PHP's Biggest Problem(stitcher.io)
stitcher.io
PHP's Biggest Problem
https://stitcher.io/blog/php-biggest-problem
3 comments
FWIW, i'll disagree. i was a _huge_ PHP fan from around 1999 until 2010, but two things finally killed it for me:
1) It's apparently impossible to cleanly structure large projects with PHP. i've been a part of many such projects across a large handful of teams, and every one of them eventually became a complete mess, structurally speaking, no matter how conscientious the developers were. Java, for example, forces a directory structure onto developers. PHP is more C-like in that regard - "do whatever you want" - and i've yet to see that end well for older/larger projects.
2) Breaking age-old working features with new releases, like when they removed split or join or implode or explode (two of the four, but i don't recall which). When they did that, and my decade-old website suddenly stopped working, i lost all motivation to continue maintaining sites in PHP. Dropped it that very day (for my personal sites) and have not looked back. i still had to use it in commercial projects off and on through 2014, and each one only help cement point #1 for me.
i don't miss it.
1) It's apparently impossible to cleanly structure large projects with PHP. i've been a part of many such projects across a large handful of teams, and every one of them eventually became a complete mess, structurally speaking, no matter how conscientious the developers were. Java, for example, forces a directory structure onto developers. PHP is more C-like in that regard - "do whatever you want" - and i've yet to see that end well for older/larger projects.
2) Breaking age-old working features with new releases, like when they removed split or join or implode or explode (two of the four, but i don't recall which). When they did that, and my decade-old website suddenly stopped working, i lost all motivation to continue maintaining sites in PHP. Dropped it that very day (for my personal sites) and have not looked back. i still had to use it in commercial projects off and on through 2014, and each one only help cement point #1 for me.
i don't miss it.
The PHP website is indeed one of the worst parts of the whole ecosystem. Just look at the landingpage (https://php.net) and compare it with those of other languages. There's not a single piece of PHP code on the page. No "what is PHP", no "why should I use it", and no "that's why PHP is great". It's just a news page showing the latest releases, and a small section for downloading PHP.
And speaking of the website: the docs always have been a pain to use. Even the man pages of most decade old CLI tools have a better documentation. The docs basically only list the functions, explain what they do in cryptic one-sentence descriptions and list the parameters. There are no user-friendly explanations and no curated examples. There's indeed a lot of room for improvements.
And speaking of the website: the docs always have been a pain to use. Even the man pages of most decade old CLI tools have a better documentation. The docs basically only list the functions, explain what they do in cryptic one-sentence descriptions and list the parameters. There are no user-friendly explanations and no curated examples. There's indeed a lot of room for improvements.
All of this for the second lowest wage after JS devs. Not worth it.
I'm using Go and i'll never go back in the immediate future: go run . and i have all of it. Period.
PHP should strive for something like this.