https://htmx.org logo
Join Discord
Powered by
# htmx-general
  • m

    https://cdn.discordapp.com/attachments/725789747212976259/1095838372397846538/image.png▾

  • interesting graph of the stars of the various projects
    m
    interesting graph of the stars of the various projects
  • trpc has absolutely rocketed up
    m
    trpc has absolutely rocketed up
  • holy smokes
    m
    holy smokes
  • I believe it gets constantly shouted out by theo, who has a very big internet megaphone.
    r
    I believe it gets constantly shouted out by theo, who has a very big internet megaphone.
  • this is very much in line with my priors, but I think the premise of RPC—it's just a function call!—is fundamentally broken, even if typescript is hot right now. network calls has wildly different failure modes from function calls, so you do in fact need to think about them differently.
    l
    this is very much in line with my priors, but I think the premise of RPC—it's just a function call!—is fundamentally broken, even if typescript is hot right now. network calls has wildly different failure modes from function calls, so you do in fact need to think about them differently.
  • Crockford is a real grug type though, I'm a big fan
    l
    Crockford is a real grug type though, I'm a big fan
  • Hi, dumb backend dev here. I've been enjoying updating parts of the page with the ease of htmx. I have a question though. What is the good practice for adding JavaScript, that interacts with these parts of the page that are swapped out in this single page type architecture? Should all the JavaScript be loaded on first load, or is there some way to parts of it when we load the part of the page that needs it.
    b
    Hi, dumb backend dev here. I've been enjoying updating parts of the page with the ease of htmx. I have a question though. What is the good practice for adding JavaScript, that interacts with these parts of the page that are swapped out in this single page type architecture? Should all the JavaScript be loaded on first load, or is there some way to parts of it when we load the part of the page that needs it.
  • I think organising your code around the idea that network calls are just rpc could work for small apps pretty well, and maybe it is even working, so that’s why trpc is getting more attention right now.
    e
    I think organising your code around the idea that network calls are just rpc could work for small apps pretty well, and maybe it is even working, so that’s why trpc is getting more attention right now.
  • Theo’s YouTube channel is probably doing its job also
    e
    Theo’s YouTube channel is probably doing its job also
  • On the other hand I am really cautious when it comes to popular tech and popularisation of tech through social media. We did it with micro services and serveless and a lot of small enterprises took it really serious, so they are ended up in distributed software architecture hell, where simple monolithic backend with Postgres could work flawlessly for them.
    e
    On the other hand I am really cautious when it comes to popular tech and popularisation of tech through social media. We did it with micro services and serveless and a lot of small enterprises took it really serious, so they are ended up in distributed software architecture hell, where simple monolithic backend with Postgres could work flawlessly for them.
  • It’s hard to establish bs filter, most of the people can’t, so they are struggling with needless complexity.
    e
    It’s hard to establish bs filter, most of the people can’t, so they are struggling with needless complexity.
  • never thought of it that way rofl
    r
    never thought of it that way rofl
  • ya, in general optimize the big things (avoiding full page reloads) and don't bother with the small things (singular kb js), code complexity is a silent killer of projects.
    s
    ya, in general optimize the big things (avoiding full page reloads) and don't bother with the small things (singular kb js), code complexity is a silent killer of projects.
  • b
    Hey all. 👋 Just upgraded to HTMX 1.9.0 from 1.8.6 and noticing a strange situation where having an 
    hx-trigger="click"
    with no 
    hx-put
    on a Submit button within a form (parent element) which defines the 
    hx-put
    no longer fires. There are no errors and no AJAX requests. In order to fix this, I had to copy the same code from the form (parent) for 
    hx-put
    and add it to my submit button to make this work. Is it no longer possible to inherit this behavior from the parent in 1.9.0 like was possible in previous versions? I couldn't find any mention of this in the release notes so wanted to learn more.
  • m

    https://cdn.discordapp.com/attachments/725789747212976259/1096085878042017792/image.png▾

  • m
    hmm, that's probably a consequence of "naked" 
    hx-triggers
  • m
    hx-trigger wasn't intended to be inherited, so the behavior you are describing seems like what I would expect, very sorry for the unintended consequence
  • s
    They could have just stopped after the words "People don't like fat js clients".
  • b
    Thanks, no worries. Also didn't realize you weren't suppose to rely on inheritance like this. Interesting. I liked that you could save some duplication by being implicit but maybe the explicit nature is better from a readability standpoint anyway. Hmm.
  • b
    Actually, in going back to the documentation, maybe this needs a slight update: https://htmx.org/docs/#inheritance. Knowing what attributes can and cannot be inherited would be clarifying and helpful I think.
  • r
    Y'know how you get your pet cat its own tiny keyboard because it wants to mimic you at work on the computer? That's the "vibrant ecosystem" of JS distractions.
  • b
    Thanks. I'll go with the load everything on first page load. However, I think I read about something in htmx that would load only once. That could be worth looking into. Unless I imagined it 😂
  • m
    I like the "load stuff up front" approach for vanilla js, and the inline approach w/ Alpine/hyperscript
  • b
    For alpine, I currently load a function in x-data that returns an object. I guess this is the same as inline, but I've been wrong more times than right when it comes to front end.
  • l
    Inheritance is specified on a per-attribute basis, so hx-trigger for instance cannot be inherited: https://htmx.org/attributes/hx-trigger/ But hx-target can: https://htmx.org/attributes/hx-target/
  • l
    I might be misunderstanding your use case, but if your goal is to trigger the form on clicking the button, it should work if you simply 
    type=submit
    your button and remove the 
    hx-trigger
    from the form entirely
  • d
    Hey
  • d
    Been using intercooler/htmlx for a while, awesome piece of work!
  • d
    Became a sponsor today...
1...109210931094...1146Latest