Talk:Fyne (software)

Regarding notability

The toolkit in this page is (and has been for some years) the most popular GUI toolkit for the Go (programming language) (which Google reports as around 4 million developers). The GitHub project is starred over 27'000 times, an order of magnitude more than some other projects listed in the Template:Widget_toolkits list.

As well as content that I have helped created (2 books, 3 magazine articles and multiple podcast interviews) it has been used in 15 issues of the renowned "Linux Magazine" by their own writers. Tutorials have been written by community members in English, Portuguese, Russian and Chinese.

The project is in use by many commercial organisations including TailScale, MindPoint Group and Docker. Would it be helpful to include that information in the article (perhaps with endorsements?) - I wasn't sure as it seemed a little commercially focused. Andydotxyz (talk) 07:18, 17 October 2025 (UTC)

You are the creator of Fyne. You should very much not be editing this page at all due to basic COI. That's you're the only substantial author of this page is not a good look. Arp242 (talk) 23:21, 26 March 2026 (UTC)

In support of Fyne's 'notability'

Hello Wikipedia Editors,

This is just a message from a random Fyne user — aye, one of the 27.000 who 'starred' Fyne on GitHub.

Of course, I'm not a professional programmer, and I'm just a very amateurish newcomer to the Fyne environment. But I can say with absolute conviction that Fyne is the GUI toolkit with which I'm most familiar with. That's not saying much, since, as said, I'm not a developer. The first time I toyed around with "GUI development" was when Windows 3.0 came out (aye, that's correct). That was roughly at the end of one of our semesters, and we were already insane enough to do the assignment on Windows 2.x, as opposed to our colleagues, who developed their own mini-GUI environment from scratch. We started our work in 2.x, and Microsoft 'decided' to launch this so-much-more-advanced 3.0 out, and of course we had to test it out. So... that was the first time I toyed around with GUI development. A few years later, I had to face development in the X Window System, and learned how to do simple applications on Motif. And that was pretty much it. The graphical WWW came into existence soon afterwards, so that was my last brush with GUIs.

I'm also a Go fan — which makes sense from the perspective of someone usually tinkering with back-end tools. Go, although designed as a systems development language (modelled after C, after all), is actually a quite simple language to learn, which nevertheless implements a powerful development environment (I call it 'IDE-in-a-CLI'), so it can tackle practically everything — even replacing JavaScript on a webpage, thanks to its native WASM cross-compiler. And, of course, like every 'modern' programming language today, it can call external libraries which have been compiled for C or C++.

This led all sorts of volunteers to wrap shims of code around their favourite GUI-building environments, especially the popular ones like wxWidgets or Qt, hoping to bridge the gap between the lack of a good GUI-building toolkit and a much simpler, safer, and (perhaps) more powerful programming language than C++.

Almost all such efforts were given up at some point in time. The last maintained Qt binding (heavily forked from previous repos) hasn't received any upgrades since 2023 or so (most of the others had been abandoned much earlier).

Why?

Fyne.

If there are 'only' 27,000 stars on Fyne right now... it's just because there aren't that many Go developers doing full-blown, GUI-based apps. This, I believe, was the main reason for all those 'wrapper' libraries around existing, popular toolkits to fail: there is no critical mass of Go developers for sustaining such efforts. And, at the end of the day, what do you get? A cross-breed between easy-to-develop-in Go files at the core of a massive constellation of accessory files of all sorts, which require to be assembled into a unified whole — and that for each different platform/OS that one wishes to support. But while Go has native, built-in cross-compilation (all you need to do is to change one or two environment variables — nothing more is required), the C/C++ part has not. Which means that a developer will need to worry about having all possible combinations of C/C++ compilers for each platform they want to compile, and a nightmare of multiple dependency trees to keep in sync — even if there is no need to worry about the Go part of the issue.

Sure, platforms such as Qt incorporate plenty of IDEs and all kinds of tools to help to automate everything, leaving the developer free to start immediately working on the Go programming at the 'core'. But all of them assume that the developer is targeting a codebase written in C/C++ — which means they won't get any real benefits from using Go at the 'core'. They'd probably be much better off just generating code for C/C++ instead, and forget all about Go — it's too much trouble to go through for only a very small amount of potential benefits to reap.

Fyne, by contrast, is fully implemented in native Go — there is no need for any other 'external' set of tools, environments, IDEs, and so forth. If you already have your Go setup in place (and that mostly means installing Go from a package management repository), there is nothing else to install, or to learn about.

No wonder, therefore, that several Go-based GUI developers have just onboarded to the 'Fyne development environment' — which is minimalist, compared to other toolkits, and thus quick to learn — and abandoned all other efforts of writing Go code while using popular, state-of-the-art, C/C++-based toolkits.

There's where Fyne's 'notability' comes from. Not from abstract metrics, compiled by who-knows-who. Not from being in a list of the 'Most Popular GUI Toolkits in 2025'. Not for having earned trophies and awards of excellence. Not for grabbing the attention of the whole world, being mentioned as a 'new kid on the block' with the highest rating of developer desirability. Not for ostensively featuring as one of the most-used toolkits on the Stack Overflow yearly survey as the 'must learn next' environment.

And then... it's a question of statistics.

Qt boasts of having a million users (developers, not end-users!). This is a claim that is hard to fact-check — only the Qt Group can know for sure — but it's possibly accurate, given the immensity of applications built with Qt. And, well, 'one million' of anything is always eye-catching. But how much is that compared to, say, the number of .NET developers world-wide? Certainly there must be more; but how many more? In other words, considering the number of developers in the whole world, what is the market share that Qt has? Perhaps it's just 1% — is that 'notable'? In fact, would anything that represents 'only 1% of the total' (not only software; anything) be worth mentioning on the Wikipedia?

One might counter-argue that such a comparison (Microsoft vs. Qt) is unfair, since they target completely different markets: Qt is cross-platform, .NET is (practically) Windows-centric. One should therefore compare the 'notability' of Qt relatively to other cross-platform toolkits — such as wxWidgets, for instance. I have no idea how well Qt fares, compared to wxWidgets. I would probably guess that Qt has more than 10 times the number of developers than those who remain faithful to wxWidgets — but that's just my intuition, I have no facts to claim or refute that guess. Nevertheless, I'm sure that the Wikipedia page for wxWidgets is not being 'threatened' to be removed due to lack of notability! The toolkit, after all, has been around since 1992.

And it is' on GitHub as well, where it has... 6900 stars.

It would be unfair to rate Qt, BTW, because it only has a published mirror on GitHub, they run their own repositories; the Qt 'core' only gets 1900 stars, and even if one adds all the stars for each of repos under Qt — there is a sizeable number of those! — together, they wouldn't come even close to wxWidgets.

So, does that mean that wxWidgets is more than three times more popular than Qt?

More to the point: can Fyne therefore claim that it's about 10x more popular than Qt?

Anyone could easily argue that such comparisons are utter nonsense and offer hundreds of very good, valid reasons for not consider that somehow 'stars on GitHub' corresponds to 'number of actual users' or to 'market share', or, in fact, to any well-known, independent, unbiased, and peer-reviewed metric. Anyone claiming the contrary would have to bring extraordinary proof!

But we haven't exhausted the market of 'GUI toolkits'! Qt and wxWidgets are C/C++ libraries. Even though they have lots of language bindings (including, as said, Go — even if mostly abandoned by now), what about all other programming languages? Surely they have their own, native GUI toolkits as well, right? Let's start with, say, the currently most popular programming language ever: Python.

According to Python sub-Reddit (aye, I know, not the most reliable of sources), the 'best' Python GUI toolkits for developing standalone, native apps are... well, Tkinter (a wrapper on top of Tcl/Tk), PyQt (a wrapper around Qt), possibly wxPython (wrapping wxWidgets), and... Kivy, the only one that I've found which seems to be fully 'native' (even though compiling it from the sources requires several non-Pythonesque tools and libraries) and comprehensive enough to be taken seriously. I had never heard of Kivy before today, but it seems to be pretty popular. It definitely has its own Wikipedia page. After all, it even has 18,700 stars on GitHub. Not bad! 😏

None of those have any 'notability issues' (they might have other issues, though). In fact, the only Wikipedia page with 'notability issues' is the one for pyglet (2,100 stars on GitHub), the only one vaguely known to me, since it's been around since 2007 (I hate Python, but sometimes you simply cannot avoid it, especially nowadays), and continues to receive lots of updates and contributions (the latest commit on GitHub was the day before yesterday, at the time of writing). It used to be the de facto game engine for Python developers by the early 2010s; which means, in turn, that it was most likely to be the best-known 'GUI/games Python library' back then. Nevertheless... today, the Wikipedia Editors consider it to 'not be notable enough'!

I would therefore conclude that there is a rather wide margin for 'defining' notability (the official page for that does not use objective metrics), and one of those criteria is its scope, or environment, in which notability is being described. And this varies, first and foremost, in its size.

Let me give you a good example of what I mean with 'scope'.

Consider: I'm a native of Portugal, where I live — a small country (pop. 10 million) in Western Europe, which most people in the world would not even be able to point on a map (it would be nice if they placed their finger on Europe and say 'somewhere in there'). Nevermind that it is the currently oldest country in the world with the same terrestrial borders — established around 880 years ago. Nevermind that it is the currently extant country with the oldest established permanent miltary Navy of the world (est. in the early 13th century). Nevermind that its capital, Lisbon, has been around for 3,300 years of continuous inhabitation as a city — the oldest city in Europe outside of Greece, and even in Greece, there are few continuously-inhabited cities that are older (Athens, of course, is older). These are great factoids for trivia questions, but not relevant whatsoever. How many Nobel Prize Winners does Portugal had, in over a century of granting them? Well, just two — and one was for lobotomy, and there are, in fact, many pressure groups wanting that the prize is removed from the list. More infamous than notable, one would say. On the other spectrum of world interest: how many Soccer World Cups has Portugal won? None. Well, it happens to currently have the most famous soccer player of the world — Cristiano Ronaldo — but, once he's retired, there would be little reason for anyone from Portugal to be considered 'notable'. How often did Portugal win the Eurovision song contest? Just once, and (arguably) it was due to very unusual circumstances.

That said, if this country is so insignificant, why does its Founder King, Afonso I, have his own Wikipedia page? I mean, what exactly made him 'notable' enough to be worth listing here (as opposed to just relegating him to the Portuguese version of Wikipedia)? Surely, considering the world's population, an absolutely unknown monarch of an insignificant country is unworthy of having his own Wikipedia page? Why waste disk space with a Wikipedia page that is not even useful for, say, trivia night?

Well, the argument might simply be, it's true that Afonso I of Portugal is hardly 'notable' outside of Portugal, but one might counter that, in the context of Portuguese readers of the English Wikipedia, Afonso I is relevant, and would be recognised by any Portuguese — so, for that very restricted scope, counting just a handful of people, Afonso I is certainly 'notable'. It might be slightly relevant to those odd, curious people, who just happen to be writing something about Afonso I for some reason (e.g., writing a book about "European Monarchs of the 12th Century"). So, Afonso I stays — even if he is just relevant to a very tiny slice of the Wikipedia reader population.

In other words: this 'notability' criterion, in the specific case of King Afonso I of Portugal, reflects only the 'notability' for the community that happens to consider him relevant — no matter how small that community might actually be. Note that in case of Afonso I, the total percentage of Wikipedia readers that might show some interest in this 'notable' historic figure is less than 1% (probably not even over 0.3%). So, what the Wikipedia Editors decided was that the criterion of 'notability', as related to the page for Afonso I, cannot be constrained by the percentage of the global readers of the English Wikipedia; or about how 'popular' such a figure is at the world's stage. I mean, even if you'd list the "hundred most influential European monarches of all times", King Afonso I would not come even close to the last position.

Nevertheless, his page is still there (with some problems, sure, but 'notability' was never questioned).

Now let's get back to Fyne.

Since we don't have actual numbers for Fyne's 'market share' or 'popularity' among developers actively using GUI toolkits, how exactly can a criterion of 'notability' be established?

Consider that such a community must have at least two requisites, severely constraining the scope within which the Fyne page on Wikipedia is relevant:

1. They must be Go programmers; and
2. They must have a strong interest in **native** application development (as opposed to do web-based applications).

Now, historically, Go has been used mostly for backend development. It is also likely that many so-called 'full-stack developers' might have an interest in anything that relates to JavaScript/TypeScript, HTML, and CSS.

There are less Go programmers 'in the wild' than, say, C/C++/Java or Python programmers. The most cited reasons for not using Go is its apparent limited number of popular libraries and frameworks (which is arguable, but we're talking about perceptions here) and the lack of complex and convoluted syntactic sugar (it's true, Go has been designed with simplicity and code readability in mind); consequently, the number of companies actively using it, or recruiting Go programmers, is considerably low — again, that is a perception and not necessarily the truth.

Additionally, there is this perception that you can develop complex applications from scratch and easily port them to any platform by essentially wrapping a web application embedded in Chromium — either via the popular Electron/Native React approaches, or similar alternatives. This means that most programmers and so-called full-stack developers are far more interested in developing complex JavaScript/Typescript applications and release them as 'native apps' — they don't want to learn the conceptually different way of building GUI native applications, and their employees could care less. Even if some of these webapp-turned-native Frankenstein monsters require to contact a backend server (to get upgrades, install plugins, or simply to save data remotely and securely on the cloud), such programmers can simply use Node.js — there is no need to learn a 'new' programming language.

And the other side of the coin is also true: C/C++/C#/Java developers have little or zero interest in developing 'web first' applications. They use their own set of toolkits (such as Qt or wxWidgets for C/C++ programmers) to deliver natively compiled applications, and nowadays few are interested in Web-centric approaches — the goal is mostly sheer performance. Even though originally CGI was first developed for C and quickly followed by Perl — thus allowing natively-compiled C applications to act as web service providers — nobody uses C/C++ for such purposes any longer. Similarly, although Java applets used to be common to develop embedded applications on web browsers, that approach has been essentially abandoned, by substantially overhauling JavaScript and integrating it much deeper with the HTML and CSS rendering engines.

So, what does all the above lead to? A relatively small community of Go developers of native GUIs needing a widget toolkit that allows them to easily 'write code once, deploy everywhere'. While there might be several other similar toolkits out there, I dare you to find a single one that is found by a search engine before Fyne, and which has more 'popularity' and 'notability' than Fyne, as measured by its number of stars on GitHub, word search popularity on Google Search, or any other metric that you might wish to come up with.

If that were the case, I would very likely not be writing this essay here. I would be on the page of that hypothetical toolkit instead (or creating it, if Wikipedia didn't already have it).

And to further show the popularity of Fyne, here's another bit of anedoctal 'evidence': when doing a Google search for a simple-to-use GUI builder for Fyne — since that's perhaps one of the handicaps it has, everything must necessarily be built out of individual components, there is no nice QML page description language for Fyne — the webpage that Google directed to was... this very same one on Wikipedia! (FYI, the GUI application is named Apptrix, not to be confused by a dozen other products and services using the same name.)

This shows that not only there is some intrinsic 'notability' to Fyne — it has tens of thousands of people using it, even if the degree of their usage migt be severely limited — but also because the Google web crawlers, finding this page on Wikipedia, considered its contents to be worthy of being listed firat.

And the best part of it, of course, is that I did find that simple-to-use GUI builder — there is a reference to it in this very page! Thus, in spite of its alleged lack of 'notability', it has already been useful to a least one person.

Well, it's a start!

— Gwyneth Llewelyn (talk)

P.S. My own 'clash' with some of the definitions of 'notability' on Wikipedia started two decades ago, when my own page was promptly removed. Not even the argumentation that there were good reasons to sustain my claim at 'notability' — and I used similar arguments back then, namely, that my personal 'notability' was confined to a niche environment (more specifically, Second Life, the only still-extant metaverse). I was given the option to move the content to my own user page instead. And so I did.

Not shortly afterwards, I got another 'warning' that the content on my user page was deemed to be 'inapropriate' — not because it wasn't NSFW, or offensive, or racist/xenophobic, or any of those things. No, the reasoning behind the 'warning' was a claim of 'not being appropriate', or 'not fitting in the context of a user page on Wikipedia'. So I edited it again, this time listing even more citations from independent sources. I've even endured accusations of plagiarising myself (gosh, that was the last drop... seriously!). I was criticised for criticising how the Wikipedia was managed. I was accused of using my user page as some kind of personal 'revenge blog'. And the list goes on and on; fortunately, I guess I remained around, while Wikipedia editors come and go.

And I've been shown to be 'wrong' in almost all those cases 😅

Disclaimer: I mentioned Second Life because the page for Second Life itself has been targetted for deletion (or at least to be forcefully subject heavy redaction). Again, the claim of 'lack of notability' was at the forefront of the reasoning for deletion. It was even argued that the text read as if it were a self-promotional page, edited by the company that develops the Second Life infrastructure. It was not; the company didn't even know there was a page for them on Wikipedia! And the arguments in defence of the page was that it had 1½ million regular users, who interacted with the virtual world environment at least once per month; furthermore, the virtual world content-exchange economy. back then, was already generating a quarter billion US$ in user-to-user transactions per year, with the company behind it just getting some fees out of that — enough to turn a small profit every year.

Isn't that 'notable' enought?

And then again, I keep coming back to pages for software that is relevant for small communities (some is open source, some isn't), and where I contributed with all sorts of things, from logos to screenshots, to explanations on how the software works or how it compares to others. Over half of my edits have been reverted, for all kinds of reasons (copyright violations ranking highest). At some point, I just gave up — it requires more patience and time than what I have.

But that's just digression. These days, I'm accused of using LLMs to write lots of useless digressions. I find it extremely amusing (of course, the 'proof' that I've always written in this way is on my personal blog — which goes way back, long before LLM's have been invented)!

P.P.S. And now, for the 'proper' signature of this talk post: Gwyneth Llewelyn (talk) 00:34, 16 November 2025 (UTC)