The best version of GNOME was Gnome 2. Before Gnome 3, I used Gnome for years. They really burned a lot of users with that reduction in functionality and consistently user-hostile attitude.
The closest modern desktop environment to Gnome 2 is KDE’s Plasma. It’s easy to use out of the box and doesn’t require any configuration (unless you want to do that kind of thing, and then it’s available — unlike in Gnome 3+).
It is not a "user-hostile attitude" to remove features. Every project that does some kind of user study will do this eventually, because:
1. Not every feature is equally useful
2. No one is using the feature, and so maintaining it is a waste of time
3. Sometimes developers discover a better way to implement a feature
4. Some features can turn out to be just harmful and misleading
5. There is no one available who wants to maintain the feature and so it slowly deteriorates and breaks over time, and must be removed to prevent further breakage
And so on. If you program in your day job, I bet you can think of a few other valid reasons you have removed a feature at work, too. GNOME has not done anything out of the ordinary here, and it is incorrect to describe it this way. If you believe they have taken specific actions that were hostile towards you, then you should say what those actions are. But the mere act of changing or removing a feature that you happened to use is not a display of hostility, it is just the normal development process working as intended.
>and then it’s available — unlike in Gnome 3+
This is incorrect, Gnome 3+ can be configured with shell extensions.
>> No one is using the feature, and so maintaining it is a waste of time.
I wish they would drop more features that I don't like. I have a 55" monitor, so I dont use workspaces. I also dont need every window to shrink and move about when I want to bring up the launcher. Also, drop the hiding the launcher feature.
In the other direction I would like the application menu back ;-)
There are application menu plugins, but I'm almost certain you are aware of them. I wish they'd fix performance degradation issues, but I think the issue might require a holistic effort across a deep stack of components. That, and make the Wayland compositor restartable.
You are making assumptions that are wrong. Just because you think a feature isn't used, that doesn't mean that there aren't users of that feature. Not every user participates in developer mailing lists to advocate for working features they currently use. Gnome has consistently made user hostile changes, like the time a minor release update changed key bindings from Unix style to Windows style for cut / paste / edit. Was nobody using keyboard shortcuts? I sure as damned hell don't think that was the case.
I used to change the window decorations to something smaller than the default, because I like having more of my screen real estate available for actual content rather than useless UI padding. The current iterations of Gnome are unusable to me, when Gnome was visually appealing and highly usable 20 years ago. Window transitions and icon docks have done nothing to improve usability, and bigger borders are not better borders.
Thankfully other desktops have far better feature retention rates.
No, I do not work for Red Hat and I have never worked for Red Hat. I am just a person who uses GNOME and systemd, and while I do have my own complaints about them, I have enough experience to respond to most of the common complaints. Are you aware that there are many Linux distributions besides RHEL and Fedora that are shipping GNOME and systemd?
I also do not appreciate you looking through my comment history to try and pick something out to seemingly dismiss what I am saying. Is it inconceivable that I do not have a financial interest, and I just happen to be knowledgeable in these areas?
I liked 2 the best, especially with a dock. I understand what they were attempting in 3, but it was such a letdown.
Writing everything in JS was a weird decision. I remember for the first few releases, having to constantly restart the shell because the JS engine leaked so badly. You could watch RSS grow indefinitely everytime you moused over the top left corner.
It's a lot better now, usable with a few extensions...but I still think what a waste it was tossing 2 like they did. Extensions are an OK idea, but because they're mostly JS and CSS, there doesn't seem to be much safety there. It's very common for them to break each release because of some obscure attribute change.
It's perfectly reasonable to prefer KDE because you like Plasma or it has the functionality you need. It's a nice desktop environment with a lot of functionality so I can see why it appeals to many people. What I don't get is why so many people feel the need to diminish open source projects they don't even use.
What is your end goal with a comment like this? Are you just expressing your opinion so that no one forgets that many people do not like Gnome 3? Are you hoping to better understand why the Gnome community has a different development philosophy than the KDE community? Are you trying to convince someone of something?
Gnome has pissed a lot of people off, and this is the end result. I mean I think it's pretty reasonable for people to have ill will towards the project and I think that politely interjecting with the occasional "we don't like gnome" is reasonable. If only to encourage new projects to use a toolkit that isn't so closely tied to the gnome experience.
That being said I do have to give them some respect for XDG-desktop-portal, which has made my main complaint (gnome's filepicker being bad and non-optional for even basic software like firefox) no longer an issue.
I get that their design direction wasn't everyone's cup of tea, and I get why many abandoned Gnome 3.
> politely interjecting with the occasional "we don't like gnome"
From my perspective these comments seem neither polite nor occasional, and they result in Gnome threads devolving into partisan arguments rather than technical discussions. Of course Gnome isn't the only lightning rod -- posts that mention Mozilla, Apple, Microsoft, US national politics, systemd, etc also seem to inspire an inordinate number of posts that are argumentative and off-topic.
It's mostly just disappointing. I was honestly hoping for a discussion about what kind of people contribute to a large open source project instead of rehashing the endless circular debate about whether the Gnome Community is user hostile or not.
I'm not sure I agree that would be another rehash. A comment that draws from the data in the original blog post to make the case that Gnome is driving away new contributors seems like it would be interesting and on-topic, and could lead to a discussion about ways to attract and retain open source contributors.
As it is, most of the points made in this thread could have been made in any prior thread with "Gnome" in the title. And they will probably be made again in subsequent Gnome threads, like some kind of Sisyphusian performance art.
>Gnome has pissed a lot of people off, and this is the end result. I mean I think it's pretty reasonable for people to have ill will towards the project and I think that politely interjecting with the occasional "we don't like gnome" is reasonable.
I would say that it is not reasonable. I have been seeing these type of "we don't like this" comments for around 20 years ever since GNOME 2. They have done nothing but incite more flame wars, making the developers even less likely to want to engage with any constructive comments that might be underneath. Please find a more productive way to contribute, either by volunteering to pick up some of the development load yourself, by forking the project, or by contributing to something completely different. That is how you get things done in open source. Making the same drive-by complaints repeatedly is not a way to change anyone's mind.
>If only to encourage new projects to use a toolkit that isn't so closely tied to the gnome experience.
This is exactly the opposite of what you want to do. If the desire is to improve the toolkit, then you want to encourage more outside contributors to join and bring it in a direction they like, not push them away.
Users don't have time to continually advocate for features they use in their desktop environment. If I'm a developer using a Linux based desktop, I have a whole bunch of work of my own to do. I don't have time to spend hours arguing in favour of retaining a feature inside Gnome.
The particularly awful disease the Gnome project has is the "rewrite everything" disease. No software project can retain features if they do large rewrites from scratch every major release. There simply aren't enough hours in the day to keep rewriting the toolkit every few years. The end result is that features the developers personally don't care for don't get implemented in the new release. That is why so many of us that once used Gnome are pissed off and have had to write off the project.
>I don't have time to spend hours arguing in favour of retaining a feature inside Gnome.
Then perhaps you can see how developers of a desktop also do not have time to listen to every person arguing for every feature, or to maintain every possible feature that ever gets suggested, or even to maintain the current feature set indefinitely without removing something at some point.
>The particularly awful disease the Gnome project has is the "rewrite everything" disease. No software project can retain features if they do large rewrites from scratch every major release.
You are confusing refactoring with rewriting. Very few parts of it have been rewritten from scratch. Refactoring things isn't a "disease". Refactors are actually necessary to keep the project in working order and to pay off technical debt so you can actually have a chance to implement new features while retaining older ones.
>The end result is that features the developers personally don't care for don't get implemented in the new release.
This is the same as any project. If there are no developers to maintain the feature then it stagnates and eventually gets cut. It makes no sense to be pissed off at this one particular project about this, every project is bound to do it eventually. If your company ever went through layoffs then you can probably name some times you have had to do this at work.
Because it is far and away the most feature complete, polished, and stable DE that exists in the Linux ecosystem. It requires zero learning curve, has an app suite that isn’t riddled with bugs, everything for the desktop user can be done in the GUI.
Like there is no alternative that does what GNOME does. It’s GNOME and forks of GNOME, a chasm, then KDE, another chasm, and then the DIY DEs/WMs dejour.
Umm.. what does Gnome have that KDE doesn't? To me they are equal in functionality, just that KDE isn't forcing the opinions of their developers on me and lets me choose what and how I want to do stuff. And has more features (file selection with image preview, yes, it exists!)
In the beginning, the licence and being C instead of C++, that is why it was started in first place. KDE is older and Qt licence was different in the 1.0 days.
Plasma seems like the worst part of KDE to me. It looks nice and has good functionality when it's working, but unfortunately it's riddled with bugs. Panels crash all the time and don't come back until I restart plasmashell.
I gave up on Plasma. I still use kwin and several KDE applications, like Dolphin, but now I use those with LxQT instead of Plasma. It doesn't look as stylish, but it seems to be rock solid. Plus I still get all the nice compositing effects and configurability of kwin.
Worth noting that this approach would have been impossible to execute on GNOME Shell, which integrates its shell, compositor and window manager in a single component.
> I feel that I am very efficient using it, more so than when I was using tilling window managers (wmii, i3, dwm, etc.).
Gnome/GTK is the only software I can think of, which slowly but surely reduced the end-user efficiency.
One starts working one day, and finds out that mnemonics (menu shortcuts) have been disabled. Another day, the file manager side-by-side panel has been removed. Another day, a convenient filemanager shortcut (new folder) has been forgotten in the transition to a new version (and nobody cares). Another day, and the file search has been replaced with an incredibly slow, and buggy search. Another day, and one finds that tiling capabilities of the terminal are severely limited because the devs are hostile to APIs that would allow such functionality (see Tilix debate). Files sorting in the file manager? Overrated¹. And let's not get into the Gnome branding debate².
In the FOSS land, I find Gnome/GTK to be the most user-hostile approach to design I've ever experienced. The statement "I have no idea what XFCE is or does sorry" says pretty much everything about the project(s) attitude².
In general, going to developers with a list of demands for patches to revert is not going to be helpful, it is much more likely that behavior is seen as hostile, coming from a user. It would be more useful to ask them why they have removed those features, figure out what their goal is, and then discuss it so you can reach consensus.
>because the devs are hostile to APIs that would allow such functionality
Apologies, this has been an incorrect definition on my side (but correctly understood by the GP).
What I referred to with "tiling capabilities" in:
> tiling capabilities of the terminal are severely limited because the devs are hostile to APIs that would allow such functionality (see Tilix debate)
is actually the keyboard navigation/selection support in the terminal.
There has been debate between the terminal maintaner(s) and the GNOME VTE maintainer, and lead to nothing.
Without providing any technical information, the twit means nothing, and IMO, in the context of the GNOME devs culture, it's just another, typical, "F* you".
> It would be more useful to ask them why they have removed those features, figure out what their goal is, and then discuss it so you can reach consensus.
That's not a productive approach; it would make the GNOME devs slaves of popularity contests.
These are not my words. They're the words of a GNOME member¹
> as a question was there any kind of community review
No, maintainers are not meant to become slaves of popularity contests.
Discussions with GNOME devs seems not to lead anywhere. Here³'s the opionion of a Pop!_OS maintainer:
Something @GNOME NEEDS to realize is that Pop!_OS, Ubuntu, and others are NOT
going to stop theming. What we will do instead is replace GNOME components
entirely if it becomes too costly to apply our style to them. Come up with a
way forward for custom styles, or lose users.
The theming problem has been very much discussed - if with "discussing", you mean lots of users (devs) talking to a wall. It's worse actually; the GNOME team actively prevented devs from developing theming, and advocated against it.
I actually had a link where a GNOME dev was urging Debian (or Ubuntu, can't remember) to revert a patch that restored terminal transparency (or something like that). Could you help me finding it?
Here's an interesting reply from Christian Persch²:
> 1. Open gnome terminal
> 2. Go to Edit -> Profile Preferences
> Actual result: Background configuration tab is missing.
> It was present in 3.6 version, please return it back.
No.
The above was actually discussed on HN⁴.
Regarding this:
This one is blatantly incorrect, see this twitter thread from a GNOME developer: https://twitter.com/tobias_bernard/status/145743855381055898...
The "*lot* of work" has never been discussed with users (I've checked the multiple requests in the bug trackers). Given the amount of egomania in GNOME's background, and lack of strong (actually, any) technical grounds in the twit, the twit has exactly zero validity.
Unfortunately, I share the impression, even the most basic "office work" use cases are not covered, like switching audio devices or UX decisions that make no sense, like auto-ditching notifications (like what the heck, the UI knows better what notifications are important and which aren't). The GTK API is like the wierdiest pseudo-objective UI API ever, and the number of modern GTK applications and their complexity can atest that. Gnome's stability makes it the best Linux native DE right now, but the gap between Mac OS and Windows is wider and wider every year. No wonder why people choose WSL as their everyday driver, they have Linux and work environment at the same time.
That is also me. I used i3, dwm, awesomewm etc. I spent just too much time maintaning these frankenstein configurations. By setting just a few shortcuts in gnome I get 90% there. My graphical shell is anyway just used for launching emacs, terminals and a browser.
> I spent just too much time maintaning these frankenstein configurations. By setting just a few shortcuts in gnome I get 90% there.
Curious what you mean by "maintaining these frankenstein configurations", and what you were trying to do. Was the WM under heavy development with user breaking changes or something?
I can't argue with your experience, but I've only touched my StumpWM config a handful of times in the last 10 years or so, usually to add a new defcommand or keyboard shortcut.
Once I had my WM running I had to clobber together lots of different pieces together (with scripts and whatnot) to get the remaining desktop functionality. For example:
Some people might argue this part of the unix principle to have separate tools that do one job well but frankly I don't care about the unix principle as long-time emacs user (the anti-thesis of unix). The end result was a frankenstein thing: random tools glued together to emulate a basic desktop experience.
What always did, going through awesomewm, stumpwm and xmonad to now sway, was to used whatever of those services were available from the OS. Some I had to launch by hand, but that was a one time change always (some of them across WMs). Nowadays it is even easier, since most of what I needed to launch manually are now user-level service from systemd.
In the end, I am running most of the normal services a GNOME desktop would be running, but without using the GNOME shell, and that has worked very well.
Really? Most of that is pretty simple to do in i3 / sway at least.
Every one of those things worked out of the box for me the last time I installed (which, admittedly, was via Manjaro, so a fair amount of config came from them). Even prior to that when I rolled an i3 setup by hand, external displays was the only troublesome bit.
It's definitely possible that some distros improved WM based setups. Last time I as down the WM rabbit hole was over 4 years ago. I spent quite a bit of time to get something usable (and it was fun!).
The time required to set up any environment with robust options for configuring and modifying the way the environment works and multiple different components for any given job is less related to the minimum or reasonable time to achieve a functional environment than it is to how long it takes to explore the many different better ways one could achieve.
Shall install vim and be happy or shall we explore the entire environment of plugins and pick the 50 we like best after trying 150 then install Emacs then find our 50 favorite addons there.
This is my experience as well. I use i3 and change a couple of lines every now and then. The other components like dbus, portals, pipewire, just work. But I am not doubting the parent's experience either, since people can have wildly varying setups and issues across distros and hardware.
When someone says "that works out of the box" what I hear is "I don't want to spend time customizing things to work better for me", which is fine, but surprising coming from someone who claims to have used tiling window managers.
The core discussion here is: "how much should the computer adapt to the user vs how much should the user adapt to the computer".
On the one hand, the benefit of customizing a system for yourself should be obvious.
On the other hand, adapting yourself to a system also means adapting to other people. This allows improving the shared understanding of how computers should be used:
- Designs can push users to more efficient workflows.
- New features, use-cases and apps can be developed based on shared assumptions.
- Other developers can integrate with or expand on the existing designs. eg: MacOS launchers [1] build on various system features.
In IntelliJ, I regularly discover new functionality that I wouldn't even know existed had I used a text editor with extensions.
Customise your system too much, and you cut yourself off from this "conversation".
Perhaps this is why some users still live in heavily customised terminal environments, despite massive improvements to GUIs [2].
It means: "I don't want to spend time customizing things that may not even work better for me."
I acknowledge that some people like crafting their own computing environment[1], and it can be really interesting and rewarding , but some people like to work on other things, or simply lose interest in fiddling indefinitely with config files and all (like I have).
> "I don't want to spend time customizing things to work better for me"
That's me.
I have tried every major DE, and something or the other is not right with all of them!
I used to use gnome with a lot of customizations.
Over the past couple of years I decided to adapt my workflow and muscle memory to what's provided by default. Works well for me now, even though it was a little awkward at first.
I'm currently using gnome with no customizations.
I'm not claiming gnome is the best DE or anything.
I also "claim" to have used tiling WMs (in fact I'm using one right now) and Gnome is my favorite DE as well. I've tried pretty much everything and in the end it doesn't matter what you use as long as you can define useful keybindings.
And imho dynamic tiling is the wrong approach. Automatic rearrangement of windows never works well in real situations that just don't fit a small selection of fixed layouts. Not sure why static tilers are so rare, but Gnome offers static tiling and is very polished, so I guess that's why I like it.
A tiling window manager is one which by default automatically assigns each newly created non-dialogue windows to a non-overlapping share of the screen according to a set of rules.
A lot of environments have a function to relocate the focused window to a predefined portion of the screen. For example left and right half. This makes it a stacking environment with limited tiling features.
I think it's more like "I don't want to be overwhelmed with lots of decisions right at the start, but I want an ability to change things incrementally to serve my needs as I go". Which is why I like Plasma too for its out-of-box readiness and configurability.
Sorry I missed this reply. I tend to usually have full window panes for most tasks and then introduce others ad hoc when necessary, usually running scripts or testing code I'm working on in a repl. I don't do anything too fancy.
I used Xfce for years after Gnome 3 happened, but eventually found KDE to be better in every respect. I like that Chrome reopens windows in the correct virtual desktop, for example — something that never worked in Xfce.
KDE is great, would like to use it full time, unfortunately I find it kind of buggy and not polished for my use case. Two blockers I stumbled upon last time I used it.
I use the app that is not showing proper app icon but instead generic icon, found some bug/discussion somewhere, basically KDE developers won't workaround when app has bad metadata despite every other DE doing it.
Sometimes notifications don't pop up when app receiving notifications is opened, only when it's minimized in system tray, also works fine in every other DE I used.
Am I the only one who feels that Gnome, somewhat hilariously, is unable to get basic consistent UI over 2.5 decades - something that KDE nailed down at ... what was it? ... KDE 1.1?
I mean it is kind of sad, but educational - you get the project without clear goal, with zero consistency, without any regard to backward compatibility - and after 25 years you end up effectively with a mediocre window manager and a bunch of programs with wildly inconsistent UIs (which predictably break next time your UI toolkit goes from 2.4.8 to 2.4.9 making a migration to the new minor release a multi-year saga which invariably loses features that worked before)...
Say what you want about rms but he indeed has a talent of ruining a successful (and free) software project with his religious views (Gnome vs KDE, Guile vs Tcl, the list goes on)...
> Say what you want about rms but he indeed has a talent of ruining a successful (and free) software project with his religious views (Gnome vs KDE, Guile vs Tcl, the list goes on)...
To blame rms of all people for what you've described demonstrates a lack of understanding his scope of influence and participation in gnome especially in this era.
I can get behind blaming rms for gcc's missed opportunities, ultimately ceding its dominant position to clang/llvm. He clearly overplayed his hand there.
But when it comes to gnome I recall little more than mailing list bickering over gnome being a GNU project or not and what the G in GNOME stands for. Perhaps my memory is failing here, but it feels crediting rms with anything more than wasting a bit of gnome devs time and frustration would be overstating it. Not that any FOSS devs need more of that with how entitled and obnoxious users already can be, but they tend to have developed thick enough skin to ignore rms's aspy silliness.
My memory might be failing me but the whole raison d'être for GNOME project was the call by RMS. The premise was as follows: Qt was not truly free (despite promises by TrollTech at the time to address) and thus KDE could not be considered free desktop software.
rms wasn't alone in this view that a non-gpl DE was unacceptable at the time.
Also don't forget we're talking about an era where C++ was still largely despised for being a terrible and poorly supported language. The folks working on gtk/gnome were going to be living in C regardless of what fell out of rms.
I use the latest Ubuntu with Gnome and it seems… OK? I’m not super picky, I can find my way around - and tolerate - anything as long as it’s not Windows.
That said, I have a Framework laptop for my personal use (my company gives us Mac devices), and they say Fedora has the best driver support. Since I’ll be evaluating a new distro, I’m also considering taking another serious look at KDE after a long time using Ubuntu’s Unity and Gnome. Concretely, what would I gain, and what might I lose?
I still miss Unity sometimes :( One thing I don't particularly like about the "new" Gnome is that it makes the searchable launcher[0] fullscreen. So whatever context you're in is completely covered up by the launcher when you simply want to search for something to open. Remember when Windows 8 did the same thing? Microsoft moved away from that pattern since then. For all of the issues people had with Unity (the Amazon integration circa 16.04), at least it got the launcher somewhat right[1]
I have the same attitude as you do. When I was younger I spent so much time tweaking and changing Linux distros and desktops. Now I just install any Debian based distro and enable the dark theme. I pretty much leave everything as default. I spent most of my time in the apps I use, DBeaver for database work, Firefox and whatever editor I am using. I am bemused by the passionate partisan comments in this thread. I am long past that but thankful to all developers working on Open Source desktops.
Yeah, I have a hard time justifying hyper-customization. I understand that folks want to work efficiently, but as a dev, I need to stop and think about what I’m doing. I also have a tendency to rabbit-hole (actual feedback from my peers and manager), and if I leave myself unchecked I’ll probably spend more time “optimizing my workflow” than doing actually useful stuff.
These days, the only real changes I tend to make are installing oh-my-zsh (thereby outsourcing the low-level tweaks to someone else) and Sublime with a handful of plugins (does that count?).
I’m the same way, not picky as long as I can figure out where to click. I just did the reverse of KDE -> Gnome, I miss KDE here and there but I’ve been able to take the pieces I miss and replace them with gnome extensions.
Gnome was faster to get started with but KDE felt deeper. KDEs UI is denser, which imo makes it feel less “modern” but both KDE and Gnome are usable.
Best thing you can do is to spin up a VM and try it out yourself! Everyone’s different.
I'm really glad to see this because I think we can now have a smarter conversation than the typical one on why people dislike Gnome so much and why it is okay, even good to criticize it. People get angry at me for insulting it because of the ridiculous "It's open source so you don't have the right to criticize" argument.
As this shows, there's a lot of corporate contributions to the codebase. For better or (mostly) worse -- that means the project will likely have much more of a "proprietary-like" mentality than perhaps what we think of when we say Free/Open Source.
This plays out, right? The thing people don't like about Gnome frequently boils down to backward-breakage and not being able to meaningfully contribute.
And yet, Gnome enjoys a lot of popularity in the space that affects the "Linux" ecosystem, and I think the way they do things messes it up. This is why, for now, I am explicitly anti-Gnome. I hope it changes, or I hope it gets less popular.
A couple f years ago I installed linux on a machine to do some dev that wasn't supported in windows. My first time exposed to gnome. I have never used such a mis-designed clueless UI, and I could not believe the number of outright bugs I encountered. It was like it was designed by kids.
It's funny because that's my exact experience using Windows. Buggy and incredibly bad defaults, completely mouse driven with insufficient keyboard controls. macOS is bearable and I can be productive on it, but I think GNOME is the perfect desktop for productivity. To each their own, I guess.
GNOME is quite decent in its current form and the most user-friendly desktop out there at the moment. People who hate on it just have some historical gripe with it, I feel like.
Ideally I'd be using something like Sway, but I can't be bothered to spend time on making things work the way they need to. GNOME gives me everything out of the box and doesn't get in my way. KDE is much more customizable, but less friendly on the eyes. Windows is probably OK if you're used to Windows (I'm clearly not). I have to use macOS at work and it's just a worse version of GNOME - bad defaults, weird design choices, less features, minimal customizability.
It's perfectly OK to mistake a subjective preference as objective user friendliness its most people's definition of the term. What is negative is to confuse disagreement with your preference with being unreasonable.
> People who hate on it just have some historical gripe with it, I feel like.
The way it handles virtual desktops with multiple monitors out of the box poorly thought out. With 3 monitors having changing workspaces only effect the primary monitor is very poor user experience. that leaves virtually pun intended without the affordance of virtual desktops at all.
There is a setting to expand virtual desktops to all monitors and now you no longer have to dig through something that looks like the windows registry to enable it which is indeed a nice upgrade but it misses the vastly superior third option of being able to independently switch each monitor.
An obvious affordance instead of digging through a settings menu would be a little iconic padlock beside a representation of the virtual desktop switcher that when unlocked enables you to manually switch a singular monitor or when by default locked allows the monitor to change with every other monitor.
Trivially enables not only all 3 possible workflows but allows one to discover this organically at the cost of a small amount of screen real estate.
This is a singular issue but their entire history is rife not merely with subjective differences in user preference but objectively bad design.
Bugs? In software? Never! By all reasonable accounts GNOME is a well-run broadly-scoped project with a perfectly normal amount of bugs given it’s size.
This is disingenuous. The implication of the grandparent comment is not that the persons experience happened 2 years ago instead of right now its that its automatically invalid because of it. In context the word historical doesn't mean anything useful.
Title of post: "Gnome Turns 25". So maybe and maybe not. I just ran into so much WTF I find it hard to imagine it's fixed now, plus some others here rather agree. It's enough to put me off anyway.
The odd things for me, having tried more or less every single Linux desktop under the sun, including several that no longer exist, is that there's no one definition of "user friendly" that holds true for everyone.
In this story's comments, there are people saying Windows is the best, others saying certain particular versions are best, others saying they find it unusable or at least hard.
Yet this has been the best-selling desktop OS in history for about 35 years now, used by _billions_ of people, so it must be getting something right.
Counter to that, there are also people castigating Macs and macOS. That's normal; there are as many biased fanboys against as there are for.
And yet, again, for nearly 40 years now, Apple has been THE ONE COMPANY to resist the rise of Microsoft, and has a fantastically loyal fan base and makes a lot of money.
I also have a number of blind friends, and they mostly tell me that Windows is the most accessible OS there is, that it has the best selection of assistive tech, that the apps are more accessible, and so on.
Some favour macOS. What macOS provides out of the box is way better, it's true. If you're a casual computer user -- bit of surfing, bit of online chat, very occasionally write a letter -- it's better for blind users than Windows.
If you have a job to do, in business, and need rich powerful apps, and need them to be accessible, my working blind mates tell me Windows easily trounces the Mac.
I am not blind so I must take their word for it.
But I can make Windows and macOS and my preferred Linux desktops, Unity and Xfce, stand on their heads and do back handsprings for me. I regularly read people telling me that any of these OSes just can't do X or can't do Y, when X and Y are things I do on a daily basis.
What this really means is: they don't know how to do X or Y, and they haven't bothered to look for instructions or guidance. It doesn't do it -- whatever "it" is, it varies a lot -- and so they decide it can't, it doesn't work, and they move on.
Don't believe me? Look on Quora for the dozens of idiots asking "why can't Macs do cut and paste?"
In terms of the mass market, outside Xerox PARC, Apple invented the industry-standard method of C&P and defined the keystrokes every other OS now uses... for the Lisa and the Mac. Of course they can.
What the idiots mean, but are too dim to know they mean, is that the Finder doesn't do cut and paste. No, it doesn't, for excellent very solid UI and HCI reasons that cause millions of dollars of data loss every year on Windows and have done since 1995.
But it's symptomatic.
People mostly don't know how to drive Windows and Windows-like interfaces with the standard keystrokes. They don't know how to search it, how to manage windows with the keyboard, how to manage virtual desktops, stuff like that.
Because they don't know, most don't do it.
Therefore most of the desktops for Linux are half-baked rip-offs of Windows that don't implement the clever stuff, because the people that implemented it didn't know the clever stuff.
So it doesn't work.
Along came GNOME and ripped all that out. If most people don't use it, then clearly, it's unnecessary so let's bin it. So it forced users into accessing the limited remaining functionality via defined keystrokes and gestures.
As a result, people have to learn the commands, and they can because there aren't many.
And the end result of that is that they then praise GNOME for being "powerful" and "efficient" because they were forced to learn stuff.
Windows did that better 27 years ago, but because of good design -- and I am no MS fanboy! -- you didn't have to learn it. You could point and click your way and stumble across a way to do it.
It's sort of Perl vs Python. One gives you a dozen ways to accomplish something; the other has one way that's encouraged as being "natural" or "pythonic".
Perl fans loved it for its power as a result... but they can't read their own code, let alone anyone else's. In the end that's doomed the language.
Python is easy enough for almost anyone with Clue № 1 but many already-skilled people hate it as a result of its enforced rules.
Programmers know this stuff and accept it. They rail about it, but they accept it.
Programmers typically do not know desktop tech well. I lived with 2 and was engaged to one. They could out-program me drunk, but they were not techies. It's a different skill.
So when someone comes along and says "hey, you know what, I am an expert in this stuff and environments A, B and C have this large feature set and cover 75% or 80% of the functionality of the OS they were copied from," they are probably right.
Then someone who knows just 10% of that functionality uses Desktop D, which only does that 10% but forced them to learn how to use it properly, and they say "no, Desktop D is better because it works and it's really efficient and has 100% of the functionality I need, and I'm a programmer, I know this stuff, so this is all anyone needs!"
It is amazingly frustrating to head this kind of advocacy, know that you know far better than the person doing it, but not be able to explain to Mr Loud-Confident-And-Wrong that there is stuff he hasn't considered and the big picture is a lot more complicated than that.
But what's worse than that is when the advocates of Desktop D are so loud and so confident that they persuade billion-dollar corporations to standardize on their fairly poor product... then they throw conferences where they pat each other on the back for their cleverness, and they patronize people online for being stupid from their position of smug, entrenched ignorance.
This is an informative and interesting comment but I would like to reply to a few points.
> Programmers typically do not know desktop tech well. They could out-program me drunk, but they were not techies. It's a different skill.
As someone who grew up in the 80s and 90s this comment is strange as hell. In the first place possibly because if you weren't at a point in time a techie you weren't liable to be a computer user let alone be programming one. But honestly compared to programming being an expert computer user is such a microscopic topic comparatively. It's like saying although he was a master automotive engineer who oversaw the development of countless cars he never actually learned to park one. I know a lot of people are rather bad at it but its really a microscopic topic.
Regarding finder and cut and paste I was curious so I looked this up thinking to find people complaining instead I found this. Seems like you can "move" something to another folder with a only slightly different shortcut. This doesn't seem much different than cut and paste. It's not like the cut operation instantly deletes it from the present folder on any other desktop interface. If not paste follows it just doesn't get moved.
> Therefore most of the desktops for Linux are half-baked rip-offs of Windows that don't implement the clever stuff
Gnome seems more like a rip off of Mac, and the majority of other environments have their own "clever stuff". If anything Windows 10 looks more like a rip off of some of the better features from Mac/Linux after they tried and badly failed to go their own way with Windows 8.
> As someone who grew up in the 80s and 90s this comment is strange as hell.
I was working from the late 1980s, and living with a professional developer by about 7Y later.
It surprised me a very great deal as well, but I 100% stand by this observation. Your follow-on statements are entirely on the same theme:
> But honestly compared to programming being an expert computer user is such a microscopic topic comparatively.
To quote Pauli, "that is not only not right, it is not even wrong."
It's also highly patronising and frankly offensive.
As a professional tech-support person from the late 1980s for about 25 years, I and my peers and fellow pros in this field knew many many times more about the general spread of desktop and server OSes, their compatibilities and more to the point their differences and incompatibilities, than 95% (not a random guess, a genuine estimate) of the developers I ever met.
And the majority of those developers though the reverse too. Many developers just assume that they are the gods of IT, shaping the raw clay into systems.
They are mostly completely wrong, but it is the pervasive belief.
> It's like saying although he was a master automotive engineer who oversaw the development of countless cars he never actually learned to park one.
That is in fact almost correct, yes.
One, yes, it is the case, yes, if overstated. It is more equivalent to saying "most car designers are not also racing drivers" while sneaking in a dig that says that racing isn't a job and racing driver isn't a real job, but just a stupid hobby for idiots.
Hi. In this metaphor, I'm a racing driver. Nice to meet you. I know how to get your car to do things you never dreamt of, I know how to customise it, and I also have spent longer than your lifetime fixing up other people's cars, teaching them to drive, and winning races.
No, in fact, I do not think it is reasonable for you, as a car designer, to mock my work. I think it is rude and ignorant.
Yes, I know many car designers. Yes, I know more about how their cars perform than the designers. Yes, I know how to drive those cars better than their designers. Yes, I can fix them too, but I would not presume to tell a professional mechanic that they are doing it wrong.
But if you wanted to know what was wrong with a car, then YES I would definitely rate the opinions of racing drivers and pro mechanics over the designers, yes. 100% of the time, by far.
In fact, if I wanted to know what was wrong with a car, I'd never ask the designer and wouldn't be terribly interested in their opinions.
Sadly, FOSS is the domain only of amateur designers-cum-builders-cum-racers-cum-repair-bodgers. There are almost no pros in this entire sector and they have zero respect for the real pros in the commercial market, who know 1000 times more than anyone in FOSS.
> I know a lot of people are rather bad at it but its really a microscopic topic.
More patronising rubbish.
Newsflash: the paid professionals deploying any complex machines know more about that software, how it works, how it doesn't work, how it fails, and how it breaks than the people who designed it.
This is true of everything from bicycles to cars to operating systems to desktops.
> Regarding finder and cut and paste I was curious so I looked this up thinking to find people complaining instead I found this.
Good. Glad you took my suggestion.
> Seems like you can "move" something to another folder with a only slightly different shortcut.
Deep sigh
The reason I suggested people reading my comment look this up is because I already knew this, and that's because I do in fact know my job and what I'm talking about.
I know you can move stuff. I know this because I've been deploying, training on, and fixing Apple Mac kit since 1988, and I was good at my job. I left it because 25 years of dealing with rather stupid customers is too much for anyone to have to bear.
I picked this example because it is a good example. Macs do have the functionality and have done since 1984.
However Windows only got a poor version of it in 1995, but that is all most users know, and when they find Macs don't do it that way, their response is not to find out the Mac way, it is to complain that Macs don't work.
> completely mouse driven with insufficient keyboard controls.
Certainly not. I suspect you aren't familiar with win, it has a lot of these although MS seems determined to wreck them.
I'm no MS shill, I wish they would DIAF, but I do find windows highly reliable and the UI was pretty good until they decided to fuck with it. By comparison I found gnome actively user hostile. Even small niceties that windows has are missing (eg one of very many, go up a folder in windows and the folder you were on will be the one selected, very handy). I would like to sit down with you and compare experiences, but ain't going to happen. Glad it works for you though.
Xfce (Thunar, Xfce Terminal, Mousepad) also uses CTRL-PGUP/DOWN to cycle tabs, as does MATE (Caja, MATE Terminal, Pluma uses CTRL-ALT-PGUP/DOWN). Neither uses CTRL-TAB for tabs but to cycle through buttons/icons on the GUI.
As far as I can tell CTRL-TAB for tab cycling is not very wide spread. KDE uses it, but that supports CTRL-PGUP/DOWN as well.
edit: And I'm not sure why CTRL-TAB even works in KDE because the default shortcuts to cycle through tabs are CTRL-PGUP/DOWN and CTRL-[/] (tested on KDE neon).
Funniest moment in UX design was probably when Windows invented the lock-screen curtain (the screen that you have to drag up to unveil a password input box), and the Linux desktops copied that.
> completely mouse driven with insufficient keyboard controls.
Is this really true now? If so, it's sad how far windows has fallen. It was very easy, almost pleasant to use Windows with no mouse, from 3.0 all the way through XP (which was the range of my regular windows use)
Yep. I find it’s absolutely horrible since they went bananas post gnome 2. All the discoverability and consistency is gone.
Honestly this is the biggest barrier to adoption of you ask me. I can’t hand a windows user or a Mac user a gnome desktop (or any of the other Linux options) and just leave them to it.
I tend to recommend a mac these days for people. They just work.
Cinnamon and KDE, and probably Mate, look quite similar to Windows, with the bottom panel having a "start up" menu at the left, the tasks, the status icons and the clock. The settings manager is arguably simpler than the numerous messy UIs on Windows.
I don't think the UI is the issue. I even think some DEs are better than both Mac and Windows on most topics. I don't like Gnome but it is okay I think, especially on Ubuntu. The rough edges are the issue, and above all, the incompatibility with Office (you might have to interact with people using it) and with the various domain-specific proprietary programs.
Windows also has - different - rough edges (and probably mac but I don't know it enough), but you'll find help online or from a nearby acquaintance.
But if I had to set up a computer for a complete newbie, that would be Linux. In any case I would have to help them, but on Linux there are just less disturbances, less risks, easier for them, the UI is better structured, everything is more predictable and since I'll need to help, I'll have better time with Linux.
My issues with a WM usually comes down to UI scaling. Modern laptops (and desktop monitors) with a 4K display render things too small at 1x scale, and too large at 2x scale. Windows and OS X solve this problem by allowing 1.25x, 1.5x, 1.75x scaling with no blurriness; it's not a simple linear resolution scale. Today on Linux I can only find this on Wayland, so I'm already restricted to certain WMs (like KDE, swaywm, Gnome), and even then it doesn't work right across all apps. And when this topic is brought up [1], it's met with "Well, let's argue about the definition of 'sharpness'." If I was a 4K display user trying Linux for the first time this would affect my perception of Linux on the desktop negatively.
My laptop has a 4k screen. My monitor is a 4k screen. I use Linux in a VM. Extending my laptop screen onto my monitor screws everything up with Linux. Windows is fine. And then KDE and Gnome have different UI scaling logic. It’s a real mess.
I'm using hidpi scaling on 13" 4K screen with Plasma/X11 just fine - that is, until an external screen that wants a different scaling factor is connected. This is still fine though; the only cases where I attach an external screen are presentations or other fullscreen content where the mismatched scale doesn't matter at all. However, if you want to generally work with multiple screens with various scaling ratios, you need to go Wayland indeed.
The problem with Linux on the desktop is that all polish, quality and consistency is "soon" and has been for 20 years. No one wants to do the last 20% of the job when you can rewrite the interesting bits again for the 10th time.
For a polished GNOME experience I recommend Fedora (once you know how to install the missing media codecs). Even Firefox there runs natively under Wayland and you also benefit from the latest advancements such as the Thunderbolt or performance managers.
It’s because Red Hat has the engineering talent and bodies to make a cohesive system by carrying a frankly insane amount of patches to bend all software to the RHEL way and tackle extremely ambitious userspace plumbing projects.
I’m so sad at what IBM will eventually do to them.
I think the proposal process works very well to modernize the distro and drive forward the improvements of the default installation, while in Debian the teams seem to do their own thing and innovation is very slow and even controversial (e.g., the systemd or merged /usr/ migration are sad examples but even small things like a good zram setup, systemd-oomd, and btrfs by default make a large difference).
Gnome was created in part as a response to kde using the Qt toolkit which at the time had some licensing issues. This is why gnome was based on the gtk toolkit which was free as in speech.
KDE is more powerful and configurable in every way, unlike of forcing the user into a particular paradigm like GNOME does. Also the software suite is stellar.
It's also full of bugs. Kwin_x11 regularily crashed on me requiring a manual restart of it. I'd rather have something that works than a lot of configuration options that don't work. I'm using gnome nowadays (without any extensions)
I tried KDE recently b/c that new Nordic theme is beautiful. But setting it up was a major pain, and there’s apparently dconf-like utility that can export the settings into a save file. Went back to gnome afterwards.
Some are, not all, and they're spread out across multiple config files. dconf by comparison puts them all into a single file, and provides import/export functionality that updates the UI without a restart.
I think both Gnome and KDE are pretty nice... Powerful is probably not the right word. KDE has a lot more features, and a lot more of those features are user configurable. Gnome has taken the path of a well executed, smaller feature set with fewer places the user can customize. KDE has taken the path of more features, but at the cost of more code, and more things that could have bugs. Personally, I love customizing everything, so I prefer KDE. But both Gnome and KDE are fantastic and give OSX a run for its money on usability, and both are superior to Windows 11 (which by the way is the best Windows UI since about XP).
As a long time user, and a front-end developer, Gnome 2 and its derivatives don't make much sense to me. "Vertical taskbar? Not on my watch!"
Sometimes I feel like I am the only person to setup a vertical bar and complain about them using a repeating 1x30px (don't remember the actual number) image for background, rendering the bar barely usable. Something MS nailed back with windows 95!
Gnome 3 is mostly fine but not before you reach for at least a couple of basic add-ons. I truly believe believe they are trolling people with the default behaviour of Gnome 3
I liked Canonical's Unity very much, although it's no longer the prettiest. I realized that last part due to the recent reboot
«
As a long time user, and a front-end developer, Gnome 2 and
its derivatives don't make much sense to me. "Vertical
taskbar? Not on my watch!"
Sometimes I feel like I am the only person to setup a vertical bar and complain about them using a repeating 1x30px (don't remember the actual number) image for background, rendering the bar barely usable. Something MS nailed back with windows 95!
»
Thank you for saying this, because I very often feel like I am a lone voice crying in the wilderness when I point this out.
«
I liked Canonical's Unity very much, although it's no longer the prettiest. I realized that last part due to the recent reboot
»
Also this.
This, though:
«
Gnome 3 is mostly fine but not before you reach for at least a couple of basic add-ons. I truly believe believe they are trolling people with the default behaviour of Gnome 3
»
This, however, less so.
It is possible to give GNOME 3 a partial makeover into something better, with a lot of addons, some now getting very hard to find.
The snag is that addons are tied to that particular version of GNOME Shell, and thus inevitably, an OS update that updates the shell breaks your entire desktop.
The misplaced confidence of the GNOME team is such that they provide but do not embrace the addons scene, and so there is no support for a safe mode, for example, so that if your addons crash the desktop can recover.
I would guess that the dev team use it unmodified and do not test for such things.
"If you break it, you bought it." IOW: we didn't supply the extensions, so if they broke your desktop, it's not our problem.
I love using Gnome but I hate the Gnome developer mindset. Everybody but them is an incompetent fool because they dare use their precious code in a different way. Every time one of these issues come up I consider just switching to KDE because although the UI doesn't sit as right with me, the developers and culture around the KDE project just seem so much better.
If users have to use debug flags because your software has no customizability whatsoever, the fault isn't with the users using those flags, the fault is with the developer not offering proper solutions.
The GTK file picker is practically unavoidable. It's in important applications like Inkscape, Gimp and was unavoidable with electron apps too until recently.
Moreover, most users don't actually follow the discussions with the developers. They would not know about their philosophy. Also, this is not the only criterion when choosing software. And given how Gnome and GTK are (still) everywhere, I'd find it worth having this kind of things fixed even if I don't/didn't personally use the thing.
The closest thing to a solution is to blame the distros for shipping the apps incorrectly and encouraging users to do unsupported things like combining several apps from several different desktops together without testing them, when this has never really been a thing that ever worked correctly. The correct way to get portal support is to use the flatpak versions of those packages, because those are the only packages that have actually been designed and tested to work with the portal. And if you are a person who really does not like flatpak, then you should ask your distro to start integrating sandboxing and portals into their own package managers and testing/supporting the packages that way.
The Correct Way™ is actually NixOS/GuiX these days.
I had major issues when I used Snap GUI apps. Like it was not clear where did the config file saved to or sometimes the file picker couldn't even detect my drives (due to sandboxing). It was doing its own thing regarding theming.
If Flatpak is anything like this, then it's not a viable solution.
I do not believe NixOS or Guix are shipping these packages with a sandbox that uses the portals, but I could be wrong. I am not sure of anything about Snap, I have never used it.
In Flatpak, all the config files are saved to ~/.var/app/. The portal should be able to handle exposing a file from the host, but the app has to be built to use the portal correctly. You can have themes but you must install the flatpak package for them, pulling themes from the host environment is a security issue because technically themes are equivalent to doing code injection.
In this case, the arch wiki was actively telling users to do things that could break their system. It was a debug flag that was only intended to be used by developers and causes a number of other side effects. So it is no surprise that they changed it. Technically it was not removed, it was only moved into a flag that is only available in debug builds.
So why don't they add a proper flag? I'm now running debug builds of gnome to get that functionality back.
I wish everyone would abandone
GTK and instead switch to Qt, but as long as that hasn't happened yet, I'll have to use such environment variables to get GTK apps to look and work consistently on a Qt desktop.
The short answer is because any flag will break other things. There is another suggestion to not require any flag, but it probably will not be implemented because of this: https://gitlab.gnome.org/GNOME/gtk/-/issues/5026
It is not possible in open source to force everyone to use one toolkit, in my opinion you are wasting your time by wishing for that. There are dozens of toolkits to choose from and the number seems to be growing. The fragmentation has only gotten stronger over the last 25 years, if Qt was the clear winner the proliferation would have stopped long ago. Now there is pretty much no chance that you will be able to hack every toolkit to get all of them to look and work consistently. In particular, the way an app works is decided by the HIG it follows, not the theme. This is reflected in the entire design of the app, to change it requires redesigning the whole app. It cannot be changed just by setting some environment variables.
It’s mostly an issue with games or java apps or electron using Gtk dialogs. Those apps don’t feel native anyway, but the areas where they interact with the OS at least should feel native.
I don’t care about Gtk apps themselves, because I just don’t use any proper Gtk apps (I want my apps to be customizable, after all, and I build my own programs to be customizable too).
Why is it always Gnome developers and fans who hate any attempt to make Gnome apps look and feel more like other toolkits? No one expects Gnome apps to work 1:1 like Qt apps, but it feels like every step in the right direction is met with constant hate and objections.
Huh, looks like I've been using Gnome for most of its history (since Feisty). Replacing the default window manager with xmonad has gotten a little more complicated over the years I've been doing it but I'm still happily writing this from my Gnome-monad setup thanks to the magic of autostart and `--replace`.
The maturity of a desktop project that drops the GUI designer, expects devs to simply type raw XML, while the upcoming designer uses Web technologies to draw what are supposed to be native UIs.
This is not a "bug" or an "issue", it is a matter of taste. Some users prefer search, others prefer a simpler "jump to file" function. The maintainers have decided to implement the former. Arrogant comments like this really make me wonder why people are still putting work into open-source projects.
The newer issue proposes to provide both methods, similar to windows. The “bug” seems to be that the maintainers prefer one style and refuse to let others add the alternative. That would jibe with what I’ve read about other anti-user attitudes in Gnome.
This is a project that officially says things like "unless it's an accessibility issue users shouldn't be selecting system fonts" and "why are we even talking about allowing themes?"
Adding more and more confusing options is generally what is seen as being anti-user. If you really care about this, it is trivial to use a different file manager with GNOME. It is not going to be useful to keep insisting that designers capitulate to a set of demands, after years of them saying that it conflicts with the stated goals of the system.
Shouldn't that then be left to user choice? The attitude that there's only one true way to do things and that's the developers' way is what put me off Gnome.
In general, it is better to leave nothing to user choice, if you can get away with it. By using any developer's software, you are conceding that their way is the correct way to do things, otherwise why waste your time with that software? There is no software in existence that is able to do every possible thing that anyone would ever want to do.
Hard disagree. And no one is arguing that _everything_ should be user choice, but there is certainly a meaningful accommodation that makes power users and tinkerers happy. We've had it more in the past and still have it with some software in the present. I'd argue this friendliness and synergy with the user is always the case in the most powerful and beloved software.
It just feels smug and talked-down when someone comes along and sands over a reasonable amount of control and complexity that previously existed. I hate having my workflow disrupted because I like getting work done.
Would you argue that we should have no shell aliases? No BIOS and boot order? No desktop wallpaper? No ability to set homepages in the web browser? Etc etc.
The computer is like anything else I own and the more it responds to my wishes, the better. Small tweaks left to my discretion remind me of that, losing control piecemeal provokes disquietude. If it doesn't accommodate my needs then it feels less mine and for that reason I reject software that restricts my choices.
I do not understand what you are disagreeing with. Not every project is aimed at power users and tinkerers. You seem to be suggesting that they all should be, that is not realistic. If you feel talked down to, please consider that you may not be the target of discussion.
>Would you argue that we should have no shell aliases? No BIOS and boot order? No desktop wallpaper? No ability to set homepages in the web browser?
Those could be nice to have. I can say I do not actually need any of those things when I use my computer to write documents in a word processor. Do you see where I am going with this? Any "control" you want to have is completely up to what you are trying to do, if all I am trying to do is write my novel then all those other things can get in the way.
> In general, it is better to leave nothing to user choice, if you can get away with it.
I'm glad there are people out there that do not follow this user-is-an-idiot philosophy and created essential software like Emacs, Vim, ranger file manager, and i3wm.
Surely you can see that those are a very small subset of the software available in the world, and most software is not like that. It is not "user-is-an-idiot philosophy" to have developers actually do their job, which is to make the tough decisions about how the software should be designed for purpose, instead of offloading that task onto the user.
If I wanted to use software that works for most people and doesn't believe in allowing users to make choices, I'd use MacOS or Windows. I can't see a reason to use Linux or BSD in that case.
It's a bug when it stalls out your computer for 2-3 seconds because the file picker is doing a recursive search through all of your files.
Thankfully, patches that fix this (along with some other bugs and anti-features) are actively maintained by community members, despite being ignored by the GNOME project.
>If they hate it, they can stop. Work on a project that has a higher proportion of praise if they need to. Work on proprietary stuff.
Please stop, this is gatekeeping. What is even the point of open source if we are going to insist that only projects that reach a certain "proportion of praise" are allowed to exist? That would mean that only the projects that were already popular will succeed, and everything else will fail.
If that is the case I am curious what motivates you to chime in on news about Gnome then?
Gnome threads seem to bring out a lot of anti-Gnome commenters and I've never quite understood why so many people who dislike Gnome are nevertheless compelled to continue to talk about it.
1) It's not just Gnome that's affected, it's also the gtk3 FileChooserDialog, which is used by many applications (e.g. Firefox), independent of desktop environment.
2) I do actually want Gnome to succeed because I think it's healthy for the wider free software ecosystem to have a selection of high quality desktop environments available. I want to be able to recommend mainstream Linux distributions to friends without them getting tripped up by things that don't work the way they expect. I think the devs (or the HIG architects) have blundered in prioritising search over find and I want to see it corrected. I am remaining engaged by adding my voice to the growing chorus. If I disliked Gnome I'd walk away and let it stay broken.
>1) It's not just Gnome that's affected, it's also the gtk3 FileChooserDialog, which is used by many applications (e.g. Firefox), independent of desktop environment.
I do not understand this. Firefox and those other apps are not beholden to GNOME, they can implement any file chooser they have an interest in implementing. If they have chosen the same one as GNOME then it is likely because they prioritize implementing with GNOME over other things, so you will have to accept that those applications are not as interested in bothering to be independent of desktop environment. Some applications will always be like this, for example Thunar is designed for XFCE and it does not fit in so well with Plasma.
>I think the devs (or the HIG architects) have blundered in prioritising search over find and I want to see it corrected.
It is not useful to say this without having some kind of larger justification for it. Just repeating this is an anecdote, it is not going to change minds. If you want another anecdote, I have never liked find, I found it worse than search in every way. If I keep pushing from this angle and telling you that I want to see your opinion corrected, would that convince you? I doubt it would.
I appreciate that you still care about getting these specific issues fixed, and I also hope they fix them.
It still feels a little odd to me though, because the initial blog post is an analysis of where Gnome's committers come from. I'm not sure how these bugs tie into that. Do you make a comment to draw attention to these issues whenever the terms "Gnome" or "GTK" show up in a thread? If so that seems like a lot of dedication to a desktop environment you don't even use.
The closest modern desktop environment to Gnome 2 is KDE’s Plasma. It’s easy to use out of the box and doesn’t require any configuration (unless you want to do that kind of thing, and then it’s available — unlike in Gnome 3+).