I'm unhappy with the removal of GPIO

[parautenbach (Pieter Rautenbach)]

@CaptTom

In future if I feel personally attacked, I’ll assume it was because they misunderstood what I was trying to convey, and perceived it as just another uninformed rant.

I agree with this sentiment. I’ll also acknowledge that it can be hard to get one’s message across.

@CaptTom

Four months is plenty of notice. That wasn’t the point being made. The point was, perhaps the group of people who were knew about this before that posting in the Release Notes wasn’t large enough to generate the discussions which led to someone ultimately taking over the integration.

I hear. I was genuinely curious to know how much lead time would’ve been sufficient and wanted to point out (indirectly) that someone will be left behind no matter what. If I consider many of the comments on this topic, many would like to postpone this almost indefinitely (that’s my impression) in order to avoid the future removal. I think, if a clear stance wasn’t taken, it would’ve been taken less seriously and the eventual blow-up would just have been postponed. There are always leaders and laggards. I will bet money on the fact that even though there are now three major threads on this topic (including the release topic and this one) there will be someone coming June that will be caught by “complete surprise”.

@CaptTom

The problem is, nobody who just reads the release notes knew any of this. They saw the deprecation and abandonment as an immutable, final decision. When you hit thousands of users with a notice that the function they depend on will no longer work, even if it’s four months away, you have to expect that decision to be questioned.

Personally, I first became aware of this when I read this post: Removal of GPIO support. This was in December. Granted, it’s in the developers section so it doesn’t have the same reach as the release notes, but to me that made sense, since if anybody was to take this over it would probably be a developer. One can also subscribe to notifications on the ADRs. Here is where the original issue was proposed (already in August): Stop supporting Integrations that use GPIO in the core. I’m just trying to say this has gone through several stages, but I understand your emphasis is on the general users of HA and how they perceive the change.

I agree that the community are allowed to ask questions, but they/we should also try hard to understand why that decision was taken and understand that this is an open-source project, not a commercial one so making demands will never go down well.

@CaptTom

If anyone reads this far before firing off an angry retort, what I’ve been trying to say is that it could have been broadcast in the Release Notes (or anywhere else users will see) that developer support was needed to help avoid losing the integration. I think most of the community would respond positively to a request for help, to keep something they use alive. The negative response was because of how it was presented, not when.

Some fair points. Perhaps including links to prior discussions (like the ones above) might’ve helped to give more context and make it a but more user friendly for those that don’t yet know where to go look. Some sensitivity acknowledging that there is an impact on some users might’ve helped, but here’s the kicker: As it stands (even before thecode put up their hand to make this a custom integration), there was zero impact at that point in time, but reaction was as if the world was ending, and that was mostly for two reasons: Misunderstanding what deprecated means and not any willingness to adapt or get involved. It took surprisingly long (in my eyes) for someone to put up their hand. Even so, the lack of a maintainer wasn’t the only consideration.