A point on Linus' videos though: his criticism is objectively not fair to the community.
I have been using linux for everything (work/gaming/devops etc) for over 7 years now. To this date I still stumble uppon simple issues that a normal user would probably never figute out how to solve. Linus is right in this regard, there is waaay to go, linux distros are not yet plug and play for normies.
However, and this is my problem with Linus' criticism: He has taken an existing subset of the people in the linux community (toxic gatekeepets who will also claim limux is noobie friendly) and he is replying to the whole community in frustration, as if we all are these people. So the thousands of volunteered effort, delivered for free to all of us, gets a slap from Linus because some specific people set some arbitrarily high expectations. He actually disclaims this fact, but I don't buy the disclaimer.
It is not polite/fair to say "this sucks, the other sucks, X thing is horrible" when people have volunteered so much free work to get it to this state.
For that reason, I am annoyed at his take on the challenge. I see luke's stance as waay more appropriate.
I am grateful to everyone who has contributed to this wonderful ecosystem. Yes, there is way to go.
I recommend checking out Linus's WAN-Show videos, they discuss it there with more context and nuance, something which the shorter cut videos can't do that well:
Oh so he's only pretending to be one of the people who know just enough to be dangerous while expecting everything new to have zero learning curve and live up to all the hype?
I don't even know if that raises or lowers my opinion of their content.
I don't think I get your point exactly. The point of this challenge was to show how an average Windows gamer would fare if they tried to use Linux for gaming(without having prior knowledge).
Yeah, it's not like I don't get their angle. It's valid, drove a lot of engagement with their content, and I can respect the commitment to the bit. But at the same time I just see a single point belabored over and over again instead of really digging into anything really interesting.
I can't predict how their series will continue, but I think they've missed a massive opportunity to use the same premise and display just a little bit more self-awareness about how people experience different levels of software skill (or graduated steps of competence in any endeavor, really).
I thought for a few seconds that such a change could only help the content, but after snapping out of it I realized: no, that would only give it appeal to me. Of course it would hurt the content overall. This content is bait for people who want to laugh/facepalm at linux masochists and people who want to laugh/faceplam at a noob struggling.
> just a little bit more self-awareness about how people experience different levels of software skill (or graduated steps of competence in any endeavor, really).
They show more of this in the WAN show discussions, but I think you are right that a lot of people would find such a conversation boring. So, I can't really blame them for seperating this into the show format.
Good resource. I have actually been watching those as they came out. It still gives me a bitter taste on the challenge videos. Most probably the reason for it is that linus was being frustrated at the time of filming the episodes but was calmer and more thoughtful while doing the wan shows.
Good point. I believe he mentioned something about doing an additional video where he wants to put the whole challenge into context, by comparing it to Windows from the point of view of a novice user.
I was very shocked after watching the first and second Linux challenge videos. In the third video Linus uploaded yesterday[0], it seemed like he was being a lot more diplomatic, and distinguished between a small number of toxic gatekeepers vs. the wider community and their positive contributions.
> Linus is right in this regard, there is waaay to go, linux distros are not yet plug and play for normies.
Really depends on the hardware and the use case. When using completely standard hardware for Web and a bit Office the installation is very easy. Seen it myself how someone who never used Linux and is just average with Computers installed Ubuntu and it just worked. But yes, something as simple as a Webcam can make things very difficult.
Same sentiments but I have my reservations. I wish we just have enough Linux users to drive open standards. A lot of proprietary standards became proprietary because companies just doesn't care about a few Linux users.
The premise is that the Linux desktop developers were just just follow these concepts or to follow that strategy, Linux would all of a sudden achieve desktop supremacy.
There's no such thing as an 'intuitive interface'. It's been claimed that "The only intuitive interface is the nipple", but as many a parent knows, not even the nipple is intuitive enough by itself and every baby needs guidance.
Every desktop has idiosyncrasies. Windows' UI sucks in many, many use cases. This hasn't stopped it from continuing to be dominant in the Desktop area. The reason is ... guidance.
Everyone can wreck their computer by doing stupid shit and ignoring all the warnings thrown in their face. They will not read it and complain if something breaks.
With Windows (and to some degree OS X) this is not so much of a problem. You can just call that one friend that "knows computers" and after some begging and crying, you'll get help. Or you can bring your bricked computer to the next support store and they'll try to fix it.
(Almost) every elderly person above a certain age is lost when it comes to computers. Most people alive cannot be trusted to install an application - not on Windows, not on Linux and not even on OS X (well, maybe on OS X nowadays, because you cannot install unsigned software these days) because all that software might be compromised. Every time you let a software on to your system, you can potentially brick it. And then you need a solid understanding of what's going on.
For most people, computers are black boxes whose internal state is a mystery that is far too complex to be explored. They have no interest in exploring these mysteries. They have work to do; forms to fill, letters to write. And these people need the aforementioned informal support they can turn to whenever they need to diverge from the well-trodden paths. It doesn't depend on the operating system, but on the surrounding support system.
The Linux as a Desktop operating system is simply lacking the market share to boast such an informal support system.
> There's no such thing as an 'intuitive interface'
I can't remember where I read it (maybe Design of Everyday Things?), but something that really stuck with me was "intuitive is just another way of calling something familiar".
Exactly, that is why Gnome is not as intuitive as it could be, it does not work at all like Windows or Mac OS. Sway works very differently to Windows and MacOS, but at least brings potential productivity advantages in doing so.
In an established Windows-based desktop network where each office worker has their own PC & apps optimized for that user's particular workflow, this can be one of the most productive arrangements that can be leveraged to make the difference between a low-productivity bureaucracy and high-productivity, whether it remains paperwork-intensive or works completely paperless.
The early adopters of DOS who followed directly to Windows in their offices were most likely to have this bit well established (though it was far from universal), and enjoyed its advantages quite a number of years in advance of the internet appearing on the horizon. Regardless some offices still had one dedicated DOS PC for email purposes and directly logging in to clients' or vendors' private servers by dial-up modem, just like it was before there was any consideration to network certain desktops together much less join them all to a future internet if any.
With the arrival of built-in Windows Networking in Windows 98 and the dominance of Ethernet for both office LAN and internet WAN, the clear advantage of having the dedicated office workers and their apps remain fully productive without relying on dial-up or later broadband cannections to outside resources became more & more blurred.
Linux was still in its infancy. It has been made to accomodate Windows Networking ever since.
Remember email came out in DOS way before the WWW came around, and in that respect webmail is the killer app.
It is still considered too challenging for an IT administrator to switch this whole shebang over from Windows to Linux, it's like stepping off into the unknown to some extent or another, and there's too much at stake at one time. But one approach is to backtrack into known successful configurations and see if you can fork differently from there.
One of the most viable approaches is to insure the local Windows network has maximum functionality and productivity without requring connection to the web. Somewhat like the old days.
Then treat the web as a special case just like it used to be when you had to wisely set aside a whole dial-up session for a potentially unreliable connection.
Only now its lots quicker to just reboot to Linux for the occasional web use, then back to Windows for regular office work.
Very few offices actually benefit from having all the office machines be online to the WWW all the time.
People find Linux much easier if the only thing they have to use it for is email and browsing the internet and other web activities only. Not much training or support burden needed for such limited utilization.
They are already using things other than Windows to be online all the time on their own devices now, so it's much more realistic to get them to do the web function exclusively in Linux on their office machines before even trying to migrate the vast remainder of integrated Windows workflows into a hopefully-compatible Linux equivalent.
This is just a way of functionally recognizing that Windows has continued to become less suitable for the internet for some time now, and Linux is becoming more suitable for high-productivity office work but was not suitable enough fast enough.
They did not meet in the middle where one can completely be substituted for the other, but can be handled separately to some advantage.
Everyone would be running full desktop Linux each time they went on the web which is easy to get accustomed to, they could further become familiar with more extensive Linux capabilities at their own pace without fearing a show-stopper which could break the carefully-honed Windows workflow.
Those few workers who do need to be on the internet almost all the time would be the ones who need to be able to use Linux for almost everything.
Even if Linux was perfected, everyone else can only adopt it at a limited pace so there needs to be clear priorities but this is one possible step toward the light at the end of the tunnel.
> With the arrival of built-in Windows Networking in Windows 98
This is not correct.
Windows 95 had built-in networking from launch and an email client on the desktop. (It did not have a bundled web browser at launch, but it had networking including TCP/IP and dial-up.)
Built-in networking arrived with Windows for Workgroups 3.1 in 1992, and became mainstream with Windows for Workgroups 3.11 in 1993, because of the performance enhancements of 32-bit File Access. From '93 on almost all PCs shipped with WfWg 3.11 as the sole default version.
Also, Windows NT launched in 1993, with built-in networking including TCP/IP over wired and dial-up networks.
But networking does not and did not equal TCP/IP. DOS and Windows 3.x defaulted to NetBIOS, with optional IPX/SPX for Novell Netware, which was the dominant PC networking standard from the late 1980s. Until the mid-1990s, TCP/IP was a niche protocol only needed if you wanted to communicate with expensive RISC-based UNIX™ workstations.
Microsoft used NetBEUI. Novell used IPX. Apple used AppleTalk. DEC used DECnet. IBM used lots of protocols including DLC but didn't use Ethernet all that much -- it had its own network system, Token Ring -- so you needed special hardware to talk to IBM kit and only IBM-centric businesses used it much.
TL;DR -- this is wrong; MS OSes had built-in networking as standard from 1993, a full 5 years earlier.
Thanks for filling in lots more of the accurate historical details.
Especially the way Windows Networking in general was all over the place but not primarily TCP/IP over Ethernet until Windows 98 when anyone could finally just add an Ethernet LAN adapter having RJ-45 connections and connect to most anyone else from then on not much differently than can be accomplished still today.
Which is why late '90's is when so many offices first started any networking to begin with, it could not very well be considered mainstream until after that.
LANs in office networks were certainly entirely mainstream in the first half of the 1990s. When I started my first job in London in 1991, we had just one client who _didn't_ have an office network. That was considered unusual but it was an intentional management decision, intended to slow the possible spread of malware and increase real-life face-to-face staff communication.
What wasn't mainstream was them being based on TCP/IP.
These days, networking and TCP/IP seem synonymous, but that's just how it happens to be this century. Networking, mostly over Ethernet, initially Thin Ethernet (10base-2), in wide use as a common office tool predated the rise of TCP/IP by a good 15 years or so. Some early adopters were using it 20+ years earlier.
Network protocols then were a bit like OSes are now. Many people use Windows but lots use Macs, some use *BSD, a few still use commercial UNIX, etc., and there are things like ChromeOS, thin clients over RDP, and stuff. It's not at all homogenous and it's hard to even say there's a clear majority for any one OS: Windows has the edge, but not by a lot any more.
Well, in the era of MS-DOS, Novell was the server OS of choice for almost everyone, with rivalry from 3Com and its 3+Share MS-DOS-based server OS; 3+Share was related to MS LAN Manager, which was in OS/2 and led to 3+Open. They all used NetBEUI. LAN Manager also ran on VMS thanks to DEC Pathworks, running over DECnet, which was handy because it also supported terminal sessions -- remember, this is before SSH -- and X.11 and DEC email and more.
Focussing on big businesses was Banyan VINES, with its own protocol derived from Xerox's Alto and so on. This had the first network directory. Novell designed Netware 4, with NDS, as a direct response to Banyan's StreetTalk, and in the NT 3.x era it kicked Microsoft's behind in the market; the tide only turned with NT 4.
Speaking of big enterprises, email was common long before LANs or TCP/IP. All big DEC users and IBM users had those companies' email systems. Small firms used dial-up to pre-Internet service providers -- I used CIX, which dominated in Britain. My 1991 CIX email address is still live and still works. Americans favoured CompuServe, AKA Compu$erve, but it was too expensive in Europe where we pay for local calls too.
This stuff is, ballpark, a quarter of a century older than TCP/IP and Internet-based networking. And given that that is only about 25 years old, what I am saying is that widespread LAN use didn't begin with Win98.
At the time, Win98 wasn't even a blip; it was nicknamed "GameOS" in enterprise IT circles and few companies even considered it. NT 4 was where it was at, and it launched with full TCP/IP support two whole years before Win98.
So no, Win98's networking didn't begin anything at all. It wasn't significant in any way, then or now. Win98 was a home OS for standalone PCs with dial-up, but it merely took over from Win95 which created that market.
The rise of TCP networking in business LANs arguably began with Windows NT, but NT 3.x wasn't very significant, and NT itself arrived about half way through the lifetime of business use of machine-to-machine communications, email, groupware, etc. from its beginning to now.
If you want to argue that integrated networking in Windows was a significant turn, that I won't argue with. But it began 5 years before Win98, with Windows for Workgroups and Windows NT.
The fact that _now_ it looks big and significant that it's when TCP/IP became the default is an emergent artifact of the current focus on IP. It wasn't at the time.
Email is a 1960s thing. The Internet started to become significant in the 1970s, long after email. Corporate LANs rose in prominence in the 1980s and by the 1990s were almost a given. Macintosh-based companies (mostly in design, print, repro etc.) did direct peer-to-peer comms over ISDN.
In the 1990s, for most people, TCP/IP only ran over dial-up modem connections, and it was contemporaneous with the industry moving to 10base-T: Ethernet over UTP replacing Ethernet over Coax.
For a time, the obvious successor to 10base-T looked to be ATM, which is a protocol at well as a cabling system; TCP/IP had to be tunnelled over ATM, but it looked clear for a while that ATM was the future. 100base-T (Fast Ethernet) was just one contender among several.
But actually, as it happened, TCP/IP rose vastly in importance, and networking switched to 100base-T and then wifi.
LANs switched to IP in the 21st century but they were a roughly 20-year-old, established, totally normal technology then.
I totally understand your point and I think it's one of the reasons why GNOME's goal to become the "perfectly intuitive desktop" is utterly misguided. However, one thing which separates the interface of Windows from MacOS and Linux is continuous improvement. I'd rather have my desktop adapting to current times rather than stuck in old times and cite "compatibility" as an excuse. Linux desktops and MacOS has an incentive to experiment with their UI and actually improve. If some feature doesn't really click with its users after deployment, developers could just scrap it on the next release.
TL;DR Windows is also trying to have an ideal desktop which isn't driven by innovation but by how it was intended to be used in the past.
Do you really want it to succeed? Popularity among average users comes with greedy businessmen who tend to spoil everything they touch. Do you want apt-get to display ads or send em.. telemetry to god knows who? Do you want a "clouds powered" bash terminal that's an electron app rendering a remote terminal running on AWS? How about ffmpeg with a mandatory DMCA module for "safety reasons" or MS arbitrarily adding and deleting packages on your Ubuntu and rebooting it because you happened to be in some experiment group? Because that's where the money are.
>The most frustrating moments for a user is when the software they’re using does something inexplicable
Lol, speaking of... I tried installing sway a few weeks ago on ubuntu because I thought it sounded cool and because I still get somewhat nostalgic for ion.
It logged in to a blank screen with zero indication about what to do next. I guess it expected me to already have all of its keybindings memorized? Or maybe I was supposed to write a config file first? Was it a bug? Maybe I was supposed to already be an i3 user who knew what to do next?
well, you lost interest because it wasn’t what you expected, which is good, because it is not what you expected. I’m not sure who swaps their desktop environment with a windowing manager without reading the manual, but it’s certainly not professionals, and as stated this is a windowing manager for professionals. professionals measure twice and cut once.
I could see an argument that Sway could do better to advertise that it is not for casual users.
Uh, it's software to rearrange where my text editor and pictures of internet kittens sit on my desktop. I think in your elitist fever you might have confused it with software controlling a nuclear power plant ;)
If sway chooses to be a toy for basement nerds that is absolutely fine.
I saw it. I'm also a professional expert level linux user. I'm well aware that I could throw myself into researching how it works and figure it all out I just figured that the starter implied that the main course wasnt worth holding out for.
I don't object to this level of user hostility since it is, after all, free software and he has no obligation to me. If he writes it just for himself thats absolutely fine. I'm not aggrieved.
I just found it a tad ironic that he is lecturing us on UX for grandmas given the level of user hostility his own software exhibited to experts.
(That is, assuming it's not a way out of date version or a debian bug or something that caused this - im trying to be generous here)
I was eager to try Sway out, but if they don't provide some kind of "telescope" where I can type keywords to find the keystrokes of a command that I want to do, then I'm out.
I wouldnt object. Clever intro tutorials are what got me into vim.
Maybe as a minimum it could have a display showing some sort of quickstart guide even if it was just 50 lines of text showing how to get a terminal and a browser and a link to more docs.
I pretty much only write software for expert consumption and I consider my software deficient without a quickstart. Onboarding matters.
Most Linux users don't care if Linux succeeds in the mainstream. You're fighting an uphill battle if you need to get the general community on-board, and they literally just don't care.
Users who don't care, don't care because they went through the trouble of becoming Linux experts, and they had fun doing it. They don't care about anyone else's experience. In fact, maybe changing Linux to be more beginner-friendly will cause it to be less fun for this type of user.
The ones that DO care, overwhelmingly just want to see Linux "win" (whatever that means) without compromising a single thing about how Linux currently is (see the blowback against systemd, snap, flatpak, nvidia drivers, Ubuntu in general, System76 making a new DE in Rust, etc.)
The remainder of users who do care, just want to help people regain their privacy by leaving the big tech ecosystems. These users don't really care if Linux is hard to use, they will go to great lengths to maintain privacy. And they expect newly converted Linux users to do the same.
So if the userbase is hostile to what you want to do, what do you do? You can change the minds of the userbase (hard to do when everyone is so dogmatic), or you can expand the userbase and bring in a new segment of users who care about different things. Usability over privacy and "winning".
System76 is quietly doing this already. And that mirror's Drew's conclusion in this post. To make something with mass appeal, you'll have to charge money for it.
I hope it never succeeds in the mainstream. Reason: Look at windows and mac os and the enormous user exploitation and disregard for privacy and choice.
How anyone can enjoy tech using those platforms is beyond me. I guess we have very different ideas of what my computer and my privacy means.
Just because it may not be on your desktop doesn't mean anything.
Linux is on my desktop and that's all that matters to me. I drive a Mercedes, the fact you might drive a Ford is not going to impact my life in the slightest.
And if all of the invisible Linux machines in your personal life suddenly disappeared you would be helpless.
"I reckon that we could use a commercial, general-purpose end-user Linux distro."
There are any number of those already available. How many more do you need?
> It is silly to suggest that Formula 1 vehicle designs ought to accommodate non-expert drivers, or that professional racecar drivers should be driving mini-vans on the circuit. However, it is equally silly to design a professional racing vehicle and market it to soccer moms.
Another good example would be a commercial truck or construction vehicle. You wouldn't expect a regular person to be able to climb up into a big rig or an excavator and be able to drive away in it.
This is somewhat an odd read. As examples, I don't recall ever having my desktop crash. It's way more stable and performs better than Windows.
Then his point about the install of ODS doesn't make sense. Yes, the text of the button says build instructions. But the page it's a link to has easy install instructions for every distribution under the sun. You have to scroll past them to get to the instrucitons to actually build the package.
"Linux is one of the best operating systems for professional programmers and sysadmins, to such an extraordinary degree that most programmers I know treat Windows programmers and sysadmins as the object of well-deserved ridicule."
I make video games on Windows for Windows. I don't see why I'm well-deserving of ridicule for that.
It's not clear if it's the author's opinion as well as the opinion of programmers he knows.
Either way to me this sentiment is part of the reason Linux is seen as unwelcoming and this article makes me less interested in desktop linux than ever.
It's always shocking to me when otherwise good programmers, smart people, can repeat such obvious tritenesses. He managed to fit two into this sentence:
1) "Linux is one of the best operating systems for professional programmers and sysadmins"
2) "Windows programmers: the object of well-deserved ridicule"
I'm sure DD can't back up either of these claims with logic. Probably a case of playing to one's audience a bit too much.
Who cares? All of my friends use Linux as their desktop. They have been for over a decade. Additionally, I made others Linux users just by installing it on their laptop instead of installing Windows. It made no difference to them. One of them told me she wants to have Opera and whatever else on it. I installed them. Everything works, and she is now a Linux user.
In any case, Linux is already doing pretty good. All my packages get updated, my favorite programs along with libraries I use are maintained, and I hear a lot about Linux-related stuff. It is doing pretty damn good!
With great power comes great responsibility. Users who want more powerful tools need to understand that there's a steeper learning curve.
Desktop Linux is already fine for a lot of people, even non-technical users. My father doesn't even know what an operating system is and he can still work his way around Ubuntu well enough to use the web browser and office suite and manage photos from his phone. He also has enough common sense to read error messages and would never answer yes to a prompt saying it could hose the system.
It is by far the dominant desktop/laptop Linux in the world. ChromeBooks run nothing else and have been consistently selling 20 to 30 million units per year for half a decade or so.
Roughly as many ChromeBooks sell annually as the historic total desktop users of all distros put together since desktop distros first appeared.
It is eminently usable by the general public without training or documentation.
Maybe it should be accepted that the dominant form of "desktop Linux" will be Windows Subsystem for Linux (WSL, especially WSL 2), now that WSL accommodates Linux GUI programs. One gets all of the functionality of Windows and much of the functionality of Linux. (I understand that many Linux users on HN do not want all the Windows functionality for reasons of privacy and security.)
If distros would just turn off the Caps Lock key by default, all would be good--especially, on Live images that can't remember settings from one boot to the next. Make it work as Ctrl.
And, make Linux Libertine the default font. There is no reason why people should have to suffer with stupid typefaces.
I'm using mine as Escape but almost everyone I know use it as Caps Lock. To be honest, outside of the people using Vim/Emacs, I haven't seen anyone remapping Caps Lock.
Usually users with remapped key know how to change their keymap to match their needs, users not remapping don't. It is a small inconvenience for a minority of the users that they can easily handle. Changing this to Ctrl/Escape/Whatever else by default would be a huge inconvenience for most users.
And the Caps Lock key is not useless either, my older relatives can't keep Shift pressed while typing an uppercase word or sentence. Even for single uppercase letter, it makes sense for them to sequentially use Caps Lock > Letter > Caps Lock.
There is no legitimate use for Caps Lock. If you think otherwise, what you lack is education, or are a cockroach named Archie. Using it to shift one letter is much worse than silly.
It is what is called, in court, an Attractive Nuisance. Courts often order those to be disabled, so there is ample precedent.
Sayrus's remark demonstrates firmly that a "Linux Desktop" that appeals to windos users would be a disaster.
Good for you. Try to remember that lots of people don't.
Some don't have 2 hands. Some do but one doesn't work. Some do but have difficulty coordinating both. Some do but are using one to hold the device, or using one to support the other hand. Some do but it hurts to use both at the same time, or it hurts to stretch their good hand from shift to some of the letters.
Try to think just slightly outside of your own use case. Try to remember that not everyone is able-bodied.
There are a lot of devices designed and intended for people with disabilities, but which are affordable only because they sell most of their production to non-disabled people, such as the Jarkey. You don't think about it but you benefit from it.
Learn to think about such things, and you will become a better, more pleasant human being, and that means that you will enjoy your life more.
to be honest i don't really think linux need to do anything to accommodate those unable to read any doc, they already get an os free of charge, and honestly i was there when Ubuntu came out and i doubt Linux communities need another wave of individuals who can barely tie their shoes
Desktop Linux succeeding (if we're talking about mainstream consumer use) seems like it would be too late at this point. We need mobile Linux on phones, tablets, and AR/VR devices to succeed.
And this is where it gets confusing to me. 'Linux' to me is for general purpose computing, a creators platform. Mainstream consumer mobile devices are not that, because many/most users don't need it to be. The closest thing that might be a mobile creator-capable device is the iPad Pro or Surface Tablet. Desktop Linux was an elusive but seemingly achievable/well-defined target. A mobile Linux to catch up to an iPad Pro for general users seems much less well defined or achievable. Even starting with Android (AOSP) doesn't seem close to a creators device.
A good device in the meantime could be a (open/degoogled/jailbroken) Android phone that works well in some desktop mode connected to a USB-C display and bluetooth kb/mouse running standard Linux server and development software. This view (and I) may be too old and instead that device only needs to be a client for cloud-based development.
TL;DR the Linux Desktop ship has sailed, aim for the next target.
I used linux on the desktop full-time until 20 years ago, and even back-then it was already going to be "year of the linux desktop". It's simply never going to get there, and I'm fine with that.
A point on Linus' videos though: his criticism is objectively not fair to the community.
I have been using linux for everything (work/gaming/devops etc) for over 7 years now. To this date I still stumble uppon simple issues that a normal user would probably never figute out how to solve. Linus is right in this regard, there is waaay to go, linux distros are not yet plug and play for normies.
However, and this is my problem with Linus' criticism: He has taken an existing subset of the people in the linux community (toxic gatekeepets who will also claim limux is noobie friendly) and he is replying to the whole community in frustration, as if we all are these people. So the thousands of volunteered effort, delivered for free to all of us, gets a slap from Linus because some specific people set some arbitrarily high expectations. He actually disclaims this fact, but I don't buy the disclaimer.
It is not polite/fair to say "this sucks, the other sucks, X thing is horrible" when people have volunteered so much free work to get it to this state.
For that reason, I am annoyed at his take on the challenge. I see luke's stance as waay more appropriate.
I am grateful to everyone who has contributed to this wonderful ecosystem. Yes, there is way to go.