[Interest] Thoughts on a 'proper' (or standard) file/io API for QML?

[Alejandro Exojo]

On Tuesday 16 August 2016 07:43:49 Vlad Stelmahovsky wrote:
> bad performance is a sum of bad design decisions.

"Bad performance" is overly general. I gave two real world cases where the IO 
being done in C++ helps nothing. And it wasn't about performance in general. 
It was about completely blocking the UI for seconds that a real user notices.

Obviously I found solutions to those two problems (moveToThread a helper 
object with emitting a signal to the rescue), and I did it in C++.

But I don't think I've ever found a single time where the issue about a slow 
path was a slow language.

Note that this comments of mine are in response to your replies about the 
"slow and non effective (comparing to C++) JS". I obviously don't recommend 
people to write their whole Qt applications in JavaScript. But there are pure 
JavaScript/QML applications in the Qt demos and examples, and I don't see the 
problem at all in those.

> especially in embedded world. and problem not only with IO speed on
> embedded, but in memory footprint, GC procedures etc. actually, since you
> are in Qt forum, you can read recommendations for using JS (as less as
> possible), you can ask people, why Qt now have Qt Quick Controls 2 and so on

If you mean the fact that QQC2 are implemented using more C++ than QML, yes, 
I'm aware of that. But note that is a complete redesign of everything. Just 
changing the language is not a panacea.

-- 
Alex (a.k.a. suy) | GPG ID 0x0B8B0BC2
http://barnacity.net/ | http://disperso.net