I don’t think that would have mattered.
The amount of ranting and reasoning in the ranting has been on such a level that I think even if Frenck himself would do it for them there would have been ranting and complaining.
This is still “day 2” from when the news actually came.
Since custom components is not something HA support then the news first have to hit the fan before the custom component can be created.
Just like any breaking change there is no “here is how you solve it”, it’s just a breaking change.
Some come with no notice at all, this came with a long notice time to fix it.
I say that there is zero correlation between one and the other.
At the end of the day, the “product” is still “owned” by one or multiple people, who might or might not take the users demand into consideration.
Simple example: MySQL is OSS (plain GPL2) but that doesn’t imply that Oracle (the owner) will take any of your considerations into account.
No. Depending on the license, it is a moral or contractual obligation to propose the changes you made to an OSS program to all users.
You have a strange approach to FLOSS. Not only do you get software for free, but above that you’d want to have a say in the organization of the project?
Nope. Doesn’t work that way, never did. Sponsoring is a token of appreciation, and maybe an incentive for the devs to continue working on the project, nothing else, and surely not a ticket to have your voice heard above any other.
It’s actually that kind of attitude that might discourage maintainers. Not only are you providing free software during your free time, but users are complaining you don’t comply to their demands? The hell with it…
That’s definitely NOT the FLOSS spirit, but you are welcome to do so. Just don’t complain about what you get.
You seem to have this strange idea, because you pay for Nabu Casa or contribute to developers you have some sort of influence.
This is evident in a number of your posts this one quoted, an earlier one in this thread and in a others not related to this thread.
Your theme seems to be “I pay for Nabu casa”, and i was going to increase or sponsor people but based on this or that decision i may not bother.
Sorry but if you think just because you open your wallet means you have any more influence than someone who doesn’t then your wrong.
Absolutely agree, GPIO should not be ’ depreciated’ there is no devices on the market place as far as I can see that provides status inputs.
Before deciding to remove anything the developers should see what would replace it. A very high handed approach.
Doesn’t work that way. No developers have interest in supporting a feature? It’s dead.
They don’t have to spoonfeed you a replacement for your use-case. You have 4 months to do that on your own.
EDIT: Actually, you have until the end-of-times if you don’t update HA, btw.
Same is true for all the other users bashing the devs because they want both old unmaintained stuff and best and shiny of new releases.
He means, place 121 out of 886 integrations. Saying numbers like 0.8% makes it look tiny and never used, you can easily say that as “It’s used more than 85% other integrations”. Should the team be removing integrations that are in top 15% used ones because they’re… not used often?
I’m personally relived that Home Assistant’s future is not decided by votes or influence from the greatest contributors (financial/community support/documentation) it would grind to a halt! The disgruntled are often the most vocal but listening to the most vocal is not the way to make good decisions.
Decisions have to be made and those decisions are, more often than not, going to be swayed by the most active or in the case of RP-GPIO, least active, developers (the number of people using an integration has no relevance if it isn’t being developed by anyone!).
The project moves on, some of the changes take time to adjust to but ultimately, it’s free and awesome and there are means and ways to accomplish anything you want to.
Thanks very much to the core team and all the great contributors, the vast, VAST majority really appreciate it
Please stop with those numbers. Everybody knows they don’t represent anything, the devs in the first place.
Be assured that the stats had zero influence on the decision.
Sigh
Did you read anything above, its not just about how little it used or how many people use it, there are other factors, the number one being it has no maintainer
As has been explained ad-infinitum in this topic already, the number of users was not the reason the integration was depreciated. It was depreciated because no one was maintaining it.
ha-rpi_gpio is now on HACS, thank you Shay Levy and Joakim Sørensen!
It works great for me.
It is still not great for those who have never even heard of github, SSH or understand what “depreciated” means. Yes, there are users who are scared of even trying to install HACS.
Come June, and there will be noise.
It has become even more clear now that there was way too little understanding behind this decision to remove rpi gpio. On the release party video Frenck is telling us that it is only used by 0.05% of the users… Well HA analytics show 1% for Raspberry Pi GPIO alone and there are many other integrations using it. That is actually 2% of the Raspberry Pi users. It is number 121 of 886 integrations…so TOP 15%
It might be fun to laugh at braking stuff but it is not fun when your house gets cold or your kids can’t open the door because you “made a decision”
While open source community understands how this all works, “normal” people don’t. If you bought HA Blue and you are paying for Nabu Casa monthly, you might think you bought a “normal” product and you have the right to demand stuff. These users will never visit forums or read release notes until something brakes.
You have to start thinking and acting differently if you want “normal” people to join.
You seem to continue to beguided by statistics when it not the case, by the way it was 0.8% yesterday, so i guess those users who didn’t upload their stats in the thread decided to to try and get the numbers up, hardly a huge jump is it to 1%
It being removed, install the custom integration and move on
On the positive side, custom components don’t rely on waiting for the monthly core release and can be updated at any time. So any improvements or issues can be released promptly.
Also for the last time, it’s not just about the numbers.
There are nearly 2000 integrations in HA core, and many of those integrations have far fewer users than Nest. I can’t believe that they all have an active maintaner.
If something breaks an integration, and that integration has no maintainer, then I absolutely support it being removed from core.
But what is the rationale to remove perfectly functioning integrations?
Also to be clear I absolutely respect and admire the huge amount of work that goes into home assistant, especially from the unfinanced contributors who contribute and support niche integrations used by small groups of people. I’d hate to see all these integrations disappear.
Removing when it breaks makes for a worse scenario, they do not have any time to try and find alternatives. deprecating and giving people the deprecation time is a better solution
I really don’t see what the fuss is about, you can take the existing integrations and make it a custom integration yourself or in this case someone forks it and allows it to be installed via HACS or manually
Those number conveniently omit the core integrations (that are maintained by the main team). The original person who came up with those numbers did not include them in the calculations. This kind of miss information just leads to the argument that we are seeing.
Everyone:
Can we all please stop this?
There’s a solution. Regardless how hard it is with HACS, there is a solution that will get you exactly what you want. There always has been. Not to mention, you don’t need HACS in order to get there either. HACS is just an installation integration. You can always install these manually.
Repeating the same tired crap over and over again does not help anyone.
