https://htmx.org logo
Join Discord
Powered by
# htmx-general
  • b
    Why would it not work with htmx?
  • w
    My understanding (which is admittedly very shallow) is that Google requires the gtag to be on every page to capture activity. Since the tag would presumably only be loaded once, and not with fragments during e.g. an hx-boost, that activity might not get logged?
  • b
    If your question is specific to boosting pages then there is lots of literature out there on how to make Google Analytics work with SPA's which would apply here too. One such example is https://tech.webinterpret.com/how-to-make-google-analytics-work-in-a-single-page-application-spa/
  • w
    Thanks 🙂
  • b
    At its simplest, it will involve manually registering a page view.
    Copy code
    js
window.ga('send', 'pageview', location.pathname);
    But according to that article, there's an easier way via Google Tag Manager.
  • w
    Yeah, I figured it would involve something where I listen for 
    htmx.onLoad
  • w
    Oh wait, that Tag Manager seems a lot easier
  • w
    Thanks 🙂
  • w
    Actually, so it looks like detecting history change might be enabled in the new version of GA by default.
  • b
    i feel i have gone down a dark path.... htmx on links in svg
  • b
    HTMX is not preventing default on the events because svg is case sensitive and the browser is therefore reporting the element's tagName as "a", so I am writing an extension to hopefully just call preventDefault on the event
  • b
    i'm really enjoying HTMX but i'm doing a lot of weird stuff and not sure if i'm finding bugs or if i've actually messed up my htmx.js file
  • m
    sounds interesting! please don't be afraid to explore and ask questions, the code base isn't that big!
  • b
    the site i am working on has a web component, so i have modified htmx to work with a shadow root. the modification i made was pretty simple but i don't think y'all would be interested in these changes as require writing javascript to instantiate an htmx instance with a root element
  • b
    but after working out some kinks and getting the responses ironed out, the web component basically works.. i was pretty pleased
  • b
    supporting the web component made my django code get kinda ugly, but it's getting back the way it's supposed to be, clean views and templates
  • l
    Swap html fragment load is fast at first call, sub sequent calls html fragment load at client longer than first time. Looks like swap will go through all the elements recursively and remove. Is there any way we can replace without recursive?
  • m
    which swap are you using? can you give us a sense of the target content before and after? are you comfortable w/ profiling in Chrome?
  • h
    Thoughts on switching scrollBehavior default to auto from smooth?
  • h
    100% of the designers I've run it by do not like the smooth, and I think auto is closer to non-boosted behavior
  • m
    def possible for 2.0
  • h
    do we have a 2.0 branch yet or should I throw against dev?
  • m
    No 2.0 branch yet
  • h
    alright, tossed it up against dev with a note to wait until 2.0
  • v
    Hi, I have a websock created with htmx. After a while, it closes with reason : "Connection idle timeout". There's no client->backend messages yet. All I do is send oob messages from backend to client. Is this a reason? Server is Jetty (Using Clojure)
    r
    • 2
    • 3
  • t
    oh exploitable
  • r
    if you're allowing your users to insert their own javascript in the first place, you're gonna have XSS issues. You're better off having an attribute whitelist if you're allowing HTML as user input at all.
  • r
    re: @thankful-addition-60522
  • t
    just do a 
    sanitize()
  • r
    on the server maybe, dunno what api that is from.
1...935936937...1146Latest