10m bt range
I think you are missing my point. You can dream up all sorts of scenarios but at a certain point Apple’s responsibility ends and the users begin. I can today hook up my eve smart plug to my toaster and Automate it, imagine if my AppleTV dies my house would burn down, should we not ban Eve smart plug automation? I don’t think so, I’m a grown-up, let me make the decision.
Look at how Apple run everything. Right or wrong in the purely personal opinion of non-Apple decision making employees aside, the company is famous for everything being closed and it being their way or the highway.
Fundamentally, the decision is use Apple’s offerings or use something else. That’s the decision you get.
When it comes to security I’m going to guess Apple have collectively given it more thought than the rest of us put together over the last few months. Anything we have is based on assumption and guesswork and should be tempered as such and be described with appropriate caveats; unless someone here is an Apple insider and wishes to publicly breach protocol and probably dozens of NDA.
One thing I’d like to suggest is that it may not even be a technical reason they have not opened up the automation triggers further at this point. I can certainly imagine that if some “grown up” didn’t really understand the implications of something, did it, and then that caused a security issue that this could reflect badly on a company who publicly push security as a core tenant and have so much value wrapped up in their brand. Outside of tech enthusiasts, tech journalists and IT professionals, the public would, as usual, pick things up as a company issue rather than the end user. Apple may simply have decided in their usual meet the needs of the many not the few approach to err on the cautious side.
Hopefully it won’t remain this way forever, and at some point we may get some insight into the reasons behind keeping things with a manual confirmation as they currently have. Until then, we just need to do what we can with what we have and feedback to the Apple decision makers on what we’d benefit from in future automation triggers.
Then do not implement it, focus on something that brings more benefit. Currently its like selling an answering machine that you have to turn on when someone calls.
I’m assuming you are not a decision maker at Apple, and so I refer you back to the details of my previous post regarding how Apple ‘should’ do things. Without the details we simply don’t know the reasoning. It is pure conjecture and should be couched in terms of likes and wishes.
In terms of perceived benefit, once again that is purely opinion and the comment must be assumed to be independent of the actual road map Apple will be working to.
In terms of the answering machine analogy I wouldn’t buy that answering machine if I were looking for an answering machine. If I were looking for a call recorder, then I might well buy it. The point is there is functionality there that some people can use, and it may simply be an iteration to build on. I would ask why hold back limited functionality that can be useful. Particularly if it could be a preparation for Joe Public in terms of what’s planned next? After all, we don’t know what’s planned next.
For me a closer analogy might be a car. A car is an automated form of transport (it’s in the name in many languages and of course we have the term automobile) whether or not it is self driving. But let’s consider the starting process. Some cars are enabled with keyless entry and ignition. Some are not. It doesn’t affect the functionality or the ability to use the automation. It is still functional and it is just the ease of triggering.
I would absolutely have loved to see more potential automation triggering functionality delivered in i*OS13, but at the end of the day I’ll simply make the most of what I get on my devices.
1 Like
Let´s not confuse the issue. My point was all along that Apple did not build Automation for Shortcuts, simply Automated reminders.
While speaking of “perceived benefit”, at what point does it become objective? well, let’s go to the dictionary and find out.
Automation = noun. the technique, method, or system of operating or controlling a process by highly automatic means, as by electronic devices, reducing human intervention to a minimum. a mechanical device, operated electronically, that functions automatically, without continuous input from an operator. act or process of automating.
Now, after reading above do you see any grand “perceived benefit” with the automation tab in its current implementation?
From what I currently understand, the automation system allows for easy set-up of the triggers for shortcuts, so I would have to say that for the majority of i*OS users, yes, I do perceive a benefit for ease of use in triggering Shortcuts.
again
" without continuous input from an operator. act or process of automating
"
Tap on something, shortcut (the automation) runs. One brief interaction from the user at the point of initiation. Seems like discrete rather than a continuous operator input to me.
I’m obviously missing something more subtle in what your point is on that one 
So let us take a use case that Apple currently does not allow, that I can do with IFTTT today.
Every time I leave home for work and leave work to home, make a timestamp in Evernote.
Now imagine that I have to pull up my phone each time to execute a stupid run command, say I work 300 days a year. That is 600 times I have to pull up my phone.
Let us also say that this takes 10 seconds to log in and press ok, that’s one and a half hours a year on just one automation. That is 29 million hours a year for America’s iPhone users alone. Counting the fact that the average length of a US citizen is 27375 days that´s about 4 lives saved a year on just one automation.
So if you have 10 time-based automation Apple style it kills 40 people a year.
Wow.
That is the first time I have ever read a response for a mobile phone/tablet operating system feature addition be distilled down to what reads as an accusation of murder; and with that I think your credibility as someone to have any sort of meaningful discussion with simply vanished for me.