WordPress Doesn't Matter for the Future of Web
molodtsov.me
molodtsov.me
If WordPress dies a slow death as it's phased out over decades, that gives us plenty of time to develop the next generation open source replacement. Given time, those of us who still want to have control over our web presence (which is a significant number of businesses in addition to those here) will develop alternatives.
But if it dies a fiery death over the next year (as Matt seems determined it shall), then the author's prediction will come true—the proprietary locked-in stacks will welcome the WordPress refugees, many of whom will be businesses that will have soured on the idea of using "open source" if that means they're subject to the whims of some guy they've never heard of and have no relationship with.
That's why the drama matters. Currently 43% of the web runs on an open source platform. Matt is putting that short-lived victory for open source in jeopardy.
I think what's being missed here is that yes there are sites being built using Webflow / Shopify / whatever. And there are solutions for landing pages, or simple blog sites. And yes, flat file stuff with a GruntNodeDockerWhatever deploy pipeline is always going to make dev types happy.
But this is missing a big old chunk of people - those with relatively complex sites with a fair amount of content (say, I dunno, 100-1000+ pages), the requirement to link between pages, custom taxonomies, custom posts, have custom fields, multiple authors, different roles, requirements on things like member access, an API, feeds...
I think you'd be hard pushed to name a solution that isn't 1) Drupal (cos no-one wants to go there) or 2) Bespoke and stupidly expensive.
WordPress gets a bad rap on HN, some of which is warranted. But I also think there is a lot of hate which is unfounded, possibly from people who haven't touched it in years and years.
It will be a great loss for some who are hardcore plugin developers, but in reality they're in a far better position to pivot than WordPress/Automatic themselves.
I really hope Wordpress dies off to the history books of the internet experience, it can be a nice play experience to quickly make a website and see some results for people learning, but honestly without its plugin developers it's nothing. Another player could easily arrive and nobody would be upset. If anything this petty drama has hopefully expedited that scenario, I know I would be willing to support an alternative.
WP started out as blogging software and expanded the mission to include pretty much anyone whose main focus is publishing content on their own website. It's still pretty good at the things that matter, namely that doing the above is still very easy for a totally non-technical user, and the plugin ecosystem is absolutely massive and adds almost any kind of functionality you can think of (if your experience that there is never a plugin that will do what you need, you're definitely atypical).
Mullenweg's antics have been ridiculous - but outside of the WP community itself I don't think they've having a huge impact. Like the people whose businesses depend on a website that a developer set up on WordPress for them - 99% either heard of this issue in passing (maybe because their plugins wouldn't update for a few days) and have now forgotten about it, or they didn't even become aware of it in the first place.
Developers constantly underestimate or misunderstand what this space (publishing content) requires. I don't know why so many of them think you can replace a CMS with a bunch of static HTML files - Dunning-Kruger I guess. The plugin ecosystem is an absolutely massive moat. It's large enough that you would need to spend a billion dollars building an equivalent. Yeah WP has its problems but the most likely competitor to spring up is probably a fork of WP that can stay plugin-compatible at least for a while. The other big open source CMSes that shriveled away did so largely because they could never match the breadth of functionality that plugins made available to nontechnical users. NASA, the White House, Rolling Stone magazine the New York Post, and basically the entire News Corp empire all run WordPress just to name a few off the top of my head. Maybe its days are numbered but more likely it is big enough to be like say IBM and live on in some massive form or another forever. It never needed to be cool in Silicon Valley to exist.
What was telling to me is that even new media websites are using other CMSs or building something of their own these days. WordPress was almost synonymous with the online news media for a time. This is exactly the area where you need multiple authors, taxonomy, and other things you mentioned.
WordPress itself pushed heavily into ecommerce, because it was a more promising area. But at the end of the day, the fact both Automattic and WP Engine earn enough money from enterprise orgs doesn't mean the tech situation here is too different from IBM. Yes, it's big. Yes, lots of people use it. Will it define anything in the future of the web? Not really.
The reason why the Automattic/WP Engine fight has become the tech industry's daily soap opera isn't because we're all passionate about website hosting solutions. It's because the ramifications for open source are dire.
This fight is essentially the legal version of the xz-utils supply chain attack. It's a shock to the system that rocks the fundamental assumptions of open source to its core, here, "you can build your product on top of open source and as long as you follow the license you agreed to, you'll be legally fine."
WordPress.org could self-implode tomorrow, and that doesn't matter in the slightest, because the good will of the legal community (if there is such a thing) that the open source community spent decades trying to build up is being destroyed by one unhinged man. Matt may have ended up doing to open source what Bill Gates wasn't able to do in decades at Microsoft.
There are very good overviews out there, one I linked to in the very beginning. This post is about something else entirely.
What the author actually means is that WordPress isn't trendy or popular.
The problem though with all these managed solutions is that every time you want to setup a new site you need to start paying a new monthly fee. Setup Wordpress (or any other open source CMS) on a web host and you can host as many websites from that one installation as you'd like for the same small hosting fee. If one thing takes off then dedicate resources to it.
Speaking from experience, many customers prefer not to manage their site content themselves. They want to hire a local person they can call and have things done for them on their website. These small web shops will always choose something like WordPress over say Webflow because of cost, control, and not having to worry if the investors give up on the trendy startup you tied your business to.
Open source still has many advantages over the paid services. But just like the Linux desktop, open source CMSs need to work on UX and on telling their story better.
Wordpress will not grow from here on out, it will shrink.
Likewise, that “40%” number never says how many of those websites have been updated in the past year. My guess is less than half of currently hosted Wordpress sites are “active” in any real sense.
I see no reason any anybody starting today would pick Wordpress. The drama is only going to accelerate the inevitable decline.
Those who do not like the SaaS business models for whatever reason (and there are many) still have incentive to use WordPress, a fork, or some other open source CMS.
I don't understand why the social death of WP/Matt M. equals the demise of open source content management in general.
If anything, with AI empowering more people to do technical things I see a bigger opening for open source CMSs going forward. Training data for open source stuff will be all over the web, while training data for WebFlow and squarespace etc. will be largely limited to their support site which might even be gated.
Wordpress is past its prime, has always been a wrapper over a hideous dogshit codebase, has bad leadership, is in a crisis, and will slowly disappear.
It’s natural, and not a big deal. Wordpress pushed out other options, and now they too will fall. Something else will come along.
The article is spot on.
Spot on in observing this, everything else is not so tangible.
It does not really go into what the future of the web will look like, except maybe even more shifting to hosted walled gardens, but that's not exactly breaking news.
Unknowingly(I asssume) the author makes the best argument why Wordpress matter for the Web. None of those tools are free and open source.
It looks pretty free and open source to me, in practice!
While it is possible to use WordPress for free without any paid plugins, most professionals/businesses using WP do so for the extensibility and that, in practice, comes at a cost of either paid plugins or external developer time for custom coding.
Problem is, the market for people who want an easy website doesn’t have a lot of money! Not everyone needs a site these days, and many just stick with a single page “link in bio” type thing. So the money is in businesses who need something more… which is by definition more complex. So Automattic has some massive, massive enterprise clients (including government agencies), but there are other more powerful enterprise tools too. Plus, at that point you’re selling a hosting platform and cloud resources, not managed WP.
Automattic has even tried to make a newsletter subscription like Substack and it didn’t really catch on (https://wordpress.com/support/paid-newsletters/)
On the other end of the spectrum are headless solutions with your pick of framework, let’s say Contentful + Next.js. While far less flexible for a non-developer, the outcome is a rock solid, repeatable, maintainable template. You can democratize editing much more reliably. Although, yes, you need a dev for every customization.
Different solutions for different scales I suppose. I’d love to see the marriage of these two approaches. Something that gives the greatest experience to both editors and developers. I’m trying to tackle this problem in my free time, but it is a massive problem. The more I work on it, the more I recognize why it hasn’t been solved. And admittedly, WP has done a decent job of solving it with its massive ecosystem. There is a need for innovation, but it’s a shame to see the baby being tossed out with the bathwater.
They may not be. But they are competition. It's the railroads arguing trucks aren't competition all over again.
(I am not affiliated with Ghost, just a fan)
If Wordpress can't be easier than a static site generator then they are too complicated for their target market which is less technical people than the crowd who renders their blogs out of markdown files.
Anyway, I transferred us over to 11ty without a real plan how she was going to manage content. She's smart and computer savvy, but never opened a command prompt for example.
In the end, I had her install a Markdown exporter for MS Word, as well as Github desktop, and VSCode for mass editing files and editing front-matter.
Its a learning curve, and she'll still get front-matter syntax wrong occasionally breaking the build, so it's not perfect by any means.
But in terms of deploy, she just presses the push button on Github desktop, and then our free Netlify account handles the CI and deployment automatically. I share my AWS account with her which is where we store our images.
Anyway, like i said, it's not perfect, but I feel like it wouldn't be hard to package up these loose ends into a friendlier UX.
I'm also really liking that our 3 blogs have gone from a $15 a month EC2 instance to around 12 cents for AWS S3 hosting for the images
I agree with you that a CMS is easier to use for nontechnical folks, but Wordpress has so many managed hosting solutions it is clearly too hard to set up for nontechnical users.
The challenge is getting a similar low friction way to deploy content with a static site generator. Github pages is kind of ok - but it requires the person to have a base level understanding of git (commit and push).
To dethrone WP, there needs to be a similar experience. I'm reminded of iWeb of old ( https://en.wikipedia.org/wiki/IWeb ). This isn't saying it can't be done - but the underlying technology isn't the issue.
That gets the 70% solution. For the next 20%, you're targeting people that have some php skill and can tinker with templates and databases but aren't... Software Engineers. A sibling of mine falls into this category. Needs more than a static site generator but doesn't have the time / experience / background to design a bespoke site for the use case. Cobbling together WordPress that was inherited from a previous iteration of the organization with the occasional ChatGPT helping out question is enough to handle the needs of the site and its users.
For the remaining 10%, you've got the people who use WordPress and are familiar with how to make it really work. Static sites are an option for some cases here, but their experience is deep in the WP stack rather than crafting the site by hand themselves. It's easier to modify the templates and tools they've developed to work in WordPress than it is to create a new one. And it gives their clients an easy way to update the content.
For me? A static site would work... and it's in the bottom of my list of things to do to migrate to Hugo or the like. Or maybe even make my own as a personal project (I have a fondness for the old accessible web).
However, I don't believe that the static site generator works for the majority of the WP userbase in its current form (as much as I'd like it to be the case).
> Before you begin this tutorial you must:
> 1. Install Hugo (extended edition, v0.128.0 or later)
> 2. Install Git
> You must also be comfortable working from the command line.
Yeah...this is not by any means easier than WordPress.
That's why I wrote https://github.com/Siilikuin/minimum-viable-hugo way back when - to give myself just enough of a clue of what was going on that I could pick it up from there. Hugo makes an excellent stepping stone between "I know how to write HTML in Notepad" and "I can code up my own little web apps in Django", etc
I would also add that the static site generators have little if any decent templates to work with. Most of them are so basic that to get them to any level that looks good, it will take a lot of work. For instance, if you use Gatsby, then you're going to need to know ReactJS to be able to do change anything or improve on it. This is true with virtually any static site generator.
Contrast that with Wordpress which has literally thousands of really well made, well designed templates that take minutes to install in your CMS. Because there is a such huge variety to choose from, finding one that fits you the best is relatively easy which also means the changes you would have to make would be minimal.
You are correct that if you plan on using your local machine for development then yes, you would have to know PHP and MySQL. Then you would need to know a bunch of other stuff in order to do that which opens another can of worms.
In my experience, this is what also makes Wordpress attractive - the low overhead to get up and running. Most people/businesses I know who used WP, just got a hosting account, transferred their domain and did all their development right on the production version of their site.
"Nobody" except the people paying the revenue to support a $216B valuation. Hello HN bubble, you are all clever and innovative people and I love all of you but you are not everyone.
Also, I hate the future where the web is powered by managed solutions .
> they are also dramatically easier to launch and operate
They are also are easier to get kicked off for no reason that can be appealed to a real human being
Look at MS today. Not so irrelevant.
I do think the original post is right, that WordPress is plateauing and they're scrambling to maintain their hold on a zero-sum market. But, if they can figure out a new direction, they're also in a unique position to lead.
Getting there is not going to happen through dot releases and a squeaky new blog template named for the current year. It looks to me like they need some fresh leadership.
Me too, but it makes launching startups easier and it seems that’s what tech is all about nowadays… maybe it always has been
Oh man don’t look too hard into the medical industry, auto industry, game industry, or appliance industry.
It really seems like they’re all looking for rent-seeking type opportunities, subscription services, and planned obsolescence.
I think Serif (makers of affinity photo) is the only software shops that come to mind that sells a useful product not tied to a subscription. EDIT: oh actually bitwig is too. I think a lot if DAWs don’t require a subscription.
I’m sure there are others, but they’re few and far between.
Yes, there are people who are stuck on the CMS, and people will get roped into building WP sites due to the sheer momentum. It will be around for a long time. But it's not the future.
Exactly those companies are literally pumping out "thought leadership" that people and companies with $$$ are lapping up.
Are you referring to IBM here?
Just start banging out some HTML and check it into version control for a simple website. Add styling as needed. Optionally set up website deployment to happen on commit to `main`. Iterate from there. If you're going to have a lot of content for sure, use a static site generator. Solving problems you don't have creates ones you definitely do have.
I tried rewriting an absolutely dreadfully written, bastardized version of a bootstrap based static website from scratch. After writing the raw html5 pages, I included a bunch of small CSS files and emulated a CSS Zen Garden methodology. It was all done in a few hours and included some more advanced options like a background video with opacity as the hero image on the front page… and the site went from megabytes (!!!) to just under 200kb (including hero video).
It ran incredibly fast, and super responsive. It was easy to read and debug.
Umm, profound insight.
Here is a secret: its even easier to have a social media page and vast numbers of digitally dispossessed have no other option.
What matters for the future of the Web is foremost the degree to which small entities will have any agency.
There is no doubt that Wordpress stagnated, but its conceivable that the future of the Web will be propelled in a positive direction by a visionary project that aims to be equally impactful and empowering.
Well except they sold their PC division years (decades?) ago to Lenovo, so yeah, no-one cares at all about IBM PCs.
I'm honestly not sure whats being said here because frankly im not sure who, before seeing this, was thinking WordPress was anything other than a very widely used piece of legacy software. The reason this chaos matters is because of the number of actual people that jerk's hijack negatively impacts. Further, the post seems to unknowingly prove that actually WP still does have value because alternatives are all closed source.
You can make bad architectures in any language you like!
The only truly bad language is JavaScript. ;-)
In no way is PHP better than JavaScript, and the simple fact that JS runs natively in both the front and back end is astronomically more important than a point-by-point comparison of language features that PHP loses anyway.
https://x.com/Rich_Harris/status/1844744999272935596
Rich Harris @Rich_Harris:
A few takeaways from this remarkable data:
1. Yes, SSR matters. SvelteKit users sometimes get frustrated that it's too SSR-centric, and we have plans to address that in the next major version. But SSR is the default for a reason! Our ultimate responsibility is to end users
Ben McCann @BenjaminMcCann:
Some data from Google (big thank you to @rick_viscomi and @imkevdev !)
Client-side-only SPAs are extremely detrimental to latency as they add a network round-trip before rendering begins.
You can win at LCP by making it harder to build client-side apps
I was at WordCamp 2014 or 2015 where Matt was asked about improving security in WordPress and he literally laughed at the question. So, yeah...
The illusion that everyone's under assumes that WordPress Core is secure, and vulnerabilities are quickly patched and pushed out via the dot org update system.
Plugins are a whole different mess that people often exclude, but plugins are also what makes WordPress the success it is today.
Wordpress core is pretty secure. The problem is people who are using WP just install the core and then add in whatever plugins they need to get what they want to site to do. Nobody ever sandboxes and tests their plugins, they just install them willy-nilly - which is like taking candy from a baby if you're a hacker.
And its already happened last year:
Thousands of sites running the WordPress content management system have been hacked by a prolific threat actor that exploited a recently patched vulnerability in a widely used plugin.
The vulnerable plugin, known as tagDiv Composer, is a mandatory requirement for using two WordPress themes: Newspaper and Newsmag. The themes are available through the Theme Forest and Envato marketplaces and have more than 155,000 downloads.
https://arstechnica.com/security/2023/10/thousands-of-wordpr...
Automattic lost out to these folks and now instead of looking at their own strategic mistakes wants to punish hosted solutions in their space.
Like many VC backed firms Automattic took waaay too much investment relative to actual ability to generate revenue.