Libadwaita 1.0 Released For Kicking Off A New Year Of GNOME App Development

[ssokolow]

Depends on what they want. If they want a "just working design", where they can focus on features it's libadwaita. If they want a more flexible, user-themeable app or a different hardcoded style, it's pure GTK.

...or, if they want both, they can use Qt, which gives you both and has had "give both developers and users what they want" features that GTK either still doesn't match or is only just getting now since back in the Qt3/GTK2 era.

(eg. Persist the window geometry and toolbar/panel state in six lines of code, toolbars and panels have support for repositioning, tear-off, and hiding/showing via context menu by default, etc.)

[ssokolow]

I know this is tangentially related at best, but I just need to say "Huh. Would you look at that." to someone.

I just realized the common factor in a "CJK glyphs render as placeholder characters in some Flatpak apps" bug I just reported for several apps...

All the ones that Just Work™ are Qt-based, Java-based, or use custom toolkits (Blender, Firefox, Godot, Thunderbird, Tor Browser, Ungoogled Chromium) while all the ones that are broken are using GTK.

(Though, to be fair, there are other custom-toolkitted apps like ScummVM that don't have the problem because their text-entry fields rejected my attempts to paste in my test string.)

[arQon]

I find all this obsession with desktop styling or customization through extensions ludicrous. Maybe for hobbyists but as a professional I want my DE to be confident about the experience its providing and just work.

Then, wow, you have a novel interpretation of "professional" (not to mention a clearly unwarranted sense of superiority, eesh).

What I want "as a professional" is a desktop that works for my workflow. And since my workflow isn't "Watch YouTube on a tablet, then go skim Twitter, etc", that means that even a "good" set of DE defaults is not going to be right for me.

> If I was a member of the GNOME team this would infuriate me and would easily push me to remove all support for extensions period. You get what you get, don't like it then just use a different DE. It seems they may be moving more in this direction and I applaud them for it.

That's exactly what they're doing, yes. The fact that you think it's a good idea though only shows how little experience you have in this world.

> this type of fragmentation is what I believe to be killing Linux desktop adoption for the mainstream. T

"what you believe" carries no weight at all, since the reasons are well documented and most of us already know them. Hint: you're hopelessly wrong. (Though you're by no means the first person to try to blame it on that).

> Styling your DE to [...] having a dock bar vs no dock bar isn't helping innovate on the Linux DE experience.

I, uh, what? So your take is that if GNOME ships without a dock, nobody should be allowed to add a dock to it, because doing so wouldn't be innovative? I can't really make sense of what you're trying to say there.

> Maybe these larger distro projects could focus their efforts on improving things like compositor performance, software installation workflows, more GUIs to help people less terminal savy, BUGFIXES? These are things that make a real difference.

Bugginess and poor performance make up a sizable portion of GNOME's bugtracker issues. They haven't been particularly successful at - or even interested in - improving either one, ever. Given that more users surfaces more bugs, and more distros means potentially having more resources TO fix bugs, performance, etc, your insistence that the "right way" forward is to reduce GNOME's user base and reduce cross-distro development cooperation from "almost none" to "absolutely none" seems rather counter-productive.

> Oh, and the last thing you should do is waste valuable FOSS momentum by trying to write your own DE in a new catchy language because your feelings were hurt.

Heh. Yeah, you should learn about the history of GNOME someday. Except the "new language" wasn't catchy, it was so staggeringly bad that their own team abandoned it. (Vala, if you're curious).

[ssokolow]

> Oh, and the last thing you should do is waste valuable FOSS momentum by trying to write your own DE in a new catchy language because your feelings were hurt.

Heh. Yeah, you should learn about the history of GNOME someday. Except the "new language" wasn't catchy, it was so staggeringly bad that their own team abandoned it. (Vala, if you're curious).

More specifically, GNOME was a "with blackjack and hookers" response to Qt's original license (after Xfce already existed, by the way. Xfce was rewritten on top of GTK later.) and KDE being written in C++ rather than C.

Then, GNOME developed Vala (a compile-to-C language) as a way to have a higher-level language on top of GObject Introspection without resorting to C++ and gtkmm. However, as someone who used Vala, I'm qualified to say that it had a PHP-ish smell to it, in that its API wrapping with so thin and automatic that you often had to work around "this API expects you to dereference a pointer and act on the raw memory to do X"-isms from the C APIs being wrapped.

(In my PHP comparison, I'm specifically thinking about how you can't get the raw pixel data from PHP's libgd bindings because they didn't provide an equivalent to "just dereference the pointer that acts as an object handle".)

Vala basically got killed when Rust came along and caught their interest as a better implementation of what they'd been trying to do.

[Debunez]

Bugginess and poor performance make up a sizable portion of GNOME's bugtracker issues. They haven't been particularly successful at - or even interested in - improving either one, ever. Given that more users surfaces more bugs, and more distros means potentially having more resources TO fix bugs, performance, etc, your insistence that the "right way" forward is to reduce GNOME's user base and reduce cross-distro development cooperation from "almost none" to "absolutely none" seems rather counter-productive.

Who said anything about reducing GNOME's user base? The point was that distros should use, contribute to, and ship STOCK GNOME and stop screwing around with other nonsense. Fedora would be a prime example of a growingly popular distro that takes this approach. This would ultimately grow GNOME's user base substantially.

Heh. Yeah, you should learn about the history of GNOME someday. Except the "new language" wasn't catchy, it was so staggeringly bad that their own team abandoned it. (Vala, if you're curious)

Vala is a failed language and basically on life support. No one would suggest or recommend writing code using it, and If i recall GNOME maintainers themselves have said something similar to this effect. My hope would be any projects that do use this language port themselves to C immediately. My point was actually that trying to write a DE from scratch in (insert that catchy new language) would just waste the years of effort that has gone into developing GTK, Mutter, and other core pieces of the GNOME desktop. But, surely they probably won't actually be doing that and will still use pieces of this pie which in that case they are again just leeches.

[Mez']

Both distros and users want theming.
Nobody cares about recoloring the ugliest of themes that adwaita is. You can plaster a wooden leg all you want...
Solus devs will both drop Gnome support in Solus and GTK in Budgie while we all know how Gnome closed-mindedness made it mandatory for Pop!_OS to lead their vision (the one Gnome lacks) elsewhere.
Solus will just leave the Gnome spin vanilla to avoid any issue, and vanilla being ugly and against what the users expect of a workflow, it's going to lose momentum. Not everyone is a Fedora cultist.
Distros want theming for branding and to give a nice and modern feel (that adwaita lacks) to their UI. And as users, we want to pick, and theming is part of what attracted us to Linux in the first place.
Gnome devs should make it easier to theme and not just complicate it further and force down their terrible theme to everyone through recoloring. Once again, they didn´t listen, and are going against what everyone wants.
They´re fragmenting Linux desktops once again, as these kind of dumb moves force others to leave the monolithic block that Gnome is. It started with Unity and now Solus and System76 have no other choice either if they want to materialize their own vision.
Let´s hope they deliver, we really need another strong counterweight (beside KDE) to give us solid workflows with more focus on covering the variety of user workflows and customizations.

[Debunez]

Both distros and users want theming.
Gnome devs should make it easier to theme and not just complicate it further and force down their terrible theme to everyone through recoloring. Once again, they didn´t , and are going against what everyone wants.

Ahh so thats what 'everyone' wants really? So people currently on Windows or macOS aren't using Linux because they can't customize their DE enough? Hmm I wonder what Microsoft or Apple are doing so well that allows those people to totally customize their DE experience. Oh, yeah nothing because you don't in Windows and macOS. So this is a bunch of nonsense.

You know what people do want? To be able to boot into their OS, install apps without friction, and have all their hardware devices work. I would say there has been major strides in Linux just this past year related to this, especially with Pipewire (another shoutout to Fedora doing it right).

Stop spreading this absolute horse shite saying the majority of people who use computers will only use a Linux DE if they can customize how their buttons or shell look. Its beyond nonsense and if it were true, Valve would have been pouring their engineering efforts into writing stylesheets or coloring API's instead of things that matter like runtime compatibility layers or drivers.

[ssokolow]

Ahh so thats what 'everyone' wants really? So people currently on Windows or macOS aren't using Linux because they can't customize their DE enough? Hmm I wonder what Microsoft or Apple are doing so well that allows those people to totally customize their DE experience. Oh, yeah nothing because you don't in Windows and macOS. So this is a bunch of nonsense.

Why else would we deal with all the warts and papercuts on Linux when we could just be running Windows (which we probably already have an OEM license for) or macOS (which is "UNIX, but friendly")?

Your straw man "So people currently on Windows or macOS aren't using Linux because they can't customize their DE enough?" argument is functionally equivalent to "So people currently on iPads aren't using GNOME because it's too customizable?".

(Spoiler alert: Alienating your existing demographic in hopes of appealing to another that are already comfortable on a competitor is generally a dumb strategy.)

[Mr.Elendig]

Back in the days of blackbox users had a set template they used and styled it using line colours, patterns, fills and gradients to the extent that many desktops looked totally unique and different. One the Amiga they had MUI gui that used the same principles but added the 8 image borders to the mix, so why in 2022 is gnome/gtk still limiting users in theming when even the older apps had a better system of use that could make not only the desktop but each program unique looking ?!?!

Literally uncustomizable

[finalzone]

Why else would we deal with all the warts and papercuts on Linux when we could just be running Windows (which we probably already have an OEM license for) or macOS (which is "UNIX, but friendly")?

By we, you mean you who choose to post on a topic you care less and yet think the world evolved around your view. Strongly suggest to full read the details about libadwaita without your own bias and with an open minded approach. Until then, avoid posting when you have nothing to contribute.