Client focused hot takes on the state of frontend
- Next.jsHot takes
This is written for clients, not developers, if you're a developer go read one of the endless voids of in-depth, ultra-granular articles you see on the web. This isn't it chief.
Here's my absolutely tepid and not-controversial take on Next.js. I think it's going to stick around and absolutely dominate the market because it's an almost update parity version of React if you want less of a headache with hydration.
What is hyrdration? Well it's basically how a website gets passed to your customers browser. The short-hand way of understanding it is that working with forms, accounts, live news feeds etc is computationally expensive, anything you can make static, is a bonus.
This comic from xkcd has never been more relevant
So if you're absolutely terrified on betting on the wrong horse with frameworks, do yourself a favour, pick Next.js, it's not going anywhere anytime soon.
Tailwind is like truffles, I know it's supposed to be good. Lots of chef's go nuts for using truffles in Italian dishes, but it still doesn't sit right with me... Fully.
It's not the naming convention that I find bad; it's just that sometimes when this gets added to a project, it shortly becomes unmaintainable as you're having to look through hundreds of characters of horizontal space.
Note to devs reading this
Edit: So shortly after writing this out, we actually invested even more time into Tailwind (we're talking hundreds of hours reading through the documentation, Youtube videos, and GitHub issues), and we finally cracked something that had always been bugging us. How do you wrap the content in a consistent way and avoid the 500m horizontal scrollbar sprint.
It's these two lines, add them to your Visual Studio code, and you're in for a much nicer time when working with Tailwind. It wraps after 80 characters, no matter the screen size and makes it significantly easier to read for any horizonta-phobes such as myself.
"editor.wordWrap": "wordWrapColumn", "editor.wordWrapColumn": 80
Tailwind actually seems great; it's still disgusting to look at for a developer, but right now, it seems to be the best compromise you can make to get the best performance and support the latest features, although I'm fairly certain this is the weak point for scalability. It's bitter-sweet for us.
Oh wow, Shopify bought this shortly before we started writing this article (it's been a long time coming). Well, I guess Next.js kind of has all of the features from this now in some way or form... But it's still a pretty cool piece of tech.
It's going to be pretty hard to talk about Remix now without talking about Hydrogen. For those of you that are not nerds and have better things to do: "Hydrogen is basically a platform to build Headless Shopify shops, and it's pretty good when we've worked with it". We'd still rather use Next, but it's definitely the coolest from what we've seen when it comes to e-commerce. But it's hard not to mention it, because it's now paired together, and probably when you'll actually use the tech.
Also worth mentioning Kent C. Dodds is involved with this, and he's a fountain of knowledge and an excellent community leader with this kind of stuff. It definitely has legs, I just wouldn't necessarily use it for an enterprise project.
In the race for the fastest on the list, I think Astro wins it... at least if you believe their marketing. But... Frankly, I do. It is the fastest from what I can see, but from our perspective, it takes the metric of "fast above all else" far too seriously.
If you've not done this yet, click on any of the "See it in the wild" links. You'll notice that the page pops and shifts around more than my back pain after a 12-hour coding session. It loads stupidly fast, but it doesn't seem graceful in doing so.
To us, something like Next.js just flat out seems better... even though:
- yes, we are partners
- yes, we have a vested interest
- yes, almost all our sites are built in Next.js
But... it still provides the most elegant way of actually building a website that doesn't look like the loading process of an early 2000s Piczo website. It's smooth, and there's a whole host of solutions to stop text and images and content shifting all over the place.
However, we'd recommend Astro if you're looking to game the metrics on Google Lighthouse because it probably will. I'll let Rich Harris say it best...
Despite sounding like the swedish way of saying Velvet, this Framework is pretty interesting. I'm going to upset many people and say it's basically React but better, but nobody uses it that I've met yet.
I really want this to do well so it becomes a challenger to React, because anything that promises less boilerplate is a winner for us. Healthy competition is always a good thing as well. Svelte could be the butterfly wing-flap that caused the Tsunami, so let's hope it has a bright future.
Plus, I like Rich Harris; he seems less of a lizard person in a suit and more of a person that cares about the work he's doing. So yeah.
God damn, do I hate Material UI, the actual UI framework is pretty good, but the whole god damn design system that Google made is an inconsistent mess. They have Google Material You, which is a step in the right direction, but overall, V1 material design is a plague that has gone on for far too long. It's just so soulless and uncharismatic.
lt alI feels very "I Have No Mouth, and I Must Scream".
That's me done for the day. I've finally calmed down once again and got my heart beat below 100 bpm.
Even with all the shade thrown above, I do want to say that, whichever platform, framework, or language you end up using, they're all pretty much good in their own way, we just like to poke fun at the industry and vent. So try out whatever you like, and tell us how you get on.
Get in touch
Send complaints to...
Fancy sharing your own experiences? Or just calling me a "mug" directly to my inbox. Either way, get in touch, we love hearing from you, and your path through the wild west of frontend technology.