GNOME3

really? I know they had an SJW program, but did it really make them go broke?

it's the default of fedora and debian, two of the biggest distros.

You mean the gnome 3 fork unity?

unity really has a lot of differences from gnome3... and is it really any more flexible?

KDE devs are just as fucking bad:

forum.kde.org/viewtopic.php?f=285&t=122416

but the hamburger is removable.
On the panel, it goes away when you lock widgets. On the desktop, you just get rid of it with right click or something, I forget what.

so yeah, some prick at KDE thought he knew best, but eventually they actually listened to people's complaints.

They really did spend all their money on it, something like 300k USD in one year. It ended up being spun off into a separate entity so the parasite wouldn't drain its host (spin-doctored as it being REALLY POPULAR!!1)

I agree. Whether in OSX or Unity, global application menus are an abomination before god. That's one of the very few things I still don't like about Gnome 3.

That said, I've worked with Gnome extensions in the past. If you point me to that extension I could try to bring it to life for a buck.

extensions.gnome.org/extension/545/hide-top-bar/

No, it was a seperate entity that GNOME was managing and it was being paid for by other organizations.
The problem is that it was growing so fast that they couldn't wait for the time period in which other organizations paid their dues.

I think someone needs to fork GNOME3 and throw it into the garbage bin

We have XFCE, Mate, Cinnamon. All developers involved with bloatware could instead improve desktop environments that don't eat away at 1/8 of your ram

So rather than wasting vertical space on one bar for all windows you would waste vertical space on one bar for each window individually? Makes perfect sense.

lxde master race

Besides, having it at the top of the whole fucking screen is awkward as fuck. I'd rather it be close to the active window.

Was actually thinking of forking gnome 2 and making it a more adaptable, Human-interface using, Qt-based solution(without SJW bullshit, of course)
Still have the repo somewhere.
The GNOME devs are turning GTK into their personal toolkit tied to their DE and making it increasingly uncustomisable. Qt is cross-platform, supported even by ancient platforms like Symbian, and lots of programs, and desktop environments have migrated to it successfully.
LXQT has recently beat LXDE in frugalness, so no.

If you hide the menu you might as well have a hamburger menu then. Do you also hide the scroll bar?

...

Well, if you want extensions for a C program you need an interpreter of some kind, and JavaScript and Lua are the most common choices.

I assure you I've never seen a desktop environment that would eat 4GB of my ram. Could you point me to an example?

Why not do it like rio and have menus appear on right click.

You have to go out of your way to find a mouse without a scroll wheel nowadays.

Good luck scrolling 50 pages down with a scroll wheel.

Preaching to the choir, OP. It's a damn shame too, because Gnome 2 was quite good for its time. The "we will intentionally go out of way to make things less customizable" attitude is pure insanity. I personally suggest KDE4 until KDE5 is farther along.

it's about basic cognitive ergonomics. menu bars conceptually belong to window applications, not to a system-wide panel. that's what I mean when I say a global application menu is abominable, not that it's more consuming of screen area than the alternative solutions.

macfags brag about Mac OS being more coherently designed than MS Windows, except of course when Mac OS gets it wrong like the global menu bar and those Fitts-law-violating microscopic window buttons.

moreover, this is not a problem with modern GNOME applications. The GNOME UI guidelines don't even have menu bars. The main application menu is appropriately hidden inside a button in the header bar, so the button uses no horizontal space of its own. Oh, and I use a 4:3 screen, so I don't really need to worry about saving space in one direction or the other because I don't use slantedscreen abominations.

s/horizontal/vertical/

pleb

This is the problem. They autistically think every single program should be developed exclusively for GNOME3. Libre Office has a menu bar. Hell, actually, the GNOME terminal still has one.

KDE5 is pretty usable, again, it's what I use on my main desktop machine. I do wish they'd revamp their "show all windows" and "show all desktops" tools though. In my ideal version of KDE, I could configure it to to an "expose" view just like GNOME3 does, where it shows all the windows, desktops, and gives a search bar.

It seems we went from
>You have to go out of your way to find a mouse without a scroll wheel nowadays.
to
>You have to go out of your way to find a mouse with a good-enough scroll wheel nowadays.
pretty quickly, didn't we? That's good-enough for me. :^)

Embrace/Extend/Extinguish is not limited to Microsoft at all: Freedesktop, GNOME and GNU all regularly practice it to varying degrees and they are similarly cancerous with their faggotry. Remember when GTK broke themes all the fucking time because gotta build muh brand, gotta be apple?

Or shared libraries...

I'm kind of disappointed that the color manager didn't get redshift-esque functionality in 3.22 but that's about my only gripe with GNOME. And global menus, as someone else pointed out.

But I don't really much care. It works for me, and it works great as a Wayland shell.

Meanwhile KDE is still shitting the bed (although it's getting close) and a lot of KDE applications have their arcane dependencies. For example, just by installing their premiere qt mail client I needed a RDBMS as a dep. No thanks.

And I'd rather not see tearing/slowness anymore by going back to X.org with all of the other dead DEs.

not suggesting the superior in all aspects GNU extension language

I've actually programmed with Guile as an extension language before. It depends on global state, is not thread-safe, and is far fucking slower than either Lua or JS. It's also far less intuitive to extend with C functions and the documentation is fast worse. I like the language itself, but the implementation is far from the best, especially compared to the dream that is the Lua C API.

how does their favouring of their own aesthetic for the GNOME platform mean they want to force third-party programs to comply?
In other words why should GNOME (or any other desktop for that matter) be forced to put interoperability above originality and scratching your own itch?

Stop whining bitch.

You are missing one crucial difference in there. Freedesktop, GNU and GNOME (part of GNU) are libre. When Microsoft does EEE they do it with the explicit intent of monopolizing the market, and their extensions aren't free software so competitors can't re-implement them without going through pain. Apple intentionally breaks interfaces to their shit closed platforms (only god knows why anyone would like to enter that nightmare) to outmarket compatible products. It's not like GNOME enjoys or benefits from breaking GTK+3 applications; they broke GTK to create a better GTK.

For instance, GCC extensions to the C language have been copied by Clang, and some have even made it to the standard. There's people implementing basic systemd's logind functionality in OpenBSD and FreeBSD so that GNOME 3 can run.

How do you think de facto standards are born? Are you against any form of development and original features? You obscurantist whiny bitch ain't entitled to telling other free software programmers what features to bake in their programs. Interoperability and standards-compliance is nice and all, but it's just one aspect of quality software, and it's often at odds with other lofty goals. Such is life

t. gnome internet defense force

you have autism m8

The idea was to make extensions untrusted and incapable of introducing security holes. Making them compiled binaries executed in-process completely defeats that purpose.

Imo, GNU should just stick with making libraries, utility programs like GIMP and whatnot, and making sure "GNU compliant version of bios/whatever" exist.
Their attempts at making a DE and toolkit have created a closed-ecosystem, uncostumisable, systemd dependant, forked mess, and their attempts at making a kernel have resulted in vaporware, or a gutted linux kernel.

They broke themes all the time because API changes, and then when they decided to stagger those unstable API changes, Holla Forums bitched because "omg semantic versioning why are there going to be so many versions of GTK."

Not to mention the thread discussing this change included my all time favorite autist: the faggot bitching about dynamic linking and how GNOME was obviously incompetent for trying to stop the breakages that people were experiencing.

They broke because they weren't using semantic versioning in the first place. Yes, versioning was a problem for them, but their new release schedule does zip to fix it.

wat

I am using ROSA with Plasma 5 and the cashew is no longer a thing, or I removed it simply, I honestly can't remember which. At any rate, no cashew problems here.

This is what I did:
vizzzion.org/blog/2015/02/killing-the-cashew-done-right/

Except you're wrong.

herein I declare my victory over this argument

what is "closed" supposed to mean?

GNOME Shell extensions and CSS in GTK are the customization gods.

factually incorrect. *BSD and Guix run GNOME 3 without systemd; perhaps Gentoo with OpenRC does too.

Call me a shill, but Gnome 3 alone is the most popular DE.

Old KDE was forked too:
en.wikipedia.org/wiki/K_Desktop_Environment_3#Trinity_Desktop_Environment

Hurd's been usable for over a decade now, it's been waiting to be adopted by a large crowd of kernel developers who would implement drivers, a plethora of IPC mechanisms, more networking stacks and whatnot. But of course it's never gonna happen. Most people have no compelling reason to prefer Hurd over Linux.

linux-libre is best linux. Binary firmware is for peasants

semantic versioning is largely irrelevant as far as API stability goes. suppose they had increased the major version number every 6 months with every API breakage, in accordance to semantic versioning. Unless distros had been willing to maintain 10 different major GTK versions with ZERO upstream support, applications would have broken anyway. semantic versioning tells you how to name your versions so that people know when you break the API; it doesn't tell you when to break the API