Beating the dead horse, as you are: the Bun Rust migration cost millions of dollars in tokens and took an additional 3.5 months of labor after the widely-marketed "11-day rewrite" before being shipped as a release, and in the process introduced hundreds of new issues.

Sure if you ignore the many governments around the world that use PHP, the (still) most popular blogging software (which also has several ecommerce solutions bolted on top), wikis. The list goes on.

A pet peeve of mine is when people offer simple solutions that are only simple by virtue of ignoring all use cases that don't matter to the author.

Every interconnected system of concerns has simple solutions when you reduce it down and ignore the ordered effects.

What's easier, maintaining PHP or migrating/breaking every single PHP-based site?

And who's to say whatever sites are migrated to won't be "sunsetted" in a few decades too?

Seems like a no-brainer to maintain it.

How cheap are code migrations nowadays? Vibe-migrations, you mean?

Sounds like you’ve been snorting too much iron oxide.

migrating to what? just take a look at the newest CVEs, https://www.tenable.com/cve/newest , and you will see that no technology is spared by high or critical reports.

Better, call the company that uses PHP most, and tell them than unless they give you $100K a year, you will sunset PHP.

(You need to address one specific company, because if you address two, neither will give you a cent, expecting the other to pay.)

I agree a lot of PHP applications can be ported to other tools, but that only covers that much. Just look at banks and COBOL code.