> Most pantens are so vaguely worded that any independent solution to a problem can be claimed to infringe on a patent.
No longer true, over the past ~decade, quality of patents issued has dramatically improved. Most patents are actually very narrow. Just look at the sheer quantity that are issuing (in software in particular) - everyone is coming up with something different.
> And even if you don't actually infringe on the patent, you better have a couple of years and a few $100k-$millions to dedicate to the court battle to prove that.
Or before recklessly entering a market, evaluate the landscape. Do you have something new to offer, can you yourself protect your innovations, narrow down your product scope to minimize encroachment on others property, etc...
Think less about building the whole kitchen sink and more on being the best at some tiny improvement (texture of the handles, connectors to the pipe to minimize leakage, etc...). Think writing the perfect 50 LOC that others would want to use rather than building 50k LOC. The less code you ship, the more you've minimized your patent infringement risk (along with proper evaluation of competitors protect at the outset). Additionally, if you've found that 50 LOC that nobody does yet and others want to use, then you've likely also found something valuable to patent yourself. BECOME the innovator rather than sued by the innovator.
Business requires managing risks, patents are but a small factor if properly handled.