[OT] browser and congrats about livekeys

[dinusv]

Actually I think it would be possible, but with some extra work. Currently qml is used as the main language in the application, and I see there's this project called qmlweb, which comes with it's own engine for processing qml, and it's written in js. https://github.com/qmlweb/qmlweb I think if you would link that to your projects you could get something similar, that's if you'd prefer that approach over using json as the declarative language. As for this application, we currently rely heavily on c++ and qt, so porting it for the browser would be difficult.

Thanks! Be sure to hang around as we have a new version and video coming up in the following days

[cancerberoSgx]

Ajá, so currently opencv and qt do run in the browser and officially support webassembly versions. Do you use much more c++ besides those two libraries ?
Anyway, I'm not interested in the format itself, but in the language and its semantics to represent opencv complex commands. Now researching a little I realice that you basically are implementing qml which is a language for layout and boxes (like xul for mozilla?) using opencv as an implementation , am i right? My intention is the opposite: try to define a declarative language with which I can represent complex sequences of opencv operations, for example.

[{cmd: 'load',name:'m1', path: 'foo.png'}, {cmd: 'roi', src:'m1', dst: 'm2', rect: [10,10,100,80]}, {cmd:'write', src'm2',path:'foo2.png'}]

I will start trying to make an equivalence between cv.CC() operations and json objects and soon enough I will figureout what's missing..... then use a template language if you need to iterate or parametrize, but the objective will be an operation as a declaration and not as code... with that idea was that I reach your project, .... thanks... sorry.... chau - fell free to close it and thanks for the link.....

[cancerberoSgx]

although since I'm totally beginer and I would like to draw something with opencv your program should be the best tool