4 ms·
Like most deep in the SPA bubble, this author cannot imagine any other outcome than SPA-everything. Meanwhile, any alternative view is put aside as being an ant
by fleddr 6y ago
Like most deep in the SPA bubble, this author cannot imagine any other outcome than SPA-everything. Meanwhile, any alternative view is put aside as being an anti-JS crusader (which misses the point entirely), simply being old, or even barbaric whilst he continues to high five his bubble inhabitants on Twitter.
SPA as the status quo is a myth. It's a Fermi paradox. Where are they? Almost nothing of any importance is an SPA. You'd think that if they have such superior experience, they lead to superior business metrics, thus they are everywhere. Asides from true app-like experiences like Gmail, Spotify, the like, they're nowhere to be found. Not in places of significance. It's 2020 and Amazon continues to be old school page refreshes. Last time I checked they're doing quite OK. Even big parts of Facebook, the poster child of complex interactivity, continues to do full page refreshes in big parts of the experience.
Instead, SPAs are a startup default, not a web default. Big difference. And the way most are implemented, they don't live up to their potential benefits, instead deliver UX that is even worse than static:
- breaking back buttons
- breaking scrollbars
- rendering things that aren't interactive
These things don't have to occur, they can be solved in an SPA, but they do occur a lot in the wild. This matters. We live in the real world.
As somebody else here already mentioned, the article takes a completely wrong approach. It assumes some binary choice or outcome. The original point instead was that the web is not a single thing. It's a huge thing where different stacks are to be used for different use cases. This point is largely ignored, instead it is implied that SPAs should simply solve such use cases better, whilst continuing to be a default.
On the particular point of SPA links being typically faster/smarter: there's about 500 drop-in solutions to speed up static links. Which are usually not needed, because the point of such links being bad is overstated and lacks any user evidence. See my earlier remark on virtually all major websites using plain old links. They wouldn't if it was so bad.
Finally, a huge part lacking in this discussion is the accessibility of web development itself. The SPA web is an engineers web, it requires advanced programming skills, infrastructure, increasingly complex tools chains, and so on. If this was to be a true default (luckily it isn't), it would lock out the vast majority of people wanting to contribute to the web.
It's sad to see this sub culture of hardcore engineers with Macbook Pros thinking they own or represent the web or web development.
- robertoandred 6y agoThe idea that some bad implementations of an idea serve as an indictment of the idea as a whole is utterly ridiculous. Some images on the web are poorly optimized, but that doesn't mean we should not have images.
- fleddr 6y agoNo, it's not ridiculous, it's sane. The whole point of a SPA is to give a user a better experience. If on average this does not materialize in the real world, and often achieves the exact opposite, the entire concept fails. Bad SPA implementations are the standard. Saying that it's possible to make a good SPA implementation doesn't change the real world outcome that most often, they don't deliver.