News for hackers. Stuff that matters.
Re:React Native for OS X by Sir_Cmpwn on Tuesday March 29, 2016 @11:40PM (#11386810)
Have you drunk the react kool-aid for the browser yet? I have, and I was initially skeptical. Performance can be plenty fast but more importantly, your UI and your application as a whole is more testable, reliable, and easier to reason about if you use React (and importantly, the functional mindset that the React ecosystem comes with). I've made many desktop UIs without this sort of thing and they can be fiddly little bastards to get all of the interactions right, but React makes it effortless and reliable.

I'd love to see React and Qt talking to each other.

Re:React Native for OS X by hellofunk on Wednesday March 30, 2016 @02:36AM (#11387358)
QML on Qt is also a declarative and reactive technology, it behaves very similarly to React for the browser, but for native widgets. I think that is what the question is about, and I'm curious too... how does React differ? Qt's QML automatically updates UI elements based on state changes. I use React for web programming and they seem like similar concepts, so I'd be curious for anyone to shed light on their differences. Qt falls more into the "reactive" domain, whereas, name aside, React is not a reactive tool, technically, but the use-case is similar for both, no?
Re:React Native for OS X by Sir_Cmpwn on Wednesday March 30, 2016 @09:54AM (#11389257)
I would disagree with this. I haven't worked with QML, but I have worked with WPF which also supports two-way data bindings in a manner that I would guess is similar to Qt. The major difference between the approaches is that React embodies the functional approach. By describing your UI as a pure* function of your application state, you get a lot of important benefits.

* https://en.wikipedia.org/wiki/Pure_function

Re:React Native for OS X by hellofunk on Wednesday March 30, 2016 @10:08AM (#11389352)
Several academic papers on functional reactive programming specifically discuss QML implementation, so I suspect a comparison to WPF is superficial. You may want to take a look at QML before dismissing it. I've seen react developers who have used QML say that the experience is very similar.
Re:React Native for OS X by Sir_Cmpwn on Wednesday March 30, 2016 @10:11AM (#11389365)
That's why I included the disclaimier that I hadn't actually used QML. Thanks, I'll check it out.
Re:React Native for OS X by PeCaN on Wednesday March 30, 2016 @11:50AM (#11390157)
I used React to write an in-browser REPL with live data browsing for a programming language that I was writing and it was admittedly quite nice. It's what reminded me a bit of Qt/QML actually. But that is my only experience with React.

I have my fair share of Haskell experience so I get the benefits of statelessness. But I don't necessarily think it's a silver bullet for UI.

Re:React Native for OS X by rplnt on Wednesday March 30, 2016 @03:56AM (#11387617)
FWIW I think React is horrible for web. It brings back the (bad) feel of flash web apps from over a decade ago.
Re:React Native for OS X by untog on Wednesday March 30, 2016 @05:49AM (#11388124)
...what? React isn't even a UI framework, there is no "feel" to it at all.
Re:React Native for OS X by rplnt on Wednesday March 30, 2016 @08:56AM (#11388872)
"App startup". I.e. loading screen after the webpage was supposedly loaded. Same for transition between states (compared to how regular web pages behave).
Re:React Native for OS X by mikewhy on Wednesday March 30, 2016 @09:22AM (#11389021)
Aren't those the same problems all SPAs face? At least with React you have the option of server rendering to remove your first issue.
Re:React Native for OS X by untog on Wednesday March 30, 2016 @11:28AM (#11389995)
React is actually fairly unique among JS view libraries in allowing you to fully render the site on the server-side. If a site doesn't do this then it's a choice they have made.

But at the very least this is not React-specific.