possible race condition: `unicorn:model` looses data when saved too quickly.

[nerdoc]

Especially when connections are slow, but even on localhost, the model is not fast enough to send the data to the backend, when the model saving method is called shortly after the last keypress. See this example:

class PersonView(UnicornView):
    person = Person.objects.none()

    def mount(self):
        self.person = Person.objects.get(
            pk=self.component_kwargs["pk"]
        )

    def save(self):
        self.person.save()

<div>
  <input type="text" unicorn:model="person.name" unicorn.keyup.enter="save">
</div>

When you run that component, write a longer string like "sdhfkjhdsfkjhkjdsf" and quickly press Enter, the saved model's name field will only contain approximately "sdhfkjhds".

I want to use unicorn for "inline" editing - a <input> replaces the text when in edit_mode and is saved automatically if in any of the input fields the user presses Enter. A quick and responsive UI is important, and data loss is very bad, in any case.

[adamghill]

My hypothesis has been that multiple requests are getting sent and come back from the server in a different order, so forcing requests into a queue was my attempt to solve for this. Have you tried enabling serial?

I have always felt like there should be a more elegant solution in the JavaScript of unicorn (maybe an event bus or something?), but replicating this and building a test suite which ensures everything works as expected has been a challenge, unfortunately.