zlacker

PHP 8.2 will get official security updates until December 8, 2025, how much longer do you need?

>>Y-bar+(OP)
It's not the same, but MySQL offers like 8 years https://endoflife.software/applications/databases/mysql. Obviously very different software with different goals and monetization models (none vs some). Not apples to apples but figured it was a good example of what someone might have in mind.

>>Y-bar+(OP)
Thats just 3 years, and likely the next release will be again full of incompatible changes. That's not a long timeframe for all but the most active (or smallest) codebases.

>>Y-bar+(OP)
I see 5 years as a good minimum for LTS. Perhaps it's a personal bias, since our OS of choice (Debian) offers LTS for 5 years.

Edit for context: the small business that I work for takes on all kinds of random web work. It's not uncommon at all for us to be asked to rescue an application or website that's still running on PHP 5. For projects that we have initiated, by the time we migrate them all to PHP 8.2, it will be End of Life.

>>jjice+64
PHP 7, released December 2015, 7 years ago, still gets bug and security patches. Seems similar to me.

>>Y-bar+UH
It doesn't though (granted, support for PHP 7.4 only ended 10 days ago, so you can be forgiven for not knowing): https://www.php.net/supported-versions.php

But PHP 7.4 is not compatible in many ways with PHP 7.3, so lumping them together is not OK. As you can see, PHP 8.0 which was only released two years ago, is no longer receiving active support. Definitely, database LTS is not comparable to programming languages. However such rapid releases of breaking versions is a relatively new phenomenon (thanks NodeJS), and I get the feeling that you've never actually supported a software project for 5+ years.

>>Y-bar+(OP)
Well, try working on enterprise and you know how much long support they need. Move fast and break things simply doesn't work and doesn't scale properly.

>>Y-bar+(OP)
A decade would be nice.