It's also important to acknowledge that there are some places where both commercial and communuty based projects struggle, such as long term support (e.g. 10+ years).
In the case of commercial products, they have a strong financial motivation for you to buy the next model, so working to extend the life of existing products isn't a strategic investment of their time and money. Plus it takes a huge amount of sustained effort for years to build a reputation of supporting products for a long time.
In the case of #OpenSource projects, it's a little different, and some projects are great at long term support. The challenges are that the #software and hardware that they depend on frequently stop getting supported. At that point, do they spread people thin to stay doing long term support for all the things they depend on?
If it's #hardware, that's probably not an option even if you have dozens of volunteers.
It's still a cost problem, but the cost is people's time, not #money.
So what can help with the long term support (LTS)?
There's really no helping the commercial world as long as they're trying to make a profit. The limited exception would be companies who not only make this their central business model, but also establish a track record of delivering on that promise.
There are almost no business who have managed to pull this off. They used to be more common, but people are either unwilling or unable to buy the higher quality stuff these days. I could say a lot more about this, but I digress.
What can be done about the #OpenSource world to make things more #sustainable?
The biggest thing, by far, is to make things as simple as possible! That means you will inherently have fewer dependencies.
Depend on libraries that have long term support.
Contribute upstream whenever possible. Mainly code, but if you are selling #hardware, money too.
hestiahacker - LTS for hardware is brutal. CPUs get discontinued all the time. The same is true for small board computers (SBCs). For example, the oDroid SBCs looked promising but they never got their kernel patches upstreamed nor did they port their patches forward to newer kernels. So you could only run them with an old kernel that's no longer supported. They ended up abandoning those models of hardware all together. Now they have new models. How long do you think they will support those? Probably not long.
#RaspberryPi may have their own problems, but they've been solid on supporting both their #hardware and #software.
They even publish a minimum number of years they are going to keep producing each model of hardware. This allows projects to plan around that, which gives rPi a leg up when it comes time to decide which hardware to use.
I hope you enjoyed my reflections on the state of long term support and comparing commercial IoT to #OpenSource #IoT.
I've spent a lot of time not just thinking about this, but living it. I've had the core of the #HestiaPi (#OpenHAB) abandon support for all versions of Java that can run in a Raspberry Pi Zero W. 😩
So now our project is at a crossroad. How do we proceed? It'll run on the Pi Zero 2 W, but what about all our original users who have a pi Zero W soldered onto their PCB?
I think I see a path forward that will let still use the existing #hardware & OpenHAB. It'll be a lot of work, but I think we can do it.
If this is the kind of project you want to support:
1. Build or buy a #HestiaPi
2. Tell your friends
3. Contribute: bug reports, beta testing, documentation, #code, answering questions on the forum, or #building hardware for others
We don't currently accept monetary donations (but you can #zap me personally if you'd like).
I have been supporting this #hardware for over 5 years now and want to keep doing so as long as I can.
Having more people contribute makes it easier. Having more people buy hardware let's me buy in bulk & cut prices.
