https://htmx.org logo
Powered by
# htmx-general
  • p
    but if the browser is restoring from cache after restart, it forces a full reload.
  • f
  • g
    The 'examples' pages have this weird thing going on in the title.
  • m
    fixed in dev, will be out in the next release
  • m
    🙂
  • p
    This issue happens when you have hx-push-url or hx-boost return a fragment vs full html response. So for example using hx-boost on body with hx-select to pull out the fragment is fine.
  • p
    I.e. the server returns the full html page instead of just a fragment.
  • p
    Perhaps another cache-control header might work, certainly no-cache does not.
  • p
    Basically the browser doesn't care - it just pulls out the result of your last GET request, xhr or not.
  • p
    Apparently just chrome has this behaviour then Firefox "fixed" it and does the same. In any case you cannot safely use hx-boost or hx-push-url when returning html fragments in the response.
  • p
    My particular use-case is that I want to avoid full page swaps as there is a part of the page that cannot be swapped out (hx-preserve doesn't work for this use case)
  • p
    so only solution appears to be (when returning result of hx-boost/hx-push-url) is always return the full page and use hx-select on the result
  • p
    (unless you want to hack around with a script to reload the page which is a bit meh)
  • p
    also note: in either case, hx-boost wrongly preserves the title on back-button. Even swapping on body it appears.
  • p
    another issue: it appears that if you have hx-select issued from an element in the DOM, and that element is removed after the click, hx-select will no longer work
  • p
    so: I have a modal with a link in it. JS removes the modal body containing the link on click of that link. hx-select does not work. If I just hide the modal instead, the hx-select works.
  • p
    which makes sense, as htmx must be checking what to do after the response is returned by checking the element that started the action, rather than pulling out all the required hx- info from the element before initiating the action.
  • p
    the problem is that there is no guarantee the element will still be around
  • p
  • p
    This looks like it might be the solution: setting no-store+max-age in the http header. I'm testing that now.
  • a
    I know I have already mentioned it, but this are important topics to be lost in a chat. There is no transfer of knowledge for the general community. Isn't it better to have a forum or use stackoveflow?
  • a
    just my 2c
  • p
    I've added these as repo issues
  • s
    mocky.io is good for simple examples. But I am looking for a way to run some simple code on the server. For example, I would like to add a delay. So that I can call an URL .../wait?t=5s. My background: I want to create some jsfiddle examples so that new users can play around with htmx. Does somebody know a SaaS which could be used?
  • g
    Yes. There is a stack overflow. But getting questions directed there is hard.
  • m
    I feel like we need a FAQ or something
  • m
    tbh I'm so slammed I don't have a lot of time to think about it right now, but that's what it seems like to me
  • m
  • w
    Noob question here. Amy I understanding this right? HTMX could replace tools like Django Rest Framework? I could just use HTMX to do my "API" calls to views that return HTML?
  • m
    Hi dave, not a Django expert, but yes, you create end points that just return HTML, and don't bother with JSON
1...267268269...1146Latest