WordPress Playground: A WordPress that runs entirely in the browser

https://developer.wordpress.org/playground/

by disadvantage • 3 years ago

506 213 3 years ago

developer.wordpress.org

edoardo-schnell• 3 years ago

Next up: a WordPress plugin to host WordPress playgrounds

em3rgent0rdr• 3 years ago

Its WordPress all the way down.

fredley• 3 years ago

WordPress was where I got my start, a while ago now. From poking around with PHP files, to building my own things, to Django, and on.

It would have been inconceivable back then (mid noughties) that the whole thing could be done client side. The Web has come on a long way.

These days I tend to hate WordPress with a passion. Occasionally I will be asked to peek at some technical problem with a WordPress installation and it's never not like opening some awful Pandora's box of out of date, brittle, scary plugins, these days most of which seem to be trying to upsell via ads in the admin.

However I will always be thankful for WordPress and what it enabled in its heyday, and say a silent prayer of thanks every time I think of the people tasked with keeping creaking old sites alive.

whyage• 3 years ago

What no-code alternative would you recommend for pro web designers?

kidfiji• 3 years ago

Although it isn't open source, Webflow seems to be the current most popular thing for that specific niche (low/no-code platform for designers)

hakube• 3 years ago

Webflow

frankzander• 3 years ago

Lern to code (at least a little bit) and if not ... Contao. CSS isn't considered "coding"

8n4vidtmkvmk4• 3 years ago

Wouldn't say "pro" unless you get strictly mean design and not functionality. In which case I'd have recommended Photoshop which is what I used for design only 20 years ago, but I think the UX folks like Figma or something these days.

imilk4• 3 years ago

Wordpress is still great for downloading a theme from Themeforest to get brochure style sites (restaurant, simple small business, etc) up and running pretty fast. It is not great (possibly horrendous) for revisiting a few years later when all the plugins/theme/core needs to be updated and a bunch of different people have added custom JS/CSS all over in random places.

It's hard to reconcile "pro web designer" with no-code though. Having said that, my preferred stack for something that involves more than a simple site is Sveltekit, Tailwindcss, & Sanity. But everyone has their own preferences.

carlosjobim• 3 years ago

You can't design anything in WordPress without code. Unless you count installing a theme as "design".

If you're a pro web designer you should know CSS. Then there's some good alternatives.

Giorgi2• 3 years ago

I would recommend Kirby, it is not all there, but is getting there.

infp_arborist• 3 years ago

Wordpress + Oxygen/Bricks/...?

bamfly8• 3 years ago

Wordpress is one of those things I've plenty of experience with, but leave off my résumé and avoid mentioning to bosses and colleagues, so I won't get roped into that bullshit if it comes up—and it often does, sooner or later.

"Wordpress? I can take a look if you need me to, but I don't really know much about that. I think Fred said something about Wordpress the other day, maybe ask him...."

I've known others who confided they do the same thing. I think it's fairly common.

Plus the PHP market, and especially the Wordpress market, doesn't tend to do great in the compensation department (short of making and selling a successful plugin yourself, anyway) so I'm not exactly eager to put more weeks/months/years of experience in that column when it's not worth much.

jchw• 3 years ago

Similarly, I do not mention to anyone that I have touched Drupal in the past.

cyrialize3• 3 years ago

I have thought about making a plugin.

The plugin documentation/architecture is a bit gnarly (at least from what I remember). That being said, I feel like most WordPress plugins are awful.

I've sometimes wondered if I could make a better version of X plugin, but I fear I'm missing something - like providing customer support or covering every version of WP.

devmor• 3 years ago

Plus the PHP market, and especially the Wordpress market, doesn't tend to do great in the compensation department

WordPress work certainly doesn't pay a lot, but for PHP in general there's a large cavern in the middle. You can get very, very high salaries/contract rates as someone experienced with PHP if you are extremely experienced with PHP - specifically because so few people are willing to stick with it to that point. There are just a lot of legacy systems in very large companies that run on it.

So if you already have the experience, it can be worth it. If you don't, then trying to get mid level or even "low senior" PHP jobs probably isn't worth it compared to another skillset.

throwawayxyz567• 3 years ago

I have spent a significant amount of time developing a product (developer tool) related to WordPress. I have put it on hold recently and I'm trying to evaluate if perhaps I should not do something else instead. I have had this feeling one year ago, that the WP development market may be shrinking and it was further amplified by ChatGPT release. Does anyone feel that way? Is the time still right to develop something, launch it in some 2 years and be sure that if the product has any potential, the market is still ripe in 5 -10 years from now?

ravenstine3• 3 years ago

I had a love-hate relationship with Wordpress from the beginning.

Yes, I too basically got my true start in dynamic web development by downloading Wordpress, running it under a WAMP/LAMP stack, toying with the code, and eventually writing some plugins. My coding skills back then were TERRIBLE, but even at that time I had a sense that Wordpress was actually quite a mess under the hood. The database design was pretty bizarre in particular. Fast forward almost 20 years later and I'm a senior web developer, and my opinion of Wordpress has not changed. That doesn't mean Wordpress is bad, because on the surface it does its job very well, but it was (and probably still isn't) the pinnacle of well-written software, and it's not a coincidence that there were vulnerability reports being issues all the time for it.

Wordpress was awesome in high school, the times before StackOverflow, MDN, YouTube, etc. It was a "framework" before frameworks were cool. One of the big selling points for Wordpress was that it was far more approachable than competing CMS softwares like Joomla and Drupal. I tried to figure both of those out back then, and I just utterly despised them. You could do a ton with Wordpress out of the box because, for the most part, you just wanted to have pages for things, a blog section, CSS, and a way to extend functionality. That said, I'm so glad I never have to use it again.

RobotToaster• 3 years ago

these days most of which seem to be trying to upsell via ads in the admin.

This is my biggest issue with wordpress.

So much that should be in core is left to plugins, some of which are semi-official like jetpack, that it feels a lot more like "open core" than "open source".

photomatt10• 3 years ago

We do need to figure out as a community how to deal with that upsell / aggressive plugin problem. I hope we can come up with some norms and standards for what's allowed in our plugin directory that is most customer-centric.

RadixDLT3• 3 years ago

yep, only allow high quality plugins that adhere to coding standards

also, WordPress should be behind a framework( symfony, laravel)

momojo• 3 years ago

Genuine question, can anyone explain what the practical applications of this are to someone who's never developed with Wordpress? Is this a gimmick? A game changer? A nice-to-have?

bdcravens• 3 years ago

There have been other solutions to run Wordpress locally or hosted without having to setup PHP and MySQL; this appears to be the first such first-party tool for doing so.

tyingq• 3 years ago

It could be pretty handy for WP training...fresh environment with whatever setup you want, separate for each trainee.

dawnerd2• 3 years ago

Seems like a great way to test plugins and themes in a clean environment. Premium themes could use this as their demo too instead of having to keep a demo Wordpress running.

vorticalbox• 3 years ago

If it's call client side couldn't someone just take the theme and not actually pay for it?

bamfly8• 3 years ago

Wordpress operates a plugin & theme marketplace. This could be very nice as a way to demo those products without adding meaningful infrastructure development and operational expenses.

[EDIT]: Very nice for WP/Automattic and the people who develop plugins to sell on that marketplace, that is.

8n4vidtmkvmk4• 3 years ago

The infra cost can be very tiny tho. Digital ocean has 1 click installers for WordPress, which you can use for a few hours and then shutdown and only pay a few pennies.

sourcecodeplz• 3 years ago

I think this is great! Congrats to the Automattic team for building it.

hbcondo714• 3 years ago

Very modern how they got this to work:

  PHP runs as a WebAssembly binary

  MySQL is replaced for SQLite via a WordPress plugin

  Web server is implemented in JavaScript as a Service Worker

calibas2• 3 years ago

Neat, and here's more info on how they did it:

WebAssembly PHP: https://wordpress.github.io/wordpress-playground/architectur...

Service Workers: https://wordpress.github.io/wordpress-playground/architectur...

radium3d2• 3 years ago

Interesting, seems the dedicated worker in chrome task manager is utilizing 1,135,948k of memory

revskill3• 3 years ago

When will i have useState in PHP ?

capableweb• 3 years ago

You have it already :)

    $_POST

revskill3• 3 years ago

_POST is enough ? Why $ here ?

calibas2• 3 years ago

Yes, it's outrageous that it's taken so long to implement React-specific functions in PHP!

slig• 3 years ago

Just use `$_SESSION`.

toddmorey• 3 years ago

"Whether you want to build a note-taking application for mobile devices, an automated testing environment, or demo a live WooCommerce store on your site – Playground will support you."

The last two make a ton of sense, but the idea of building a mobile app that's powered by Wordpress running PHP in webassembly in the browser is wrecking my brain.

coliveira• 3 years ago

It makes complete sense. Php web stack is very light compared to Python/node, so it makes sense it can be embedded in the browser.

SahAssar• 3 years ago

Except this is already running on a js environment in the browser (usually v8 if you use chrome, edge or any of their derivatives, the engine behind node). So you are running something to not run the thing you are already running.

WorldMaker• 3 years ago

There are plenty of Python apps now running in WebAssembly too. A lot of Node's stack directly runs in browsers (browserify is "ancient" tech at this point and influenced so much of modern "bundling" including Webpack and many things post-Webpack). (Deno's doing its own work with browser-compatible Node APIs, too.)

gnz112• 3 years ago

The part that doesn’t make sense is using Wordpress as an application framework.

bunga-bunga• 3 years ago

Finally some RAM competition to Electron apps!

lvncelot2• 3 years ago

We need an LLM layer for this to truly be web-scale.

photomatt10• 3 years ago

The notes app is a real thing! https://wptavern.com/blocknotes-app-runs-wordpress-natively-...

mst• 3 years ago

This is ridiculous and also somehow wonderful and I mean that entirely positively.

My compliments to everybody involved.

nickserv• 3 years ago

The elegant architecture of WordPress and its plugins coupled with the low resource use of client side rendering, what's not to love?

orkj3• 3 years ago

Genuinely curious: does this mean you can run WordPress as a cloudflare worker (ignoring database for now)? https://developers.cloudflare.com/workers/runtime-apis/webas...

miahwilde• 3 years ago

Note sure, but I like how you think.

there is this: https://ma.rkusa.st/store-sqlite-in-cloudflare-durable-objec... and this: https://developers.cloudflare.com/d1/

So I'm thinking it could (be made to) work.

orkj3• 3 years ago

I did know of those, I'm just assuming its going to cause some sort of problem with network calls and sockets and what not (they mention this specifically as part of the caveats)

modernerd2• 3 years ago

The team behind this tried to deploy to Cloudflare but hit memory constraints with Workers due to the 128MB limit there:

https://github.com/WordPress/wordpress-playground/issues/69

There’s likely more work that could be done to make this viable, though (or Cloudflare could raise limits).

_vOv_• 3 years ago

So what's the min limit required to make it run?

handzhiev2• 3 years ago

WordPress is an utter crap in terms of DB design and code. Most plugins are even worse and the whole thing is a total mess.

Yet, if you have a client needing this and that you can have them started really quick with WordPress. Need an e-commerce? There are plugins for that. Need an event management? There are plugins. Need an obscure app for truck drivers to share their CVs and communicate with clients? There is probably a plugin for this. If there isn't, you can use one and modify it - it's all open source.

It's definitely wild and ugly and crap, and sometimes is totally pathetic but the ecosystem is huge and has everything for everyone.

imilk4• 3 years ago

The backend is a terrible mess. But most of the time WP sites are served as cached static assets to the end user so you can still get the load times pretty good (despite the dumpster of a backend/DB).

handzhiev2• 3 years ago

When they are content sites, yes. When there is ecommerce and other dynamic stuff - not so much. I'm often optimizing WP sites after the arsenal of the typical developer - caching, redis object caching, etc - has been wiped off. The db structure and poorly optimized queries are a real pain.

graiz• 3 years ago

No need to dunk on wordpress but it has a lot of history and issues. It would be great to have a modern alternative self-hosted solution that doesn't require static site generation or complex hosting.

ehnto• 3 years ago

There are many functionally perfect substitutes, unfortunately the WordPress value proposition is actually it's ecosystem of plugins, developers and advocacy of it's users.

We've all met the client who wants WP and nothing else. And there are countless times we've rolled a clean new solution out, but the client goes back to WP.

I have seen value in hybrid approaches, a custom app written in some other framework for complex requirements, and a WP instance and connector for content, so the client is comfortable.

zztop44• 3 years ago

What are some of the best alternatives you know of? I know Statamic and Ghost both have their strong points.

bamfly8• 3 years ago

Any competitor has a nasty chicken/egg problem to overcome. Wordpress is still hard to avoid because there's a plugin—usually with paid support available—that can do damn near anything. The codebase isn't great, developing against it sucks, and the way it's designed and the official store curated does nothing to prevent some real messes & risks on the database side—but, two days of set-up and $80/m in plugins/themes is just a much easier sell than "OK, first we'll need at least two months of development time..."

There's also the scale-benefit that you can hardly throw a ball without hitting someone who's done Wordpress development, if you do need development work, and Wordpress-focused agencies are abundant, if you'd rather not hire to get the work done. Not true for any up-and-coming competitor.

Designing a system—even an extensible one with plugin and theming architecture—that is a lot better than Wordpress isn't a small task, but is also far from impossible and wouldn't take some super-genius team to accomplish; however, that's the easiest step on the path to actually displacing Wordpress in the market.

ipaddr• 3 years ago

There is a plugin that replaces wordpress itself

btown• 3 years ago

See also Salesforce and Shopify, whose app store/plugin ecosystems give them a network effect in their spaces in much the same way - but equally tie them to antiquated API surfaces due to their commitment to backwards compatibility.

quickthrower25• 3 years ago

Damn right. Forget nocode platforms, you can build sophisticated apps by choosing the right plugin combo. And edit some PHP files if needed (or add your own plugin for bespoke stuff).

blowski• 3 years ago

There are CMSes developers want to work with, and CMSes companies want to use. To get into the latter bucket, you have to make a lot of pragmatic choices that causes you to fall out of the former. Also, companies like things that have been around a long time, and such things tend to be old-fashioned by definition.

EGreg4• 3 years ago

Ok. Here it is:

https://github.com/Qbix/Platform

rapnie2• 3 years ago

Oh no. Still the mouse movement animation on the website? It made me not read further down the page.

throw_m239339• 3 years ago

Wordpress is the success it is today because over the years, while other CMS kept breaking API, Wordpress remained mostly stable, allowing the development of a large ecosystem of plugins.

Another thing is that Wordpress PHP API uses very little OOP. It's mostly functions + arrays. So people who know very basic programming could write their own plugins, themes and co.

Reptur3• 3 years ago

Its cool and I get it, but how is this useful other than good marketing? If an app's code is run on a device, any rooted device could just modify its memory and do whatever it wants. Just ask a game developer about the infinite anti-cheating war that they deal with.

desas2• 3 years ago

Could you expand? I don't follow.

WordPress is open source, you can already download it and do whatever you want.

Reptur3• 3 years ago

My point was no one would ever run a web app's backend in the browser. Not speaking specifically to Wordpress.

wzy2• 3 years ago

...any rooted device could just modify its memory and do whatever it wants

Like what? Change a heading? Or, create a blog post that only exists on this device?

Reptur3• 3 years ago

My point was no one would ever run a web app's backend in the browser. Not speaking specifically to Wordpress.

dboreham5• 3 years ago

Almost like a real computer.

wouldbecouldbe2• 3 years ago

Php & WordPress were fun. With that ugly php tag you could get pretty much everything working on the web.

All you had to do was edit, save & refresh. No rebuild of code, server restart, compiling Just edit & save.

Did it have the tendency to get ugly? Sure. But it was easy, quick and fast. Not every blog and needs a git flow, united, e2e, type-safe double encrypted dev environment.

Both PHP & WordPress got shamed for their dirty nature so much that they are in identity crisis. PHP tries to mimmick Java as much as possible. WordPress plugin structure is a disaster. Having to run an entire npm project for every plugin, no easy data saving helpers, etc. Let alone the ux-disaster of the block Editor.

Feels like the php decade is truly over. And it was php that killed itself.

Laravel is a joy btw.

8n4vidtmkvmk4• 3 years ago

Laravel is not a joy. It moved everything in the wrong direction by adding a ton of magic (which murders IDEs) and a ton of version churn.

redox99• 3 years ago

All you had to do was edit, save & refresh. No rebuild of code, server restart, compiling Just edit & save.

The majority of webdev is either exactly like that, or effectively like that with very fast rebuilds/restarts. It's not really something unique to PHP.

In fact in your modern stack you'd probably have autosave, and most importantly "fast refresh", so you don't need to manually refresh and you don't lose state.

wouldbecouldbe2• 3 years ago

Nah you don't yet it.

Nowadays there is so much build config you need 4gb just to run wepback.

Php is just a save of a file. That's it's.

So for instance we would use remote file save. Edit & save via ftp. Most editors have such a plugin. Then you could work remote easily.

But also hosting php is 10 times easier and stable then node. Php servers don't run stale or crash. Because there is no server.

gamblor956• 3 years ago

WordPress exposes a huge divide in the HN crowd:

- one group recognizes WordPress for what it is: a value-multiplier used by non-coders for websites so they don't have to deal with the coding and can focus on their core competencies

- the other group hates WordPress because of the ugly codebase, and meanders off to work on their own perfectly formatted, ultra-scalable code that has no little to no value because it's never actually going to be used by anyone outside of their startup.

WordPress powers more than half of the Web. It might not be the best tech, it might very well be the worst possible tech, but the lesson that so many programmers continuously fail to take from WordPress is that NOBODY CARES HOW THE SAUSAGE IS MADE.

scarecrowbob2• 3 years ago

This is a reasonable position and with the utility sites I build for, say, bands I play with I totally agree. I toss some shit on commodity hosting and hope the band breaks up before the gcal integration breaks.

-=-=-=

However.

I will now take this opportunity to vent a bit, though it has little to do with your imminently reasonable post. Mostly because I am procrastinating on porting a 3rd-party SSO system from a very old and shitty WP site to a new and shitty WP site and it's hurting my brain to decide how much of the mess of jQuery and bootstrap I really need to keep.

So hopefully this will be understood as venting and not a personal attack against your reasonable observation.

-=-=-=-

Consider that there are those of us who have to fix the first group's problems and make things keep working so they don't have to care how the sausage is made.

That is to say, there is a very real third group of folks: those of us who support the first group.

=-=-=-

I've been doing that work for about 14 years for all kinds of folks. I host 300 or so WP sites for a mid-sized university, I built and maintain a .gov site for a three-letter-agency, and I've built plenty of react-based blocks (across 3 majorly different patterns because the GB team doesn't give a fuck about anything, as far as I can tell).

I've worked on every element of WP from deploying servers to fixing CSS so it works in IE6, to scraping non-availible CMSs so that I can migrate thousands of pages and tens of thousands of images to WP. I've written plenty of WPCLI commands to do dumb tasks, I've written plenty of shit-tier code to make, say, calendar events propagate across the bad-for-most-purposes multisite functionality.

=-=-=-

Someone has to know how the sausage is made. I do.

And I know the difference between the teetering pile of steaming crap that is WP.

I've worked on actually useful codebases in my life with tooling that is easy and databases that are grokkable. I am aware that other platforms have issues, and have encountered them in my work.

However, WP is really bad.

And I have no problem bitching about it- I'm burnt the hell out and am about ready just quit and live in my truck and play music.

It's a shitty platform, and people do not pay well to keep it working.

=-=-

So anyhow, yeah, you've got a fair point, but there those of us who, because of circumstance, keep heading back into the burning building that is WP and our collective hatred of the platform is well-founded.

photomatt10• 3 years ago

I'm happy to pitch in and help migrate those university sites, we can get them on a WordPress stack that's much more stable and forward-compatible.

rchaud• 3 years ago

I think the divide is more between the new generation of full-stack JS developers and those that cut their teeth on PHP + MySQL.

dgb232• 3 years ago

WP is fantastic for the first group you mentioned. I put it up there with Excel, Filemaker, Visual Basic and similar software that democratize software and make it accessible to everyone.

Naturally all of these things have limits and quality issues that some percentage of users eventually face. That is in of itself fine.

The problem arises when it is expected of a professional to integrate or build on top of these.

People come to web developers with quality expectations, unique problems and requirements. But the legacy that is WP puts so many stumbling blocks and gotchas in the way of these goals that it's not worth it to put a square peg into a round hole.

(Similar can be said about common CRM and ecommerce platforms and solutions.)

sircastor• 3 years ago

meanders off to work on their own perfectly formatted, ultra-scalable code

Or stops working on their current, imperfect codebase to start a new project that will be perfect… that’ll show ‘em.

dboreham5• 3 years ago

I'm a serious software engineer (I think...) but I use WP for my corporate web sites.

timeon2• 3 years ago

ugly codebase

Ugly codebase is often popular choice - ie. Tailwind CSS.

ravenstine3• 3 years ago

This playground, which is running PHP in WASM, is much snappier than 95% of garden variety websites.

noahtallen• 3 years ago

I think the coolest thing about this is you avoid so many round trips to the server. Every single database call and asset request happens on the client! (After you accept the several seconds of loading time.) I think it's also hilarious, because WordPress is probably the most-used MPA solution out there, but this basically turns it into an SPA. And it performs even better than most SPAs, because it doesn't need to do API requests to a server constantly.

I also think the SPA parts of this (the post/page and site editors are very complex React applications) feel snappier in this demo than they do on my local dev environment. Really shows how good SPAs can be when you cut out server round trips.

Disclaimer: I work for Automattic, and this has been a super cool experiment to watch.

wg0• 3 years ago

Reporting from a budget Android Tablet with Firefox, the response times aren't exactly instantaneous but aren't any different than running full LAMP on a reasonable cloud instance or VPS which is surprising to say the least.

Also imagine the software layers at play here. Whole PHP Interpreter plus whole thousands of lines from WordPress codebase plus SQLite.

Amazing when that's all so much totally useable.

Saris2• 3 years ago

I've always found WordPress performance doesn't change much on lower vs higher end hardware. It seems to be limited by itself.

WhackyIdeas2• 3 years ago

Absolutely. The other day, I took my slow Wordpress server hosted at Vultr on bare metal (£200/pm) and put it on a server at £500 per month and even after doing all the performance tuning I could with MySQL and PHP I couldn’t notice a difference in speed when I was on the back-end where Cloudflare wasn’t caching it.

I then tested on a lower performance server which costs £80 per month at Vultr and again couldn’t detect any difference when in the back-end.

No matter what performance tuning I done, I couldn’t make it work faster for us. But at least I am on an £80 server instead of a £500 server I guess.

judah• 3 years ago

What's the answer to making WordPress sites faster?

I too have tried better hardware, CDNs, plugins like WP Super Cache, and others means of optimizing. Very little impact. Is the problem PHP?

I'm thinking of trying one of those plugins that convert the WordPress site to static web site. I'd imagine that'd have a big impact, at least for sites that don't need e.g. login, shopping carts, etc.

joshspankit• 3 years ago

Yes, but: Number of concurrent visitors can change dramatically on lower vs higher. A caching plugin mitigates this somewhat, but I find that most companies don’t test whether their site can handle surges, or even if it can handle the projected traffic load.

sublinear2• 3 years ago

Can we just throw wordpress in the trash instead already?

It sucks at everything. It's not a good CMS, its plugins are awful, PHP is a dumpster fire, every web developer hates it, etc.

This is also pointless as a tech demo.

wzy2• 3 years ago

That must explain why WordPress' market share has been trending downwards for the last 20 years, right?

sublinear2• 3 years ago

Yes! Ignorance absolutely explains its "market share". There's no reason for anyone to be running this trash in 2023. I'm surprised Shopify hasn't completely wiped out Wordpress for ecommerce yet. For all other kinds of websites, they don't even need a CMS. Blogs are dead so it's all niche sites that demand custom web dev by now.

geocrasher2• 3 years ago

I'm curious as to how useful this will be. Tying it to SQLite makes it portable, but abstracting it so that it can be used with any db would be far more powerful. Why? Because the SQLite implementation is enough to get it working in the browser, but it does not make WordPress 100% SQLite compatible. For example, a WordPress site built on MySQL can't be migrated into an SQLite powered WordPress using a common migration plugin. I know, because I tried.

I installed WordPress in a standard PHP environment, and use the db.php dropin and SQLite compatibility plugin, and actually got WordPress installed on the file system. It works, but when I tried to migrate using all-in-one-wp-migration, it failed.

A neat concept with niche use, but not a panacea for agencies and other devs who need an extensive playground, even if it's self hosted on a persistent PHP server.

dmsnell3• 3 years ago

the Playground can connect to a MySQL instance when running in a Node environment. it's not supported in the browser because there's no existing mechanism to relay the necessary socket connection there.

danielbachhuber• 3 years ago

For example, a WordPress site built on MySQL can't be migrated into an SQLite powered WordPress using a common migration plugin. I know, because I tried.

Out of curiosity, how recently did you try?

We're using a new SQLite integration that seems to be pretty good at translating SQL to SQLite: https://github.com/WordPress/sqlite-database-integration

We'd be curious to learn more about how it failed. It's definitely an area we'd like to improve.

geocrasher2• 3 years ago

Ok, didn't expect that response! I tried this today.

I admit I didn't try super hard- I have a lot of other things to work on, was a 20 minute curiosity mission. I just downloaded the zip file and grabbed the files out of that. Don't get me wrong- it worked- but I couldn't do an import as mentioned.

If you want lots of detail, I can spend more time on it. Email me at my hn user at the google mail.

Giorgi2• 3 years ago

As a side note, I wonder if WP will still be in demand after all these chatbots come into daily usage

butz• 3 years ago

I wonder if this could be useful for penetration testing, considering the ease of setting up different plugin combinations on different PHP and Wordpress versions? Or stack used to run it is way more different from real one?

jacooper• 3 years ago

This would be killer for static websites, make all your modifications on the /admin endpoint and then deploy to your static page host, using a plugin like staatic.

synergy20• 3 years ago

Gosh this is truly snappy, does that mean we shall run all the stuff on WASM now? very impressive.

ComputerGuru• 3 years ago

Really cool tech behind this demo! I wish I could see the realtime php error/warning log (configured to log everything, of course); it would be interesting to see how perfect this simulator is, especially with the SQLite db rewrite.

It's also awesome how easy it is to use this to test out a plugin (or a theme) - you just need to add &plugin=plugin-slug-from-dir where the slug matches the plugin's WordPress.org url and it'll be automatically downloaded and added to the sandbox!

The full query api used to configure the sandbox (tech and loaded env) is documented here [0].

[0] https://wordpress.github.io/wordpress-playground/query-api/

EDIT

I'm absolutely positive there has to be a way to open or show a debug log (if only because I wouldn't have been able to develop this without that)!

EGreg4• 3 years ago

The question is — how does your WebAssembly in the browser run a server that can listen for incoming requests and serve the page? I don’t get that part at all.

sylvinus• 3 years ago

I think this is the part done with the Service Worker, which intercepts network requests: https://developer.mozilla.org/en-US/docs/Web/API/Service_Wor...

EGreg4• 3 years ago

But how can anyone other than the user actually access this server? Like if I serve wordpress off a phone?

inportb• 3 years ago

Ooops! I suppose it does not work in private browsing mode.

adamziel2• 3 years ago

Ooops! Would you file an issue on GitHub? This would really help us track it and fix it.

nilslindemann• 3 years ago

Apply the following styles, and it will become a practical tool:

  ._wrapper_5didy_26 {
    padding: 0;
    height: 100%;
  }

  ._window_5didy_31 {
    max-width: none;
  }

  ._experimental-notice_5didy_9,
  ._toolbar_5didy_64 {
    display: none;
  }

adamziel2• 3 years ago

Or add `?mode=seamless` to the URL:

https://playground.wordpress.net/?mode=seamless

See https://wordpress.github.io/wordpress-playground/query-api/ for more cool options!

web3-is-a-scam• 3 years ago

Just because you can doesn’t mean you should.

photomatt10• 3 years ago

I tried to make a big deal about this in my State of the Word keynote in December 2022: https://wordpress.tv/2023/01/04/matt-mullenweg-state-of-the-...

It kind of blew my mind as an engineer to see it work, and know all the layers that had to interact for that to happen.

It's fun to see it blowing up on HN 6 months later! :) Thank you to everyone for checking it out.

quickthrower25• 3 years ago

Nice talk! For anyone curious, 48:33 is the timestamp where Wordpress Playground is introduced

dboreham5• 3 years ago

It's very interesting to me since I have an interest in porting various hitherto "server-side" applications to run in-browser. Doubly interesting because it's the first real product I've seen running SQLite -- so far I've only seen toy demos.

But, wondering about something -- poking around in the code a bit, and realizing I know next to nothing about present-day php, I realized this may be not using the "traditional" SQLite-over-OPFS approach to get persistence?

Actually, after a bit more reading it's not clear to me that persistent storage is available yet via Emscripten.

And after yet more poking around I found: https://github.com/WordPress/wordpress-playground/issues/19 which I guess means that persistence isn't yet possible for WP in the browser :(

lioeters3• 3 years ago

There is partial support for persistence, for browsers that have origin private file system (OPFS) implemented.

OPFS is only supported in Chrome-based browsers at the moment like Edge, Android browser. Safari and Firefox users won't be able to benefit from this feature yet

Persistent Playground: Sync between MEMFS and OPFS - https://github.com/WordPress/wordpress-playground/pull/546

quickthrower25• 3 years ago

RIP static site generators. Just run this in your browser and export a static site. No more complicated setups. No more running ruby in docker or fighting hugo templates. Install the themes and plugins and go nuts!

/hottake

Tijdreiziger• 3 years ago

Tools for that purpose already existed, e.g. WP2Static.

quickthrower25• 3 years ago

Sure but you had to install wordpress on your PC: hard or server: easier but cost $ on top of the static hosting bill

snickmy2• 3 years ago

how would you export a static site if you are running on the browser? you can do some cache management with local storage, or just leverage the browser built in system, but as every rendering session is on the edge, you would not have the benefit of a static site.

Or are you talking about, developing it on the browser, export it and upload it somewhere as static files for internet access (eg a cheap host with no server side languages)?

neets• 3 years ago

What a time to be alive

pier25• 3 years ago

I don't have a very good opinion of Wordpress but this is super cool

astatine• 3 years ago

I use WP all the time for a variety of nearly static sites that I can let someone else non-technical continue to add content. It works great for this purpose. There is nothing close to WP for this purpose.

It does need careful curating of plugins that one uses, set up suitable cloudflare frontends and preventing users from doing something seemingly simple but dangerous.

I get the feeling that a good chunk of the hate comes from people who tried it many years ago; or had needs which don't get met by WP; or received it as a legacy.

lvncelot2• 3 years ago

A combination of the second, third and a client who insisted on using it for a project where a headless CMS + custom code would have worked a lot better did the trick for me.

Still not really WP's fault, but it doesn't change the fact that I recoil a bit in horror when it's mentioned.

snickmy2• 3 years ago

Hijacking the spirit of this link to just saying that I'm very sad about what Wordpress has become. In my mind it was a very good website / blogging platform, with a clear and functional markup / raw code editor.

Building themes was easy, and so was the plugging ecosystem.

I installed the latest version ~6 months ago, and I found myself overly confused by the million of options, the poor functionality of the block system, and the overall bloating of the software. It just run slower than 10 years ago. The templates and plugin available out there are extremely complex and a lot harder to read, code wise.

I assume is a 'dead by 1000 papercuts, or customer demands, in this case' but from a product standpoint, I'd really love to know who are the personas that this new interpretation of wordpress cater to.

globalise83• 3 years ago

Just tried installing woocommerce and adding a product, but got a bug:

Warning: require(/wordpress/wp-includes/post.php): Failed to open stream: No such file or directory in /wordpress/wp-settings.php on line 2

Fatal error: Uncaught Error: Failed opening required '/wordpress/wp-includes/post.php' (include_path='.:') in /wordpress/wp-settings.php:2 Stack trace: #0 /wordpress/wp-config.php(11): require_once() #1 /wordpress/wp-load.php(2): require_once('/wordpress/wp-c...') #2 /wordpress/wp-admin/admin.php(2): require_once('/wordpress/wp-l...') #3 /wordpress/wp-admin/post.php(2): require_once('/wordpress/wp-a...') #4 {main} thrown in /wordpress/wp-settings.php on line 2 There has been a critical error on this website.

lioeters3• 3 years ago

It would be helpful if you could report this to the issue tracker.

https://github.com/WordPress/wordpress-playground/issues

flaxton• 3 years ago

This looks interesting, but a better option in my opinion is to use the Local app (on Mac) to install a lightweight VM running WordPress in seconds. Then you have all the tools, for example phpmyadmin and MySQL database management, all running locally on your Mac, and with Internet access. Open Source and free to use.

https://localwp.com

apesnape• 3 years ago

I have checked out the approach. As well as the guidelines. I agree to all. More sites should use this.

WordPress Playground: A WordPress that runs entirely in the browser

Related Stories