> But learning from the past is not a weakness.

What's the lesson you are learning from?

The lesson I see is that any language, even PHP, will give you success if you "build something people want".

I read lots of startup "shutdown" stories - and I've not read one that mentioned the choice of "PHP" (or any language) as the reason for shutting down.

So if your comfortable with PHP - punch out what you need to get started...

Businesses usually don't fail because of picking a wrong programming language. If the tool that's best for the job is not the same as the tool you can make what you want the fastest, I'd say go with the tool that gets you where you want the fastest, even if it isn't the best. If you're successful you'll have money to fix that later on. If you're not successful then it doesn't matter which tool you picked anyway. If Zuckerberg would've learned another language first, there might not have been a Facebook.

Or you use sane development practices, well supported libraries and frameworks, and the latest version of PHP and run very successfully, because it's more than fast enough, shared nothing is a viable paradigm and being able to quickly see the effect of code changes without compiling and redeploying into a container is something you miss when working in the Java world.

Is it suitable for EVERY application? No. But then again, nothing is a one-size-fits-all solution for all problems.