@andy my team have been using it for years (a decision I deeply regret signing off on) and I’ve found that if you know what you’re doing with the Web Platform, you’ll just fight against it.
It feels like it sets an ‘okay’ (i.e. not ‘good’) baseline for people that would otherwise deliver ‘bad’, but if you treat ‘good’ as a minimum and strive for ‘great’, it’s an uphill battle.
@andy In my previous job a new starter argued that we weren't following "best practices" because we were using <a> instead of <Link> in our code base. Left me feeling that NextJS is giving people some distinctly unhelpful ideas about HTML.
@andy been there… you’ll also have to reinvent on-page anchor links in case you need them. 😉 Also, good luck with lazy loading images etc 😅🤞
When I did a project with Next.js a while ago, all images were showing up with a huuuuuge delay on live. It was unusable, basically. The problem was that Next uses (used?) Squoosh to resize images on the fly which was just too slow in production. Installing the additional “sharp” library fixed it…
@andyhttps://astro.build is what I'm looking at for what will be the "next generation" of something beyond nextjs
Not sure it'll be literally that, but I think it has a lot of the right ideas (same for https://remix.run). A lot of which feel very similar to 11ty
Which, I love 11ty, but component authoring and consumption feels leagues better as a UX cycle with JSX than with most html templating. (I just wish JSX didn't currently imply "throw out the whole ass web browser" 🙄)
@andy that said, everyone seems to slowly finally be moving to a healthy blend of
Static first is good actually
Generate and strip out what you can
Server side render what you can't
Make simple server logic very lightweight to integrate
Only ship the runtime for the last-mile of what actually needs to be interactive
And nextjs has a ton of legacy stuff in it making the dev experience rough for sure. But I think it's a sign there's better things coming soon (I hope)
@hazelweakly yeh I’ve sorta thought maybe Astro might be better too because clients wanna use react components, which I totally get tbh, but if the framework around it makes things silly for folks in the studio — designers — I ain’t interested
@andy yeah that's super fair. I've continued to be disappointed in how little everything has focused on a11y, designers, collaboration, and making things amenable to low-code integrations (CMS + page builders as concrete examples)
Like... WordPress is still eating the world in a lot of ways, and for good reason. I just don't see that changing until people start figuring that out and actually addressing it 🤷♀️
@andy yeah I actively recommend clients to avoid it. I have had at least 2 projects to move out of Next and not only was Next the reason to move out, it made the process hard af.
I see what attract people to it, but it does not deliver.
Add comment