https://linen.dev logo
Join Discord
Powered by
# haxe-ui
  • b

    bright-gpu-74537

    08/11/2022, 8:59 AM
    well, it defo insta crashes with that exception
  • l

    loud-salesclerk-7438

    08/11/2022, 9:01 AM
    I mean 5ms is a very small number between clicks, if you set it to 50ms then it will run for a lot longer, but sooner or later it will crash as well
  • b

    bright-gpu-74537

    08/11/2022, 9:02 AM
    either way, id like to understand it, even with 1ms
  • b

    best-agent-85158

    08/11/2022, 9:02 AM
    drag clicking tho
  • b

    bright-gpu-74537

    08/11/2022, 9:02 AM
    the code looks sensible... so i dont understand why it thinks something has been destroyed (and is trying to invalidate it)
  • b

    best-agent-85158

    08/11/2022, 9:03 AM
    it lets you click an insane amount of time in less then a second, and most of the clicks are clumped together
  • b

    best-agent-85158

    08/11/2022, 9:03 AM
    im pretty sure i myself can get to clicks 15-20ms apart (not consitantly tho)
  • b

    bright-gpu-74537

    08/11/2022, 9:04 AM
    heh, i doubt that... 66 clicks a second?? 😄
  • b

    bright-gpu-74537

    08/11/2022, 9:04 AM
    either way, its irrelevant, something is going wrong
  • b

    bright-gpu-74537

    08/11/2022, 9:05 AM
    i dont think the
    if (this._isDisposed) return
    is a valid solution (though it may make sense eventually), as the real question is "what is causing this"
  • b

    bright-gpu-74537

    08/11/2022, 9:10 AM
    ok, i can see whats happening... the dialog tries to center itself 1 frame later... and in that time, its already been destroyed
  • b

    bright-gpu-74537

    08/11/2022, 9:10 AM
    so its a dialog bug really
  • b

    bright-gpu-74537

    08/11/2022, 9:12 AM
    fyi, this is a minimal repro:
    Copy code
    haxe
                var dialog = new LoginDialog();
                dialog.show();
                dialog.hide();
  • b

    bright-gpu-74537

    08/11/2022, 9:16 AM
    OK - fixed (git haxeui-core), so this is actually a good example of why the exception is important...
    if (disposed) return
    was there it would have hidden this bug
  • b

    best-agent-85158

    08/11/2022, 9:17 AM
    My best is like 42
  • b

    best-agent-85158

    08/11/2022, 9:17 AM
    (Consistant)
  • b

    best-agent-85158

    08/11/2022, 9:17 AM
    Wait a sec
  • b

    bright-gpu-74537

    08/11/2022, 9:17 AM
    42 CPS?
  • b

    best-agent-85158

    08/11/2022, 9:18 AM
    With drag clicking
  • b

    bright-gpu-74537

    08/11/2022, 9:18 AM
    oh, right, yeah - didnt realise thats what you meant... thought you meant manual clicking
  • b

    best-agent-85158

    08/11/2022, 9:19 AM
    Bruh I recorded myself and it's bigger then 8 megs -_-
  • l

    loud-salesclerk-7438

    08/11/2022, 9:48 AM
    seems to be fixed for me, now I can continue my quest to find the "other" one 🙂
  • b

    bright-gpu-74537

    08/12/2022, 11:02 AM
    @loud-salesclerk-7438 - so i was kinda intrigued by your haxeuim project, and more importantly the way you dispatch events via lime to haxeui, so i thought i would have a little play around and see whats possible:
  • b

    bright-gpu-74537

    08/12/2022, 11:02 AM
    thought it was quite fun / cool 🙂 - food for thought maybe
  • l

    loud-salesclerk-7438

    08/12/2022, 12:11 PM
    nice! I haven't tested more complex mouse actions, but the building blocks are already available in the library. I think things like drag & drop could work in current version, they might not be the most convenient to use, but that will improve over time. of course every mouse event will have to be sent through the socket connection one by one. which might not be quick enough for whatever needs testing.
  • l

    loud-salesclerk-7438

    08/12/2022, 12:11 PM
    sending more than one mouse event in one go should be doable (kinda like Selenium's Actions API) in a future version.
  • l

    loud-salesclerk-7438

    08/12/2022, 12:11 PM
    I started work on this like two weeks ago, so everything is still a bit WIP.
  • b

    bright-gpu-74537

    08/12/2022, 12:54 PM
    i would think local websocket should be fast enough
  • l

    loud-salesclerk-7438

    08/12/2022, 3:23 PM
    so far speed has been sufficient, but I'm not testing against a real world application so who knows what testcases users will want to run.
  • b

    bright-gpu-74537

    08/12/2022, 3:32 PM
    yeah, i would have thought sockets locally would have been near instant, obviously you have message processing etc to take care of, but i would guess its negligible overhead
1...120812091210...1687Latest