GNOME - State of Affairs

Redhat DE is has always been arrogant but following Ubuntu's switch back to GNOME, it has begun asserting itself as the one-and-only GNU/Linux desktop.

Other urls found in this thread:

boards.4chan.org/g/thread/64527834
youtube.com/watch?v=Bwiln7v0fdc
kolibrios.org/en/
twitter.com/AnonBabble

redhat being redhate
^ that typo was unintentional but it works

not meant to sage

This is disinfo fake news. The IgnorantGuru article is over 6 years old and LvS is correct. There's a lot of time wasted on things that don't matter (like "discussion"). I think there's a clear divide between ideas people who've never maintained a codebase themselves and those who have and know what "discussion" brings: a lot of bikeshedding, wasted time, useless GitHub notifications, and drama.

Sage goes in all fields, thank you.

Heh heh.

This same post is a general on cuckchan.

Cuckchan doesn't have Gnome generals nor does it have double images.
Don'yt you have anything of substance to post?

Are you saying you saw this on cuckchan? Why were you there?

boards.4chan.org/g/thread/64527834
If people want to discuss this, there's nothing wrong, but OP is a faggot who didn't disclose his cuckchan membership. Saged and hidden.

You're the one linking there faggot.

Jokes on them, I'm migrating to tui only software.

protip: people who make real investments set the direction.

If you want to set your direction, you need to make a real investment.

gnome sucks

mate is where it's at

...

It's hard to get upset at something I haven't used in almost a decade.

This is why I'm never going to use wayland.

Wayland =/= GNOME.

But this is not posted by a GNOME employee?

You sure seem to know alot about cuckchan

If you go back and read OP's caps you'll see that that's not as true as it should be.

But what is the answer to Linux GUI ecosystem? Qt + KDE?
This makes me sad.

You will never find anything that satisfies you if your basis of judgement is a strawman.

don't misrepresent the argument bud

right back at you pal

nice straw man

y-you too

That list of Gnome technologies sounds like a terminal illness syndrome which is sapping Linux of all non-CIA niggerdom:

All different diseases to be cured. Wayland is another new cancer which CIA niggers will be pushing. It's fucked.

GNU/Linux is now GNU/systemd/dbus/wayland/etc. Thank fuck for a few sane distros like Slackware and Devuan.

These faggots sure are full of themselves. There were better desktop environments 30 years ago that ran on 7 MHz computers with 256 KB RAM and a floppy drive. It was so damn smooth, you could be playing a protracker MOD while formatting a floppy disk and still use the GUI without any mouse lag/jumping or graphical glitches. Gnome is dogshit. These people are fucking deluded.

*high fives*

It was also harder to do things back then. When you wrote software, you targeted the machine itself. UI latency was small because the data volumes were tiny. You can still get this kind of software today by following the exact same programming practices of targeting the machine itself and outputting data on low resolution displays. I suspect that you won't be developing this kind of software.

Bullshit, most platforms came with development tools, or they were easily available, and assembly programming was easy enough that it made sense to optimize even non-critical functions just for fun.

You sound like you've never programmed a computer in your life.

youtube.com/watch?v=Bwiln7v0fdc

I'm not the user who you replied to, but I definitely would code such software if I wasn't a retard:
I have a dream, of an open-source desktop OS which is almost realtime with no GUI latency, and with a maintenance phase of 50 years. That means 50 years of bugfixes, security updates, and hardware compatibility updates. The way you run it today, your grandchildren will be running it 50 years from now, even if on radically different hardware. It would look a lot like KDE3 but it would have functionality at all 4 screen corners and certain programs would be "militarized" in design to start fullscreen by default, use high contrast colors and put all the important buttons at the 4 corners for the easiest possible access.
But of course nobody would use it because that's the way things happen.

There's a difference between fulfilling your specific processing logic and then completing a full GUI application. I don't doubt that it was easy to write applications in assembly. I do doubt that it is easy to write a full suite of applications with a GUI front end in assembly while working within the hardware limits of the computers of the time. The hardware limits are different in modern hardware so now it's easier to achieve your demands.

As for me, I believe code reuse and software libraries. I believe composible code snippets. I believe in targeting soft(ware) platforms as opposed to targeting hardware platforms. I believe in letting other people writing platforms instead of me reinventing their work. Software platforms are a solved problem, I don't need to reinvent that kind of work. I focus on my focus which is the unique functionality that makes my application unique.

kolibrios.org/en/

I tried that a while back, was impressive. I even had MenuetOS on a bootable floppy one time. Thanks for reminding me.

I can't speak for Atari ST, but on Amiga you used an ANSI C library to interface with Workbench, much as you do today. The resolution wasn't necessarily low-resolution either. Obviously the very first model released in 1985 was limited, but later on they added more graphical modes in ECS and AGA chipsets, and eventually 32-bit RTG cards with very high resolutions like on MacOS. But it was still the same exact library and Workbench doing all the work, and you got the same exact, flawless performance.
As for development tools, well there were expensive and cheap options. If you were a poorfag you could just download from BBS or mail order the Fred Fish disks with DICE for a few bucks. Same dude who's now the head of DragonFly BSD wrote that compiler btw.

I should also mention that Amiga programs made this way (that made library calls, rather than directly accessing hardware) are very portable and can be compiled today and run natively in AROS.
So this idea that everthing back then was bare metal programming is false. That was largely a demoscene and video games thing. Even a lot of simulation/strategy type games used the libraries so you could multitask while playing them.

Fantastic. Now how much work does it cost to port the one application for that platform onto my x64 machine today. Now how much work does it cost to port the one application for that platform onto my rPi machine today. How about my (theoretical) Java machine. Is the cost "not so big", well that's perfect.

Now let's say I want pretty vector fonts in my applications and I want hiDPI support. Let's update the Workbench platform to include this feature. Let's say I want to run audio in my laptop speakers but then it automatically reroutes the audio to headphones when I connect my headphones. Let's update the Amiga software platform to deal with this specific issue. When you add up all the numerous little quality of life issues that we take for granted in Gnome, KDE and Windows, I'm willing to bet that the software platform you have will start to weigh just as much as GNUstep does today.

Install AROS, most of the work is already done for you. On ARM you might have to run it hosted on top of Linux, but the programs will still run. I doubt you'll get very far in Gnome on an RPI board anyway. Amiga stuff is much leaner so stands a better chance on small hardware. Hell, even Firefox crashes these days if you have only 1 GB memory, so that rules out most ARM SBCs right there. Modern software is bloated junk, I can't see what you think is so great about it. I guess you just have no frame of reference, or have completely forgotten what good software used to be like. I mean before everything got a CoC and needed constant fixes and updates, because they can never get their shit right.

Also, I should mention that you reply was a classic, by-the-book "move the goalposts" tactic. I had already addressed your issues about direct hardware access and incompatibility, but you came right back with a different set of "requirements". I'm sure that if I address those, you'll just come up with some other shit. Don't think I didn't notice this. I'm watching you closely, and so is everyone else.

Btw, what "quality of life"? Are you mad? There's endless bug reports and threads about how X sucks in Gnome, KDE, or other bloated desktop environment. You make it sound like everything is perfect, when in fact it's an endless treadmill of upgrades, arbitrary changes, and bugs.

The subject I was addressing is about so-called "bloated" software platforms. These examples that I expressed are part of the argument that simplistic software requirements make for simple to write software. When people start demanding reasonable things like the ability to port software from computer to computer and little, these things contribute to the code base size. People want software that will do more today than what was capable in the 1980's. You would call this flexibility as bloat.

I'm not talking about perfection, I'm talking about the cost of portability and flexibility. Making software platforms portable means reducing the cost of application development. Making software platforms flexible means reducing the cost of application development. You would call this bloat.

In the old days people had GUI toolkits and APIs too, my young Indian friend. They were just incredibly more efficient.

They were also limited to the bare minimum i.e. more efficient.

In the specific case of modern Linux DEs, even the most "lightweight" ones like Xfce or LXQt, is there really anything they do that's impossible on a GNOME 1 or KDE 1 environment? Those older DEs felt a great deal snappier than already fast contemporary mainstream OSs (Mac System 7 or Windows 98), whereas every modern DE short of meme-tier tiling WMs are slow as mud, even compared to sluggish macOS 10.13 or Windows 10.

Even ignoring things like poor hardware acceleration, or software that's inherently bloated (modern browsers running too much JS, LibreOffice being written in Java), simple operations like using a filebrowser, text editor, or terminal session, feel amazingly laggy on computers hundreds of times faster than late-90s machines.

Don Ramón would easily fall for every troll out there, though.

In the bad old days, devs and users used the same computer. In this brave new world, the dev uses a 16GHz 2000-core CPU with "lightspeed" SSDs and 2PB of L1 cache functioning as RAM. The user has a strip of toilet paper with 'computer' components badly rendered with God's pencil.
This has two effects, both leading to sluggish UI: the software is shit because not enough thought goes into making it otherwise; the software does too many things that it can afford to on the dev computer but which disrupts core functionality for the user.
That second effect combines in space with "lots of tunable dials are bad" and "unimpressive defaults are bad" to form SLOWTRON.

You can't ignore that though. That is the answer. Your file browser used to browse files. These days it might have an extra feature like constantly scanning your image assets on disk for CP. The extra shit has costs which aren't always obvious on a fast disk with not many files.
I once accidently deleted an important project due to interface lag that I only realized later was due to hundreds (because the next would be fired under the unchecked assumption that the previous run completed) of filesystem-trawling shit that had no other purpose but to find random images in the extensive documentation on-device that I added, to populate a "Your Photos" UI that I never looked at.

with 20x persistent RAM disks in RAID 0*

I don't know have a single friend who actually uses GNOME. Everyone who starts with Unity of GNOME eventually switches to something more functional or lightweight like Xfce, KDE, LXDE, etc.

Yes there is. LXDE/Qt has configurable pipe menus, a powerful idea that Enlightenment is practically the only other window manager/DE to expound upon. Xfce for instance can embed an XDG applications menu within its standard desktop right-click menu. KDE forces you to choose between one or the other.

Sometimes I wish I had friends, even if they used GNOME.

In the first screenshot, he seems like an outsider bitterly recounting his shit experiences with the fishface niggerfaggots at GNOME.
In the second one he appears to be speaking from the POV of some pretentious shitcock inside of GNOME.
This leads me to believe he's shitting on GNOME and you're all fucktards.
Am I too autistic or not autistic enough or lacking context or what?

just use MATE lmao

Just use Cinnamon, the only good DE.

That is going to hit a wall sooner or later with the shit Gnome is doing. There is a reason Budgie defected to Qt and plans to replace the other Gnome components.

I was not talking about desktop environments but GUI toolkits. Cinnamon is still based on GTK+ and it is basically an early GNOME fork that now became its own beast.
I am more interested in growing the whole platform not just end product. Nothing wrong with having 10 different desktop environemnts as long as those are based on the same and good foundation.

Sorry, fix: early GNOME 3.0 fork

Or you could just use Xfce, which has more features better customizability with similar or smaller footprint than either of those.

Some additional info on reasons why Budgie is switching to Qt:
- Ikey (the author of Solus and Budgie) is primarily a C programmer, but Budgie is written in Vala to reduce boilerplate code
- the problem with Vala is that it transpiles (right word?) to C and then to binary executable and its final C code is not that great. This sometimes brings some weird behavior bugs that are annoying to solve.
- if you want to develop with GTK+ your best bets are: C, C++, Vala, Python and JavaScript. Budgie aims to be somewhat performant and not use too many resources, so Python and JavaScript are no-go. C needs shit load of boilerplate code that gets really annoying to maintain and reduce chance of other people contribution to your project, since they are not gonna deal with your spaghetti code. I already explained the disadvantages of Vala, so in the end only C++ remains.
- Ikey (Budgie author) doesn't like C++, but it is the necessary "evil" because it is sadly the best tool.
- While Budgie wants to remain independent (not based on anything else) it is still a fully featured modern desktop environment so it needs to leverage the time-tested technologies to reduce complexity and NIH.
- I think Budgie uses libmutter (Mutter is GNOME WM), which is inflexible and requires you to do things in GNOME way (basically pretend to be gnome-shell). This is probably true for any other desktop environment based on GTK+ stack!!!

So end result is C++ and something other than libmutter. So basically Qt falls into this category. Ikey did some basic experiments and decided to use Qt Widget instead of Qt Quick (QML). He will also probably leverage some KDE Frameworks (but not all). As for the WM it is still unknown, since the real development of Budgie 11 still hasn't begun. Solus team is still improving the distro stack (configurations, improving boot time, packing tooling.....).

We have essentially 2 different issues here;

-GNOME devs don't care about popular DE forks, to the extent they would want to fuck up whole DE's because it might make their repo look minimaly cleaner

-GNOME devs, but especially Red Hat, do not care about making Wayland universally relevant since they only care about their implementation, essentially fucking with KDE which has far less resources

Just because some individuals want to force Linux as a mainstream OS doesn't mean one should sacrifice freedom of choice over "universalizing" an Ecosystem. And its not only GNOME; just look how Systemd and pulseaudio is shoved down everyones throat although both have been shown to have tons of horrible issues. Programs should be writtten with extendibility in mind, neither of these Software deployments should have a walled garden in mind. And if you know that software is fucked either way from ground up; stop, think and do it properly again instead of hotfixing millions of lines of code.

The only people sacrificing freedom are the people who refuse to invest into writing code. You don't have to use Systemd, Pulseaudio, Wayland or anything else for that matter! The reason why you feel that you're forced to use these is because you refuse to invest your time or money into developing the software you want to see. Don't whine about other people doing some work you don't like when you yourself refuses to do the work you're responsible to do.

Fuck you. GNOME is refusing to play nice, and that's a danger to free software itself.
They gain nothing by designing their software to so contrived that other projects can't use the sensible parts of it.

I have no idea what you're talking about with anything you've written there.

Hello lennart.

I'd argue that GNOME is actually harming Free Software by levering it's connections and market-share to siphon resources from other more promising projects.

GNOME is a controlled ally.

I don't believe this is true but assuming it was, what's stopping you from injecting your resources into these "more promising" projects?


My name is Fred, not Lennart.

Did you even read the post you're replying to, asscheese? GNOME is embrace, extend, extinguish.

Redhat is very close to owning copyright on all the vital GNU/Linux userspace. Maybe they wish to do to XDG/Linux what Apple did to OPENSTEP..?

I don't understand how Gnome is embrace, extend, extinguish. Redhat owning the copyright to all the vital GNU/Linux userspace means nothing to me. The reason is free software means users will always have the permission to modify it whenever they wish.

It's a subpar product that leverages auxiliary means to remain relevant and earn resources.

Redhat is a singular entity and having the modern GNU/Linux userspace belonging to a singular entity is concerning.

There's nothing stopping Redhat from waiting until GTK, Flatpak, pulseaudio, etc become more-and-more 'required' and then re-licensing as proprietary or otherwise less free.

Gnome leverages auxillary means to remain relevant? Gnome leverages auxillary means to earn resources? I'm sorry but I still have no idea what this means.

For you maybe. I don't care at all.

The GNU GPL means that on the event that this happens, it isn't going to harm the users of this software because the users have all the permission in the world to fork the GPL software at any time.

This. Just keep the latest GPLed versions of these softwares around and keep developing openly off of that.

Nobody wants to keep any version of these software at any time since they're broken beyond anyone's wildest definition of broken. Software that fucked up simply didn't exist even 10 years go, it's downright insane. And that's only on the technical level. The fact they're almost as useless as plain bricks on account of having fuckall features, and even less each release, not to mention going incompatible with any and all plugins (an officially supported extension method) with even a bugfix release with no warning, is ridiculous to say the least.

The problem is that software that worked independently of that ecosystem for years or decades (like udev or upower) suddenly overnight got eaten up by the ecosystem and is now completely unmaintained outside of it. Packages that have 0 functional dependencies on the ecosystem (the software literally compiles and works identically by changing exactly one line of code) still declares dependencies on it.
And the devs are proud to say not only that they actively don't care about anyone but themselves, but also that writing specs or stable apis is "not something they do", so it's literally impossible to write a compatible system in the first place (not that anyone would want to do that) for the programs that should be trivial to replace from that ecosystem.

Hi lennart poettering how are you doing today. why haven't you chocked to death on your own dick yet?

So you're saying that because you don't like how the code works so therefore nobody is going to use it? If that's the case, why do you care about Gnome when you should be writing other software that does other things? There's no need to make it compatible with Gnome or systemd or udev or whatever. Go do you own thing with your own software!

I'd go one step further: Linux's intentionally unstable ABI was a mistake, the root of this evil.

Why? Unstable ABI means Linux can continue to evolve without the need for needless backwards compatiblity. If you need backwards compatibility, you should stick to the stable API that will hardly change even after being deprecated.

Because it encourages needless, pointless change of the sort that subtly introduces bugs.

Good software isn't software, but a specification, and production code should only be written to implement that specification.

Yeah, that's why there's Free Software MacOS even though Darwin and Cocoa are both FreeSoftware, right retard?
Android is never locked down either right?

So why can't people write according to the stable API? Why must they write according the ABI?

They want to make proprietary software.

If parts are missing, then it's our responsibility to fill in the missing parts.

They can continue to struggle with the unstable ABI for all I care. The right way is to target the API which will be stable.

Red Hat is cancer and people like you are the proof.

Lennart has enough time to shill his warez on the chans, Reddit, IRC, mailing lists, and still crank out thousands of lines of systemd code every workday. He is industrious.

there is none, Linux is designed for a command line shell, GUIs are a crutch
use TUIs and curses in the meantime with a light window manager like fvwm

...

Wow. Completely rekt.

Never understood the hype for Mint. Ubuntu already fills the "linux for retards" niche, only it does it competently and actually works out of the box.

As I understand Mint, they based the distro on having a windows-like start menu.

yep. After Ubuntu switched to unity, Mint was fairly successfully sold as the most Windows-like Linux. Cinnamon and their way of organising Mate played to all that.

I would never suggest mainline Ubuntu to a Windows users trying out Linux. Interface is too alien.