I think part of the challenge is that no ground is knowably stable. Since very few of us are shipping products that we've built from the atoms up we do have to rely on platforms provided by others. Everywhere you look there's uncertainty:
- relying on big hardware, OS platforms, or cloud platforms, e.g. iPhone/iOS, Android, AWS you're at the mercy of bugs, changes in behaviour, deprecations
- relying on walled garden ecosystems your risks are app review processes and changing priorities, e.g. the App Store, console distribution
- relying on open source you're at the mercy of project governance and contribution. Yes you have the source, could your business model really fund the maintenance and development of the project if independent contribution ceased or you needed to fork?
At some point you just have to place your bets and play the game. Good architectural decision making can, of course, mitigate your risks - and there definitely are 'bad' decisions.
You extend even more with project governance, extending to copyright, user agreements and product lifecycles you can not influence.
That's the thing right? You don't need to extend it if it worked at the time. You have the source and you can employ someone to make it work. The risks are much less than a any other option. Copyright, access to source and even the ability to continue functioning almost make it a no brainer.
I do agree that at some point a decision has to be made. We often shortcut the right thing to do for the immediately profitable.
I should not be surprised, but I do get disappointed.
- relying on big hardware, OS platforms, or cloud platforms, e.g. iPhone/iOS, Android, AWS you're at the mercy of bugs, changes in behaviour, deprecations
- relying on walled garden ecosystems your risks are app review processes and changing priorities, e.g. the App Store, console distribution
- relying on open source you're at the mercy of project governance and contribution. Yes you have the source, could your business model really fund the maintenance and development of the project if independent contribution ceased or you needed to fork?
At some point you just have to place your bets and play the game. Good architectural decision making can, of course, mitigate your risks - and there definitely are 'bad' decisions.