It is not flawed thou. I have different routers, but one DHCP server and I could track everyone on my network (even guests to activate “guest modes”) regardless of them installing apps or whatever. I had just fine-tuned the scan_interval and consider_home variables to the devices. Just a few ICMP packets running around and it’s all good.
There are different scenarios and use cases.
They can put the default to whatever they want, but it should just be configurable, to be able to change scan_interval to anything the user desires, ie. leave it as it was. it wasnt bothering anybody.
Apparently, someone is bothered with this 
Please stop yelling.
I just find this whole thing so weird… Like if you’re technical enough to use ping in HA. You could find a tool to ping spam. It just seems like such a fringe case, where the devs think what they are doing is waaay more important then what it actually is 
Mate the decision is part of a bigger plan. Ping just got caught up in a plan.
If you make plans for an organized software ecosystem you stick to the plan, or you end up with inconsistencies.
It appears the consider_home was left out. That is or will be fixed.
Relax.
Taras, I have a question unrelated to ping issues.
Assume we need to trigger an automation every 5 minutes.
Then we can probably use a “time_pattern” trigger - with “minutes: ‘/5’”.
For more complex cases we can use a template trigger like you suggested.
I wonder which trigger is more resource-consuming - “time_pattern” or “template” with “now()”?
The Template Condition Trigger is evaluated every minute because it employs now().
The Time Pattern Trigger is executed every 5 minutes.
Beyond that, I don’t know how they operate ‘under the hood’ so I can’t comment on which is more efficient.
They are both a single listener, I doubt there would be any difference between the two if you’re just using now() in the template. If you start iterating the states table, that’s a different story.
Taras, Petro, thanks for answers.
Just installed the beta. Consider home is back. Awesome. I don’t quite understand yet what the reasoning was to omit it from the integration in the first place, but I am glad it’s back.
There’s nothing to understand, the volunteer who made the original changes didn’t bring the functionality over. This isn’t rocket science, there are people who volunteer their time to make changes and sometimes they omit features willingly or accidentally.
I can only add to my predecessors - this is a step in the wrong direction. Although I appreciate all the effort that is put into the system, it has always be carefully evaluated if a new feature is really worth to be implemented or if it only clutters the system. Sometimes I am in fear that HA will just become too overload to be run on a reasonable embedded system in future.
What is a step in the wrong direction? Moving to the GUI? Making a mistake? Correcting the mistake? Implementing a decision made ages ago?
“Wrong Direction”? That will depend on were you are standing and what direction you are looking.
In the case of ping I agree. It’s a step in the wrong direction.
Im not against moving to the GUI. What im against is the loss of functionality in such a basic thing.
Every time anyone configure a ping to check something they know how periodicaly they have to check and every device have diferent needs.
For example:
The point is: users should be able to set a “scan_interval” or an equivalent when configuring ping, because every single device have diferent purposes. Some are critical, other not so much.
Most likely devs will say: just use the automations, and stop moaning. For that, I have a question: Everyone had the interval and all necessary data in the YAML. Then, why didn’t HA created the automation automaticaly?
Why the hell would you ping a battery sensor to get the battery level? Ping does not report battery level.
He’s probably confusing the ping integration with zwave and zigbee’s ping. Which ironically has only ever been usable via an automation.
Not Zwave, its wifi. My sensor does not report the batery status, and I ping it once a day as a workaround. Its just to check if it is online or not. If not, time to replace bateries.
Maybe its stupid, and I need to replace them, but it works for me.
Then ping them once a day via an automation.