https://htmx.org logo
Powered by
# htmx-general
  • h
    yeah 100%
  • m
    tests are the worst possible way of ensuring correctness, except for all the others
  • m
  • m
    neat hyperscript + htmx solution to a UI problem, really shows off the event-driven ideas of both!
  • g
    This would be great as a widget in the cookbook.
  • m
    🐝 🐝 🐝 🐝 🐝 🐝 🐝
  • g
    No, I just 7b-Ed you first.
  • m
    i just pointed at the example
  • m
    i made no other suggestions
  • m
    😑
  • h
    I think this is probably my ignorance of JS, but reading through issueAjaxRequest how does the returned promise get attached to the xhr request? IE when you call xhr.send what makes this function not blocking
  • m
    this is some serious jank in the library
  • m
  • m
    basically, when 
    htmx.ajax
    is called from javascript, we create a promise and capture its resolve and reject functions.
  • m
    we don't do that in "Normal" html-attribute driven requests, because javascript will complain in the console if resolve or reject isn't called on a promise
  • m
    but I don't think that matters too much for what you are doing
  • h
    it doesn't
  • h
    I just want to understand it better
  • h
    this is much better than the avg js code I read so I wanted to take advantage
  • m
  • m
    lol, better
  • h
    yeah I've got it setup already
  • m
    👍
  • h
    I just like understanding everything
  • m
    I wish I did!
  • h
    it make the next phase of yelling at a bug I don't understand less painful
  • h
    ok so I think I got it - basically you make the xhr.send call immedeatly which is async by default and the default callback for that is onload where you call resolve
  • h
    the trick is that resolve is null when you first call it
  • h
    so you are expecting the next step in the promise chain to have a chance to change the ref to resolve from null to a real callback before the send call completes
  • h
    I've never seen that before, it's neat
1...526527528...1146Latest