If you try again, be less ambitious. Don't try to get KDE or Gnome going.
Do a minimal install. Add just enough packages to get X working.
Configure a simple window manager like openbox or dwm.
I uninstall the networking managers. I wrote a simple bash menu program. In one terminal I used that to launch wpa_supplicant for the current network. In another window, I run dhcp. I have a workspace just for that.
When I stay in a hotel, I need to set up a new wpa supplicant profile, and that takes a few minutes. It's klunky. But system upgrades pass by without me noticing.
It's like running an old car with carburettors and simple electrics. You learn the machine once, there's hardly anything to go wrong.
Yes, don't try to actually get any work done, install anything you care about, or use it in a way that improves your life at all. Just reorient your life around learning this clunky system and don't expect to get any actual benefit from it until you've spent years learning it.
This is the standard Linux pitch, followed by "why doesn't anyone use it!?"
That's why I use Linux Mint, even if some Linux guys might think less of me.
Switched from Windows 11. My work system is a Mac. I just love how the Linux gets out of the way, everything just works, UI much better and consistent than Windows.
And I'll probably keep receiving security fixes and app improvements in 10 years for my old Thinkpad.
I also like old carburetors and simple electronics, but we're in the vast, vast minority. Most people want a car they don't have to think about or understand, looks pretty, and will take it to a shop when something goes wrong.
It’s also possible for a system to both be a “carburetor type” while also being well-considered and elegant in design and function so it’s easy to tinker on if you want to, but doesn’t need that 99% of the time. I just don’t think that’s ever been a serious goal of any desktop-oriented Linux distro or DE. They all embrace the complexity and add to or multiply it.
From my ~25 years of using Gentoo Linux, I'd say that it's exactly what you describe in your first sentence. The complexity is there if you want it, but only if you want it.
Gentoo is quite a bit more work to get set up than other distros, but
* Once you get it set up, it just keeps on going
* If you're already fairly familiar with Linux system administration, the setup process isn't hard... it just involves a lot of waiting, unless you use the distro-provided precompiled packages
* There's really good documentation that walks you through the installation process and leaves you with a computer with a functioning bootloader and kernel, Internet access, and some pointers towards what you might want to install next
I agree about system design, but I think - especially in the context of your networking example - the complexity lives below the OS and not a case where it's a lack of effort.
Networking protocols, connections, error states, etc are inherently complex and require equally sophisticated handling that linux may not always do. Some software chooses not to handle at all and expose directly to the user - but others like KDE simply lack the millions of dev-hours of polish that Windows and Mac have.
Is it that simple though? There could be so many other reasons. Maybe the particular issue is related to a specific hardware or network configuration that the developers have never run into or can't reproduce.
Linux needs polish for sure, but given a finite amount of time, energy, and money the reasons are likely more complicated than "they don't care"
I’ve tried the minimal tiling WM thing several times. It’s not for me. Too many little QoL holes that require manual setup to fix and the disjointed patchwork of small utilities that don’t work with each other or share config conventions quickly becomes grating.
I spend a disproportionate amount of time tinkering with Linux desktops trying to make them act the way I like anyway, but the amount of tinkering is exponentially higher with a minimal WM setup. I can’t get any work done under them.
I also don’t really care to administer everything via the CLI and don’t find it a good use of my time to write Bash scripts for everything. If I’m going to do that I’d rather just write my own DE, because even if that’s a wildly larger task than writing a few scripts I’m going to get vastly more satisfaction out of it since I can make everything work exactly as I think it should.
To be clear - I agree with your statement that Linux is full of sharp edges that won't get fixed. I got abandoned by a series of desktop systems. Like you I had gone to and from the linux desktop but had found it unpolished. I was with Mac OS X for a while, and then a major upgrade broke my workflows. Finally I gave up on having a nice desktop. Once I had accepted that, I found a kind of contentment. It has been 15 years, I wouldn't go back now.
I don’t think I’ll ever give up on having nice things. I’m very firmly of the opinion that my computer should bend to me and do its absolute best to serve me and my needs. A machine that sporadically requires manual twiddling does not meet that requirement.
I've been in and around hobbyist computer topics for decades, and I'm still not 100% sure if this comment is straight serious or a really well executed humorous trolling. They'd read essentially the same, so IDK...
"Frustrated with how critical details are missing from the GUI, requiring you to learn console commands? Don't be so ambitious, stop using a GUI for these things at all and write your own CLI for them!"
No, I replied because I think cosmic_cheese should be more cynical. Linux is hard, and won't get better. The X-Windows clipboard design is awful, the wayland team made a decision not to improve on it. I replied to the frustrations of trying to create a daily driver.
The fact that people care a lot about Unix clones is significant, though. nine_k could have been more effective in arguing the point, but it seems like a strong point to argue. Do you think you Go is flexible enough to write a Unix clone with performance equivalent to a C-unix? If so, why has it not been done?
Besides the sibling Biscuit, maybe because no one bothered to do it?
As simple as that, not everyone of us is a Linus.
I should also point out that if you are using a Mac, changes are its iBoot Safe C might already been replaced by Embedded Swift, a GC enabled systems language.
Chapter 5 of The Garbage Collection Handbook, or A Unified Theory of Garbage Collection paper for the incoming replies related to RC.
People care about UNIX clones because they are lazy, UNIX has the source code available, and an existing ecosystem that they don't want to replicate, so it always ends up being yet another clone, thus throwing away all the possible innovations.
We see this happening even with Haiku, Genode, Redox OS, or Windows now shipping alongside Linux on top of Hyper-V.
Unless one is an Apple or Google, with the money and will power to push something out the door, using Objective-C, Swift, Java, Kotlin, with plain C and C++ standard libraries, and even then people will bend backwards to put UNIX into those systems, even when the platform owners went to great effort to hide it under the official userspace APIs.
> As simple as that, not everyone of us is a Linus.
I’m pretty sure Linus could not have written Linux today. Too much hardware to support, too many drivers to write, before it’s remotely useful. Well, except perhaps on some specialised servers.
Yes, not everyone is Linus. Otherwise we’d be using GNU Hurd. But he also came at the right time.
> People care about UNIX clones because they are lazy, UNIX has the source code available, and an existing ecosystem that they don't want to replicate, so it always ends up being yet another clone, thus throwing away all the possible innovations.
Which is a real shame, because people thinking outside the UNIX clone box have come up with some incredibly innovative operating systems like Plan 9 and OS/400.
Because of the 30 million lines problem. Writing a serious OS kernel nowadays is flat out impossible: the hardware is just too diverse, it requires too many drivers. The best we can do right now is just add to an existing kernel. Kind of a vicious cycle: the more drivers we accrete to any given OS, the more impossible it is to get feature parity from the ground up.
The only solution to that is for hardware vendor to standardise their interfaces (at the hardware level), and actually give us the fucking manual. And I mean the real manual, the one that tells us voltages and pins and baud rates and the actual wire formats required to talk to it. A proprietary Windows driver is not a manual, even if it comes with an SDK. Heck I’m not even sure the source code of a Linux driver would count.
But they won’t do it. They just won’t. So unless we do something drastic like forbidding hardware vendor to ever ship software (to force them to standardise and document their actual interfaces), writing another OS that can actually compete with the existing ones will remain flat out impossible.
On a system of that era, some scenes look far smoother with the GUS than SB16. With the GUS running, the CPU could focus on graphics because the audio was offloaded to the GUS. This was particularly true of the 3d flight scene.
Something that undermines the operating system angle is that emacs does not implement hardware drivers. All interaction with the computer go via system calls.
Norton Commander contains a text editor. Emacs operates at that level, whilst being reprogrammable.
What you say seems true. But a comment - there are more effective ways to achieve it than a bill of rights. Australia doesn’t have a bill of rights but does have decent due process, as a result of deliberate legislation. The bill of rights leads to the US Supreme Court being highly politicised because it is a nasty undemocratic backdoor to synthetic legislation. Australia does not have a politicised high court.
I'm Canadian, and as a fellow commonwealth citizen I don't think relying on common law precedent is stable. The common law, after all, is just whatever a judge says it is, and judges can be bought or unduly influenced.
As we're seeing in the United States, the Rule of Law itself is being fundamentally eroded. Laws in the USA are worth essentially nothing now, because the Executive Branch is brazenly and openly ignoring the law and Congress is either too inept or too corrupt to do anything about it. That culture of lawlessness is not going to just go away. It's already started and will continue to "trickle down".
What's more, neither the USA nor either of our countries are immune from political appointment of justices, and it's my understanding that Australia has in fact had some supreme court justices who were previously parliamentarians (sorry if these are not the correct terms), so that seems politically-motivated to me. The USA is merely ahead of us in crumbling, but I think we're in trouble too because their fucked-up political culture is so insidious in its spreading.
We've seen similar in Britain, with its much-vaunted "uncodified constitution". These systems, much like the common law itself, only work when everyone's more or less on the same page.
But when you have an entire political party that revels in shattering constitutional/governmental norms and conventions to the detriment of its perceived political enemies, the whole system gets ugly real quick.
> That culture of lawlessness is not going to just go away. It's already started and will continue to "trickle down".
Only for as long as it's allowed to continue. When the people who are acting lawlessly and those enabling it are finally held accountable and an example is made of them things can get back under control. It's just a question of how long it takes for accountability to happen and how painful that process will be for the rest of us. I'm still hoping it'll be as quick and painless as possible, but the longer it goes on the less likely that seems.
I have built large systems on python that use classes, for more than ten years. I came to it from Java, ten years.
As a rule, I avoid implementation inheritance. Occasionally I need to facade a library that assumes implementation inheritance to avoid it spreading into my codebase.
When the codebase hits a certain size, I hand-roll some decorators to create functionality like java interfaces. With that done, and a suite of acceptance tests, I find it scales up well.
This is a rough history from an outsider: the original developers (“DevTeam”) went quiet, and would not take in new talent. There was some new talent in the community, and momentum for code tidy up and new features. One group forked and called theirs nethack 4. There were other forks with a similar spirit, such as unnethack. Eventually, DevTeam decided to reach an accommodation with talent in the fork groups. Release 5 is a DevTeam release with input from what was new blood fifteen years ago.
The story is a bit more convoluted. After the 3.4.3 release in 2003, the DevTeam stopped releasing new versions. They were still responsive when contacted by e-mail, though. But then we also didn't know how the development version looked like.
In 2014 the dev version was leaked to the community and in the following discussions with the DevTeam on how to handle this, the DevTeam got the first shoot of new devs in a while which lead to the release of 3.6.0 in 2015. This version was a polished version of the dev version, also incorporating some of the popular community patches at the time. The 3.6 branch received regular bugfix and security updates (3.6.7 was released in 2023).
Since 3.6.0 there's a mirror repository of NetHack on GitHub. So the development version that was internally numbered 3.7.0 which would become 5.0.0 was always accessible and, contrary to the 3.4.3 era, could be played anytime and was also installed on the public servers to play.
> It's only certain strains of right-leaning governments that
> figure out you can grow the pie so rich and poor alike become
> wealthier.
Credit to a few. Roger Douglas in New Zealand. Contemporary Peter Walsh in Australia, the Hawke finance minister, also got it. Keating somewhat got it, and put his neck on the line for politically-difficult but structurally-easy growth-pie macro reforms as treasurer, but did not follow through for the politically-difficult and structurally-hard reforms, like wholesale sales tax, and then became a fixed-pie prime minister. Walsh was gone by then.
Howard/Costello era in Australia. Reagan 80s. Pinochet - fits your criteria.
Arden is indefensible. She increased the size of government, decreased social cohesion via critical theory, housing promises went nowhere. Worse balance sheet, worse outcomes, across the board.
>Howard/Costello era in Australia. Reagan 80s. Pinochet - fits your criteria.
Ok, no one really needed NBN and Howard didn't destroy the housing market completely to name just two lasting legacies of the Howard era. Lets not leave out GST either.
Well for a start, he outright lied about the introduction of the GST. Not once, but twice. First that he would never introduce one, second that it would replace other sales taxes to simplify the system.
Well, neither of those were true, and gst we got was used to cut taxes to the wealthy and as a bargaining chip to reduce the power of the states. It is inherently regressive, the implementation increases the tax burden on businesses, and it did't even raise enough revenue to allow them to simplify the tax system.
For all your words, you have dodged the only question of my last post.
By the late 80s, the wholesale sales tax was creaking at the seams. Toys were taxed at 24% but luxury goods at 0%. Also it was complex and expensive to administer. The wholesales sales tax was awful public policy.
Keating knew the GST was good policy, but lacked the conviction to stand up to “jellyback” Hawke (Walsh’s characterisation) and his caucus for it. Keating had taken it to the Tax Summit as his preferred policy “Option C”. Lacking meaningful policies of his own, Keating won the 93 election on a platform of opposing the GST and could not engage in reform as a result.
In the aftermath of the 93 election, Howard said never ever to a GST. Then, during government, cabinet and treasury looked at the indirect taxation mess and concluded that the GST was the optimal policy.
They could have done several things at this point. They could have done nothing, and focused on holding onto power, as Keating had done. They could have dressed it up as a VAT. Or they could have just introduced it with their majority. Instead, Howard gave a speech where he plainly recognised that he had said never, and said he had made a mistake, and his conviction was it was the right policy.
He then called an early election, in full knowledge that he was bad in the polls, and made the GST cause the centrepiece of that campaign.
This was the greatest act of political courage and decency of our lifetime. They risked everything on that conviction. Costello then ran a meticulous publicity campaign in which he made not a single mistake to open ground to the rerun of the ALP scare campaign. Against those odds, the Coalition won the election and made the reform, which now has bipartisan support.
But if you think there was a better reform to the indirect tax system available, let’s hear it.
Howard may have talked a lot about decreasing the size of government, cutting red tape, and reducing legislation and the cost of government. But all these increased under his terms.
Most of the early economic gain was due to the opening up of Australia in the nineties along with the floating of the dollar.
Dude was a dog whistling neo con, so I never liked him. But what is really telling is that the shitshow that is the current Australian housing crisis was foretold and discussed at length in the late nineties when he introduced the changes to cgt and ng.
He and everyone else knew what would happen even then with these changes. The liberal party thesis, openly discussed, was to prioritise legislation that would promote individualisation in order to break unions and get people to vote against their interests.
Plenty written about the other two you mention. Maybe you should read some of it.
arden is indefensible, but you like pinochet? your barometer for a good right wing government improving the quality of life is an actual dictator who tortured and murdered thousands of people?
"did exactly what you needed and nothing more" You can still do that. Build a config for openbox or dwm. While the wm still compiles you can ignore the fads.
Do a minimal install. Add just enough packages to get X working.
Configure a simple window manager like openbox or dwm.
I uninstall the networking managers. I wrote a simple bash menu program. In one terminal I used that to launch wpa_supplicant for the current network. In another window, I run dhcp. I have a workspace just for that.
When I stay in a hotel, I need to set up a new wpa supplicant profile, and that takes a few minutes. It's klunky. But system upgrades pass by without me noticing.
It's like running an old car with carburettors and simple electrics. You learn the machine once, there's hardly anything to go wrong.
reply