Live data from Hacker News

Technical Vision for Qt 6

blog.qt.io

101–110 of 184 posts

Re: Technical Vision for Qt 6

#101
post #32

I'm looking forward to doing a comparison of Qt 6 and Electron in terms of both usability and performance.

Unless the QT guys do something completely of the charts, it will be a lumbering wheelbarrow vs an Aston Martin.

I'm not sure which is the wheelbarrow in your analogy.

Re: Technical Vision for Qt 6

#102

I'm looking forward to doing a comparison of Qt 6 and Electron in terms of both usability and performance.

Evolus Pencil is a cool desktop app for creating wireframes (open source). It really offers great functionality.. it's now based on Electron (was based on Mozilla XULRunner) ... Performance is terrible on a quad-core i5 notebook with 16gb of ram and a gtx 1060 graphics card... I am talking about dragging the first element on an empty canvas and resizing it. Hopefully Electron will improve - more choices are always ni…

I'm quite skeptical that this is a problem with Electron.

I use nw.js which is similar to Electron for a UI that also allows dragging shapes and text around. It's SVG-based and it's stupid inefficient-- for every mousemove event I'm sending coords over the socket to a backend and receiving back a message that triggers a callback to do the actual property/attribute updates to the respective DOM elements. In other words, I'm using the browser's slowest graphics API in such a way that any layout changes are guaranteed to trigger the least performant graphics updates.

Still it performs just fine on a Lenovo from 2006.

What you are probably seeing is that the Electron GUI toolkit allows devs to throw together a GUI so quickly that they can ship a product in the time it takes to read the documentation of other GUI toolkits. You may thus be comparing a class of cross-platform applications in Electron which simply don't/can't exist in other GUI toolkits given the same amount of effort.

Re: Technical Vision for Qt 6

#103
post #12

Thanks for supporting CMake and deprecating qmake. Not that cmake is the best tool ever, but having to deal with several build systems in the same project has always been a nightmare.

I hope that in future the module system that will be introduced by C++20 will eventually kill CMake. Having said that I don't know the scope of the functionality brought by it, but if not with C++20 the next one will probably be good enough.

I wonder what's the Qt 6 road-map as C++20 is around the corner and is supposed to bring an amount of change similar to Qt 6, I presume. But basing on C++17 is already as new as it gets with their client base in embedded sector for example. Though it may be a bit too new. They can't lay dormant so it's a good deed after all, although timing seems unfortunate.

Re: Technical Vision for Qt 6

#105
post #49
post #5

I hope QML/js crap will be abandoned, and C++ API will be prioritized.

For better or worse, pure C++ APIs have lost the war on GUI frameworks. Microsoft is the only consumer OS vendor that still supports C++ for doing full stack GUI apps, everyone else moved into a mix of C++ plus managed language or layout DSL.

Who is „everyone else“? GTK‘s primary API is C, Apple‘s is Objective C or Swift, none of them are „C++ plus managed language“, only SwiftUI is „layout DSL“ and it isn’t even released yet.

So for targeting X11/Wayland or Windows, you can bet 100% on C++, only Apple platforms do not offer native C++.

Re: Technical Vision for Qt 6

#107
I hope for the overlap of their TypedQML with TypeScript. I am a C++ dev by career and had recently an opportunity to use TypeScript. It's quite nice. It has it's warts as it's structural typing, because of limitations within implementation goals - type erasure and being JavaScript but with types. As much as I would like to see TypeScript Native it's probably to difficult to marry JavaScriptness with relative ease of compilation to C++. Hopefully Qt effort could be extended beyond Qt itself and that would be something I would look forward to.

Re: Technical Vision for Qt 6

#108
I wonder whether Qt could get rid of the C++ moc preprocessor. There are wonderful native signal-slot libraries for C++, and RTTI, introspection and reflection are either there or about to be standarized.

Re: Technical Vision for Qt 6

#109
post #49

Earlier quoted context omitted.

For better or worse, pure C++ APIs have lost the war on GUI frameworks. Microsoft is the only consumer OS vendor that still supports C++ for doing full stack GUI apps, everyone else moved into a mix of C++ plus managed language or layout DSL.

Who is „everyone else“? GTK‘s primary API is C, Apple‘s is Objective C or Swift, none of them are „C++ plus managed language“, only SwiftUI is „layout DSL“ and it isn’t even released yet. So for targeting X11/Wayland or Windows, you can bet 100% on C++, only Apple platforms do not offer native C++.

Metal is written in C++. Try to display graphics in Objective C or Swift without Metal, going further.

Which UNIX OS SDK uses a C++ GUI full stack?

One could argue for Kubuntu or SuSE, maybe, except KDE isn't tied to a specific OS and then again it uses Qt alongside QML since Plasma, so no pure C++ stack here.

And since you missed to fully read my comment, here it is again:

> Microsoft is the only consumer OS vendor that still supports C++ for doing full stack GUI apps

Re: Technical Vision for Qt 6

#110
post #64
post #43

Earlier quoted context omitted.

Easy just watch React Native for Windows talk at Microsoft BUILD 2019. Electron has 700% performance lost over UWP/C++, which is the baseline. Expect similar results when comparing against Qt.

I had to search for the video, but I guess you are talking about this? https://youtu.be/TZnajQqqKV8?t=2894

Yep, that one.
Post reply on HN