My only observation here, and I could be wrong, is that you use 'boost' then go through an awful lot of trouble to deactive it and do other workarounds to get custom behavior.

When in realite boost is just a shorthand for a url push and swap anyway and you can construct it entirely out of those attributes and custimize them as you like.

I think hx-boost is cute, but it leads people down the wrong path of 'boosting everything' and then working backwards to get custom events

when they should make custom events and only boost in few places like navbars.

I see your point. We will probably arrive there later on as we adopt more of htmx... The problem at hand is that we are implementing a sort of permanent banner that is partly updated with websockets and also can have a video playing. For that to work most of the site will need to use boost or push / swap. So the easiest way was to add boost to the body element and only skip it for the few cases where we don't need it. The thing with "unboosting" when a new deploy has been noticed would have to happen with push / swap as well. With only having one hx-boost="true" it was easy to just flip that to false if the git revision was changed from what we had to start with on afterSwap.

And I wouldn't say 'awful lot of trouble' ... Atm it is about 15 lines of js

Any reason for afterSwap to only trigger on ajaxresponses that does swapping?

when restoring from history, a swap is done with swapInnerHTML but that won't trigger the event

Looking at https://github.com/bigskysoftware/htmx/issues/409

We have js that needs to be run after manipulating the dom, for ajax calls that does the swap the afterSwap event works nicely. But when using the back button and the history is restored there is no afterSwap or any other event that is easy to listen for... the htmx:historyRestore event triggers before the actual restore occurs which is a bit to early as the swapping hasn't occured yet

that seems like a bug

want to take a crack at fixing it and creating aPR?

Sure. I tried adding an afterSwap but that seems like it should be handled at the swap method level instead. How about just triggering the htmx:historyRestored after the history is restored instead of before?

that makes a ton of sense

like, way too much sense

Hmm... When I try to make a PR against the dev branch in your repo I still see all the commits in dev in my PR as it would like to merge into master

Not really used to make PRs from forks 🙂

I based my branch on the dev branch as per the readme. Correct?

My bad... I think I based my branched originally on master...

🙂

Welcome to git, enjoy your stay

Heh... Been using git for over 10 years... Still need to read the docs

clearly the users fault

😑

Seems like the CI action on Github never completes?

i don't know what's going on w/ CI plz halp

Setting up the test environment locally to see if that works better... Looked at the workflow file. Nothing that is obviously wrong

Github actions is a bit tedious to work with...