PHP Foundation 'under-funded' with one full-time staff member, director says
j / k
|
Down or up a comment |
h / l
|
Previous or next link |
?
|
This help |
https://opencollective.com/phpfoundation/transactions?search... and https://opencollective.com/php-foundation-eur/transactions?s...
(seems like "work" selects most of that)
And it's not cheap to pay that many people, especially not full time (which is partly why they don't have more full-time contractors). According to https://opencollective.com/phpfoundation#category-BUDGET they only have a balance $350k currently, sinking.
It'd be really awesome, if more companies would sponsor the PHP foundation (especially with thousands or tens of thousands of dollars per year. A lot of small contributions is nice, but not enough.). The work of a few core maintainers is really crucial to the continued development of PHP.
Plenty of companies can donate 0.1 engineers to work on php, if for no other reason than to make php better for their own use cases.
And asking harder? How does harder look like in comparison to regular asking?
Better than running on fumes would be to honourably phase out and give people time to move on.
It overwhelmingly seems like the language of the cheap and not very technical or providers of services to the cheap and not very technical.
I worked at a major CDN with roots in the dot com era. PHP errywhere. You wanna log on and see your bill you're invoking like 40 different (modern containerized) lamp stacks.
Wordpress is the obvious answer to your question, but its codebase is admittedly still stuck in the past in many ways. Nobody would write it that way if they were starting from scratch today.
There are plenty of “serious” companies using Laravel or Symfony. Take a look around those codebases if you want to get an idea of what solid modern PHP design patterns look like nowadays. Nothing “cheap or not very technical” about them :)
Like if I were to sell PHP in 2026 to a new team, how would I position it?
Laravel is really excellent. PHP’s package manager, Composer, is great, and Laravel makes excellent use of it. The two main front end tools, Livewire and Inertia, are fun, with good starter kits. It has an elegant built in AI SDK, an official LLM developer integration, good payments support. There’s a new typescript binding. It has, IMO, one of the few truly well-designed ORM implementations.
There’s an incredible community around it. Lighthouse, the schema-first GraphQL package, is one of the neatest I have seen.
And PHP 8.x is a much more capable language than people expect. LLMs speak it well and are typically very up to date with its capabilities.
ETA: I think one of the things that sets it apart from other environments is the sense that it has a mature clarity of purpose, to be honest. In this AI-slop-documentation era, have a look at the documentation, the clarity of it, the ease of onboarding. Laravel as a project tends to execute well on ideas, has sensible defaults etc.; it just feels grown-up. It moves a little slower than other frameworks in terms of innovation, but sweats the details. It also has a well-managed upgrade process.
Unfortunately, I think PHP, and ultimately Ruby, Python, Perl, and Node.js, will all eventually be proven to be inappropriate for web development. It feels to me like we are heading away from a world where every application sits on top of an interpreter and a huge attack surface.
For myself, I really like Laravel; I have deployed apps in it that did good things for people and it was never once in my way. It has been fun, simple, even quite poetic to use. But I think I am done with these sorts of programming environments, and I am going to force myself back to things with a smaller footprint.
I certainly think that one aspect of PHP has to change, and that is the tendency to deploy with distro packages. That is clearly an inappropriate burden in the LLM-assisted hacking era.
Because we all know why COBOL is still used and why those old institutions are still important for functionality. PHP isn't on that level, but isn't far from it. many/most 20-30+ YO legacy codebases in web likely have PHP in some important part of its stack.
Wikipedia
Wordpress
Hrmmm.
They recently posted about how to patch their software from the command-line (https://www.netgate.com/blog/critical-php-security-update)
Also, not a company but Wikimedia Foundation definitely has some funds too.
There were a lot of sites affected by the later exploits, so I will say, yes.
[1] https://www.reddit.com/r/Wordpress/comments/1g6qi8i/automatt...
[2] https://techcrunch.com/2026/02/12/automattic-planned-to-targ...
https://www.therepository.email/automattics-anne-mccarthy-jo...
https://x.com/ThePHPF/status/1853390265429508427
GPL/AGPL/LGPL is the way to go for business. There are many dual-licensing success stories using GPL/AGPL, both large and small: Qt, iText, Grafana, Loki, Tempo, Mattermost, Cal.com, Baserow, and even MongoDB originally.
Qt is an especially interesting case because KDE and the Qt Company have a written contract. Qt can't just switch licenses. They have retreated from every voluntary contribution to open source, but the contract binds them and they can't do anything to it unless they do complete rewrite from scratch.
What does bother me is when the projects pivot to claiming that users now have a moral obligation to financially support the project. There is a kind of dishonesty in this move.
It's not that a successful project "should be free", it's that its success depended on it being free. If they had said all along that they'd eventually ask users to pay, that would be honest... but then they wouldn't have seen much of that adoption in the first place.
I suspect this is also the case for PHP though.
Likewise, users are free to continue using open source software without payment or donation, to fork the project, etc. If anyone tries to shame them for this, they should pay no mind.
Also it gets less fun when people start making demands of you and you have to deal with taxes and all the unfun stuff.
It only bothers me when they try to claim their new license is open source or use misleading language when it isn't. Which seems to be what 90% companies doing that do.
People can do whatever they want with their intellectual property as long as they aren't lying about it.
Until very recently, big tech pumped untold millions of dollars into obliterating FLOSS altogether, calling it a cancer that will destroy any enterprise it touches.
But if it was true: In the 90s, millions of small businesses around the world used FLOSS to host websites and offer services over the internet. That was decidedly not the route big enterprises wanted to go. They wanted to push through big standards through ITU et al, build giant, patented moats, and host it all on big mainframes with huge Oracle databases.
Even today, there exist small to medium businesses in the hosting and shopware space, and they can only do that because big tech could not push them out of that space because FLOSS protected them.
Security improved one blockbuster vuln at a time.
Closed source might have more money but they almost never invest it in security.
[0]https://www.fordfoundation.org/learning/library/research-rep...
If you like PHP, you might want to contribute. I like tmux, for instance, and I contributed code to them. I did the same for Sysdig, and a couple other projects.
Non-greedy action is by definion an action that doesn't directly and materially benefit the one acting. That is universal across all political and I would argue all biological systems.
I noticed bills for both Workspace and GitHub (admittedly small amounts for Workspace, less small for GitHub)
https://www.google.com/nonprofits/eligibility/
https://github.com/solutions/industry/nonprofits
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.
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.
(You need to address one specific company, because if you address two, neither will give you a cent, expecting the other to pay.)