My intuition tells me that #htmx's hx-swap attr will suffice for many dynamic #web behaviors but there's a point where the abstractions will start to leak and you'll want full control over #DOM again.
I speak highly of htmx, but it isn't a "replacement" for anything; I see it more as something that is intended for a specific class of applications. Probably best for thin-client applications, that allow more than just navigating using hyperlinks, and sending form data over POST requests.
If a team is using htmx, I find that the team should be responsible to make it clear that the web application is more a thin-client application that provides a UX that is very similar to paper forms.
If stakeholders want more, it's time to transition away from htmx and onto something else, IMO
As an example; my app is a mastodon client; it fetches a list of statuses and renders them. When the user has read them all, they may fetch the next page. With HTMX you could hx-swap the next page into the list. But what if the items have been reordered, what if some of them have been deleted or edited? A simple DOM append won't cut it.
Add comment