decline of Qtforums
-
There doesn't seem to be many interesting or sophisticated threads anymore and activity seems to have dropped way off over the past year or so. I'm kicking around in my neuro-divergent brain any possible reasons I can immagine. Any thoughts?
-
I just did think about that yesterday, as well. I suspect that AI is the culprit. For a lot of people AI was quite a slow adoption and I would say that it is not the past year, but only the past 3 or 4 months where it really quieted down.
I suspect that a lot of people quickly find answers (not necessarily good ones) with AI. Then they don't have to post their problems anymore. This will not go on forever because new problems will emerge in new versions of Qt. And AI will not have any training data on them. The good news is that experts might not be necessary right now (because of AI), but they will be relevant in the future. Otherwise software will freeze on the status of 2025. Experts will still be able to make software better, learn from past mistakes and improve. AI will be stuck or will always be behind the curve.
-
I have 2 guesses: AI, like @SimonSchroeder mentioned, but also Qt licensing & hiding of Open Source downloads section which keeps making Qt less and less popular, especially among young students - which previously were the source of big chunk of the questions here on the forums. Qt has focused on automotive and embedded, and falls in popularity everywhere else. At least I feel so; perhaps I am wrong.
-
I have 2 guesses: AI, like @SimonSchroeder mentioned, but also Qt licensing & hiding of Open Source downloads section which keeps making Qt less and less popular, especially among young students - which previously were the source of big chunk of the questions here on the forums. Qt has focused on automotive and embedded, and falls in popularity everywhere else. At least I feel so; perhaps I am wrong.
@sierdzio said in decline of Qtforums:
Qt has focused on automotive and embedded, and falls in popularity everywhere else.
But, what would be the alternative people are flocking to? Do you feel that they don't write platform independent code anymore? Or is Dear ImGUI stealing the thunder? Especially young people wish to develop games one day and ImGUI plays on this. I'm still quite divided if the design of ImGUI necessarily give much better performance than the design of Qt (I suspect benchmarks comparing ImGUI with Qt will show that ImGUI is faster, but I'm not entirely sold this is entirely based on the design). I have yet to see an ImGUI app that respects display scaling on Windows. This alone disqualifies it as a GUI framework in my work.
Any thoughts what people are using instead? Or are they now all just writing web apps and have no use for Qt?
-
Some thoughts from my side:
I run Somco Software - a Qt consulting company, so we are doing a wiiide range of Qt projects on all platforms. The company is already 5 years old, and I would say that only over these couple of years our typical project changed.
In the beginning, there were many desktop and a few mobile projects. Embedded as well.Currently, I have to admit that I don't remember when I had a request for a regular desktop app.
It doesn't mean that there are no Windows, Linux, or Android apps. Quite the opposite. What does "embedded" actually mean? Nowadays, it means mostly "closed system", so the OS and hardware behind it can be anything, and that's a huge advantage of Qt. Many modern medical apps, like the ones developed by GE, are Windows in reality. Some of the tablet-like medical apps are on Android.Tomek, @sierdzio is probably right about the low adoption of Qt by students. If students want to build their first visual things, they will go with React Native and Flutter for desktop. Mainly also Electron. Embedded is a niche and will remain so.
Some time ago, I compared Qt vs Flutter vs React Native in this blog post.
I've been also coding in Flutter myself for some time.My conclusions after this comparison: Qt is a great technology that can freely compete with the best on the market. But also, as sierdzio mentioned, licensing does not help here. Let's take Flutter. They have a completely different way of thinking. They're not asking for a licensing fee as there is a huge company (Google) behind it. They can afford it because they earn money on services. You develop apps with Flutter, and you later use their Firebase and Google Ads. They are gaining the market (also in embedded) while Qts market probably shrinks although we don't have numbers. Probably the best number to compare is the number of new contributors to the framework.
In the world of embedded, there are also other competitors: I just wrote that many embedded systems are in practice desktop or mobile platforms, so you can use basically anything you want.
Of course, Qt Group is the main contributor to the framework, so they have the right to be paid for that. However, the strategy of making an impression that you can't develop commercial projects with Qt using LVGL is probably one of the reasons why Qt is being pushed in more and more sophisticated use cases.
Time will tell which strategy makes more sense. At first glance, it looks like a typical way of operating for stock companies. Investors expect revenue and profit growth, so you look for low hanging fruits, just like an increase of licenses sales by making an impression that you can't develop commercial products with Qt.There are good counter-actions on the horizon that are promising to me and I am happy to say that I or Somco, somehow participated in them:
- Qt Academy - it brings Qt closer to the people. On one side it might be a sign that almost noone wants to create Qt-related content out of their genuine interest, but still Qt Academy is one of the most important initiatives Qt can take.
- Qt Bridges - making Qt Quick and QML the go-to frontend solution for Python, Rust, Kotlin, Java, .NET, Swift. If they play it smartly it can work, but it's certainly a challenge.
Also, as @SimonSchroeder mentioned, AI plays an important role in making this forum quiet. It again changed the way people get information. I think the results will be that AI might raise a generation of let's say average specialists. I see it from myself that a lot of times it speeds up the work, but it doesn't mean that you're learning. You can easily get AI addicted.
Sorry for making it so long, my last thought:
Community members: You and I have to make sure that the community stays vibrant. By making content online, by organizing and attending offline events. That's the only recipe to keep Qt floating. I try to do it (and of course combine with our commercial purpose) in various forms, on the blog, on YouTube, offline, and I encourage others to do that as well.
