HACS: restart later

“restart later” means “do not restart now, wait until the machine is restartet in a few days or weeks”.

Once a week or once a month!
And I want the orange notification to disappear because it alarms me every day.

In the grand scheme of things, war, famine, pestilence, impending climate catastrophe, etc… is it really that bad?

In the grand scheme of things, all of home automation is irrelevant. I see it as a big achievement that even within the small irrelevant universe of HA, this small thing is what people would like to see improved. It’s a testament to how good it is already.

Do I share the feeling that this would be a nice feature to have? Yes. Do I think this is the first thing volunteer devs should spend their sparse time on? No.

1 Like

I think this is the answer to the original question, and this is the place to send the request: Issues · hacs/integration · GitHub

Just to be sure - you mean this orange bubble?

Yes this is what I’m talking about…

Than this is the one and only possible solution.

Is there no possibility to build in an override into the HA component (is that core or supervisor?) that allows any repair notification to be ignored, even if the integration raising the repair does not offer that explicitly? I think that would be a reasonable feature request and in line with the idea of having the user in control as much as possible.

1 Like

The Spook custom integration has an Action that can ignore all Repairs…that’ll prob do what you want…

Forum post is here >

and specifically on the Repairs features is here >

Technically yes, but it wouldn’t make sense. It is in the responsibility of the integration to create proper repair issues, not HA to overrule this. Further it could be disadvantageous or causing other issues, when forcing to ignore an repair issue, which should not be ignored. Will say, if an repair issue is fine to be ignored, than the integration has to consider and implement this.

So still, just request this change to the HACS repository, maybe they will implement it or have strong arguments against it.

Anyway, this is not an Home Assistant, not Supervisor issue, when a custom integrations repair issues, does not meet the user’s expectations :wink:

1 Like

So it is the repairs integration that shows that orange icon. And it is the repairs integration that could offer an “ignore” option for the issue.

nope, repairs is not an integration, repairs are helpers within the Home Assistant Core, which integrations can make use of - for details, see the repairs dev-docs.
As i mentioned above, the integrations, which make uses of this repairs framework are responsible to configure their repair issues as neccessary.

1 Like

Instead of trying to get rid of the repairs that you get after updating, why not skip the updates and with that avoid the repairs all together?

I am pretty sure you can still see them in the HACS page in something like a Pending updates section. That way you can apply them at a time it suits you and then reboot after applying all pending updates.

Yes, technically it is an integration (better said a component) as almost everything is in HA, but it is an internal one, those part of the HA core features (like the sensor or switch component).
It still counts (this is the last time I’ll repat it) the integrations, which make uses of this repairs framework are responsible to configure their repair issues as neccessary.

I believe the user, and not the developer of the integration, should have the last word about what can and cannot be ignored. If the user gets a warning and they explicitly ignore it, it’s their own risk.
In the case of HACS, I completely understand that the developer thinks this warning cannot be ignored, since the update cannot be applied without a restart, there is no way around it. However, I don’t care that it is only applied later.

I’m not sure, but isn’t the functionality to skip an addon update part of HA as well? That’s not part of what the addon does, is it? Or can an update to an addon or core version be marked as unskippable? (note this would not mean you are forced to do the update, just that you keep being warned about it)
I think it would be good to keep that end responsibility with the user.

Since nobody raised this request or question to the HACS project, yet, we don’t know if this is on purpose by the HACS maintainers, or just has been implemented as is and no one ever complained about it.

Even the update entity component, allows the integrations to disable skipping an update.

The GitHub issues policy on HACS forbids feature requests so I posted on Discord:

If you are not on Discord:

Support “Ignore” for “Restart” in Repairs
Would it be possible to ignore the “Restart required” step after updating an integration?

I often install integrations when they are available but I don’t need / want to restart HA immediately.

The number badge on “Settings” is annoying and I’d rather ignore / dismiss the “Restart required” action.

It’s not preventing HA or the integration to work while they are loaded so no immediate action in required.

Other reasons discussed there (for the “dismiss notification” part, not the “restart later” ability):
HACS: restart later

And the answer of Ludeeus:

1 Like

Thank you!
If I would restart HA every time it wants to, I had to restart twice a day!