https://htmx.org logo
Join Discord
Powered by
# htmx-general
  • m
    at least in theory
  • w
    Can I see an example? I can't tell from the docs how I'd do it
  • b
    Copy code
    html

<tbody id="mailbox" hx-sse="swap:new_mail" hx-swap="afterbegin">
    So that’s what I do on the target element.
  • b
    I connect the SSE with
    Copy code
    html

<table class="dt mt4" hx-sse="connect:/newmail/{{ mailbox }}/">
  • b
    (Replace with your url)
  • w
    Ah - awesome. Thanks for the help! Still new to this so examples are super useful
  • b
    I have a memory of the SSE docs maybe not being exactly up to date. Though that was months ago.
  • w
    I think it's more just the concepts - it makes sense now that the 
    sse
    response is a "thing" that I can control with hx-swap, I just didn't think about it that way
  • b
    Hmm. Seems okay… (sorry for the delay). Might be worth turning debugging on and seeing if you get any more clues. (Or summon @mysterious-toddler-20573 to have a peek)
  • a
    I implemented and run a local test, and they seem to not be
  • a
    a load from a standard link (so no 
    HX-Request
    header) that triggers a 304 loaded the partial version of the HTML, cached from a previous request with 
    HX-Request: true
  • a
    it accounts for query strings, though
  • a
    double checked it for both Etags and last-modified
  • m
    🤔
  • m
    maybe we should always include a parameter in htmx requests?
  • m
    and not push it
  • a
    that's the workaround I was thinking of
  • m
    ?htmx-cache-buster=true
  • a
    oh, wait, I'm finding something on the docs
  • a
    let me test
  • a
    I added 
    Vary: HX-Request
    to my test, but it keeps caching the wrong response 🤔
  • a
    unless I am understanding this wrong, that should work
  • a
  • m
    awesome find, we should definitely implement this
  • g
    it's a response header
  • a
    correct
  • a
    so it's up to the server, as part of generating the etag or last-modified response headers, to set the right Vary
  • a
    good news, it's seems that the issue is not the issue as the standard accounts for it. Bad news, I can't get it to work?
  • a
    ok, Vary works for last-modified, but not for etag... will keep researching
  • a
    Ok, I think I get why this is happening with Etag. The ETag is supposed to be a hash calculated from the response. So, formally, the Etag for the full HTML and the Etag for the partial should be different. The server shouldn't be generating the same Etag for both responses.
1...823824825...1146Latest