This attitude is obnoxious. Yes, not everybody is on a metered connection or running a mission-critical system, but some are, and it is hardly unreasonable to accommodate them.
And Microsoft didn't have a choice. Given an option regular users will never update their computers, perhaps partly due to fear of what they don't understand, fear of change, or maybe due to past bad update experiences. I witness this in my mom with technology all the time. Every time there's a popup she mini-panics, and she has trained herself to click close every time she sees something she doesn't understand.
Google started the trend of silently updating Chrome and everyone including Microsoft followed after, except upgrading an OS is nothing like updating a browser.
For most parts, I think auto update is necessary for tech illiterates, especially now that everybody's jumping on the Agile bandwagon, including Windows. There needs to be a way to ensure new versions reach their users given everyone's just churning out barely working software these days.
Honestly I don't have a problem with that. But if they don't give power user the option to opt out, this is just disrespectful
The benefit of linux is that it doesn't force stuff on you that you can't disable compared to windows. If you take that away, there are very little reasons for any normal user to move.
Windows Server gets to exist so long as it acts as a funnel into Azure. Someday it will be sold off to Wipro, a la VMS or OS/2.
I'm well aware of how important Windows Server was in the enterprise, and I'm aware of how far it has fallen.
Is this a personal opinion or pure speculation?
For a normal user, the INSTANT something goes wrong in Linux, you step from “something kinda like Windows” to “100% Voodoo you are expected to know”
Think about explaining to grandma over the phone.
It’s just a non-starter for Linux. Which is fine, it’s what it is for a reason. But it’ll never have mass appeal because of it.
I knew that, but only in the back of my head. If the internet wasn’t working, you know, a WiFi problem, I would have a very tough time remembering that. Which is the point, trying to remember the voodoo.
And if I had to ask grandma to type in nmcli d then read me what she sees? Well, if anyone wants to prove me wrong, I’ll set my grandma up on Linux and give her your cell number.
I get that there are a lot of Linux fans here, but be realistic, it’s not for most people and never will be. You need autism or have never worked outside of tech lack the empathy of how mass computer users interact with their machines.
But I want to know what specific things they use that can't work on linux subsystem?
Maybe there are some and I don't know about them.
Also, it‘s way easier to work with multiple desktops and windows position better than on Windows where it seems that they are opening always there where you don‘t look.
There are also quite a lot of others, IMHO.
I don't have anything to say about your other nitpick.
A good reason to use windows is games, Adobe and office. Games using anti-cheat will not be available on linux and if you can get them to work, good luck not getting banned.
And IMHO, the discussion about which OS one should use is probably older than any OS at all. It comes down to personal preferences and experiences. I would not for any longer time want to use only Windows for a couple of reasons. I use it for some office stuff (we have SharePoint and OneDrive for Businsess does not exist for Linux).
I did a quick search and here are two references about the WSL speed:
[1] https://vxlabs.com/2019/12/06/wsl2-io-measurements/ [2] https://github.com/microsoft/WSL/issues/4197
It seems that they already improved a lot bur are still nowhere near native FS speed because of some technical limiations on how to embed the FS.
A more comparable situation is if Ford had requested his supplier send him a certain type of screw but the supplier sent him a completely different type of screw because it was “better”. It may be better. But it may also be absolutely the wrong thing to build Ford cars out of because they weren’t designed to use it.
And that’s what Canonical often seems to forget. Ubuntu isn’t just a product. It’s also infrastructure, and an individual (although critical) part in many other products and systems.
And AFAICT Ubuntu as a product is far less popular, and pays far less of the bill than Ubuntu as infrastructure, which is why their actions are doubly incomprehensible.
I look out my window today and I see many colorful cars. It's better this way. Fuck control freaks.
I believe in mutual respect
This does not reflect the hundreds of regular users that I have serviced over the years. Perhaps fewer than five would postpone updates beyond a few days for convenience / uncertainty, the rest would always install Windows updates within a day or two of being prompted.
Part of the solution should be that Microsoft Windows should refuse to install on any mechanical hard disk or eMMC and on any SSD slower than a certain benchmark. The other option would be to minimize the obnoxious amounts of reading and writing Windows does to disk but that would make sense.
Personally, I don't mind snaps that auto update. Apparently, "modern" snaps are sandboxed but "classic" snaps are not? I think this is the biggest failure of snap. What is a snap? Why couldn't we give legacy snap and real snap two different names? It isn't like there used to be snaps and then we got modern snaps. They both started existing at the same time.
I agree - when one's workflow must be interrupted by either application or OS restart, then the software must work for the user and wait for a convenient time.
After having an update happen at a bad time, one closing a file I hadn't yet saved, and another killing a process of computations I had been leaving to run overnight, I decided that I wanted to use my computer the way I saw fit, not the way someone else saw fit. If someone cannot wrap their head around the idea that I might leave a process running while I sleep, or that I might forget to save something in a program that doesn't have an autosave feature, well, they're not considering all of the use cases.
For giggles, after LTSB I did "all the things" to get rid of forced updates on one particularly infrequently-used tablet and, as an experiment, left it unpatched for slightly over a year before seeing what would happen once I finally let it do its thing. For one, the updates took about three and a half hours, which I thought was interesting; just re-installing would likely be faster.
More interesting, though, was that the world did not end because I had not patched this tablet.
I get aiming at the eighty percent of your users who just would not think to update, but being that hostile to the other twenty percent who are aware of updates and still would like that control is annoying. It is sad to see that this kind of thinking has made its way to Linuxland, which I had imagined to be the last bastion of "we're doing this for your own good."
It's just really hard to get officially for consumers. They should sell it to everyone that wants it.
And yeah, Ubuntu Desktop is kinda the "Windows Home Edition" of Linux though. Something like Arch, Debian or Alpine won't do this to you.
It's annoying having to set up a whole windows domain just to disable this though :(
You make it sound like an irrational fear but there's a real cost/benefit ratio to consider when you upgrade a machine. I personally always try to keep my machines up to date but stuff does break from time to time. Like last week an arch upgrade updated some system library to a new major version which forced me to regenerate my python dev env for work (which is not a trivial task because for various reasons our environment is fairly custom).
Windows is even worse in that regards because its upgrades are often significantly more intrusive (and they use automatic update to push new features and products, which is a great way to have people attempt not to update just to avoid them).
IMO if OS vendors want their users to update at all costs they shouldn't force it onto them, instead they should develop better transactional update systems that effortlessly and reliably let you revert to a previous version if something goes wrong. Then there would effectively be no reason to fear any update, since you know that at any point if something goes wrong you can just click on "undo update" and you're good to go.
95+% of users have no chance of even making this cost-benefit analysis because they cannot assess the scope or risk associated with the upgrade even if they wanted to. They have to rely on vendors making those assessments for them.
That's why I run Debian stable, which provides security updates and critical bugfixes for 5 years or more post-release with very limited changes otherwise. Maximized benefit, minimal cost.
FYI -- use Debian almost exclusively, been doing so for 20+ years and love it. (Well mostly, the last 5-7 years have been rough though.)
What Debian provides, is security updates for 1 year after the next stable release. That's it.
'5 years' is via LTS is volunteer support, provided by corporate and private donations, via a corporation in France. Their goal is to extend Debian oldstable's lifespan. The entire process is absolutely not the same as updates when running Debian stable.
For example, due to its volunteer nature, companies get to decide where they 'put their money'. What packages are prioritized. Rare / unused packages may never be addressed, depending upon funding.
Use PHP? Apache? The Linux Kernel? Sure, you'll see updates! Use rare package $x, and that may not happen, even though Debian Security would handle it.
I can also tell you via experience, that QA is not quite as good as Debian proper.
Still, is it a good thing? Sure! Is it managed by the Debian security team, 100% embracing all of its methods, and so on? NO.
I felt it is important for you (and others) to know this.
It's not as stable as Debian, not managed by Debian, and should ONLY be used as a stop-gap. You want stable?
Stay on Debian managed security updates!
What everybody does however is to use past experiences to evaluate the risks. Who hasn't had a system upgrade break something that took a while to fix? In these conditions, who wants their system to auto-update if the system is critical and it could happen at the worst possible moment?
As I said I try to keep my system up to date, but if I know that I have an important deadline in the near future I'm likely to postpone updates to avoid shooting myself in the foot.
Most Windows updates are security or bug fixes only. The exception being the twice yearly feature updates. These have been an issue mostly because Microsoft had been using a random subset of Home users as beta testers for new updates. However, they are being less aggressive with this now.
Oh, that might explain my experience a bit. I've never had any issues with automatic Windows Updates managing 50+ with a mix of common software. Every one of them was always Pro or Enterprise though.
Ran Arch for a couple years for dev with a nightly cronjob of (iirc) "pacman -Syuw --no-confirm". That broke regularly but I knew it was a bad idea.
The problem is that "power users" can't be trusted to use this power responsibly, or they operate based on assumptions that won't always be true. Either way they end up hurting other people in enough situations that this becomes an ecosystem issue.
Power users often help others set up their systems like they do, but those users can't manage them. Other times, power users will move on. When they were managing the systems, it wasn't a problem. Now that someone else is, the systems stop getting updated and someone has to clear up the mess. Yes, good power users help transition things properly, but don't kid yourself on the number of times this actually happens.
This phenomenon is prevalent in software. Just looks at the defaults which "power users" frequently choose (cf. JWT).
But when those decisions affect others? When your computer becomes part of a botnet and is used to attack a target then it's no longer just your problem anymore. The responsibility to prevent this must lie somewhere, either with the user or with the manufacturer. And Microsoft would rather force the updates on users and occasionally be in the news for a botched one than to constantly be in the news for botnets of hundreds of thousands of "abandoned" Windows PCs, or the constant complaint that Windows is insecure because people keep their PCs stuck on the same version they came with.
To use some very current imagery, imagine if someone walked around coughing in your face saying the decision whether to wear a mask or not is theirs.
Further your argument draws a connection but its spurious a sufficient difference in degree is a difference in kind. The way in which we comport ourselves while sick have the potential on net to kill millions of people where as the peril implied by users failing to update windows has never resulted in peril of that magnitude. Further one can reasonably suppose that one can with sufficient care design a system where updates are on by default and we don't create perverse situations which inspire many users to turn them off entirely.
For example one in which applications are updated without disturbing or interrupting the users workflow, where updates aren't effective until the user reboots, where applications are isolated from their underlying environments where users can roll back and pin a particular version if a new version is buggy or undesirable. Where major changes to entire user interfaces are rare and opt in for years.
How many are going to bother disabling updates?
Please don't make unfounded accusations and personal attacks based on suppositions. It was just an easy to understand analogy of how your decisions can have consequences beyond yourself, and does the job without any hint of "emotional weight". The rest is in the eye of the beholder.
> have the potential on net to kill millions of people
Talk about manipulative and lending undue emotional weight to your argument. Nothing gives weight to your own words like not following them yourself.
> one can reasonably suppose that one can with sufficient care design a system where updates are on by default and we don't create perverse situations which inspire many users to turn them off entirely.
This supposition didn't fare well in reality because it's easy to suppose but hard to implement. Especially when talking about a very complex system that has to be put in the hands of ~87+% of computer users out there, and work with tens of thousands of combinations of hardware, software and different configurations. And perhaps the most critical aspect is that the perverse incentives are left to the judgement of users with little to no understanding of the system or the wider implications of misusing it. They are more likely to follow terrible advice because the explanation for the good advice is too complicated. This is why the easy to understand analogy was useful.
This is true of capitalism and democracy. The worst choice for economics or governance except for all the other choices. I want a system that respects the users judgement not yours. No matter how well meaning you can't adopt the perspective of all users nor do I desire to see the mediocre results of smart people who know better than their users trying.
It's actually not that hard. If you make updates something that silently happens periodically without interrupting the users or making many changes to the UI that the users rely on they will let you update the parts they don't directly touch all you want to secure their systems.
When you want to make major changes make them opt in and test them to ensure they are actually substantially superior. After a while deprecate the old UI. People will tolerate infrequent major changes far better than constant small breaking changes to their workflow.
Again if you don't make updates suck you don't have to coerce people into doing them. If you are figuring out how to coerce users for their own good you are solving the wrong problem.
Of course, since it’s my hardware I just uninstalled Ubuntu and went with something that gave me more control.
(That kinda reminds me of "trust us, we're the government and we know what's best for you".)
Sorry, but that goes against everything that this whole "free software" thing stands for.
Or, why we have regulations to protect workers. Everyone doesn't have to share your worldview. There are plenty of reasons why these changes are happening. The people making the changes are normally aware of their trade-offs.
Change is the key word. If the world was better before from the perspective of the designer then they are unlikely to have changed anything, or introduced an approach you disagree with. The choice is for you: accept it, or find an alternative. That's they way of the market, no?
>> Sorry, but that goes against everything that this whole "free software" thing stands for.
I agree. Ubuntu is a particular vision of free software. They can do it the way they want. There are way that you can do free software the way that you want to have it.
you vastly underestimate the amount of changes which are just done because people have to justify doing things in order to get a paycheck
Nevertheless, I dual-boot my laptop and use Windows almost only for presentations. Whenever connections to external devices and other OS shenanigans force me to reboot my machine when I'm in the speaker position and the dreaded forced update cycle begins I swear I'll never touch that ugly OS ever again. And I already developed routine of checking for updates and booting in advance, but somehow this still happens from time time.
I guess there's need to be a special "seriously not now pretty please" button.
The fact that everyone is on the agile bandwagon churning out barely working software is the main reason why people do not want new versions of software magically appearing on their computers.
Once you've got a version that actually works for you, allowing an update to come through is like playing russian roulette.
I think this is a misconception.
A few weeks ago I was updating my Windows 7 install that hadn't booted for a year or so. I opened Windows Update. It looked for updates, and found some. I clicked the update button. It proceeded to start downloading, by which I mean it would hang for 1 to 2 minutes and then download very quickly.
When it was done updating, it required a reboot. After the reboot it needed to do something for another 5 minutes. The last 2 of those minutes it showed 100% progress, seemingly stuck again. Then when that was done, it rebooted again.
I started Windows Update again. It looked for updates. There were more updates. I clicked the update button. It proceeded to start downloading, by which I mean it would hang for 1 to 2 minutes and then download very quickly.
When it was done updating, it required a reboot... and so on, and so on, 5 or so times. Every iteration took somewhere between 10 and 45 minutes. Determining how long an iteration would take was impossible, because everything appeared to just get stuck all the time. This has been the Windows update experience through XP, Vista, 7, 8 and now 10.
The result of this madness is of course that nobody updates Windows. To solve that, you can go two different ways: Make updates painless, or force everyone to go through the pain again, and again, and again. Microsoft being Microsoft, they picked the latter.
Meanwhile, on a typical Linux system, I just decide to update every once in a while. I make it look for updates, get a list of everything that will get updated, and accept. It starts downloading, by which I mean it immediately downloads everything very quickly.
When it is done updating, it may require a reboot if the kernel got updated. After the reboot it doesn't have to do anything.
I make it look for updates, and there are none. Updating my system actually brought it up-to-date. This generally takes 2 to 15 minutes, almost entirely dependent on the amount of data that needs to be downloaded.
The result of this is that I update my Linux systems very often. It's painless, so why not?
The Windows 10 servicing model doesn’t have that problem, a new cumulative update gets pushed out every month and can get a machine to the latest update for that branch regardless of how long it’s been out of contact with Windows Update. The semi-annual branches can also be directly upgraded to from any prior branch, as they are essentially a full upgrade of Windows just like moving between releases of Ubuntu/Fedora/etc.
None of the other problems are excusable though.
No download should take 2 minutes to start, especially from a company like Microsoft. Sure, an anomaly is possible, but this update loop took several hours and every iteration was like that.
No update should ever take 5 minutes after the reboot, especially on an SSD.
No progress meter (that isn't dependent on a remote service) should ever be at 100% for 40% of its total runtime.
No single update should ever require 2 reboots.
19.04 Ubuntu just died on me today (I know some expert could have gotten in working). Apparently 19.04 support ended and someone took down the servers. So trying to update would tell me something about the servers having no release file. And they wouldn't let me update to 20.04 until I patched 19.04. I never modded anything. Whatever broke it broke itself. I searched the net for answers but my search foo sucked. Someone said download the 19.04 ISO and extract the sources.list file out. I did, it had different repos but got the same errors (with the source urls pointing to the new places of course)
So, 8 hours later I just finished reformatted the drive and installed 20.04
Yep, Linux is painless :rolleyes:
If you're not going to do release upgrades frequently you should stick to LTS releases (20.04 is one), which are supported for over 4 years: https://ubuntu.com/about/release-cycle
Sure they could use LTS but that has the same problem, just on a longer timescale.
I could turn on a PC with the first Windows 10 beta and it would update to the latest Windows 10 no problem. There's no reason Ubuntu can't do the same.
Do not feel bad about it at all: somehow it seems impossible to search for answers related to Ubuntu. Anything related is for something like 14.* or 12.* releases. I do not understand what is going on: are people really not asking the questions for newer versions, both Google and DDG not properly ranking or are the questions/answers getting somehow marked as dupe and deleted?
Inevitably you will miss some use cases. Given sufficient configurability users faced with substantial challenges will be able to find a way to use your software.
If you can easily search options including a description of the option this can still be usable for all.
As many others I haven't liked vanilla Ubuntu since they swapped Gnome 2 for a (what I think is a) clone of the Mac OS interface, but my desktop still uses the Ubuntu 18.04 base system under the hood so this might come in handy for a number of people like me.
https://people.canonical.com/~ubuntu-security/cve/universe.h...
So, Ubuntu's LTS support is only relevant security-wise if you mostly stick to the main repository.
https://www.debian.org/security/faq#contrib
However most packages in Debian are in main. AFAIK contrib is for FLOSS packages that somehow depend on packages that are not FLOSS.
There's way more evil things about Windows, like click to raise and menu bars.
personally, i think XDG_CONFIG_HOME + a robust config file framework with a standardized format (yaml/toml, whatever) would work as well. being able to specify a schema could be extremely useful, too, to prevent borked configs. as you've said, we've seen a lot of tools/OS's go this direction. and it doesn't have to be perfect, the 80% case would be a huge improvement.
i'm also sure the registry made a lot of sense at the time. it was probably way quicker than reading and parsing files. we didn't care so much about sandboxing/isolation and backing up. people probably had less application installed.
However, Canonical pushes snaps via deb packages (e.g. Chromium).
The best option is the use an Ubuntu-based distro, like PopOS or Linux Mint or Elementary OS.
I wish they would have based it directly on Debian instead of Ubuntu, but I just stay away from Snaps and get on with my life.
Ubuntu went totally overboard with their snaps, putting everything in it trying to force it to become a standard. But FlatPak is not great either even though it is an open standard.
It's nice for the 1 or 2 niche apps that aren't available any other way, but for the most part I prefer normal packages.
you can actually get a very stable system in the same way you would any other OS: install only the packages you can live with. you can install, say, xmonad and no DE and have a slick, solid, low resource dev setup, whereas ubuntu comes with some goofy ubuntu software store that's not removable (or wasn't when i last used it) - one wonders what other "goodies" are in there, esp. after amazon search behavior change. remember that deleted code is debugged code: the more spartan, the less chance for issues. so arch's actually a plus there.
FWIW my boomer parents run arch on 2 different desktops solidly, without much sysadmin work from me other than a few pacman upgrades whenever i remember to VPN back in, perhaps biweekly on average. yes the occasional breakage happens, see above re: website/forum/etc.
debian stable was OK as firefox was fairly regularly updated for vulnerabilities, but the actual version number was lagging a fair bit behind. ubuntu LTS just seemed like a worse debian. ubuntu non-LTS is a non-starter: probably 50% of the dist-upgrades i ran ended up failing and basically needing a clean reinstall for anything not in /home, so it's fairly admin hostile in that sense.
Also consider Manjaro if you want an Arch-based distro but don’t want to go through the trouble of manually setting everything up on first install. I’ve been running the KDE edition as my main OS and it’s fantastic!
I always shill for it.
It's like Ubuntu if Ubuntu were rolling (so much faster drivers and DE features, but nerfed delivery to minimize breakage and still really user-friendly, with a GUI kernel and driver switcher) and if instead of PPAs it had a big repo that just had everything on it (the AUR, which Manjaro's GUI package manager, Pamac, works with once the setting's toggled).
On the flip side, this is incredibly empowering. If something breaks, you can just fix it, since you were the one who put it together in the first place. That's a big jump from Ubuntu, where everything just works, but you have less control. Arch is worth it if you think of it as a future investment, rather than a short-term easy fix.
Arch with its rolling release model is prone to breakage. Every upgrade gets you closer to having to reinstall everything from scratch. On contrast, NixOS is transactional and if something breaks you can switch to a previous version easily.
(I actually went from Arch to NixOS, after using Arch for about 3 years).
Package managers (including snapd) basically just "think" in semver. Semver is one-dimensional: releases are arranged on a big line, and they'll know to auto-update based on e.g. whether a given release is close to your current release on the line.
For the cases where we do distinct security updates (e.g. kernel updates), we seem to do them using all sorts of hacks.
Maybe we just need a superset of semver, that can encode more than just the "newness" dimension, but also the "criticality" dimension?
I'm not sure if that person / team can ever be replaced by modifications to a version numbering scheme.
This is NOT how semver works, though... each section has, well, semantic meaning. Some updaters might linearize, sure, and many developers run fast and loose with versioning, but semver is a graph of varying major, minor, and patch version numbers which change with their own semantics which can describe more than just newness
Bump major -> y'all best be careful
Bump minor -> something you care about maybe changed, read the changelog
Bump patch -> we'll probably just fix some bugs
patch releases could and should be backportable to other minor releases under the same major release if people care about stability of a module. I think that the "work from master" mentality that npm and GitHub UX has lead people to is one of a handful of reasons that prople misunderstand versioning strings...
I agree that flagging criticality is useful, though. Linux packagers like YUM/DNF have had this for a while, even the ability to feed a CVE identifier or bug id in to the package manager to resolve them
If you are running "mission critical" software, you need to be using a packaging mechanism (c.f. Docker) which lives at a lower level and provides harder guarantees about what you're running. Those solutions exist, they just aren't Snappy.
That combined with Ubuntu forcing snaps as the main packaging method (i.e. you have to jump through various hoops to install .debs of snaps) is what a lot of the outrage about.
(Side note: Something about listing Docker as a packaging mechanism makes me uncomfortable. IMO Docker and containers in general are deployment tools, not packaging systems)
To me snaps seems like a desktop solution that Ubuntu forgot to disable on the server edition.
Obviously it sits at the boundary, but a "Dockerfile" is (at least when properly used) a recipe for reliably reproducing a specific version of software packaged in a format that can be deployed to all sorts of systems with absolutely minimal dependence on host configuration.
That's what people who want a "mission critical Snap" almost certainly want.
They have prioritized snaps but how have they made life easier for those of us running "mission critical" software?
> obviously used mechanism
OK...
> If you are running "mission critical" software, you need to be using a packaging mechanism (c.f. Docker) which lives at a lower level and provides harder guarantees about what you're running. Those solutions exist, they just aren't Snappy.
What? No it doesn't live at lower level. Docker uses same Linux APIs as Snappy.
The only upside of Docker when compared to Snappy is that it doesn't turn your hardware into zombie execution unit constantly pulling code from mothership (Canonical).
> In that realm, the aggregate benefit to society of having everyone running current software outweighs the annoyance of the specialists, sorry. It's always been this way.
There's in fact an easy way to satisfy both groups, but for some weird reason Canonical chose to turn every user into zombie execution unit, no matter the level of proficiency. It's not like a patch to disable this malware behavior would be hard to submit, but it's crystal clear it'd be rejected.
And it's clearly not for causal user benefit. They have documented switches to postpone calls to mothership. That's command line realm for experienced users.
I honestly wonder what their real motivation is. Those who seek to take back the control are people who want to control their own machines. The only thing that comes to mind is that sometime in future blog posts recommending to disable automatic updates would crop up and those dumb pesky users would just copy paste it into console without much thinking.
I think he meant you can't totally disable snap auto-updating.