https://htmx.org logo
Powered by
# htmx-general
  • g
    But yeah, it’s an open source project, so feel free to explore and look potential improvements. It will be awesome to have another persons insights and contributions 😀
  • i
    Yeah, I'll have to think about the whole thing. It's an idea I got when @User put that poll about 11KB on the twitter 🙂
  • m
  • m
    13ms on emerging 4g, the first time (& cached locally and in the brower's compilation cache after that)
  • m
    wouldn't mind htmx losing some weight, but once sse and ws are out of core, there isn't a lot I would cut
  • g
    There are definitely some things that could be done style-wise to imbue this essay with an air of professionalism which I think it needs. I drew up a sketch of how I'd go about it (it looks better in my head than on paper...) With the page being so prose-heavy, there isn't much I can do for styling without a custom font, if I have your blessing. As for the contents, no notes other than 
    s/<div>/<p>/
  • i
    Yeah, as we spoke on twitter, it's a good trade-off for the majority of users. This would be useful only to those who want full control over the whole thing (not so much about the size, although that can also be a factor).
  • i
    i.e. something between full vanilla and htmx
  • g
    couldn't be me opening a text editor to write my message and drawing up a design when you just asked for "suggestions"
  • m
    You have my permission to do whatever you think is best
  • s
    For me the size of htmx does not matter much. In my app the first http response a user gets from the server contains the whole html which is needed to display the content. This keeps the user happy for at least two seconds. Now htmx can get loaded in the background. That's the big benefit compared to other approaches where you need JS to generate the html for the first page. This means for me, that it is perfectly fine if htmx would be 40k big. Because I care for the initial page, and for the initial page, I don't need htmx at all.
  • i
    Agree. I don't think that size is an issue for htmx. (I mentioned that in my previous comment, but I didn't link it correctly (I don't know if that can be done afterwards)).
  • m
    but I don't think size is an issue for htmx. we should argue about this
  • m
    🙂
  • i
    Well, we can argue if you really want 🙂 I included an argument (not really, just a thought :)) in your twitter thread: https://twitter.com/devsimplicity/status/1448337793017655308
  • m
    ~1kb 😑
  • m
    \> u will never have a 1k library
  • i
    I don't know if you've seen the project, they managed to do a lot in ~1kB 🙂
  • m
    I took a look at it, very impressive
  • i
    yup
  • i
    I tend to like minimalism and "less is more" approach, but as you said, it's about the trade-offs
  • g
    No, not 1k, but I bet we could get close to 9kb. Yes, numbers are arbitrary, but single digits are very impressive on the landing page.
  • i
    Agree, 9 sounds better than 11 🙂
  • m
    \> including a base 64 encoded svg of a dancing fat man in the next release of htmx to own hyperapp
  • i
    🙂
  • b
    Given the marketshare of jquery which is 80% and that is 30kb min+gzip, I think we are ok with 11KB
  • g
    I thought we were adding an svg of a teapot to handle HTTP 418 conformantly...
  • g
  • m
    beautiful
  • m
    only comment is maybe knock down the hero font size a few ticks
1...396397398...1146Latest