I am still not over the collapse of Gatsby (for reference i nearly joined the company in 2017, was saved only by my own character flaws).
They rode the highs - being the default docs tool for React, and building a massive ecosystem of integrations you could install out of the box. But too many abstractions, divided goals between cloud and OSS, and the better stewardship/design of Nextjs brought it down.
There were the simple lessons (https://swyx.io/a-world-without-plugins-cig). its easy to say in hindsight that graphql was too much for gatsby. i also believe the company went too hard for number of integrations over quality of them, an issue I had even in my interview. this was a poorer expression of the better insight that seb markbage had; just have a small api surface area bro (https://www.youtube.com/watch?v=4anAwXYqLG8)
But the bigger lesson is bitterer. Frontend tooling isnt worth that much. the fact that vercel is pretty much the only successful frontend startup of its generation makes it an exception to the rule (there are plenty of smaller companies that are thriving, like tailwind, but it is not a venture scale startup and thats fine). People dont pay for frontend tooling. they expect it to be free, expect it to do everything, get into internicine squabbles between frameworks when they are all basically doomed compared to just betting on React sponsored by Facebook and now Vercel (and a little bit of Shopify), or going back to fullstack frameworks like Django/Rails/Laravel. all frontend tooling, nextjs included, is just leadgen, loss leaders, while investors/salespeople patiently wait until you "grow up" by... building cloud backend/ci/cd services.
5 years ago i wrote about the "frontend ceiling" for individual developer careers (https://x.com/swyx/status/1682748872047886337) - i fear this is the "frontend ceiling" for companies.
I deeply admire Astro and hope they figure out a way to break the ceiling. Their recent cloud products have been encouraging.
I am for one surprised that the initial concept of Gatsby -- a static site generator with React, GraphQL and other hype words of late 2010s -- was actually able to raise $40+ million dollars to build a platform for what to me at the time seemed incredibly over-engineered stack for static html pages and an unproven founder.
> But the bigger lesson is bitterer. Frontend tooling isnt worth that much.
This is a modern oddity, as historically front end tooling was worth a lot.
During the 80s and 90s, multiple companies made a lot of money off of front end tooling. Even as late as 2009 or so, Silverlight tooling was bringing in money for Microsoft. But think of how popular Flash was and the empire that helped build. Prior to that. Visual Basic helped Microsoft take over the world. Go back another generation and Borland's Delphi dominated for years.
And, arguably, all of those systems were more productive than anything we have now. (A topic I've written about many times!) As an industry, we may indeed be getting exactly what we pay for.
Astro is what I wanted and thought Gatsby would be. GraphQL was the wrong sized abstraction for a basic blog site like mine. And once Gatsby turned towards the cloud (which made sense as a business), everything just got so complicated. I was swept up by it too.
Cool insight on frontend ceilings -- deployment is king.
I think frontend tooling is valuable, but there’s so much good free out there that it’s hard to monetize.
Gatsby always seemed wildly overengineered to me. It’s like all the bad parts of react, with all the bad parts of complicated js tooling, and none of the upsides.
Is any tooling worth that much, really? I think the key is that backend tooling lends itself to selling hosted solutions, while frontend tooling is usually stuff you run locally or on CI to build artifacts. Sure, there are exceptions, but most people don't pay for developer-focused software anymore, at least not directly.
My bet is on Remix (which will eventually just be React Router). I think it's just the right amount of abstraction over native web functionality with a nice, simple API that doesn't deviate too much from web standards. I also think the full stack approach just makes sense for developing on the web.
Astro is fantastic. I initially used NextJS with MDX, and something simple like using relative markdown links to images was utterly impossible.
I remember going down a day long rabbit hole to understand why, and it boiled down to content layer, MDX and NextJS using different, incompatible module loaders, bundlers or transpilers of some kind[1][2][3]. Ridiculous. And don’t get me started on the image component.
In the end, Astro just works. No need for React (unless you want it), it’s simple, fast and produces a static site you can use without JavaScript.
Data fetching is also utterly trivial, so you can have a fully static site with “live” data pulled in at compile time.
My personal website was originally written this way. The compile and build script was about 20 lines of shell scripting (relying on pandoc), and could have probably been reduced further. It worked very well and I stuck with it for about 2 years before moving on to Hugo. It was a good experience to hand-roll all the components of a static site, and something every web dev should be required to do early in their career.
I don’t love js/ts but Astro is so nice. Don’t have to fiddle with react or graphql. Just upgraded to 5 beta and nothing broke. Really well managed project. Oh, and page load speed is insane.
If you don't love js/ts maybe you'd like Civet, which transpiles to js/ts and has Astro integration. (No relationship, I've just been looking at it myself.)
I have heard great things about astro, but my experience with Gatsby keeps me away. I would much rather keep an SSR framework like Remix and handle whatever pipeline is needed myself. No magic.
I felt the same way! But I came around. Take a look at the source code for a component [1]. The .astro format is pretty much just TS, CSS, and HTML. I don't even need React anymore. Way less glue code and dependencies than with Gatsby. There is a configuration file and some abstractions for things like collections, but almost everything is opt-in.
Fully concur with the points you made. I also just finished transitioning my personal website to Astro from an Eleventy setup that I got too lazy to maintain. I made a couple of landing pages for clients and my own projects with Astro and the maintenance experience was eye-opening.
Astro really is quite wondrous. If you don't stray too far from its opinionated route, it's almost magical. But some more advanced things do still take some tinkering. My sitemap and RSS with full text setups I wouldn't describe as the most elegant things in existence, but they do their job.
What I'm still lacking/wondering about is whether there's a CDN static site/serverless hosting provider like Netlify, Vercel or Cloudflare, but that isn't free or $20/mo. I'd like to pay for a good service that obviously has running costs, but I don't believe there's quite something like those.
Anyway, I like the website. It's easy on the eyes.
Thanks for sharing! I use Fireabse Hosting, which I guess gets grouped in with the other services you mentioned. It's free, has a lightweight devX, and integrates with CI via GitHub actions for preview and production builds [1].
This looks lovely! I've been recently working on sth similar: building a custom pipeline from my Obsidian notes into a website (untested.sonnet.io).
I ended up using 11ty because I wanted to stick as close to the web platform as possible, but a part of me wishes I had tried Astro as well. Mainly because I feel that it strikes the right balance when it comes to flexibility and boilerplate. 11ty is lovely, but there was soooo much stuff I had to write from scratch. I feel like it paid off, but it took me much longer than I had hoped.
One surprising side effect: I noticed that the site worked faster online than served from localhost. 5 minutes of digging after I remembered that it's served via HTTP2 w. multiplexing. I'm not even interested in adding a bundler/minify step, just plain CSS/HTML and asset optimisation w. 11ty-image is enough.
Plus one for eleventy. Love its minimalism and flexibility, although it means that there isn't a clear guidance on assets management. It's a shame that the web still doesn't have a standard way of authoring components, and different frameworks and static-site generators, including eleventy, have to invent their own.
Astro is great. I love that I can mix and match between markdown, plain HTML, and Svelte components as needed (and even intermix them within a page). Pretty much the sweet spot I always wanted for a blog or marketing site: 98% static stuff, 2% sprinkles of JS.
In early experiments it feels like Inertia.js be this same feeling but for Rails and Laravel.
I never got very far on the Gatsby train. It always felt so slow and fragile during local dev, and while GraphQL was cool it felt like total overkill for most of the data I needed to grab (like… markdown files from disk).
Not a full replacement in all scenarios. But for my personal website with a basic blog, Astro is by far the simpler and more maintainable option. I bet that is true for most people here who maintain their own personal website.
This is a bit misleading. You are only prevented from using UI libs that use a Provider/Context (like Mantine does) since Astro does not support Context.
Things like shadcn/ui work fine, even when statically rendered.
You could just slap on any library like Swup.js, Turbo, etc. and get SPA-like fast navigation out of a regular "MPA" Astro site. And that's true for any MPA, actually.
You can achieve SPA-like navigation in Astro with their transitions API, but if that’s what you’re after it’s better to use Next.js or Remix IMO.
(I helped build playboy.com/magazine with Astro, which uses the transitions API for this SPA-like behavior… as well as playboy.com/app (nsfw) with Next.js)
~4 years ago, redesign phase of our 50K a month public facing wordpress NGO website, agency strongly suggests moving to gatsby. We/I thought about it for 5 minutes and said no. I am still happy with that decision and I will push to take the same if/when another agency comes at us with Astro. Node CMS are autophagous.
I spent the entire time building a business network platform with Astro instead of WordPress or other web frameworks because it is exactly what I need to easily manage UI components and provide a safer JSX-like environment.
In comparison to Ruby on Rails, Django, and Laravel, Astro stands on the same level but much less learning curve.
Astro is a step in the right direction. But I wonder why in 2024 we still have to deal with proprietary component languages? I would like something like Astro with web components. Perhaps with the help of lit.
With this the pure markdown content could be easily enriched with components.
This is my dream also. I much much prefer lit-html syntax to Astro’s.
I even found standard declarative shadow dom style components to be a pain in the ass in Astro which left a bad taste in my mouth since that’s literally just stock standard HTML.
Great write up, will be curious to see if Astro sticks around. It seems like all the static site gens last a few years then fade away - Jekyll, Hugo, Gatsby, etc.
Me too! The closer my framework is to raw HTML, JavaScript (really TypeScript nowadays), and CSS, the less lock-in I feel. Those 3 languages are continue to get better year over year. And we are get better framework abstractions every few years too. Astro feels like it hit the sweet spot, but maybe there is more to come.
Gatsby gets shit on a lot - which I think is a but unfair. I find it to be a solid, well-documented platform. It has some rough spots, but show me a platform that doesn't.
The author also complains that:
> I couldn’t build the site locally or upgrade dependencies, leaving it outdated since 2019
This is true of many things in the javascript ecosystem. In react/node/js-land you HAVE to invest in maintenance (and re-writes) or be left with a legacy codebase that just might not work one day if npm install fails.
swyx|1 year ago
They rode the highs - being the default docs tool for React, and building a massive ecosystem of integrations you could install out of the box. But too many abstractions, divided goals between cloud and OSS, and the better stewardship/design of Nextjs brought it down.
There were the simple lessons (https://swyx.io/a-world-without-plugins-cig). its easy to say in hindsight that graphql was too much for gatsby. i also believe the company went too hard for number of integrations over quality of them, an issue I had even in my interview. this was a poorer expression of the better insight that seb markbage had; just have a small api surface area bro (https://www.youtube.com/watch?v=4anAwXYqLG8)
But the bigger lesson is bitterer. Frontend tooling isnt worth that much. the fact that vercel is pretty much the only successful frontend startup of its generation makes it an exception to the rule (there are plenty of smaller companies that are thriving, like tailwind, but it is not a venture scale startup and thats fine). People dont pay for frontend tooling. they expect it to be free, expect it to do everything, get into internicine squabbles between frameworks when they are all basically doomed compared to just betting on React sponsored by Facebook and now Vercel (and a little bit of Shopify), or going back to fullstack frameworks like Django/Rails/Laravel. all frontend tooling, nextjs included, is just leadgen, loss leaders, while investors/salespeople patiently wait until you "grow up" by... building cloud backend/ci/cd services.
5 years ago i wrote about the "frontend ceiling" for individual developer careers (https://x.com/swyx/status/1682748872047886337) - i fear this is the "frontend ceiling" for companies.
I deeply admire Astro and hope they figure out a way to break the ceiling. Their recent cloud products have been encouraging.
vladgur|1 year ago
It was truly a testament of times
com2kid|1 year ago
This is a modern oddity, as historically front end tooling was worth a lot.
During the 80s and 90s, multiple companies made a lot of money off of front end tooling. Even as late as 2009 or so, Silverlight tooling was bringing in money for Microsoft. But think of how popular Flash was and the empire that helped build. Prior to that. Visual Basic helped Microsoft take over the world. Go back another generation and Borland's Delphi dominated for years.
And, arguably, all of those systems were more productive than anything we have now. (A topic I've written about many times!) As an industry, we may indeed be getting exactly what we pay for.
epolanski|1 year ago
I think Typescript's websites uses it (or used to), and I wanted to fix a bug in the website and holy hell if Gatsby made me dislike the experience..
jwngr|1 year ago
Cool insight on frontend ceilings -- deployment is king.
dartos|1 year ago
Gatsby always seemed wildly overengineered to me. It’s like all the bad parts of react, with all the bad parts of complicated js tooling, and none of the upsides.
Never made sense.
kelnos|1 year ago
Is any tooling worth that much, really? I think the key is that backend tooling lends itself to selling hosted solutions, while frontend tooling is usually stuff you run locally or on CI to build artifacts. Sure, there are exceptions, but most people don't pay for developer-focused software anymore, at least not directly.
mmckelvy|1 year ago
madeofpalk|1 year ago
I've always seen them as a AWS-wrapping hosting company, who funds some react projects on the side and will drop them at a moment's notice.
Netlify, who bought Gatsby on its decline, still seems to be going alright without being a frontend tooling company.
victorhooi|1 year ago
Did something happen to them, in the community or project?
Googling, I saw that they were acquired by Netlify in early 2023 - but not much concrete beyond that?
orf|1 year ago
I remember going down a day long rabbit hole to understand why, and it boiled down to content layer, MDX and NextJS using different, incompatible module loaders, bundlers or transpilers of some kind[1][2][3]. Ridiculous. And don’t get me started on the image component.
In the end, Astro just works. No need for React (unless you want it), it’s simple, fast and produces a static site you can use without JavaScript.
Data fetching is also utterly trivial, so you can have a fully static site with “live” data pulled in at compile time.
I recommend it to anyone with frontend fatigue.
1. https://stackoverflow.com/questions/63957018/how-to-use-imag...
2. https://mmazzarolo.com/blog/2023-07-30-nextjs-mdx-image-sour...
3. https://github.com/contentlayerdev/contentlayer/issues/11
indigodaddy|1 year ago
remipch|1 year ago
- create a simple html template using simple.css [0]
- write markdown files
- wrap pandoc [1] in a simple bash script to manually convert markdown to html
- and that's it.
By minimalist, I mean: no script, no component, no database, no react, no SEO.
The result is a minimalist website that you write in markdown.
It's very limited compared to full-featured frameworks, but it can do the job for a simple website.
Here is mine: [2] (I'm not a web developer at all).
[0] https://github.com/kevquirk/simple.css
[1] https://pandoc.org/
[2] https://remipch.github.io/
syndicatedjelly|1 year ago
dartos|1 year ago
I like hugo.
Good tempting language and taxonomy features.
indigodaddy|1 year ago
cpursley|1 year ago
garyrob|1 year ago
https://civet.dev
jwngr|1 year ago
dimitrisnl|1 year ago
jwngr|1 year ago
[1] https://github.com/jwngr/jwn.gr/blob/master/src/components/b...
jama_|1 year ago
Astro really is quite wondrous. If you don't stray too far from its opinionated route, it's almost magical. But some more advanced things do still take some tinkering. My sitemap and RSS with full text setups I wouldn't describe as the most elegant things in existence, but they do their job.
What I'm still lacking/wondering about is whether there's a CDN static site/serverless hosting provider like Netlify, Vercel or Cloudflare, but that isn't free or $20/mo. I'd like to pay for a good service that obviously has running costs, but I don't believe there's quite something like those.
Anyway, I like the website. It's easy on the eyes.
jwngr|1 year ago
[1] https://github.com/jwngr/jwn.gr/tree/master/.github/workflow...
rpastuszak|1 year ago
I ended up using 11ty because I wanted to stick as close to the web platform as possible, but a part of me wishes I had tried Astro as well. Mainly because I feel that it strikes the right balance when it comes to flexibility and boilerplate. 11ty is lovely, but there was soooo much stuff I had to write from scratch. I feel like it paid off, but it took me much longer than I had hoped.
One surprising side effect: I noticed that the site worked faster online than served from localhost. 5 minutes of digging after I remembered that it's served via HTTP2 w. multiplexing. I'm not even interested in adding a bundler/minify step, just plain CSS/HTML and asset optimisation w. 11ty-image is enough.
azangru|1 year ago
dceddia|1 year ago
In early experiments it feels like Inertia.js be this same feeling but for Rails and Laravel.
I never got very far on the Gatsby train. It always felt so slow and fragile during local dev, and while GraphQL was cool it felt like total overkill for most of the data I needed to grab (like… markdown files from disk).
toinewx|1 year ago
Astro is MPA while Gatsby provided SPA-like navigation.
That's why Astro is not really a replacement.
jwngr|1 year ago
cmgriffing|1 year ago
Things like shadcn/ui work fine, even when statically rendered.
jaredcwhite|1 year ago
In addition, I believe Astro now has their own solution for this, though I haven't tried it personally: https://docs.astro.build/en/guides/view-transitions/
kcrwfrd_|1 year ago
(I helped build playboy.com/magazine with Astro, which uses the transitions API for this SPA-like behavior… as well as playboy.com/app (nsfw) with Next.js)
johnchristopher|1 year ago
Already__Taken|1 year ago
todotask|1 year ago
In comparison to Ruby on Rails, Django, and Laravel, Astro stands on the same level but much less learning curve.
ulrischa|1 year ago
mdhb|1 year ago
I even found standard declarative shadow dom style components to be a pain in the ass in Astro which left a bad taste in my mouth since that’s literally just stock standard HTML.
skwee357|1 year ago
It’s very easy to build a static website with zero js, or pick the desired framework you want to work with.
Pair it with TypeScript and you get almost 100% type safe templating and resulting websites.
SrslyJosh|1 year ago
So...a website. weary emoji
joshvince|1 year ago
abeisgreat|1 year ago
jwngr|1 year ago
mediumsmart|1 year ago
schpet|1 year ago
jwngr|1 year ago
hallman76|1 year ago
The author also complains that:
> I couldn’t build the site locally or upgrade dependencies, leaving it outdated since 2019
This is true of many things in the javascript ecosystem. In react/node/js-land you HAVE to invest in maintenance (and re-writes) or be left with a legacy codebase that just might not work one day if npm install fails.