https://htmx.org logo
Powered by
# htmx-general
  • m
    visual basic was my first professional programming language. The debugger for vb6 is still the best I have ever used: you could drag the program counter around and rewrite things live. If you had a loop with an error in it, just drag the program counter to the top of the loop, rewrite it and step through again. Alien technology compared to most of what I use today.
  • p
    Not so sure re HTML. Someone is going to make a framework/environment for WebAssembly or Canvas with drag&drop components and UX like VB back in the day, wired up to a cloud database. As with VB, it might not be elegant or efficient, but it will be good enough that you can throw a project at a junior dev or even tech-savvy business person and they can get some browser/desktop/mobile CRUD app up and running quickly.
  • p
    Something a bit more than "no code" but also a lot simpler than the current complexity of React/Vue/backend frameworks
  • m
    would be interesting if they tried to blend the web architecture with that drag and drop behavior
  • g
    why canvas? most vb components can be made with html, no?
  • m
    yeah, agreed, it would just be Because Google Did It(tm)
  • m
    hypercard for the web, it's coming eventually
  • g
    https://mmm.page is a good source of inspo for drag&drop responsiveness
  • m
    Yep, that, but with hyperscript.... GO
  • p
    I'm not sure. It feels like we've been trying to use HTML as an application (as opposed to document) language for a long time, and while it certainly can be done, the learning curve to get there is so steep beyond the simple drag&drop site builder stuff. Something as conceptually simple as old VB might require having more control.
  • p
    I'm not saying I would welcome a canvas-based thing, but i can see the business attraction.
  • s
    the biggest issue with html as an application language imo is that every page is reinventing the UI from a blank slate
  • g
    the applications you would build with vb are exactly the sort of thing html is good for though. I'm not saying the user should write html, but the drag&drop tool can be backed by it
  • s
    I can't believe web component still isn't something that just works
  • g
    there are some strides towards fixing that (dialog, details, soon maybe popup and select-menu), but it's not easy to land on a set of ui controls that work everywhere the web is, and belong in a long-lasting spec
  • s
    ya im already using dialog polyfill its good but a bit frustrating to use, hard to get positioning right, and im essentially still styling the ui from a blank slate.
  • m
    😑
  • g
    that's politics for ya
  • g
    html imports, my beloved
  • g
    Copy code
    * Carson(?) - Proper scoping, including in behaviors
  • g
    @User can you get this in or are we leaving it for 1.0?
  • m
    I would like to try to get it in next week
  • m
  • m
    o_O)))
  • m
    one must admire the ingenuity and drive to solve a problem entirely of our own creation
  • s
    Reminds me of XSLT. A kind of programming language in the syntax of a data exchange format. My personal point of view: Nobody wants to write this.
  • s
    jsonlogic reminds me of something else: Esperanto. The best idea ever, which would solve all communication problems for ever. But somehow it has not reached the goal yet.
  • m
    Reminds me of this classic

    https://imgs.xkcd.com/comics/standards.pngâ–¾

  • s
    I'm always amazed at the amount of tech that exists to solve a problem only to introduce an equavalent sized problem.
  • s
    like having a perfectly fine relational DB with an active record ORM on top of it and then deciding to put graphQL in front.
1...279280281...1146Latest