WordPress 18 | Hacker News Reader

WordPress 18

ma.tt

301 pointsgmays266 comments

Every time there’s a WP post on HN, there’s the inevitable haters. The ones who say things like “just setup a static site with a CI pipeline”. The understanding of audience here is pretty bad. Your average business owner wants to focus on the important thing: their content. They want to type in a box and push publish. End. They don’t care about how crap (or not) the code is. They don’t care about performance too much. They do want to have the extensibility that is offered. They don’t want to get locked in (cough, Wix, Squarespace).

The thing is, of course you can fk this up. Install Wordpress on crap hosting, install 100 plugins, fail to keep them updated - things are going to go badly. But now think about what happens if you do the same with literally any bit of software, even the hip jamstack stuff doing the rounds right now.

There is good. There is bad. There are people who claim they “do Wordpress” when in fact they can just install plugins and a builder theme. There are people who do it well, with good deployment, repo management, backups, updates.

But the end thing is always the same. Your site owners / editors just want something flexible where they can make content changes. They want familiar. They want flexible.

There are times when Wordpress is the wrong solution. There are bad ways of doing it. But, really, you can’t argue with 40% of the web.

While I agree with this, careful with the “can’t argue with LARGE_PERCENTAGE of GROUP”. Windows is used by 95% of desktop computers and yet is really crap. Or - “eat crap - 75 billion flies can’t be wrong!”

Yeh, sure - but two things come out of this "dominance". One: the enormous quantity of support online. Have a problem with littleknownbutcoolcms? You're stuffed. Have a problem with WP? 10 gazillion people have written about it. Then two: clients want it because it's familiar, they've used it before, they've got another site using it, etc etc.

So yeh, I take your point, but it's the peripheral stuff that makes this important IMO.

Windows is crap because of, not in spite of, it being used by 95% of people and therefore has to be compatible and backwards compatible with everything.

It's not crap, it's good enough. Like wordpress is good enough. That's really how life works, subject matter experts look for perfection while the average person is happy with good enough.

Do you buy the best dishwasher or do you buy the good enough one?

I actually quite like windows and use it on purpose on pretty much all of my computers, except my outdoor work laptop on debian/xfce.

You're allowed to dislike it, but for me windows 10 is a great OS with a thin removable layer of crap sprayed on top, and more importantly it's just the right balance between osx and linux.

To each his own, but you shouldn't be making such generalizations, especially in a message telling people not to make generalizations.

Windows is crap compared to what? I've just switched recently to Windows 10 because I couldn't put up anymore with the lack of quality drivers for Ubuntu. Desktop Linux for me is a lost dream. Keep in mind that I was writing Vala apps not too long ago, I was that kind of a GNOME fan. For me, Linux is only a terminal, a deployment OS.

WordPress is awful, I have had the misfortune of having to write plugins and themes for it and having to immerse myself in the arcane knowledge of WP plugin writing (their docs are seriously crap, I have no idea if I did things 'the right way').

But! My company's website runs on WP nonetheless, mainly because I don't want to spend time on creating a website when I could work on our product. WP allows my less technical employees to tinker with the website and the marketing oriented folks to do whatever marketing oriented folks do.

It runs conveniently on its own isolated VM where people can bikeshed it all day long. If your website is not your primary product, I don't see the point in writing your own stuff unless you got some leftover time you're willing to sacrifice to write it in.

I remember forking the social link plugin (#2 on wordpress plugin page), there was more dead code than code, it was magnificent

Your average business owner also wants his site not to be defaced, and security for Wordpress is pretty bad.

Why? It has auto updates for a long while now. Honestly, I've received auto update notification by emails more often before I see any blog posts or news article about the vulnerability itself. I agree it used to be worse, but this? This is awesome.

I think your comment just proved the OP's point.

I've had customers using WP in the past, and defacement was just part of the routine. It's an annoyance for the business owner for sure, but so are personnel issues, finance, legislation changes, etc.

For most businesses, there are more important issues than having a perfectly optimised and hardened website.

WP, as bad as it is, is 'good enough' for most websites.

That’s the trope. But it’s not true, provided you know what you’re doing, don’t use $50 a year hosting, don’t install a gazillion plugins, and keep stuff up to date.

Which, as I say, is the same for Every Bit Of Software Ever.

Like, consider this: your friends PC gets hacked and he loses everything. Turns out they were running ‘doze 95 and had failed to update or backup. Where is your sympathy? Right, you’re thinking “you’re a fool!”. And you’re right. They failed to look after their stuff, and it created a problem.

Why is this any different with WordPress?

Do it well, and the “security for Wordpress is pretty bad” trope just goes away.

> it’s not true, provided you know what you’re doing, don’t use $50 a year hosting, don’t install a gazillion plugins, and keep stuff up to date.

But... we started this thread with 'WordPress isn't for you, the alternative isn't a static site and a CI pipeline', so it's supposed to be for people who don't know what they're doing, often do use $50 a year hosting that comes with 'cPanel', keep installing different plugins until one kind of does what they want and then leave all the others, and never update anything until something breaks/they want something new and they start installing plugins again.

I think WordPress is a fantastic piece of technology / achievement. Just like Windows is as well. And just like with Windows, as you say, the experience of the person setting up the system is what matters. So, of course there are some super easy to use, yet flexible, excellent and satisfying WordPress setups out there. And, for complicated, large sites, there is probably hardly anything that beats a good WordPress setup.

But, then I look at my neighborhood and the 1,000 small-business WordPress sites, with blogs that have been abandoned in 2015, should be one-pagers and not seen an update since because their owners don't understand WordPress and also can't afford to pay someone for every small change and now they think the whole Internet is like that (40% as you say), and changing that is going to be one hell of an uphill battle.

I don't have to hate WordPress or knock their efforts to see ways to improve things with more modern approaches.

I like your PinkPigeon website builder! Very nice with the visual themes and modularity. And the 'Special Thanks' document is a great idea, appreciative and humble. I'm missing a vertical scrollbar though (on Ubuntu Firefox). Edit: somewhere I can follow you, like Twitter or something?

Yup, Twitter: https://twitter.com/PinkPigeonDigi

Or, if you send me a message via the contact form and tick 'submit', you can sign up to our newsletter at the same time.

I'll split that out into a separate newsletter signup in the next few days.

Thanks so much for the kind words, much appreciated!

i'm not hating wp because good use of hate is scarce, but wp frailty is something even win95a is jealous of

> But, then I look at my neighborhood and the 1,000 small-business WordPress sites, with blogs that have been abandoned in 2015,...etc

That's not a failing of WordPress. That's:

1. A failure of how the business owner just couldn't be bothered. Wordpress, Wix, Squarespace, there's a fairly good chance they'd fail at all of these. They won't pay for the design, marketing etc etc, and the reason why?

2. their business doesn't need a website e.g. local plumber or the only game in town for a specific vertical.

3. They're on their local Facebook groups and don't need anything more.

These folks also abandoned Yellow pages, Thomsons etc years ago but yet they survive through word of mouth. The internet and these other marketing services can't compete with that (especially outside certain confined bubbles). The local plumbing and heating guy in my area gets on fine without all the paraphernalia of "being on the web", as do all other trades.

Not everyone needs a "website", especially if they've a well developed funnel of incoming business that, by now in 2021, they still get on ok without needing to be internetized.

What does get these folks business without websites is other folks recommending them "on the internet", via local Facebook groups or whatever. Why would you waste time having a website when FB funnels the business to you, it's already job done.

>Every time there’s a WP post on HN, there’s the inevitable haters. The ones who say things like “just setup a static site with a CI pipeline”. The understanding of audience here is pretty bad.

It's almost like HN has a large community of people who find setting up a static site simple and would find the technical constraints of WordPress offensive. People for whom WordPress is the wrong solution.

So now the question is - what should those people say about WordPress? Obviously they have no reason to praise it, so they will damn it, if you say they should not damn it because it is not for them then this raises that old HN question - why is this here? It's here because it would be of interest to Hackers, but it seems hackers would be the people most likely to dislike it.

It's here for the 15 minutes hate.

"It's here for the 15 minutes hate. "

Most likely, but Hackers Spirit for me is also finding the right tool, for the right job.

Meaning for me, I allmost use vanilla JS and html only, because that way everything is under my control behaving exactly like I want it.

But I clearly see, that other people need a more convienient tool, to enable them to get things done, with their skillset. Why should I hate that?

Oh, but about hate - I really do hate PHP, so I could probably chime in with the choir.

But I know my hate comes from having to take over a half finished webproject, long ago, with badly written PHP and I was new to it. I never touched PHP ever since ... but I am aware, that other people get stuff done with it, so good for them.

> with badly written PHP and I was new to it

that's been a combination of my experiences with pretty much every other tech as well.

I've done php, java, perl, python, javascript ruby, asp and poked at a small cold fusion project over the last 20+ years. Every one of those had baffling parts which, after some time, and after reviewing it with people far more proficient than me in that target stack, I learned it was "crap code". "Yeah, they didn't know what they were doing." "That's n00b code." "Garbage. How did they get hired?"

I've also seen some good 'clean' projects in some languages, but not as many. I guess I can not, for the life of me, figure out how someone looks at some bad PHP projects and writes off the entire language, yet if I do the about Python, for example, I'm somehow close-minded, or a n00b, or... "just not ready for it yet" or... something. Have had conversations like this at tech conferences (back when those were a thing!). Seems like projection going on, but maybe there's something more.

FWIW, it's often not specifically the language itself which is a stumbling block, but the entire ecosystem around a stack. If there's not a good, usable patterns and tools for easy testing, or integration between FE/BE, etc, I don't care how "tight" or "sleek" or "elegant" a specific language is; it's going to be a pain to deliver a full project with.

Well, what I really dislike about PHP is the syntax. Weird special characters.

And since I care about that and there were alternatives - yes, I just wrote off PHP entirely ;)

> Weird special characters.

After you've worked with perl for years, none of those sorts of criticisms against PHP seem to hold any water with me. ;)

I never touched perl, either ;)

It's a fine product but it's obsolete in everything from tech to features. It's fine to say there are better options now when that's actually true.

The comparable replacement for Wordpress isn't a static site but something like Ghost: https://ghost.org/

Old and obsolete are not synonymous. WordPress will be obsolete when people no longer find it useful. There is little sign of that happening.

Obsolete is not defined by people finding it useful. In fact I called it a fine product.

But it is outdated and I offered a better alternative as an example of why Wordpress receives so many negative reactions at this point.

What is the suggested ecommerce play for Ghost?

Are you aware WooCommerce is an ecosystem in itself?

Curious, when was the last time you used WordPress? The WordPress editor these days is a React SPA, and the new block editor is a leapfrog over traditional MS Word-style text editors.

> “just setup a static site with a CI pipeline”

Yeah I do do this.. for my personal site. If I did this for a client then they wouldn't be a client haha

Truer words.

As you rightly point out, WordPress has a sweet spot of providing control over content to the customer.

The problem I see is that, in many instances, WP is pushed by dodgy developers running a template for what is a static site. This is easily the case for 90%+ of the small business sites I see. Many are never changed.

WordPress is a much larger attack surface than static hosting. There are fairly obvious ways of automating using WordPress to generate a static site that will be faster and more secure. I believe this is generally referred to as headless WordPress. Why does this not happen more – or do I just not see it when it does?

Did you ever do web development for customers? You build the website (10% of the work), and then the rest comes.

"Can you put our holidays on the main page?", "Can you adapt this text?", "We have a new image of our main shop, can you update it?", ...

I've seen plenty of people building a website for someone and thinking they made good money, but after all the requests that come afterwards, it's not worth the time.

On the flip side, I've also seen Wordpress specialist consultancies grow large businesses out of this inevitable influx of change requests and the revenue that comes with it.

These consultancies strike the right balance in terms of knowing their product, managing expectations, and knowing when to say "no".

It's created lots of jobs in South Africa for what that's worth. Can't speak for other countries.

> It's created lots of jobs in South Africa for what that's worth.

WooCommerce began as a plugin by a wordpress theme company in South Africa (WooThemes). And then they were aquired by WP/Automattic in ~2015. A really cool case of pivoting.

Last two WP updates I did were essentially downloading the site with firefox's "save site as", running prettier on it, and setting up a basic netlify for what's now a static site. HN is quick to talk about "the right tool for the job" about WordPress, but WP also has great branding which often makes it the choice when it's _not_ the right tool.

imho the best part of WP is the dashboard stuff, so headless WP is a good choice. It's also frequently hurt by PHP's heavy push on inheritance, which can be at-odds with what is actually a very good events-based architecture in WP. Sad part is that the very good events-based architecture is so overlooked, most people who have spent time developing in WP aren't even aware it has events.

If you're talking about WP hooks, basically every single decent WP developer has not only heard about them, but uses them extensively.

Maybe, but what percentage of those branding themselves "WP Developer" are decent? I've had to unravel a fair number of plugins where they'd crammed everything behind a single action , or under one big theme.

> Your average business owner wants to focus on the important thing: their content. They want to type in a box and push publish. End. They don’t care about how crap (or not) the code is. They don’t care about performance too much. They do want to have the extensibility that is offered. They don’t want to get locked in (cough, Wix, Squarespace).

That's IMO just a wrong understanding of the average business owner. Wordpress is much more useful to web agencies with the necessary skill set to heavily customise it than to the average Joe SMB Owner.

To them, Wordpress is just as much a black box as Wix or Squarespace is, and all the web agencies using Wordpress to make websites for that demographic modify it so much with so many mutually exclusive plugin pipelines that getting their website from one to another often involves a complete rewrite anyway, if you want meaningful changes done.

And for a long time, it was just the cost of doing business. But now? Nowadays, Squarespace et al. are probably a much better fit for them. Effectively the same vendor lock in if you're not a web agency yourself, but with much less overhead involved.

I agree with both you and OP.

That’s why I’ve completely changed my position on who should be using what tech to build sites. There is no one-size-fits-all solution.

I now recommend clients three options, based on the type of business:

1) If they’re a local small business (eg. Restaurant, flower shop) or solo creative business like a photographer or videographer, just do Squarespace. Wordpress is way too complex for their needs and they will easily shoot themselves in the foot.

2) If it’s a Saas company, tech startup, design agency, or enterprise, I strongly strongly recommend Webflow for their marketing site even though they think they want JAMstack. When the marketing team starts pulling dev time to redesign parts of the site every week, they immediately will regret the JAMstack decision.

3) if it’s a media/content business, always go Wordpress. Their employees will already be familiar with it, and the collaborative editing and SEO experience cannot be beat for publishing content on a daily basis. Combine this with plugins for newsletters and social integrations and the extra hassle of Wordpress is worth it for them.

We've recently switched out of Wordpress to Webflow and hit upon 3 snags that soured my opinion of it, though now we're "done" on it these are less of a big deal but I'd still probably have just stuck with WP with a reimagined and cleaned up theme/ACF for custom data:

* Only one person can edit code at a time - this caused us to have to split who was doing what and we even ended up losing work when someone overwrote someone elses changes https://wishlist.webflow.com/ideas/WEBFLOW-I-39 (this goes back to 2017)

* Tracking changes in VCS easily (https://wishlist.webflow.com/ideas/WEBFLOW-I-381) another from 2017, we tried out flowgater though in the end have just settled for exporting the code as is and then just committing it by hand.

* A max limit of 10k characters (edit - as pointed out in comment, that's per file, I think IIRC we had all our JS in one single place which is how we hit this our side) in site code (https://forum.webflow.com/t/embed-html-10-000-characters/443...) - this caught us on doing some embedding, we worked around it in the end and not a massive deal but still felt a little annoying having to work around it.

Webflow is definitely in need of more features (with their last funding round I bet you’ll see them soon). But even with those hiccups, you can just move so much faster inside Webflow to update or completely redesign a site than with Wordpress.

For example, I work with a Saas company that has their designer throw together landing pages for competitor comparisons and integrations in like a day, and they change them weekly based on testing.

You just can’t do that on Wordpress at that speed without making a mess with heavy plugins and bloat.

You can do that with wordpress plus a site builder.

There are sure builders that are as good as Webflow, but without losing WP power

> A max limit of 10k characters in site code

A brief investigation says it's 10K characters per file, but still how the hell does a web site enforce that? Sure seems arbitrary and silly.

> "There is no one-size-fits-all solution."

Absolutely. Unfortunately, tools create confirmation bias, and given WP's market share those voices are collectively the loudest.

The point being, the general perception is WP is *always* the solution for *every* need. And Automattic and other leaders with the WP "community" don't make much of an effort to change that.

I like WP. I use WP. ButI find the one size fits all attitude very annoying and ethically suspect.

At one point, we used Wordpress for everything from websites for small restaurants to collaborative desktop publishing platforms, but none of this turned out a great idea longterm.

We still use wordpress occasionally, but no longer for those customers.

>1) If they’re a local small business (eg. Restaurant, flower shop) or solo creative business like a photographer or videographer, just do Squarespace. Wordpress is way too complex for their needs and they will easily shoot themselves in the foot.

This (with a slight modification).

I built websites back in the Web 1.0 days. I know how to use HTML, CSS and a bit of javascript. When I needed to build a "brochure" site for my professional services business, I could have done it from scratch, I could have waded into the convoluted and plug-in dependent mess that is WP, or I could go to a simple tool like Squarespace or Wix. I am capable of learning anything I needed, but the real question is: when I need to update a page, do I want to relearn all that stuff I forgot because it isn't my day job?

When comparing Squarespace to Wix, I realized that Squarespace doesn't have backup and restore capability. That is a mindblowingly huge hole to me. So I went with Wix. Yes, I am "locked in" to Wix. If I wanted to go to another solution, I could pay someone pennies to replicate my existing site.

Contrast to my experience with WP. I inherited a WP site with the non-profit I work with. I ended up getting them to pay godaddy to manage the site and the SSL because I don't have the time to deal with it. I dread the day when the next WP upgrade breaks the janky WSIWG plug-in (elementor I think), which already doesn't work properly on one of the pages, so I have to edit the HTML. As soon as our long contract with godaddy expires, I am paying someone to replicate the site in Wix.

Like you pointed out in #3 above, I guess there are some applicable use cases for WP, but anyone who doesn't need more than a brochure site is ill served by that choice.

> If it’s a Saas company, tech startup, design agency, or enterprise, I strongly strongly recommend Webflow for their marketing site even though they think they want JAMstack

Yeeeah but no. Companies (especially those ones) can have plenty of good reasons to need more control / custom code / integrations / etc. I think a lot of people don't realize how crippling and limiting committing to Webflow can be, and how much you end up hacking around it. Raises hand

https://www.plasmic.app gives you the same (actually, a FAR better) visual editor that plugs into whatever codebase you want. More people need to know about them.

Plasmic shows more promise as a solution for designers on product teams to take over CSS styling on the product itself (eg. your SaaS app with a react frontend). But I think they still have a long way to go on the visual editor...last time I checked their editor didn't even work on Safari.

Using Plasmic + JAMstack on a marketing site would be the worst of both worlds IMO. All the static site hassles with none of the stuff that makes static sites fun for me.

It gives design and marketing the ability to edit landing pages, but they still have to bug the engineers for bigger changes to the site structure.

The biggest advantage to Webflow is its super versatile CMS, and separate simplified editor for content teams. Our engineering team literally never has to touch anything on our main domain. Marketing/Design completely owns it. Our engineers focus entirely on our product.

Like all developer-centric tools coming out of the JAMstack ecosystem, Plasmic seems to ignore the existence of content marketing and SEO (Wordpress's writing experience and friendliness for non-devs is why they won the internet), which is the entire point of having a marketing site in the first place.

Hello! I work on Plasmic. Thanks for the feedback (and thanks mosr for the shout-out!). Webflow is a great platform, glad you're on it - we're fighting the same fight.

Just wanted to share that our main use case actually has been for marketing sites (apps are a minority) and giving content editors autonomy. We do spend a lot of time on page performance, SEO concerns, etc.

Publishing new pages and other site structure changes (and even publishing new sites) does not actually require bugging engineers. Letting marketers/designers own the site and freeing up developers is our entire focus.

We are also working on a simplified content editing mode that restricts design changes.

Appreciate your following along with our progress, and hope we can continue hearing your feedback - that's how we make the product better for you.

Agreed. I know a number of small business owners who prefer Squarespace for the ease of use and integration, and they've never thought about lock-in as a dealbreaker. I know some entrepreneurs and nonprofits who prefer WordPress for the FOSS and the ability to customize. A little while ago I set up a personal WP site for a family member and they got frustrated with having to install updates seemingly every time they logged in to the panel.

^ as a former WordPress developer, this is where I'm at. If anyone with a small budget wants a website for their small business or project, I would not suggest WordPress unless they are working with an agency they can afford who specializes in it. SquareSpace, Wix, SpaceCraft, etc are much easier to work with.

Also agree re: lock-in. This is actually improving somewhat with Gutenberg, but transferring data from a WordPress site to another platform is still usually going to be a giant pain in the butt unless your content is extremely simple - no metadata, widgets, forms, custom blocks, etc.

> Every time there’s a WP post on HN, there’s the inevitable haters.

To be fair, every time there is a WP post on HN, there is also always the inevitable comment complaining about the "haters" and to just think of the "business owners" tangent.

And then there is comment (like yours) complaining about comments complaining about haters :P

Then there is my comment - now it is like the movie inception :P

It would be nice if we all just used those tools/software that works for us and simply ignore the rest.

> There are people who do it well

What are the best resources that describe how to "do it well" in Wordpress? I've made several attempts to google, but due to the enormous popularity of Wordpress, and the low technical level of most of the users, the signal-to-noise ratio is incredibly poor.

Check out https://roots.io/

From an engineering perspective it is absolute garbage. There is practically no separation between View and controller layers, which is a big part of the reason otherwise simple cosmetic plugins are security and stability vectors.

That's why a whole dev ecosystem on a modern stack exists: https://roots.io/

At that point why not just make something properly designed to start with?

At the end of the day, regardless of other usability and adoption merits, it’s just a really poorly designed architecture.

> At that point why not just make something properly designed to start with?

Of course that would be ideal development-wise. Business-wise, I doubt most agencies or small teams are going to rewrite W3 Total Cache, Yoast, WooCommerce, Elementor, ACF, Gravity Forms and all the top popular plugins... not to mention all the battle-tested security options.

This is classic sunken cost fallacy, the reason to stick with a platform shouldn't be it has the most bandaids covering deep, deep, problems and all the investment that has gone into bandaids can't be given up.

My complaint about Wordpress is always about how frequently it gets attacked. I have had multiple wordpress websites hacked even after keeping them regularly updated and even after having special "security" plugins such as Wordfence installed, the application firewall installed etc.

It puts too much of a burden on the site owner from a security perspective.

Quite frankly, Wordpress leaks like a sieve. I wish this wasn't true because it is really one of the best/most accessible CMSs around. If they fix the security issues, it would just be wonderful.

Stop using shady shared hosting companies, self-host everything and harden php/apache/mariadb deployment not just Wordpress.

Wordpress is pretty secure but putting everything in a shared hosting with lax security like dreamhost or hostgator et al its asking for trouble.

I think that's the part most people overlook.

I have many wordpress sites, some large, some small, and only 1 was ever hacked and it was on a shared host 10+ years ago. Today I have sites that I haven't updated in 5+ years, no hacks no problems. Good wp depends on good server config and good wp setup.

I am not using shared hosting. It was hosted on a Digital Ocean box using the recommended security updates, settings, file permissions etc and with as few plugins as I could get away with. I also had the "pro" version of Wordfence installed with active monitoring. Yet it was hacked.. multiple times.

>really, you can’t argue with 40% of the web.

This is because WordPress is the equivalent of PHP. Sure it can do the job but there's a very big BUT in there, always.

There is no but. It can do the job period.

And no, I don't use it either.

If you can see no BUT in WordPress or PHP I doubt you know much about them and their history (especially since you wrote you don't use them). There's a massive list to be written about their bad sides. Mind you I use PHP so this isn't coming from some hater BUT having used it in many versions I can definitely see many huge problems that mostly have not even been acknowledged until lately. As stated, it can do the job BUT.....

I used both Wordpress 'versions' - free on .com, self-hosted on Amazon. It's GREAT! The dashboard that gives you all information you want/need, how easy it is to edit posts and pages switching from desktop to mobile, having a widget on mobile to see visits, etc...

The only reason I switched to Hugo is because I wanted custom domain + javascript (for adding some embeds that wordpress doesn't support by default), and I didn't want to pay 8 bucks (I think 10 CAD) for that on .com, and didn't want the hassle to host on Amazon anymore.

But I surely miss being able to have all my drafts available everywhere, and publish with the click of a button.

EDIT: I also miss that all other bloggers from my circle except another one, all use wordpress, and follow each other on the wordpress reader, and comment using wordpress comments. I'm likely missing a good chunk of views and comments because of that

I can totally understand where you are coming from but that doesn't mean I don't also see the big BUT there. I'm sure we can agree neither PHP or WP are perfect, especially historically.

I have a few WordPress websites. Loading them is as fast as it can be thanks to a good caching plugin correctly set up, and this is what I care about. No JavaScript except for things that are faster with it. 100% on Page Speed tests. They are easy to maintain and people not into computers can post things.

I'm attracted by static site generators essentially because I like the idea that no dynamic code is run when the website is visited, but WordPress almost gets you there after the first visit. Plus, you don't have to regenerate the whole website for each change, WordPress generates pages "lazily". So, maybe WordPress can be lighter in some cases than a statically generated website when considering everything if setup this way? It still requires a potentially stronger server than when everything is static.

WordPress is like Craigslist in that software engineers hate it, but everyone else happily uses it, and both are thriving businesses providing a sustainable livelihood to their employees and owners instead of trying to take over the world. You should be so lucky to invent something as well-loved and as useful as WordPress, and to stick to your principles for as long as Matt has.

The headaches are simply not worth it. No designer should be using WordPress for client websites in the age of squarespace/wix. It is just irresponsible to build something for a client that is going to break itself within 1-2 years max.

It completely depends. I think building a site on Squarespace/Wix can very easily be considered irresponsible.

edit: For Wordpress only install reputable, necessary plugins. Turn on auto-update. Keep backups. Things should be just fine. Now you aren't locked into a single platform and you can extend your website as much as needed (unlike Squarespace/Wix which severely limit what you can do).

> Turn on auto-update

Having been paid quite well to migrate to and from WordPress, and to help maintain the WP sites of major publishers, I have to say that this is questionable advice. In most cases in my experience, the publisher has employed third-party providers (at huge cost) to develop frameworks and custom plugins that can easily be broken by automatic updates.

Updates need evaluation and testing in staging sites before they can be rolled out. For publishers, the first install of WP is free - everything after that costs plenty.

Like you said, "major publishers" will have those problems due to custom code. Not the sort of websites where Wix or Squarespace would have ever even entered the conversation. You need to keep context in mind.

And not so major publishers, including hundreds, if not thousands in the top 10,000 sites. So it isn't just the Washington Post, et al. This is why I intervene on generalized advice to enable auto-updates on WordPress. I have seen it cause publishers a ton of money, and have lost a number of weekends when called in to perform damage limitation.

'Make sure you keep everything updated!' has a ton of caveats and pitfalls, that's just the reality.

How do you know if a plug-in is reputable without being neck deep in the ecosystem for a year talking to other devs?

I just make sure the install count is something high (if it is lower than say 100k, then probably there is a better plugin out there to do the job) and it has been updated in the past couple months, and it is compatible with the latest version of Wordpress. All that info is directly visible while you are searching for plugins within Wordpress itself on the search page itself. And typically I'll also do a Google search for whatever problem I'm trying to solve and make sure it's recommended by at least two people.

Btw, I'm a senior full stack software engineer. But for my digital media business I operate several Wordpress websites. I don't talk to any Wordpress devs and do any custom Wordpress theme or plugin development (aside from what is available in the GUI and extremely minor CSS tweaks).

Download count and reviews aren’t necessarily trustworthy though.

https://krebsonsecurity.com/2021/05/using-fake-reviews-to-fi...

WordPress is an open source project, immensely supported, fully customizable to it's core and that let's you port your website to virtually any hosting service out there, that let's you export all your data however and wherever you want.

Wix and alikes are proprietary software, that lock in users on their services, have limited customization options, don't allow you to edit code and have little to no export options.

Whoever pays someone else to have it's website built on wix or alikes is getting a very bad deal, due to the above mentioned restrictions (compared to WP) they'll end up very dissatisfied in few years if things go well (meaning more traffic and more needs).

Source: experience. As a developer I feel the frustration coding for WP, but at the same time it's the best compromise to provide a CMS that the average user can manage

Scaling is famously a good problem to have. Unmaintained wordpress instances are the internet equivalent of toxic waste dumped in the river. I agree that open standards are important, but IMO it's irresponsible to leave someone with a website hosting setup that doesn't have a qualified person responsible for ongoing maintenance. (Fully managed wordpress hosting from a reputable, established company is one possible answer, of course).

Is it impossible to make a secure CMS? I'd think step 1 would be to not have the admin stuff with the same access as the public website, and step 2 would be to have a way to sandbox plugins, but are those things possible?

Step 1 is possible nowadays (with SNI) but was hard until recently, and I don't think practices have really caught up with that yet.

Step 2 is difficult to even define properly. What is a plugin supposed to be able to do, and what is a plugin supposed to not be able to do? If you can answer that question precisely, then you can solve the problem, but if you can answer that question precisely then you probably don't need plugins in the first place.

I understand wp is more flexible and from a development perspective is superior. However it's inferior if you are just hooking up some small business with a website they can maintain and they don't want to pay you for ongoing maintenance.

>Wix and alikes are proprietary software, that lock in users on their services, have limited customization options, don't allow you to edit code and have little to no export options.

If you are building a simple brochure website, none of this matters.

i am going to strongly agree.

i set up my first real blog over a decade ago. i picked wordpress, because that seems to be the default, and i didn't know any better. oh my god, i hated maintaining that thing so, so much. it pretty much killed my desire to blog, as a matter of fact.

i recently started another blog, and this time i used a static site generator. i wouldn't recommend such a thing for non-technical people, but if you are willing to get your hands a little bit dirty, it is a hell of a lot easier to fix than wordpress ever was.

That’s a pretty bad comparison. The entire reason WP is a better solution than SS and Wix is it’s extensibility, data portability and usability.

Plus: what headaches? Do it well, keep it updated ( uh, yeh, like literally ALL the software, ever ) and it’ll work flawlessly for years.

Who is paying me to keep their site updated?

It's so interesting to see just how successful WordPress is. Among developers especially. I look at small business websites all around where I live and it's all WordPress and every single one was made by yet another wordpress-shop that does nothing but setup WordPress for people. It's a huge ecosystem.

And you know what I have learned by talking to these businesses? It's that WordPress is still not easy enough to use for non techy people.

Nobody wants to touch their website, out of fear of doing something wrong. So they have to go pay someone to do it every time.

I'm running pinkpigeon.co.uk, which tries to actually make websites easy enough to use for everyone. But I am actually running into a problem where people expect all websites to be like WordPress and the first thing you hear is "oh I don't want to learn another thing". It's only when I show them the system, that they go "oh wow, that really is a lot easier". The problem is trying to convince anyone of this remotely via the Internet. Much easier to do in person.

I have the utmost respect for WordPress and used it much myself, but at the moment it's challenging to have to be compared against it when simpler solutions exist.

What baffles me is that huge portion of those small business websites have absolutely no need for Wordpress or any backend altogether. At best they need some static site generator and a CI/CD pipeline maintained by the provider to rebuild the site if it is an actual blog, which is not that common.

How are the small businesses supposed to update their websites, if there is no backend? You can't expect them to write HTML.

Not sure if suitable WYIWYG editors are still around, or if Word exports to HTML have become usable.

" or if Word exports to HTML have become usable"

I doubt there is beauty to be found.

And websites are a lot about beauty.

Businesses, even very capable with web dev, update their websites very rarely. Businesses want to update the content, which is different. Static site generators are built exactly for that.

Can I open an app on my phone to create/update a post? Can I Use my phone to create another user so a friend can guest-post on my blog? Can I send an email to a certain address to publish a new post?

Nope

Now you just take things a blogging platform can do and try to morph that into an argument against static site generators. The very point of static site generators is to drop a whole host of functionality.

How often does a construction/handyman "online flyer" webpage need a change? Once a year, maybe. The primary point I am trying to make is that websites of a lot of small business (obviously ones that are not webshops) serve more or less static content and therefore have no need for a backend at all.

You have a food blog where you publish recipes/reviews daily? Surely a blogging platform is a much better fit for you. You have a restaurant page where you publish contacts, coordinates and a menu which changes every second year? Well, maybe a static site generator is a better fit.

You say that, but what if they want to add GA? Or an image gallery? Or a chat box?

Those are just some examples that with Wordpress, it's a click and pretty much plug and play. With the static site generator you're proposing, you quickly need to do coding yourself.

If you buy website development service you do not necessarily care whether the provider writes custom code for WP or static site generator.

Why do you need a backend for GA or image gallery?

Because you don't want to keep paying ongoing fees for new development when it could be as easy as installing a plugin.

But image gallery does not need any "additional development". All it needs is frontend code to display the gallery nicely, some storage mechanism and a "manifest" of sorts with relevant metadata of the gallery.

If that is a photo-sharing website, then yes, proper backend is needed. But for a portfolio website of a photographer, which gets updated once a month at best, static manifest will work wonders.

I love the idea that "frontend code" is somehow not "additional development". Have fun with your "manifest" and "metadata".

I equally love the idea that hiring a web agency to build a wordpress template is somehow entirely different thing from hiring a web agency to build a not-wordpress template.

The frontend code which eventually renders the image gallery does not care whether fetched image list is produced by PHP code or is served by http server directly from filesystem.

> But image gallery does not need any "additional development". All it needs is frontend code to display the gallery nicely, some storage mechanism and a "manifest" of sorts with relevant metadata of the gallery

I can't tell if u missed a /s or genuinely think that isn't additional dev.

> At best they need some CI/CD pipeline maintained by the provider

My business requires a what now?

Missed a part. A static site generator with build/deploy happening on provider's infra.

> A static site generator with build/deploy happening on provider's infra.

“Say what now? Where do I login to update a photo on the site?”

In software engineering, CI/CD or CICD is the combined practices of continuous integration and either continuous delivery or continuous deployment.[1]

In short everything has to be continuous, otherwise you'll stall and die. Now go buy some CI/CD from the shop.

[1] wikipedia

Cool. Now imagine I’m a business owner that makes and sells my own socks and happen to need a website to do so. What you’re talking about would still make zero sense to me.

I should have added it was sarcasm :(

> At best they need some static site generator and a CI/CD pipeline maintained by the provider to rebuild the site if it is an actual blog, which is not that common.

What most people want is to have something that works, where they can focus on their content and not on your tools. Wordpress works for people.

If they get a CI/CD pipeline maintained by someone, how do they create pages, get updates, fix those pesky security issues, write a blog post on their mac that when saved automatically gets posted? How easy is it to apply a theme to your CI/CD pipeline? Your tools work for you, but are you going to maintain them with the same level of features as Wordpress?

Oops, it's holiday time, will you write a greeting card plugin for your CI/CD pipeline for them to purchase cards on their website and send site-specific cards to their email list by Monday? No? Sorry, they just switched off of your CI/CD pipeline to Wordpress.

When moving to a different hosting provider, people want to conduct business not learn a to use a new CI/CD pipeline to match the tools at the new hosting company.

What about when you get tired of updating their site generator code for them? They will need to start over, and they won't pick another CI/CD tool that's unmaintained and is expected to get horribly outdated after a year.

People go into business in order to conduct business.

And they will want to automatically publish to twitter/facebook, integrate shopify on the site, some forms, etc...

Wordpress has 1 click install for all those things as plugins (or work out-of-the-box)

"Joe's porchbuilding" does not send automated greetings. They want a website that is discoverable on Google by new clients and can display their pitch, everything else is negotiated on phone or by email anyway. You only need to worry about security issues if you build such a site on top of a dynamic platform. The updates and security issues disappear if you serve factually static content as static content.

Static site generators are bad for dynamic websites, but that does not mean you have to try and fit a whole dynamic platform into a static website. Different use cases, different tools.

> "Joe's porchbuilding" does not send automated greetings.

And, that's false. Unless you make them run one of your restricted platforms that doesn't have the features they want and need for their business.

Wordpress runs 60% of CMS websites. 15% of the most viewed websites run Wordpress.

> You only need to worry about security issues if you build such a site on top of a dynamic platform.

Do you actually believe that there are no security issues with static websites?

Thank you for the conversation, but I'm going back to running my business.

>Nobody wants to touch their website, out of fear of doing something wrong. So they have to go pay someone to do it every time.

i think it's a mistake to believe that this is a problem to be solved, or that the reason people don't want to touch their website is just a "fear of doing something wrong". it's not a question of how easy the admin panel is to use - the vast network of people willing to handle wordpress sites is a feature of wordpress. a business owner doesn't want to deal with their website, they want to pay somebody else to do it for them. tech people like self-serve solutions that they can log in and administer themselves, but there's a whole lot of people who really don't want that - they want to just call somebody up, explain what they want, and have it taken care of. the reason wordpress runs 40% of the internet is because it delivers that experience.

And here I am, making a website with an image and boxes around links.

I can attest to that the fundamentals of Wordpress are truly awful. Will never return and can not recommend it for anything business related. It is a damned fine blog software though.

Define business though, most business do fine with wordpress alone. Then there's woocommerce with about a 20% market share iirc...

I hate WordPress because it has bad DX. But why would I hate CL, it's a simple, useful site with minimal JS.

What do you find bad about WordPress DX in particular?

I have not worked with it recently. I recall needing to rewrite a URL and it not being possible with .htaccess - which would have taken a couple minutes. Instead I had to learn the WordPress URL rewrite system.

> WordPress is like Craigslist in that software engineers hate it, but everyone else happily uses it

Why would a software engineer hate Craigslist? It's one of the most streamlined, responsive web sites I interact with on a regular basis. No third party scripts. It's built to purpose and it works very well without any fuss.

I suppose I have some minor gripes - with noscript it looks awful and they killed off their RSS feeds a while back. But hardly enough to inspire hate!

and to stick to your principles for as long as Matt has.

Technically he subtly shifted from supporting free software to open source. Just pointing that out, not making a big deal about terminology or anything as WP is still GPL'ed.

I’m an engineer and I always recommend people to use wordpress if they want to blog.

I still remember the first Wordpress installation, I was in high school, China still had decent connection to outside world, I rented my first VPS, in west coast US, even tho the domain was eventually blocked by the great firewall after a year due to some personal thoughts I posted. But I learnt a lot from it, it’s the first time I tried to read and modify actual code to make things work the way I wanted. Oh and there was a huge community for Wordpress in China, they’d hold those events in top universities in China to talk about Wordpress, PHP, web and technology in general, always wanted to attend. I think things changed now. Exchanging links to each other’s blog was also a huge scene, I still know some of the people, some of them are still blogging, about their lives, kinda like diary, from high school to graduation from college to getting married and having kids.

From what I see PHP is still very strong in China. There are even frameworks like Swoole, with majority of contributors from China.

I wouldn’t doubt it about PHP, it’s just the fact people gather together to actually discuss and appreciate technology for the sake of it, not wanting to sell something is becoming rare. Or it could be my perspective have changed, and I didn’t realize people were selling things.

And the whole blogging community was so great, people exchange links not for SEO, but from appreciation of content or styles.

I handled the inherent security risk of WP's plugin architecture by meticulously monitoring the plugin-code/plugin publisher, Firewall, File changes, 2FA Auth etc.

I handled the performance drawbacks with best hardware, cache etc.

I even accepted the premium for above activities. But random database errors was the final nail in the coffin.

After a decade of WP usage, I have ditched it for good. My current web stack for the past 2 years has been -

Simple websites- Vanilla HTML, CSS, JS.

Blogs - Hugo.

Complex Web Applications with CMS - Custom Go framework.

I understand that none of the above is applicable to someone who doesn't have web-development experience, Even in that case website-builiders[1] are a better option than WP. At this point, the only reason I'm not nuclear on WP is because I believe there's a thriving economy based on WP especially from developing countries.

[1]https://startuptoolchain.com/#website-builders (Disclaimer: Mine).

> there's a thriving economy based on WP

a lot of digital marketing/service company use WP to quickly setup a website for their client. my current employer's website is a WP site build by a third party.

Yep, But I really don't know whether that's for good or bad overall.

Good for that 'digital marketing/service company', But there are better options for your employer in terms of cost/performance/security; Better options have been available for over past 5 years.

I’m curious what options you are considering here that are comparable and don’t take 10x more technical skill and time? Sites like Squarespace or Wix? Or Webflow?

Correct, Those are the ones I've linked to in my original comment.

Of course website-builders do come with the caveats of vendor locking, But then again the typical relationship between 'digital marketing/service company' and the client mentioned in the parent comment is often much worse.

this is just my experiences from years of working in small to mid size company.

company tend to see company's website as part of marketing thus marketing dept usually have the final say and they tend to go with what digital marketing/service company recommended.

edit: not sure if that's a blessing but at least I don't have to deal with marketing people going over some small details like color...etc. the website that I build is web application for internal use.

I can program and am very familiar with web technologies. For my own project https://rpgplayground.com, I decided to go with WordPress, and I haven't regretted it for a single day.

It started out as a landing page and a blog. But then I needed a forum for my users -> install BBPress and done!

Was getting some spam comments and posts, so installed some anti spam plugin and captcha plugin.

Seemed my users want to post their own games on my platform, so installed the BuddyPress community plugin.

Wrote my own plugin to integrate my game builder with the website.

I run a Discord server and wanted to notify Discord when a new game is published on my platform, so installed the Discord plugin to push posts to Discord. Plus, I can show Discord activity on my website.

Wanted to send out a newsletter for user retention. Very expensive options out there, so I installed a pay-once solution to run this from a wordpress plugin (50k subs and counting).

Soon I'm going to implement a marketplace where my users can sell and buy resources. And guess what, you have a WooCommerce plugin and some other one that enables this.

Oh yeah, I want to do this with virtual credits, and of course there is a plugin for that.

Absolutly awesome! And in the meantime, I can focus my developments on my RPG making tool.

I also sometimes hire a cheap developer to implement some custom things, and it works great.

Edit: plus, it's really fast to run experiments without losing development time. Just install or enable something, and if it doesn't add value, remove it again.

Did you setup a local installation to test drive plugins you want to add to your production installation ?

Most of the time I put it straight in production :D. I'm at CET time, and so if I do it in my morning, most users never find out when something would go wrong.

But I do have a custom script to create staging and test environments. Custom developments happen there.

It's basically copying all files and database, do a few config changes, and you have an exact copy of production that you can play with.

I looked into a proper development flow for WordPress, but it's very hard. Mainly because both data and config lives in the same spot. And you always want to work against the latest production vesion, since things might have changed there.

So no local setup, only multiple hosted ones. The DevOps flow could be improved, but right now it seems to work out fine.

So for custom developments, my developer works on test. Then I check it and move it to staging. There I check again if the update works fine, and do the same on production.

It's a pretty good flow for a 1 person project.

Be aware that WordPress makes it easy to breach GDPR compliance when the db is copied left and right and especially if other environments are shared with third parties (like external developers).

While I don't like WordPress either, I begrudgingly use it. My only wordpress website is a hobbyist news website where 1) 5-6 people need to be able to publish article on it, none of them (except me) have technical experience, and 2) we can only spare $10-$15 per month for the overall cost (and no, there's no way to make money here). This put me in a situation where almost the only option is self-hosted WordPress.

It makes sense and I think that's the reason WP has survived. In fact I started out like that, later building my startup product website(s) using WP; But once the tech-debt became too high I had to take a call.

Have you looked at https://ghost.org/? Self hosted on DO costs around $5/mo. Though it's much less customizable than WP

I went the static site via Hugo -> cloudflare workers sites route. It is so vastly superior to fiddling with hosting a SQL database, plugins, constant updates, etc.

I built an entire frontend around it (my username .co.uk) which tries to reshape how tech novices work with websites.

I have found small business owners I talked to (n=10) to be allergic to dealing with their WordPress or squarespace sites, because despite everything, they are just not simple to use. They just try to cater to too many use cases. Being everything to everyone comes with complexity that's very difficult to navigate as a tech-novice. Though I think squarespace do a better job than WordPress. Albeit more expensive.

The funny thing is, people have grown such an aversion to dealing with websites that I am also having quite a hard time trying to get anyone to look at it. As soon as they do though, it's like a lightbulb moment for them, to see how simple it can really be.

Naturally the trade off is flexibility, but I haven't had any complaints about that yet. I am not dealing with very complex sites though (and wouldn't want to either).

WordPress is synonymous with the progression of the web. It has allowed a lot of people and businesses to have a digital presence on the web and will likely continue to do so in the coming years.

Here's looking for more innovation and possibilities ahead!

While WP's ecosystem is nothing less than impressive, but I'm still in shock to this day, how cumbersome is site migration.

For me and I guess for many other engineers it's quite normal to develop a site locally (or at least on a staging non-public domain name) and then move it into production under the final name.

In Wordpress, the site name is baked in all over the place — in config, in database, in generated stylesheets and so on. In theory, this should not be like this (if you use barebones install w/o plugins), but then you buy a theme or some less-than-perfect plugin and boom, you need a separate plugin just to handle migration (which may or may not work 100% of the time).

Installing a WordPress site is actually very straightforward, in fact the platform is famous for its 5 minutes install. What you are describing as hard to achieve is copying state that is kept in the database, but I don't think this is an inherent issue to WordPress. But if you are creating the entire site locally and want to deploy it to production, it seems that WordPress may jot be the right solution in your case, instead a static site generator would be more appropriate.

That's categorically untrue. Almost all websites are first developed by a developer on his local machine, and then when it's ready to showcase or put online, it will move to the 'real' domain. It is no different for WordPress, and it's true that moving a WordPress site from one domain to another is a bit of a headache.

It should be "gather all things, move it to another place, change the domain name somewhere", but instead it's "gather all things, move it to another place, change all the links, change the domain name in two different places, hope you didn't forget anything or made a mistake somewhere". There are plugins to help you do this though, but all-in-all it's an annoying inconvenience.

Maybe, but I do this on a weekly basis and even with complex custom sites/apps it takes me less than 10 mins to change the urls everywhere throughout the site/database. Decent plugins indeed exist to make this super simple.

This just isn’t the case if it’s done well.

If you do need help doing it then migratedbpro or migrateguru are seamless.

It’s, like, really really easy to move Wordpress.

> In Wordpress, the site name is baked in all over the place — in config, in database, in generated stylesheets and so on. In theory, this should not be like this (if you use barebones install w/o plugins), but then you buy a theme or some less-than-perfect plugin and boom, you need a separate plugin just to handle migration (which may or may not work 100% of the time).

WP-CLI is your friend:

    wp search-replace 'http://example.test' 'http://example.com'

Don't do that.

Wordpress store some information fields and the length of the string of those fields. Use a plugin that search and replace for those and recalculate the length.

Also, search for //example.test, not http://example.test.

Edit: I read a bit too fast, maybe it does implement the string size but I can't be sure https://github.com/wp-cli/wp-cli/issues/1224

Edit 2: https://github.com/wp-cli/wp-cli/pull/1261 sweet, looks like it does ^^.

As far as I know WP-CLI also handles serialized data correctly.

By the way, I very much prefer the wpmigrate-db plugin.

Also, wpmigrate-db has wp-cli commands to export the database, I don't know if wp-cli search-replace can do that.

Edit: And it does:

    [--export[=<file>]]
    Write transformed data as SQL file instead of saving replacements to the database. If <file> is not supplied, will output to STDOUT.

Well, wpmigrate-db allows to replace multiple strings in one command and that's something wp-cli default search-replace doesn't provide.

actually if you add

  define('RELOCATE',true);

The Wordpress will be in the relocate mode the next time you visit wp-login.php. It will reset all the URLs.

Documentation: https://wordpress.org/support/article/changing-the-site-url/...

Yes, I've read this. Unfortunately that's not enough if you have 3rd party themes or plugins installed.

Why would have the site name in a generated stylesheet unless someone wrongly put it there ?

Some commercially available themes do it during installation. I was surprised to find that also as this contradicts every instinct I have as a web developer.

I agree the ecosystem is what makes WordPress attractive. I'm impressed by the efforts put into third party themes and plugins that are simply amazing but the WordPress core is a joke.

One of its function name is literally "the_content();" where you output the main content in a theme, no class, no context but just a dumb function that sits globally to force the output and there are global variables like "$wpdb" that keeps the database methods.

Of course, you don't want to look how messy the database schema is. Configuration values are just thrown into the db with serialized strings with random keys, you'll never know what's where.

Seriously a first year high school student would put out a better code and no wonder it took so long to iron out vulnerabilities and performance issues to acceptable levels but the mess is already there and people have to live with it unless someone brilliant creates a new platform with theme and plugin compatibilities to WordPress, so the ecosystem lives for people to move on to it.

Been working with it for the past 15 years, and made a good business on it. It's relatively easy to create a hardened install that will -most likely- not get hacked. Slap a caching solution on it and performance is not a problem.

Name me an alternative CMS frontend that any client can use to update/create content (without learning Markdown) and visually design a page with pre-defined blocks, that doesn't cost $200+ a month.

If only they would include a localisation solution, I could say goodbye to slow and cumbersome WPML.

Craftcms is excellent.

Craft is indeed a decent alternative. I haven't tried v3 but v2 was horribly slow on sites with a lot of custom fields.

v3 is snappy and the built in caching mechanisms can help a lot with performance.

It is flexible, fast, and secure - especially on a host that specializes in craftcms - like Fortrabbit - and combined with cloud flare.

Webflow, squarespace, etc will do that for 1/10th that price.

Both have vendor lock-in though. A WordPress site is portable, it runs on anything with PHP and MySQL. SQspace and Webflow only run on their own servers.

Webflow has zero lock-in. It spits out static sites with standard semantic HTML/CSS and you can take the code anywhere. They also have an API, so you can export collections.

Squarespace, yes, you're locked in.

Re: the portability of Wordpress, I'm not sure how many Wordpress site migrations you've done...but I wouldn't use the word "portable" to describe the process.

Webflow HTML on its own is useless because it is not dynamic. You still have to set up code snippets inside the markup so it can display the menu, or post and page information stored in the database or flat files of whatever CMS you're using.

Webflow CMS solves that problem, but with vendor lock-in.

Webflow is ok, but you still need to customise a ton by hand and it gets slow really quickly for a decent site. Squarespace is way too locked in.

$200+ a month?! What's your setup? You can easily do less than $10 per month if you cache and CDN properly: https://www.concurrencylabs.com/blog/handle-thousands-of-use...

I obviously don't pay that for my setup. I was talking about decent alternative CMS'es which are easy to use, dynamic and flexible. Like for example contentful, strapi, dato, etc.

I've recently setup a multisite solution for localisation of a WordPress project. Definitely needs a bit longer to setup properly, but it's so much faster than wpml.

I left the web hosting business in 2008 and came back in 2013. Prior to 2008, there was a little wordpress, but there was b2, b2 evolution, MovableType, and a lot of the landscape was Perl based.

I rejoined the ranks in 2013, and holy smokes. WordPress. WordPress everywhere. It's even more so now. These days, to be in Web Hosting is to be in WordPress Hosting.

I wonder what the exit strategy is for the VC's behind Automattic.

It doesn't seem like Matt has any intent to go public, yet he's raised $700M+ [0]. I've heard anecdotally it's a profitable company but not sure if that's still the case after the Tumblr acquisition.

[0] https://www.crunchbase.com/organization/automattic

Personally I think despite being unpopular with many developers (I dropped PHP/WP for most purposes many years ago due to the amount of hate it was getting and general poor reputation) and the problems of the past, the concept of plugins and what you can do with WordPress is terrific.

And not something that anyone who really understands software engineering should scoff at.

I would go so far as to invite the downvotes with a comment like this: WordPress may now be the most powerful (low-code/no-code) application development platform ever created.

What is the WordPress of the 2020s? Something built on web assembly or TypeScript or Python or whatever maybe?

> I would go so far as to invite the downvotes with a comment like this: WordPress may now be the most powerful (low-code/no-code) application development platform ever created.

And you would be correct.

The WP API is mature and stable, so much so that tools around fast WP scaffolding are already available. You can create a site in Webflow, export the HTML and use a tool like Pinegrow Theme Converter to attach WP functions to HTML elements, such as Posts loop, ACF fields, WP Customizer, etc. No PHP knowledge needed. Pinegrow will export a fully featured WP theme to your site folder, and handle enqueuing the functions and stylesheets. You never have to open functions.php again.

Add to that the vast library of plugins for things like Ecommerce, online courses, appointment scheduling, user accounts, social networking etc. and you can create a pretty complex prototype.

Trying to decide between these:

(a) Wordpress (b) SquareSpace (c) https://strapi.io/ (d) https://prismic.io/ (e) https://www.contentful.com/

The output will be a simple website, we already have all pages in HTML, and from the tech perspective we could use any of the above.

What I've seen was: SquareSpace is still rather limited, Prismic and Contentful are around $500 per month in their non-free-tier, and Strapi is just too expensive to find devs for. Plus their editor looks the same as Wordpress's if I'm honest.

On the other hand, Wordpress has a massive community of reasonably priced freelancers, has updates many times a year, and most things you want out of the box (such as SEO, or easy hosting on Heroku or elsewhere). The only thing it doesn't have is the excitement to work with it :)

Is there a good business case for any of the competitors?

My current headless CMS choice is sanity.io. you'd still need a dev for the frontend and sanity schema setup probably.

Squarespace has somehow managed to make <img> tags require 3rd party javascript to work.

I cannot take them as anything other then grossly incompetent.

Choose boring technology.

WordPress is 18 years old, there's a good chance that it will be around for quite some time longer.

Consider a static WordPress site via the WP2Static plugin: https://wp2static.com Going static solves most of the performance and security problems with WordPress. There's a prebuilt Docker setup at https://lokl.dev/ and an AWS template at https://staticweb.io/open-source/wordpress/ There are also commercial hosts (with their own plugins) at https://www.strattic.com/ and https://www.hardypress.com/

I've worked a bit with Contentful, and it seemed like a pretty good CMS, but pricey. The client was moving off of it because they wanted the more familiar WordPress UI.

Is there something up with wp2static right now or did they intentionally strip out all CSS/Style?

The appearance is intentional. The author (https://github.com/leonstafford) is very into minimalism and accessibility.

Ohh thanks for this I'm going to try this out looks awesome!

I would choose something that would let me move my site to other vendors easily or self-hosting it.

Webflow.

Webflow is basically a super good editor that spits out static sites on build.

As easy to update as squarespace for non-tech folks, but as customizable and secure as a static site generator (since you can build everything from scratch).

Also look into Frontity https://frontity.org/ if you're using React on the frontend. Check out their showcase page, a lot of big companies already using it: https://frontity.org/showcase/

Pretty awesome for a guy who studied jazz saxophone and political science. You may have to say a lot of things about Wordpress etc but I know a lot of arrogant computer science people with Phds who think they are programmers but have created nothing in particular.

Having never peeked at WordPress internals, can someone give examples of what they see as poor code at the heart of WordPress? It is 18 years old, is the code really still a mess?

WordPress alternatives like Wix or Squarespace are closed-source, but the HTML/CSS generated by these services is simply dreadful.

As for Static Site Generators (SSG) - liked by developers but by no-one else - they are particularly unsuitable for non-technical users and lacking in features (by design). They are certainly not a serious alternative to WordPress.

It has a very different architecture from what web developers are used to. It has an event driven architecture, somewhat like C. During the process of running the code on a HTTP request, there are multiple hooks (actions and filters) that get called, where you can attach your own function in a plugin or theme.

About 5 years ago I saw a lot of requests like "When will WordPress move to MVC?", but those questions seem to have died out mostly. Then there is the OOP architecture of data and communication that is loved by a lot of developers. I think Joomla and Drupal have moved into this direction.

This makes the code in WordPress look oldfashioned, it is partly still procedural code, with some classes added into it. WordPress community very much wants things to be backwards compatible, and so far that works out.

I've used WordPress for a personal blog since 2005. Over the years, I switched hosting, different versions of PHP / MySQL, upgrades to WordPress,... But one thing I've never done was exporting all my content and then importing it in a fresh installation. Yes, I tinkered with SQL statements and I went through manual cleaning if an overzealous plugin didn't clean up properly after itself. But I never ran into the wall where the entire database got into an irreparable, unusable state.

It's a simple blog. So, I've always tried to use plugins sparingly, and avoided massive database model changes if I could help it. But I like to think it wouldn't have been possible without the care put into WordPress upgrade path.

At the same time, the focus on backward compatibility, to my mind, has held it back. Sure, innovation happened at the visible level, the surface level. New themes, new administration UI, Gutenberg, Blocks,... all of that have improved the author experience. But looking under the hood, the codebase simply isn't in line with modern PHP practices adopted by major PHP framework communities. Which leaves quite a lot to be desired for.

Granted, part of the appeal of Wordpress is simply being able to grab the PHP files and (S)FTP them to a shared hosting. No need for complex CI, no need for composer, no need for complex configuration. That's what lowers the bars for many people who simply wanted their own self-hosted space on the Web.

My other major gripe is how WordPress authoring experience evolved over time into a direction which I don't agree with. The editor in modern WordPress feels heavy handed: every item you add to an article morphs into a distinct component accompanied by a contextual menu with myriad options. I feel that this just gets in the way of what's actually important: writing. Personally, I like(d) the terseness of a simple textarea form element with limited affordances as to what you can add in terms of markup. Then again, the ambitions of Automattic are beyond simple writing and moved even beyond authoring and publishing. I think that's just par for the course and I don't fault them for that. It just has made WordPress less of a good fit for me over the years and that leaves me some with a sense of saudade.

Either way, I owe a lot to WordPress. And I'm curious to see how it will evolve in the years to come.

Oh god the Gutenberg editor, ugh. The day they EOL the “use the old editor” plugin is the day I finally consider moving to something else to run my personal site. I do not need that shit and I do not want to crack open the multiple custom styles involved in my site to make them work with that shit.

Gutenberg is a joke.

Clients wonder how to use it and internally, it's a mess like all the other core part of WordPress.

They leave a bug having the previewing feature broken with any custom fields for 2 years and they don't know what to do.

https://github.com/WordPress/gutenberg/issues/16006

There are plenty of better visual editors out there.

During the early 2000's I was part of creating a competitor to Wordpress and Pyra labs (blogspot) in Sweden called blogg.se. The reason it succeeded (read: outgrew both at the time, at least regionally) was two-fold: 1) The platform was available in a language the users preferred and 2) there was a human support crew that answered to questions.

Many years later, I'm happy to see Wordpress thrive – the power of one person's voice is more important than the platform itself. This is how any platform should work: empower their users, not control them.

Localization is mostly a solved problem today but empowering the user and respecting their rights isn't. I hope us developers can help pushing the ball in that direction.

Does anyone have recommendations for learning Wordpress if you already experienced with modern practices like GitOps and immutable deployment artefacts (containers/VMs) using more fully featured frameworks like Django and Rails, which have abstractions like ORMs and Migration frameworks to look after all your database changes (within reason obviously since we all know these tools aren’t one size fits all) …

I know there’s value in Wordpress even if it’s just potential future revenue supporting a subset of the 1/3rd of the web running on Wordpress. So I’ve pushed but when I loose spend twelve hours over a whole week trying to work out how I’m supposed to manage my database (don’t even get me started on how frustrating it is that you are basically forced to use MySQL) and basically giving up because there both appears to be nothing other than do it yourself with SQL DDL if you don’t want the plug-ins doing it for you, and the fact the web is full of SEO for how to “migrate” your content from one Wordpress site to another hosted somewhere else…

Id love some recommendations as it feels like I’m making an uphill slog through a river of 18 years worth of outdated info posted by users and decaying SEO bacon written by developers.

Get your local environment set up (Linux, nginx, PHP, MySQL), then configure Netbeans with xdebug, and start stepping through the code.

I think this is probably both the most painful and most useful way to learn how Wordpress works under the hood. It will probably take you another 12 hours (and another week), but at least those hours won't be wasted.

Edit: oh, and take notes while stepping through with the debugger. Having a document you can refer to later that has some kind of outline with your own ideas and opinions is a huge help.

Use the official docs instead of going off what people say to do on stackoverflow and the other wp seo spam sites out there. The snippets are often not optimized correctly, using an outdated method, or just insecure.

I suggest going through the VIP docs as well as they have some really good advice on securing and speeding up wordpress along with what to avoid.

https://docs.wpvip.com/technical-references/

Also just a tip: https://github.com/WordPress/WordPress-Coding-Standards

https://pantheon.io/ has WordPress plans with dev/staging/prod sites integrated with git. If you want git, I would highly recommend trying their system out rather than trying to figure out how to make WP cooperate on your own. WP installs normally don't use a VCS at all.

Don't try to use WP for anything involving a lot of programming. It will just hold you back. Use it for projects simple enough that you don't mind writing a few MySQL DDL statements with DIY migrations.

> Does anyone have recommendations for learning Wordpress if you already experienced with modern practices like GitOps and immutable deployment artefacts (containers/VMs) using more fully featured frameworks like Django and Rails, which have abstractions like ORMs and Migration frameworks to look after all your database changes

Personally, I think trying to apply modern development practice to wordpress development is a waste of time. You're going to fight an uphill battle. Instead, just treat it as a legacy platform and use legacy development methodology: have at least two or three copies of the site (dev, staging and production). Use sftp to live edit the code on dev/staging, then copy to production regularly. Make sure to have a versioned daily backup in place which at least can pull the last 7 days daily backup and the last 6 months monthly backups, and you're set.

If your app is getting too complex for this development method, you can always switch back to django or other modern frameworks with modern tooling.

I agree with this approach. Fortunately, there are managed services specifically for Wordpress that handle all this stuff. Cloudways, WPEngine, there are several others too. Cloudways lets you set up and manage WP using Digital Ocean, Vultr, AWS or GCP if I recall.

I don't see why this would preferred over using something like vvv [0] to develop locally and then pull changes from a repository to a production site. That's how I've been doing it for my personal site and it works like a charm.

[0]: https://varyingvagrantvagrants.org/

What's great about this tool? since you can use Ansible to install Wordpress and more.

You use VVV for local development. Ansible is certainly something you can use to deploy.

I haven't looked too deeply into corcel [0] but it does give you some kind of interaction between Eloquent ORM and WordPress.

[0]: https://github.com/corcel/corcel

I recall installing it when it was still pretty new, freshly forked from another application.

The last time I looked, WP was running ~1/3 of the public-facing web. I use it for all sorts of things, as it makes a perfectly usable CMS. It's easy to keep it updated and secure these days, plus they've had an emphasis on security for a while now - or at least a better attitude with regards to security.

It's actually 40% now. "Running the web" might be stretching things a bit though, because a ton of sites use it just as a CMS layer for their existing front end or to host their blog subdomain.

I've switched from the Jamstack/static site generators to nocode WordPress for content sites, to focus my coding more on app UIs in the long term.

Jamstack is nice, but it's either for personal developer blogs or larger teams. Outsourcing is too expensive with coders being in high demand.

Everything WP can be easily outsourced when needed, and there really isn't much of a speed benefit with Jamstack sites if you cache your WP site on the edge (e.g. with Cloudflare APO) and use a fast DNS. Security of the origin server isn't a problem with a good host or going static.

You can generate static sites from a local WP install if you find secure hosting too expensive, here's a nice docker based tool for this: https://lokl.dev/

WP is really as hard, ugly fast or slow as you make it, you have to find the right theme and plugin ecosystem within WP.

I don't have to and don't need touch the code building sites with it, so all those weird WP code conventions don't really bother me.

Wordpress made a great impact on the net, and I was happy when clients liked its ease of use and relieved from the burden of making content changes. (Though, I've always felt that https://textpattern.com/ was more secure and better than Wordpress).

I think that's the remarkable thing about wordpress. There have been better engineered platforms consistently, but WordPress wins because it's got a diverse long-tail module ecosystem, an enormous developer pool, and a self-perpetuating userbase.

Fun fact - Textpattern and WordPress are both 18yrs old, PHP and GPL.

Software lasts way longer than you expect it will when you write it. WordPress is one of those apps that continues to endure despite lots of criticism: bad license, insecure, badly engineered have all been lobbed at it... and here we are 18 years old, 40% of websites use it... So, congrats to Matt and really, the entire WP community.

I am getting old. I remember when b2 forked to b2evolution and Wordpress. I remember when cpanel started allowing for installations.

I’m glad they are doing so well.

I personally think it’s cool you got to see all that in real time OG :)

For those who were not around that long, here's a brief summary from the b2evolution site https://b2evolution.net/about/b2evolution-vs-wordpress

I haven't been around the open source CMS world this long, I only started with Drupal 17 years ago (to the day as of yesterday) https://i.imgur.com/TsNTYqZ.png

I remember CPanel and having to set up Wordpress installations or using it to create MySQL databases and install phpmyadmin + setup FTP accounts so I could deploy PHP applications I had written myself. I got into programming during this era to create a website for my uncles store and “share” his hosting to run a forum I built dedicated to Ocarina of Time.

I definitely look at those days with rose tinted glasses and had basically forgotten about the time before I used VPSs and knew enough of bash and Linux to set all these things up myself. And even that was quite awhile ago now I write frontend and back ends and deploy them to AWS or Heroku. I guess it just shows how much the web has changed and matured compared to using shared hosting from websites that only a few hundred people max used.

This thread is a giant bucket of nostalgia for me and I’m not even 30 yet. Crazy to imagine what the web will be like when I’m 50 longing for the days of React and Flask.

Well, cPanel hosting and shared hosting is still a thing. There's a LOT of such hosts around much thanks to WordPress.

My humble experience with Wordpress is very good.

I used it for my blog and decided to not write a single line of code, because I would use my time to write articles: https://gioorgi.com

I started using it on 2004 about 17 years ago. I migrated about 4-5 times my VM provider, doing backup and never lost a bit.

After moving to a fast provider, my machine works fine, and I find out JetPack is quite a CPU eater; now I prefer a more huge set of plugins and avoid JetPack.

I used it sometime for building customer's Web Site, and, yes, Drupal was better for general purpose CMS at least on 2009.

But Ghost did not buy me and I think WordPress is still a good tool.

Does it make anyone else nervous that 40% of the web is run on software with such a poor track record when it comes to security?

Anything highly-used will be a larger target by default I'd guess.

Isn’t just the “Windows” of the internet? It’s heavily used, directly exposed to the internet, so it’s a bigger target, hence more bugs a found (and exploited).

If you'd have ever looked at the WordPress source code, you know why it's been having a crappy track record. It's one of the easiest example of "don't code like this".

Those issues went away with the automatic updates feature.

RCEs are rare enough now that the price has grown to 300k for a zero day: https://wptavern.com/zerodium-temporarily-triples-payout-to-...

Not if you choose your plugins and theme carefully.

The 'poor track record' reputation comes from the giant number of mom-and-pop sites that got compromised. I imagine 40% of the web is really just a handful of massive commercial websites that have developers on hand to take care of security, backups, devops etc.

Simple in use and simple to extend. Still powerful and showcase for continuous innovation without large disruptions. The example that gpl software can be very successful.

I wish WordPress would bake a flat site exporter directly into an install, instead of forcing folks to mess about with plugins.

New page builder just came out, may be someone will be interested in trying - https://bricksbuilder.io

Looks like a combination of Elementor + Oxygen, and looks like it is producing very clean HTML and some how compatible with Gutenberg (haven't tried this yet)

I've heard good things about it as well. Interested in trying it out.

I run a small website on wp. Every time I update it I take a backup first. One every couple of days on average I think about replacing it with something custom. But the sad thing is that I don't know enough of the CSS side to make the site look good without some sort of theme support like wp has.

How the three big CMS (WordPress, Joomla!, Drupal) compare nowadays?

Drupal has become more targeted to enterprise. For sites you're developing yourself, wordpress can do the same stuff as drupal now, especially with advanced custom fields (ACF). Another alternative is Statamic.

statamic is incredible...I can't recommend it enough...such a breath of fresh air

Thanks for the kind words!

It's empowered the web and countless people able to feed themselves. Yes, WP works does leads to messy code, but that's also how people learned what's makes things messy.

Thank you WP

Hate this kind of new, it make me feel old. It's like someone saying to you "Did you know <some child you still see in diapers> is now 18.

Yes, it's brutal. Other people's kids grow so fast.

I think I’ve been using it for my personal site for most of Wordpress’ life, wow. Certainly at least half the project’s life.

There are many critics, but WordPress was an integral part in shaping the web today and the blog era.

Blogs have gotten worse recently. It was nice for a while, so many people with little hobby sites. Not it’s all content marketing and people branding themselves as experts for career growth. It’s hard to find those old blogs now, they are buried 3 pages deep in search results

I wish more people used wordpress instead of medium or stacksomething or whatever is trendy. Wordpress works, it has an rss feed, it doesn’t have a paywall, it really is all about content.

Medium offers an audience, and it's human for people to want their work to be read and feel validated. I have friends who share their blog posts and social media and have it read that way.

Unfortunately my own blogs are quite niche in topics, and there wouldn't be much value in sharing socially as my network wouldn't really get it.

I used WP for various sites and blogs, and now I am left with one last instance, which I refuse to babysit, and will just turn it off. I had enough.

congratulations! WP has indeed come a long way, I was just checking up on what happened to Mambo, another CMS, and it seems to have fallen off the internet around 2008.

iirc Mambo was forked into Joomla which is still used but nowhere near as much as WordPress

I'm glad WordPress is 18 and successful.

I wish it had stronger open source competition.

WP is ripe for disruption in so many areas:

||| Price |||

WP succeeded in part because it was cheap and easy to host. This is still true to an extent but the story is weaker than it was.

WP hosting is a race-to-the-bottom market, but hosting secure, fast WP sites with good backup and support remains valuable. The WP hosting market matured and there are enough customers who won't pay for cheap hosting because they've seen their sites get hacked or experienced performance and reliability issues. This supports an industry offering expensive managed hosting at a high markup tiered by monthly visits.

So it's not surprising that Squarespace/Wix/Shopify price points are no longer the turn-off they used to be for many at the low to mid end, and that larger businesses are looking at Jamstack and static sites for the potential cost savings at the mid to high end.

Takeaway: A CMS built for “start for free” hosting platforms like Cloudflare Pages and Netlify could provide an even more compelling story than WP, reignite the DIY open source site building movement, and be a more natural fit for enterprise customers.

The “Jamstack” movement is the closest thing we have to this, but there's no strong frontrunner that includes both a modern and compelling admin area for publishers coupled to the mechanisms for previewing, building, and deploying a site out of the box. Devs still have to cobble pieces together into a semi-coherent whole. This is currently touted as a strength because you can “use the best tools for your unique scenario”, but it might be healthier for the Jamstack community to start treating this as a major flaw blocking wider adoption instead. That way they could just build a good apple pie instead of asking everyone to first create the universe.

||| Deployment |||

When technical users in a company get to pick the stack, it is easier, cheaper, and faster to deploy to Netlify/Cloudflare|GitHub Pages/Vercel using modern deployment practices (like the ability to preview multiple deployments — one for each PR) than it is to ship updates to a WP host. Cloudflare Pages pricing is also tiered on deploys/month instead of visits/month or bandwidth, which is more palatable for most.

Takeaway: A headless CMS that deployed the admin app and website to an edge network could outperform WP for ease of deployment, performance, affordability, and access to other built-in perks of edge hosting platforms that are hard and expensive to get right with WP today (caching, security).

||| Marketplace |||

WP.org offers free plugins and themes, but it still does not (and may never) offer paid themes and plugins. This makes commercial WP options decentralized, but it provides users with a poor experience: comparing, purchasing, updating, and getting support for paid themes/plugins is a wild west that is still harder and more frustrating than it should be. It also means anyone who wants to sell themes or plugins has to manage their own sales infrastructure.

Takeaway: A CMS platform with a built-in marketplace could be more compelling than WP for end-users. It could also pull plugin and theme developers away from WP — most plugin/theme devs would probably prefer to spend their time making themes and plugins than figuring out how to sell and deliver updates securely.

||| UX and ease of use |||

WP can be overwhelming for new users, in part because of the disconnect between the promise on its homepage (“the freedom to build anything you want“) and the journey people have to go on to achieve that lofty goal.

The main issue is that there is no real guidance — you can build an ecommmerce site or a podcasting network or a gated content blog or a newsletter or even just a simple blog with WordPress, but in all cases you're basically on your own to figure out how to achieve those things. Yes, there are a huge number of blog posts with titles like, “9 Best WordPress LMS Plugins to Sell Courses” and “We Pretended to Test the 46 Best WordPress Contact Form Plugins and Then Just Recommended the One that Gives us the Highest Affiliate Commission” that you can sift through, but WordPress itself gives you very little help.

Takeaway: A modern CMS or user-friendly application platform that claimed you could “build anything you want” would actively help you to do that. It could perhaps offer starter templates for different scenarios, tutorials, or _basically anything at all_ that didn't just dump you into an admin area that was last refreshed in 2013 containing some vestigial apparatus for blogging but no real clues as to how to even succeed with that goal let alone anything more complex.

||| Localization |||

It's 2021 and you still need plugins to translate content in WordPress. Naturally there are a lot of localization plugins that you'll have to compare and experiment with first, and you'll find that all of them offer a worse experience than you can get with a hosted CMS like Contentful or GraphCMS.

The good news is that built-in content localization is on WP's product roadmap, so this should eventually improve. The other positive is that while content localization is pretty bad in WP right now, plugin/theme localization is generally quite good.

Takeaway: there's a small window of opportunity here to present a more compelling content localization experience than WP does by default.

||| PHP and the developer experience |||

A large number of developers who use WordPress would rather not. (WP was the “most dreaded platform” in Stack Overflow's 2020 survey with 67% of developers who are using it not expressing an interest in continuing to do so.[1])

Of more concern for the health of WordPress is that new web developers generally aren't picking PHP. It's not even that PHP is especially bad these days, and even WP APIs are gradually improving — it's just that new developers targeting the web as a platform are probably writing JavaScript or TypeScript. And, because they're not writing PHP, they might not choose WordPress if a compelling alternative existed in the language they prefer.

Yes, Headless WordPress is a thing, and every WP host is jumping on that boat because they can see the writing on the wall. But WP is going to be tightly coupled with PHP for a long time, and that may be an opportunity for a CMS that can attract new blood simply because “we use your favorite language instead of PHP” becomes a competitive advantage.

||| ECommerce |||

WooCommerce is used for a huge volume of transactions ($20 billion a year was mentioned in State of the Word 2020).

But if you need to build an online store today, Shopify is in most cases a quicker and more stable path that will let you focus on your business and products, simply because it's built for commerce out of the box.

If you need to sell products casually, Gumroad/Stripe Payment Links[2] are probably a better choice than WooCommerce.

If you need to gate content behind a paywall on a multi-author content site, Ghost is increasingly a more compelling choice than WP, since it includes subscription commerce as well as audience building and outreach features by default: https://ghost.org/features/

WP can be an ecommerce platform, but it isn't one by default. Making WP do ecommerce feels very much like any other form of digital alchemy or contortion.

A CMS built from day one with commerce and monetization in mind would look quite different to WP. It could be all the more compelling for users who want those options out of the box.

[1]: https://insights.stackoverflow.com/survey/2020#technology-mo...

[2]: https://stripe.com/en-at/payments/payment-links

Thank you. This analysis of how "WordPress is ripe for disruption" was well worth spending my time, reading carefully and reflecting on your insights.

If I may attempt a summary..

- Secure, fast hosting with good backup and support

- Admin area that allows publishers to preview, build and deploy a site

- Headless CMS that deploys the admin app and website to an edge network

- CMS platform with a built-in marketplace

- User-friendly application platform that actively helps you to "build anything you want" - such as starter templates for different scenarios, tutorials

- Built-in content localization

- CMS using "your favorite language", such as TypeScript

- CMS built from day one with commerce and monetization in mind

Great summary! It shows how big a task it is for one project to take on. I don't think it hurts for more people to try, though.

Am I the only one that doesn't like WordPress anymore? The concept was great, but it has basically become a "website builder" and not really a piece of blogging software anymore.

Throw on top of it the architecture of plugins (no sandboxing!) and the culture this has developed for websites being "built" with a combination of 50+ outdated plugins and it frankly becomes a nightmare. Of course this can occur with any huge software project, so I somewhat concede this point.

The real turning point for me was the push to move to the "block" editor and effectively turn WordPress into Wix or Squarespace or whatever

I’m newish to wordpress. The Gutenberg block editor is really word presses answer to squarespace( they said so at a talk word camp Boston a couple years ago).

But it still has the ability to just type away and write stuff without ever leaving the keyboard. I think it’s not bad. And it gives the graphic designers some control.

It’s a weird ecosystem though, you need a solid theme to start and finding those is a little hit or miss. Same with plugins, it’s hard to tell the great from the ok sometimes.

Apparently, lots of people want a website builder.

We in the startup culture often forget that the market for building bespoke webapps using the latest and greatest frameworks is only a tiny part of the web. The vast majority of people who order websites just want something that works out of the box, and they want it yesterday.

I don't think anybody is particularly fond of WordPress's medieval internal structure or its byzantine plugin ecosystem. For the time being, though, I'm glad that there exists a well-known, open-source, self-hosted alternative to Wix and Squarespace.

There was a time when lots of people wanted something bloggish. Now that time is over, and WordPress has pivoted accordingly.

>The concept was great, but it has basically become a "website builder" and not really a piece of blogging software anymore.

You can definitely turn it into practically anything, and it quickly becomes a mess due to the fact that it was never meant to be an application platform. But WordPress is still best in class at what it was meant to do; a simple self hosted blog.

WordPress is highly complex... it has plugin extension points, convolutes HTML building into application logic, and regularly has new security vulnerabilities discovered and exploited.

A guy I know runs a blog for his non-profit on WordPress. At one point he asked me for help because pages were loading extremely slowly. It turned out the blog had been hacked and was being used to host gigabytes of junk pages with SEO boosting links to really trashy websites.

I used to recommend WordPress wholeheartedly for self-hosted blogs, but these days I strongly prefer something that doesn't have any back-end code that the site owner has to maintain and/or update -- a static site generator, a JAMStack-based blog, etc. Or consider using someone else's hosted blog platform (someone you really trust to host a secure platform).

Best-in-class for a simple blog is something more like Ghost I would say.

Or better yet, any static site generator (Hugo is nice) + Netlify.

Even with people moving to walled gardens, many people still want to build and collaboratively author websites.

WordPress is still best in that, even when it kind of sucks.

I myself prefer the "static builders" for small websites, like Hugo/Jekyll, but WordPress also have built-in editor and access control; with small websites it's whatever, but with bigger websites, it comes handy.

The core WordPress codebase has been polished until it's sorta OK. It's got some quirks, but many of the obvious problems have been patched over the years.

The plugin marketplace, on the other hand, is a disaster area. Plugins are marketed to non-software people who don't understand security and don't know how to evaluate products for it, even though they might care in the abstract. The result is that the typical WordPress installation is a festering mess of insecure plugins, and sites get hacked all the time.

Anymore? I stopped liking Wordpress before 2010, especially being a PHP programmer (back then) in small town Midwest where 90% of gigs I landed was fixing some monstrosity that some guys nephew threw together made up of a dozen+ plug-ins that had since been hacked on by a series of online personalities and a theme they purchases three years ago that they had paid five different people to “customize.”

Yeah sure I’d be glad to quickly change over this theme to match new designs. Wait every plug-in also has to be fought with since the templates for those have also been “customized.” Wait I have to match the HTML almost to a tee because the chrome plugin no one else has ever heard of says you have perfect SEO (despite only getting first page results with a highly specific query no human would ever type) and you think this will lower your score? :(

I’ll take a stressful 10 hours at a startup over that any day.

> Wait I have to match the HTML almost to a tee because the chrome plugin no one else has ever heard of says you have perfect SEO (despite only getting first page results with a highly specific query no human would ever type) and you think this will lower your score? :(

That's hilarious. I'm assuming it would not have gone well to try to tell them otherwise?

Most of these people owned mom and pops or online stores; they weren’t tech savvy and tended to believe what they read on google (usually blogs trying to sell them Wordpress plugins, SEO guides and the like) so yeah it would have fell on def ears. It is funny, now that I don’t deal with it anymore.

There are definitely a lot of people who do not like WordPress anymore, there are people who complain about the same things as you do. There just is no other tool that fills the same niche that WP does, semi-easy to use for people without any technical know-how, to add new pages and content.

To me WordPress is a relic of a time long gone, of internet that does not exist anymore, and sometimes wish I could get back to - it enabled so many people to just put out things that they wanted, without having to learn HTML

I suspect you’re no longer in the circles that WordPress is targeting: it’s still a fairly common “no-/low-code” platform for people who just want to put together a site for themselves. I think that what’s going on is that programmers generally don’t encounter it as much in “real programming jobs” because it has been so successful in making programmers unnecessary for a whole swath of websites.

There's plenty you can do with WordPress as a developer. There are also no-code tools like Pinegrow that attach dynamic WP functions to static HTML templates.

Automattic's failure IMO is making Wordpress.com feel like a blog platform for far too long. A lot of what users would consider reasonable functionality for a website was not provided in WP core but by the plugin ecosystem. So people just defaulted to plugins for everything, which eventually gave WP a bad name, because plugins were often janky and poorly supported.

> Am I the only one that doesn't like WordPress anymore?

I stopped using WordPress over a decade ago. I use other website//blog platforms now. I may go back if there is nothing better available, but I have a control-streak in me.

At the time I started WordPress I just didn't have the knowledge coupled with just needing to start something quick.

I moved on.

I agree, unfortunately. For me, it's been on a decline since wordpress.com came along and from then it's felt more and more like Automattic is driving development to suit their business model. I still run a large-ish blogging network that uses WordPress Multisite, but I would love to migrate away from the platform if there was a suitable alternative.

I think there is a sizeable market for a basic 'none blocks' wordpress clone with a limited range of options for themes and plug-ins.

Anyone who needs to install a plugin to add a table or to do a redirect is already well catered for.

I have discovered ClassicPress - https://www.classicpress.net/

They have forked from WordPress 4.9. I'm not sure yet about their long term feasibility.

With you 100% on the 'block' editor. Instead of offering a decent alternative to regular site builders they have changed the landscape so someone may as well us wix or squarspace since wordpress is now offering nothing better?

>The real turning point for me was the push to move to the "block" editor

Yep, this as well as seeing random ads appear in my blog while scrolling through my posts. Recently, I tried going back to WP, but only if I could use the classic editor but they made it so confusing to find it and use it that I gave up. Too bad. I was a happy customer for many years prior to the block thing.

18 years of polluting the planet with bloated and bad code, breeding insecure developers writing insecure plugins and themes.

Kudos!

Reply on news.ycombinator.com