"Use javascript to avoid using javascript" someone frame this

“We wrote all the JavaScript, so you don’t have to.”

Hi guys! Total newbie here. 👋 I was wondering what editors/IDEs you guys use that work well with htmx - i.e., any offer plugins that enable syntax highlighting, autocomplete? I normally work on a Mac, and use Sublime Text or Nova.

That’s what I call dedication to your craft 😂👍👍

Hello all! I have an express endpoint that return a partial template which will be rendered by htmx in the hx-target that I specify. The problem is that the same endpoint returns an error partial template when something goes wrong in the server which I want htmx to render in a different hx-target. How can I achieve this behaviour. Thanks in advance.

funny I just added a new suggestion about it here : https://github.com/bigskysoftware/htmx/issues/607

Thanks a lot! I will try to use hx-swap-oob but I will also keep an eye in the next release for the htmx:beforeSwap (seems more promising)

As of 1.6, you can use the

htmx:beforeSwap

event to modify swapping behavior: https://htmx.org/docs/#modifying_swapping_behavior_with_events

Also, you can use the

HX-Request

header on the server side to detect if a request is coming from htmx, and render something different

depending on how you want to handle things.

this could be implemented as an extension now, interested in trying it out?

I am a huge fan of the JetBrains editors, and I work on htmx and hyperscript in WebStorm

Hi there. I have a small site where I'm testing htmx. Everything is working great and I'm really happy with the philosophy of this language. It kicks ass! I have a question though: there are 3 navigation places on the site I'm working on; The htmx attributes are the same for the 3 of them

hx-boost="true"
  hx-push-url = "true"   
  hx-select = ".site-main" 
  hx-target = ".site-main" 
  hx-swap = "outerHTML swap:0.5s settle:0.5s"
  hx-indicator ="#loader"

I was wondering if I could define these attributes only once, and then apply them to any div that needs them (something like a hyperscript behavior for instance)? I think I could add them to the body tag and then every anchor would be boosted. However, I am not sure what would happen with external links (would they load on the main page?). Also, I am more in favor of a granular control of where htmx is applied.

You're right, doing is better than complaining, i'll try it

You can hoist the attributes up to the

body

tag and they will be inherited. boosting only applies to local links that start with a `/`: https://github.com/bigskysoftware/htmx/blob/1778ae77ca5bec9a9e437b0de114660fa5602220/src/htmx.js#L964

@mysterious-toddler-20573, Thanks so much!!!

@User had some shower thoughts today after watching your seminar yesterday. 1) I'm curious in your tale of divergent web technologies, is there a place to talk about the brief time that XML and SOAP was the rage? In some ways, I see HATEOAS and hypermedia having links to XML. 2) I feel that I don't truly buy any argument that JSON is better because it is smaller/faster. In order to render a page based on JSON data received from a web server means that the client has had to have first downloaded all the UI and template elements. This makes me think of the legacy thinkclient deployments of previous years. A client needs to download a bunch of styling and functional code (including some he or she may not even use) to make use of the JSON. Caveat: I am not a smrt man, and I may be misinformed/out right wrong about many things, so welcome to be guided towards 'the truth' wherever it may be.

heya john, it would be interesting to go back and look at the conversations that were going on when XML/SOAP was big. My take is that the reason REST was inappropriately imported into the JSON conversation was because REST/HATEOAS was still plausible with XML (although still incorrect, since REST/HATEOAS requires a consumer with agency: https://intercoolerjs.org/2016/05/08/hatoeas-is-for-humans.html), since it kinda looked like HTML, and then it just stuck when people flipped to JSON, where it was obviously inappropriate, but the thought leaders couldn't let it go.

for 2, my take is yes, the HTML payload might be a bit bigger, but latency is a bigger issue than payload size and also the browser is lightening fast at parsing HTML. Compare that with javascript executing to produce either HTML to be parsed, or just DOM objects. The browser/HTML parsing is gonna be a ton faster.

Well – gotta adapt, right? I've just super addicted to everything Django-related the last 2yrs, soo..

Oh thats a good point about time required for a JS frontend to parse and render a UI based on JS

On that note – somebody posted this on Reddit: https://htmx-django.com/

Momentum is coming

Thank you for the info - I’m finding that most editors already highlight htmx quite appropriately, which is great. The main issue seems to be that without using a

data-

prefix the htmx-specific attributes get flagged as html errors.

using htmx + some handwritten javascript for page transitions on a static site

absolute unit

\> js tho 🤔

page transitions now live on https://this-week-in-htmx.netlify.app/

Does htmx support JSON POST/PUT commands? I heard this from someone with a lot more dev-fu than me and this is his only real complaint at this point about htmx: > if you want to do “real” API-style work, with JSON bodies, you have to write script; HTMX doesn’t support JSON POST/PUT. For my posts and puts I had to back off and use fetch.

x-www-form-urlencoded: not real