Interesting

kinda feels "Citrixy"

thanks everyone!

I think an interactive map might not be at its core an appropriate hypermedia use case. The question might be what the interface between a JS-heavy map and its hypermedia surroundings should be. I’ve seen the #975839826559508510 and the blog mention events: have the map output events which trigger hypermedia requests to update HTML where appropriate. Eg if the threads are shown outside the map (different section, different page, etc), clicking on the marker could trigger a view-thread event and this could trigger an hx-get with an hx-target pointed at a div on the side, or the main page content if the thread takes over the whole page. If that marker is an actual then it’s even easier, the click event is builtin https://htmx.org/essays/hypermedia-friendly-scripting/ And so, the choice of tool or library would be also guided by how “hypermedia friendly” it is, ie if it has a “rich event model” (using the terms from that essay). Not sure if there’s an existing JS map library that fits this description. But the book has an example of turning a more usual callback into an event trigger: https://hypermedia.systems/book/client-side-scripting/#_integrating_using_events

Incidentally I may have to implement something similar in the next few months which is why I’ve already tried to think it through, and that’s the approach I think we’re going to use, seems to make sense (if using an SPA framework I would probably also implement this with events —

<Map bind:marker-clicked=…>

)

I am helping a jr with some code and he has a pattern of making a form, attaching an hx-get/post to it and then using an input submit

this ends up submitting the form and not actually triggering the hx request - I might suggest that the default trigger for forms include submit so that he doesn't have to think about submit vs button?

my fix has been to switch it from hx-get to method=get + action and letting it be boosted

default trigger for forms should be submit

or am I misunderstanding?

hm possible I'm doing something wierd

I'll try to put together a test case

this guy really did a lot of silly subtle things that broke things and it's been hard hunting them all down

does a swap throw an error if you use hx-select and target the top most element?

I will also put a test together for that

ah no nm, I think this is the same bug

the enter/vs click

which I am going to try to debug now

with click on button: http://controlc.com/b359c827

with enter in form input: http://controlc.com/71a3c121

mmm

looks like an email input doesn't support

setSelectionRange

it doesn't, and that's an error - but it only happens post swapError

it looks like when I hit enter htmx drops the htmx-request class on the form, but when clicking submit button it doesn't

😑

I'm wondering if it doesn't like swapping out the element that's trying to do the htmx request?

that doesn't make sense, nm

gonna switch to debug version of htmx

in this section: https://github.com/bigskysoftware/htmx/blob/v1.8.4/src/htmx.js#L3102

if I click the button then I get:

html
selectionInfo.elt
<input type=​"submit" value=​"Continue">​