Focus, Focus, Focus

This is just a general rant, but as someone who treads lightly making changes, I’m finding that the amount of issues I have to regularly deal with is making me wonder about the focus of the project. Let me explain…

Yesterday, at 9:00 AM, my perfectly functioning (albeit “custom integration”) National Weather Service Alerts component just says “unavailable”. Reinstall, reconfigure, restart… Nothing helps. So any automations dependent on the state of that entity don’t happen.

Also yesterday, I deleted all the events from a network calendar previously inside iCloud, and recreated them all inside the new Local Calendar integration… This way I wouldn’t have to see all my backup and ZFS snapshot schedules in my phone calendar. Looked great… Progress!

When I rebooted to try and fix the NWS Alerts, now the local calendar won’t load. All the events are gone with the calendar, so I can’t delete and create again… Just lots of garbage like this in the logs…

pydantic.error_wrappers.ValidationError: 24 validation errors for IcsCalendarStream
vcalendar -> 0 -> vevent -> 0 -> __root__
  Expected end value type to match start (type=value_error)

So the rant is that I think this project needs to focus more on keeping things working. Endless weekends trying to simply keep this thing functioning as expected are getting old. There are many integrations which provide core functionality that need serious love, and they don’t get it. We still can’t setup Emby through the UI, get no added features like how many items are in libraries, etc… Many integrations that need further development sit stagnant so we can implement “new features” that do things like make us lose all our appointments…

This is by no means a negative review. I think this project does amazing work. I’m hoping that as a long time user without coding skills, I can raise awareness about us “average” users who REALLY don’t want to spend all day keeping this thing working, and would rather just use it.

The focus needs to shift to making it a platform that a novice can use, rather than a niche platform for people who like to abuse themselves with endless frustration.

Just my opinion,
Steven

1 Like

The log file would probably be telling you that there was a problem with that cloud service.

There’s not a damn thing the HA devs can do about services other people provide.

What they can do is fix bugs in Home Assistant when people report them by opening an issue on GitHub (and by providing all the required information) - such as the problem there with the calendar. Even then, if it’s a bug with an upstream library HA uses it needs fixed there before HA can do anything about it.

Well, this is a project where most of the work is done by volunteers. If you want particular outcomes then either:

  1. Open a feature request (here, in the feature requests section)
  2. Get stuck in and develop the change you want
3 Likes

I understand the issue with the custom component not being a HA issue.

But let’s address the local calendar integration as an example. This was just released as a new and highly touted feature. I installed it, created some recurring appointments in the interface provided, and rebooted… Broken. Is this brand new feature the problem of some upstream conflict? It seems more likely that I tripped over some bug that nobody else ran into doing something pretty basic.

I’m not messing with the calendar, in hopes that it gets fixed and my appointments reappear, but it’s a sign of what I’m describing. In this case I just tried to implement what the devs gave me, and it broke out of the box.

As for excusing the rest, I’d say this… Yes, an integration might be created by a volunteer… Emby, for example. But once that is accepted into the project, it becomes the projects responsibility to ensure that they don’t implement a change which breaks it. Because once I set up the “supported integration”, I don’t expect that a future update should break it. A common API for integrations to pass information to the main platform would help. But I’m not a developer and that’s beyond my ability to elaborate on.

I think the devs do a great job, and I’m not knocking their efforts. I’m simply expressing why I FEEL that the project will get stuck as a niche platform if it’s not addressed. I would like to see it develop into the STANDARD for automation. You buy a HA device and you add the integrations you need. Wuala! But the average Joe hasn’t the patience for the current state of this.

As for me, I will continue to subscribe to Nabu Casa, submit bug reports, and tinker. But I will also continue to submit constructive criticism if I think it’s warranted.

What “national weather service alerts” custom integration do you use?

This… (Is that you?)

Filed a bug…

It was literally working, crashed out somehow and stopped.

If it matters, it’s snowing here so there is a possibility that it crashed out when an alert was issued?

yup.

I assume you are on the latest version of HA?

What is the code(s) you are entering? Has it worked prior to this?

Commented on bug report…

Post your .yaml code here so we can see what you have. I show the current alerts when I add a sensor for the NYC111 zone. What do you see for the state attributes when you select the sensor in Developer Tools - States?

  - platform: nws_alerts
    zone_id: 'NYC111'
    name: NWS Alerts Sensor (NYC111)   

I deleted the integration, now it’s showing 2 alerts. Grrrr…

Sounds like it’s working then?

Yes, sir… Thank you for the help!

1 Like