4 ms·
To me this highlights how backwards our way of building software is. Having a user display and draw on top of an image is trivial, and something our systems ha
by GuiA 7y ago
To me this highlights how backwards our way of building software is.
Having a user display and draw on top of an image is trivial, and something our systems have been able to do for decades.
Streaming video bidirectionally is also trivial and well supported by our primitives at this point, even though it's merely a decade or two old rather than almost half a century.
And yet, if you want to combine these two very primitive operations, you need to find the app that supports exactly the use case you want (not guaranteed), or to be a developer yourself.
We're still catching up to Engelbart's demos, and it's disheartening.
- sitkack 7y agoNot being snarky and I agree with you. But we have everything we need! Both of those things are accessible via web technologies. I am not a web developer, but it doesn't seem like we have a bidirectional data/component model. What if we could instantiate a component on a page that shares its state with all other components in the same group? Since they are also a shared experience between humans, it by no means has to be correct all the time. Only that hits "reality" every X seconds. Analogous to the delta compression of video [3], if we define a rate at which I frames are broadcast, we know that the view of reality doesn't diverge too far from the other participants. Think of it as sCRDT a sloppy convergent replicated data type. [1] https://github.com/automerge/automerge https://github.com/automerge/automerge [2] https://www.youtube.com/watch?v=PHz17gwiOc8 https://www.youtube.com/watch?v=PHz17gwiOc8 [3] https://en.wikipedia.org/wiki/Video_compression_picture_types https://en.wikipedia.org/wiki/Video_compression_picture_type...