https://htmx.org logo
Powered by
# htmx-general
  • g
    Just opened an issue on Github. And added this mini site to show the problem. Notice that clicks on the homepage are ignored by htmx 1.6.1, but the others do work. This is very likely a big oversight on my end. Thanks. https://mauricioarangoosorio.github.io/issues/issue/index.html
  • s
    The homepage link is the only one that has a 
    target="_self"
    attribute... is that it?
  • g
    Yes that’s correct.
  • s
    If you remove that does everything work as expected? And if so, are you able to remove that from the links in your app?
  • g
    Let me check
  • g
    Ok, you are right! I removed the target attribute and it worked. I updated the mini site on git hub
  • s
    Copy code
    function boostElement(elt, nodeData, triggerSpecs) {
            if ((elt.tagName === "A" && isLocalLink(elt) && elt.target === "") || elt.tagName === "FORM") {
    👀
  • s
    ^ that's the code in htmx, so it is checking that the target attribute is empty
  • g
    Ok, so anchors with a target will not work at all, right?
  • s
    AFAIK, 
    _self
    is the default context anyway, so is essentially the same as being empty, but htmx requires it to actually be empty.
  • s
    It appears so.
  • g
    Ok, thanks so much for clarifying this. It was driving me crazy (!).
  • g
    Specially because it seems that in 1.6.0 the target attribute was being ignored
  • s
    I can imagine! 😆
  • g
    Thank you for looking into this. Really useful!
  • s
    Copy code
    function boostElement(elt, nodeData, triggerSpecs) {
            if ((elt.tagName === "A" && isLocalLink(elt)) || elt.tagName === "FORM") {
    ^ 1.6.0
  • s
    So, yeah. You're not crazy 😂
  • s
    IMO, the new behaviour in 1.6.1 is correct. E.g. if you have a link with 
    target="_top"
    (new window/tab) - you don't want it boosted, and it probably can't be boosted (?).
  • s
    So in 1.6.0, you might have had htmx trying to boost it as well as your browser opening a new tab - leaving the boost attempt potentially failing or doing some unexpected behaviour.
  • s
  • s
  • g
    That makes a lot of sense. I wonder though how this works if you are using a text editor in a cms, and create a link to an internal page. Sometimes one doesn't have a lot of control over the way those links are generated. But I can see why there's a need to restrict that.
  • s
    Yeah, that's a good point. You could maybe do a find+replace on your side to convert target="_self" to target="", or submit a PR to htmx to treat target="_self" the same as being empty.
  • g
    'treat target="_self" the same as being empty.' I think that would be awesome, especially if one needs to retrofit older projects.
  • i
    okay, I have what is probably a "okay you noob" question, but here goes: How can I make an hx-get replace the entire page? I can't for the life of me get anything like 
    hx-target="#root"
    to work.
  • m
    You can replace the inner content of the body
  • i
    the problem is that I think I need to replace everything because it's a link to the Django admin interface, which I believe needs to have a different from my application
  • i
    and I know, this is the simplest tiniest bit of JS to write, I would just rather use HTMX everywhere for consistency
  • p
    isn't the query selector ":root"?
  • m
    This sounds like a case for anchor tag!
1...568569570...1146Latest