Any amount of service toggling and hosts file stuffing will not suffice. It just screams ignorance. As a software developer you should understand that plugging holes in a black box is a futile effort. All these tools are doing is giving a false sense of privacy, that the next update will undo by flipping a switch or installing a new service.
If you think the OS is violating your privacy, stop using it or remove it from the Internet. Or both. It's the only way.
Edited to add: If you actually like Windows (I do), just switch to the Enterprise Edition and dial Telemetry down to "Security". Here is an explanation of what little is then shared, and how to even further minimize your footprint: https://technet.microsoft.com/en-us/itpro/windows/manage/con...
Edit to address the availability of the Enterprise Edition: If you are not able to get it via your $JOB, a valid key from MSDN surplus shouldn't be more than $50 if you look around. Of course you'd then be bending the EULA in your favor, but hey, since Microsoft is spying on everyone against their will I think it is fair game, right?
Pragmatically there are reasons for some people to run Windows 10 vs. other operating systems, even if you don't/won't recognize them. Tools like this allow people to run Windows 10 in a "good enough" state and spread awareness of the problem. It will always be a cat and mouse game if the black box vendor so chooses but it's still better than doing nothing.
Edit: To address your edit, aren't you trusting their black box and using a different tool to accomplish the same? Also, not everyone has access to Windows 10 Enterprise.
I actually like Windows from a technical point of view. However, as a private consumer how do I buy the Enterprise edition that gives me full control of my system? Even with the "Professional" edition Microsoft is still in the driver's seat (e.g. http://superuser.com/questions/1110265/how-to-prevent-window...).
It's not "off". My workstation running Windows 10 Enterprise still makes a lot of network calls (even explorer.exe) not related to Windows update. I never came across an in-depth analysis of what exactly is being sent, please share if you know of one.
The lowest telemetry setting level supported through management policies is Security.
[Security Security data only. 0]
Information that’s required to help keep Windows, Windows Server, and System Center secure, including data about the Connected User Experience and Telemetry component settings, the Malicious Software Removal Tool, and Windows Defender.
The Security level gathers only the telemetry info that is required to keep Windows devices, Windows Server, and guests protected with the latest security updates. This level is only available on Windows Server 2016, Windows 10 Enterprise, Windows 10 Education, Windows 10 Mobile Enterprise, and Windos IoT Core editions.
Ah… I did some more looking around and found out that the “Off” option actually did still exist¹ in the release version of Windows 10 Enterprise. However, it seems like some update to Windows 10 changed the label from “Off” to “Security”² instead. I can only think of two possible explanations for the change:
• Microsoft removed the already existing capability to completely turn off Telemetry for some reason, or
• the “Off” label wasn’t accurate in the first place, so Microsoft changed it to something less misleading
In any case, it seems like Microsoft has no plans to include a way to fully turn off telemetry on Windows 10 Enterprise anytime soon³.
As for your inquiry, unfortunately, I haven’t seen a more in-depth analysis of what being sent than the one at the link you’ve posted (although it actually does go into a bit more detail than just the part you’ve quoted here). There is this⁴, although it’s just a list of hostnames and IP addresses; there was no packet inspection done, so it doesn’t make it clear what’s actually being sent.
The reason why people will continue using a OS that they dislike and distrust, is the same reason why people don't switch to an Free and open source OS. Too much software is exclusively on windows, and that forces the user onto that sticky platform regardless of user preference.
Its the same reason why people who dislike and tries to block advertisement don't simply stop consuming contents that contain advertisement. They don't want to turn into hermits that live on a mountain away from the web, TV, mail, email, radio, billboards, milk cartons, the sky, and practically everywhere where a company can stick a advertisement on something. It is an imperfect solution to an imperfect world.
> The reason why people will continue using a OS that they dislike and distrust...Too much software is exclusively on windows...
Is there any room in your opinion for people who love Windows and think it's better than any other OS that is currently available? Because that's why I stick with it, despite having some very minor issues...
Also, the reason that I don't switch to a Free and open source OS for my desktop is because they all suck. They're slower and clunkier than Windows and they don't have the features that I want.
All of my Windows issues were solved by simply toggling features via Settings and Group Policy though. I think there is one setting that you need the Enterprise version to toggle and that is Telemetry. However, you can disable that service manually too - http://www.thewindowsclub.com/windows-10-telemetry/ Of course disabling Telemetry causes you to lose Cortana, the Windows Store and any use of your Microsoft Account - but I don't use any of that crap anyway and anyone who does want to use that stuff wouldn't care about the basic Telemetry data that gets collected, which is detailed here - https://privacy.microsoft.com/en-US/windows-10-feedback-diag...
I really don't understand how something that can run on less than a Pi can feel slower on consumer hardware than something that requires beefier specs.
For example, I never have to wait for my file manager to open. Not half a second.
Secondly, though Microsoft details the telemetry, its encrypted before the user can see it. You have to trust a company, that have a habit of bending over backwards for the US's clandestine organisations. It can't be verified.
I'm judging by the speed of the apps that run on top of the OS not the OS itself. Desktop apps specifically.
For example, all of the browsers run slower and are klunkier on Linux.
I'm with you on the telemetry. I just disabled it via the registry though. That option works on all editions of Windows unless I'm mistaken... which I very well may be since I did not go to very far lengths to verify that my machine is not sending back anything. However, I am not worried about US clandestine operations because there's nothing I can do about them anyway. They are into everything around you, not just Windows.
In my opinion the greatest threat is not spying on you. The thing you should be worried about the most is psychological warfare. They are not supposed to be running psychological operations on US soil, but it's so obvious that nobody follows that rule. TV, movies, news...all of them are used to program people. Honestly, there's nothing you can do about that either unless you are seriously rich and very well-informed.
Spying leads to manipulation, true. But my fear is based on not living in the US. And disabling regkeys doesn't stop 5gb of telemetry going to MS a day. Which I find just a tad excessive.
Another big reason is maturity of the software. I've been using Linux and Windows in parallel for years now and even though it got better recently, I still stumble over minor bugs, inconsistencies, usability slips and such on Linux while those are practically non-existent on Windows.
It's quite understandable given the different objectives and budgets of the two, but I think for most average users this is a deal breaker for switching.
So - no effort is better than some effort? That's pretty dark, and if we're given a control of at least 20% of the holes, I would say - use that control.
> If you are not able to get it via your $JOB, a valid key from MSDN surplus shouldn't be more than $50 if you look around.
Then again, the main reason for running Windows instead of Linux is because you want stuff to just work. Once 'buy Windows' becomes 'go looking on the gray market in the hope of finding something unsupported that might or might not actually work when you try to install it' the value proposition relative to Linux has been significantly eroded.
People want to have their cake and eat it, too. The windows ecosystem is all they know and they are not ready to step outside their comfort zone to use open source alternatives. They rather pirate whatever comes along and try to patch things up as good as possible using antivirus, firewall, VPN and a whole lot of snakeoil software. I don't think we can change a lot on that front in the coming years. I don't know if we should even try.
Windows actually has a pretty nice UI, and although some of the underlying system is different than more unixy variants, that doesn't make it bad. I use Linux (htpc, and work dist) as well as Windows and macOS daily... I prefer the Windows UI on the desktop, and Unity (Ubuntu) is close enough for me... macOS is the odd one out... I mostly stick to bash and node stuff lately, so I can get by anywhere. I use VS Code for editing, so again can get by anywhere.
I agree, for example when they block IP in host file and when you check these IP as nothing to do with Microsoft...
Also they block things that can be used in a good way, like recognition of ink pen, yeah you don't want Microsoft can improve then how they will improve it? and they will complain why this pen don't recognize anything, the solution will be to use ccleaner to do more mess because we all know the script that BOOST windows performance...
All of this "help" for "fixing" Windows 10 privacy issues seems to be based on the "don't worry, I know best" method rather than actual facts explaining the effect each of the items being disabled.
I like using Cortana and Windows Store apps. I like being able to provide bug reports when something hangs.
I don't like the idea of any data going to the mothership that doesn't have an obvious effect on my day to day computing. I'm especially leery of anything that requires (arbitrarily defined) personal information.
I also don't have time to research every one of these registry keys or policies...soooo...thanks for doing some of the work!
I am actually quite happy that someone else than me is trying new drugs before they are widely used as a treatment. I have mixed feelings about Microsoft using consumers as beta testers.
I don't mind being part of the solution, I just wish it were clear what trials I was participating in ,and had the ability to opt out if it didn't feel right...
Unless you're in the insider program on the insider build, I believe what they are grabbing is mostly usage data for features and applications... Now that could be nefarious or benign, there's no real way to know.
Yeah but Microsoft should have catch the start menu hang plaguing windows 10 recently before it goes into production.
And I am not sure that "everyone else does it" is a good reason. Everyone else does nagging. Apple nags again and again for its icloud, apple pay, apple music, etc. I get some nagging for using instagram all over my facebook feed. Microsofts nags me for using edge, onedrive, etc.
But "the others do it too" doesn't make a good product.
They should absolutely fix it. But they just fall back to the "everyone is doing it, so why not us" attitude. I've seen companies include phone-home features after win10 rolled out theirs.
This is not something new. There are a lot of tools avaiable on GtiHub which are far better than this tool. Infact, the first tool I used was Nummer's DWS [1] (stil the best one out there).
Honest question: how do you know? I read the OP article and it talks about 100+ 'rules'. How can one ever make sure all this stuff is disabled (or even needs disabling)? And possible new stuff after updates gets disabled as well?
I'd really want to switch to W10 because it certainly has appealing features but didn't do set yet exactly because of this privacy stuff. But reading things like this makes me almost give up and go like 'well, yeah, f this, nothing I can do about it so let's just use W10 it and to hell with my privcay, nothing to see here anyway'
If I were to run just one of those tools, I'd pick the original article. That's because his plain English description is far better than the others. He starts off with "What & Why?" and explains what problem he's solving; the others jump in with words like "remove all spyware modules".
Even for very technical people, unless we're willing to analyze the source code, a good clear description (and reviews) is how we're going to decide which product to try.
Author of [3] here, you are right, I do skip the what & why, on purpose. I think there a far better places to learn about the whole spyware / telemetry and privacy topic, even better than the original article. [3] is made for admins which already know what they want (and why) to help them kickstart their own setup scripts. They are advised to read the scripts themselves instead of an inaccurate explanation.
I do not intend to sell you the project. (Sell in the sense of talking you into using it. Either it's what you are looking for or you need something different. And I will not waste your valuable time if you need something different.)
"What is normality with Apple since long and apparently has been accepted by OSX users..."
As far as I know, OS X asks for permission for diagnostic data collection on first boot and can be turned off anytime. Unlike Windows, on OS X the data collected is not stored encrypted and can be inspected. They are not equivalent.
Seriously. People just irrationally hate on Apple no matter how far they'll go to trumpet user privacy from the rooftops. There is tons of objective evidence of Windows spying, yet no sources for this claim about macs. SMH.
Just having reinstalled Sierra yesterday, it does. It also asks you if you want to turn Siri on, and tells you everything Siri collects. It's also easy to turn off "Spotlight Suggestions".
Whether that's actually everything, I don't personally know. It seemed like a unnecessary and out of place dig though.
I have disabled diagnostic data collection, and LittleSnitch still finds at least 6 daemons constantly trying to do requests and send data to Apple servers.
(And yes, I also dug through the settings and disabled things in a rather counter-intuitive Spotlight menu.)
It's surprising and sad how much drama people are willing to put up with rather than switch to GNU/Linux. (edit: some) People are willing to get unlicensed copies, exposing themselves to legal consequences and malware, rather than use a free (as in freedom and price) operating system.
There are too many things I can't do yet in Linux. I can't play the games I actually want to play. There are graphics tablets and modeling/animation/drawing tools that I want to use without having to spend as much time fussing with emulation layers or finding the right binary drivers.
So yes, it is simply easier to just use Windows and turn off as much as you can. It's a pain, but the relative pain to using Linux is lower.
You're financially supporting the wrong people. So those are the people who can afford to implement the tools you want. Financially support the right people, and you'll find your tools follow the money.
Outside of random crowdfunding events with specific targets or tiny details that I could bug-bounty myself there is no way to give money to "the right people" while being reasonably sure to actually get what I want, so the argument seems pointless. Donating to open-source is great, but not a replacement for buying tools that do what's needed now.
The problem there is that it really restricts your hardware choices if you want to not support Microsoft. And installing Linux doesn't financially support anything.
There are games, there are professional software toolkits only running on Windows and/or macOS (like architectural designs, photo editing, drawing hardware + software combos, and a lot more) and newsflash, people need them for their paycheck. Oh, and there are also polished, consistent and non-glitchy interfaces and system updates to be found in Windows 10 and macOS. You know, these things only the "sheep" require.
/sarcasm off
Get a grip and please go trumpet your own "freedom" somewhere else. HackerNews is a place for objectivity, not bandwagon hating.
I tried many times to use Ubuntu as my main machine but I always go back to Windows. Why? more than 4 times when I run a system Upgrade it fails and I end up with a broken system.
The other thing that holds me back is the UI. Hundreds of Themes available but I always see them as "cartoonish" themes that are not well though of.
I'm dual-booting between Windows and Ubuntu. I keep Windows for games that require it; most of my daily use is on Linux, but there I spend most of the time in the shell, Emacs and in Chrome. So I tolerate various UX annoyances. But I'm not going to recommend Ubuntu (or for that matter, any other Linux system I've used so far, and I've used a few) to normal people looking for a normal desktop - there's just too many glitches, too many UX annoyances, too many bugs in GUI-dependent software. For all its issues, Windows has a quite polished desktop ecosystem.
Windows brakes on updates also[0]. It's all about knowing the system well enough to fix it when it does.
[0] Heck, Windows breaks on reboots! Sometimes when I reboot, it either forgets my bluetooth or my wireless. And by forgets, I mean it doesn't think the hardware even exists. Multiple machines, multiple vendors, so it's not just one odd computer.
Rough count, I've worked with 30+ different windows machines over the last few decades. I have never seen the problem you are describing. I'm not saying you're not having it, I'm just saying it's not necessarily a common event. Maybe you live/work somewhere with a lot of EM interference?
No, not just my machines either. Work machines across the country out in the field have issues with Windows forgetting hardware[0] (network, wifi, bluetooth).
I think people just get used to Windows' issues and don't recall how many times they have to deal with them.
[0] At least once per week out of several hundred machines.
> I think people just get used to Windows' issues and don't recall how many times they have to deal with them.
I can recall a lot of issues that I've had with windows. I can tell you with certainty that I've never had a device go missing from the device manager.
I've had printers have trouble being discovered after they are unplugged. I've had driver updates cause hardware to stop working. I've even had windows updates completely mess up my windows installation.
But I've never had the issue you are describing.
> [0] At least once per week out of several hundred machines.
You're saying you have a greater than 1% occurrence rate per year? Google doesn't show the issue in the first three pages for me with the query "Windows forgets network hardware." Have you opened a Microsoft support ticket? I mean, if you're losing device drivers every week, that's man hours your burning! I would want to get that fixed ASAP!
I'm far more wary of Windows updates than Ubuntu updates.
I have the same install since 14.04, I have updated to each version up to 16.04 and it still works fine.
OTOH, I reverted the Win10 free update and went back to Win8.1 because Win10 did not work with my bluetooth devices, the tablet functionality was inferior to Win8.1 and it was actually unpolished compared with the Win8.1 experience.
Your second point is simply opinion, and while I can't argue against it, my own experience using Ubunty with multiple monitors and virtual desktops makes Windows feel limited in comparison and therefore much less professional.
not to be glib but the best way to fix windows 10 privacy issues is to stop using windows 10 and switch to linux, there are many user friendly, mature distros out right now and they are only getting better: ubuntu, linux mint, arch, manjaro, openSUSE, centOS, etc. as well as up and comer new distros like elementary
Who picks their operating system first? IMHO, the OS is only there to support the applications I need to use. If those applications are available only on Windows, then your advice isn't very useful.
So I would say your advice isn't helpful. Microsoft has a financial relationship with their users (especially their business users) and despite what many people think, they do listen to their customers. So I would say the best way to fix Windows 10 is to continue to let Microsoft know that you aren't happy with specific changes they've made.
And here come the "b-but you can use wine" comments, which of course everyone who would be a proponent of such would understand that it is not perfect, does not run everything and is far from user-friendly.
The most obvious incompatibility right now is probably DX11-only software. Aside from that kind of known compatibility problem, wine has worked well for several years now.
> far from user-friendly.
While I agree that wine can get nasty if you need to debug anything or if you have unusual requirements, but the standard "just run this win32 binary" case has also been trivial for years. I believe some distros even enable the kernel feature that launches wine automagically when running a .exe file.
I pick the OS first, but most of the tools that I use are cross platform. If I did need applications that were only available on one, I would likely use a VM. Of course, if the main applications that I needed (maybe ones that I use 80% of the time) were only available on one, that would be different.
I think his advice is perfectly reasonable, it just doesn't apply to everyone. I bet that it applies to a lot of people in a technical audience, though.
Aren't you people EVER gaming though, programmers or not? I am doing it a lot less compared to only 5 years ago but I want the opportunity to be there the second I figure I want to game a bit, and no, dual-booting is not a way of addressing this at all.
Elaboration: when I have 3 Chrome windows, 2 work VMs fired, a separate Chrome window for the work services, several consoles outside of the VMs, and a live Twitch stream running on the TV hoooked to my PC then no, I am not willing to reboot only to play a game.
After many years away from gaming, I built a linux-based gaming machine recently. A surprisingly large number of games, from GOG and Steam, work natively on Linux today; many others (most dx9 games) works near flawlessly under wine.
I had actually planned to setup GPU bypass on the box and run Windows in a VM, but I have more games now to last a long time, and I haven't actually bothered.
Last several months I probably have like 4 hours a week at maximum. My point is, I want all the games in my Windows 10 machine to be there and ready the moment I figure I want to get into them.
Convenience > all, including for some of us the programmers. =)
36 here, and a well-paid programmer too; still gaming occasionally although it's times less compared to in the past. I suspect you're not in the majority, having in mind the statistics about adoption of user-level OSes. :P
From a web search the best-selling PC game in history was Diablo III with 5.15 million copies sold. That's barely a blip given 2.5 billion people online. Gaming is a niche activity.
According to Wikipedia [1] the top selling PC game is Minecraft, with 24 millions sold, followed by World of Warcraft that got to around 14 million subscribers, while Diablo III is in third place with 12 million copies sold...
Sure, that's still not much compared to 2.5 billion, but that's also 5 times what you said.
According to World Bank, the earth's population exceeds 7 billion people and 65% of those people are between the ages of 15 and 65. So let's say there's 4.5 billion people of driving age on the planet.
The Toyota Corolla is the best selling car of all time and has only sold 37.5 million units.
By your measure, driving is clearly a niche activity as well.
League of Legends has over 100 million active players a month. Dota 2 has around 39 million. Steam has over 125 million active, with 9 million concurrent users during peak hours. The free to play games market is enormous.
AFAIK, macOS runs more software than any other operating system simply because you generally can't run macOS in a VM (I believe you can iff you are running on Apple hardware). So that's probably the safest choice.
This is why I have a strong preference for free software that is available for the dominant three operating systems. Only if free software doesn't satisfy my needs I look for nonfree software and then I first look for commercial software which supports free operating systems. If all that fails, it's time for a Windows 7-VM to run that program.
I'm only one customer but I'm letting Microsoft know that I'm not happy with the way they go with Windows by not using Windows 10.
I have a strong preference for using the best software available to me. IMHO, all of the major operating systems are good enough although none are perfect.
These days, hardware and software is very inexpensive compared to time. I love the idea behind free software, but sometimes it's just too expensive to choose.
The "best" depends on the context. I don't want to argue what's best for you.
In the university context I've found that there is a lot of free software that works good enough or sometimes even better than commercial products. If you factor in that you still can use that particular program after you leave, it becomes a big plus.
That's very true. For example, if I need to use an image editor, Photoshop is the only thing I would consider. It's inexpensive ($35 for a month subscription), but more importantly it's easy to hire people that know it well. It's also well supported and documented.
When the subject "Windows 10 Privacy" comes up, linux folks come running offering linux as a magic pill, it's not.
As mature as linux distro's can be there's always the hardware variable in the equation, specially on laptops using nvidia optimus, the experience still sucks. It's the truth.
And i feel sorry for it, because i recognise that there are advantages in linux , but until nvidia fixes the damn optimus issues, linux will be held back.
Even though you're downvoted, I mostly agree with you.
Most Linux proponents are people with older hardware and trivial setups: 1-2 monitors, 1 sound card and a pair of normal speakers (without subwoofer), 1 ethernet or wireless card, only 1 ISP, etc.
We get it, for very regular setups Linux works fine. But there are many people who have more requirements outside the warm little box that Linux serves and I cannot understand why do these people come here convincing us this is not the case; and why are they downvoting the people who are outside this bubble.
Linux is not a magic pill on the desktop. For all its strengths (and I do use it in my work every day) it's still a lousy desktop OS and this hasn't changed for 10+ years now.
It's a magic pill if you can RTFM. Not everyone can.
Try to keep in mind that the vast majority of internet provided services are served on linux. A large portion are developed on linux, by people that don't seem to be hindered by the parents problems.
Why should people be interested? They get an advertisement that they buy something and "it just works". Are you by any chance interested in the intricacies of the process needed to guide your sh*t to a dump or a cleaning station? I'd bet you are not.
Don't be bitter. The fact that most of the internet infrastructure is served by Linux doesn't make it a user-grade OS. Unrelated areas.
>It still reported like 1/100000 of what Win10 does daily
Oh, absolutely, but if we're going to advise would-be Linux converts who are concerned with data leakage, may as well advertise actually-telemetry-free distros.
>Afaik Ubuntu removed that module like 2-3 years ago
For those out there that haven't bothered to do a quick search, one notable story about the issue can be found here [1]. Note that the date on that article is from 2014, and only refers to the Unity service for searching from Amazon. A later article [2] shows that Ubuntu has changed directions, but it certainly hasn't been 2-3 years. Needless to say, while it took Canonical some time, at least they dialed back what they were doing with online searches and made them opt-in. I've yet to see Microsoft, Google, or Apple dial back their telemetry services regardless of what criticism they may have faced (fair credit here, at least Apple is opt-in for the most part, though not entirely).
But ~3% is also what that tutorial found, and IME, is not the norm. Most other video cards that I've tried get much less than that. I'd love some kind of stats that show what the average performance difference is.
Additionally, video-card drivers are still buggy and fragile. And it's not just video cards, but wifi and sound as well.
People suggesting dual-boot make me think they only switch on their PC for maximum of 1 hour a day and have no idea what it's like to have dozens of programs started (and you needing all of them) and what kind of inconvenience is to start them all manually or wait your OS 2-3 minutes to start most of them automatically.
You also probably don't realize that many of us stream video and music 24/7 and your main PC is de facto a programming + gaming + home entertainment server.
> have no idea what it's like to have dozens of programs started (and you needing all of them) and what kind of inconvenience is to start them all manually
You can hibernate one OS and still start the other one. This way all your apps will reopen the next time you boot the OS.
> You also probably don't realize that many of us stream video and music 24/7 and your main PC is de facto a programming + gaming + home entertainment server.
You're right, most people will have to use Windows at some point :(
It helps a lot if you encrypt the other OS's drive and use Windows as little as possible. If you have a computer with an IOMMU and a virtualization-friendly graphics card you can do even better.
yes, it helps by sending a signal:
it gives to microsoft the information that you are only using Win10 for playing. And 'something else' the rest of the time.
They run Windows in a virtual machine, and pass a dedicated GPU into the VM, the results are very impressive, but the installation is very complicated, and you won't get rid of Windows.
As VFIO matures, and the API stabilizes a bit more, I think we will see distros offering an easy "install and play" experience.
I've tried many times to run linux, it usually ends 3-4 hours later in frustration. I tend to run older hardware so a lot of times support isn't there. I will admit that it's getting noticeably better (or I'm getting luckier).
After using exclusively linux at home during a couple of years, I have bought a laptop to my wife. I wanted to experience windows 8 (new at that time). It was a complete week of frustration. Almost no application are present by default, the video player complains about all file formats, installation of visual studio express succeeded only at the third try for no obvious reason. The ergonomy was a nightmare (it became usable after discovering many keyboard shortcuts that existed already in win7).
Linux is fast to install (when no problem occurs). Software installation (and deinstallation) is a breeze. Everything is available out of the box. I maintain a diary with a list of non default packages to install:
apt-get install nfs-common wine gnome-media openssh-server squid curl filezilla curlftpfs ethtool docker.io xtightvncviewer x11vnc tmux whois git wakeonlan youtube-dl npm sshfs
Concerning support for old hardware, it was the main reason that made me switch to linux: vista64 had no support for my printer and my scanner.
Probably 2007-2010 range, I still have my budget desktop that I put together in 2008 if that gives you an idea. These were never top of the line machines so they don't get much love.
I usually just went with Ubuntu because at the time it seemed like it was user friendly. I tried openSUSE at one point. If you can name a distro better aimed at supporting these types of devices I'm game.
> Frankly anyone posting on a tech site that can't get it up and running adequately should be embarassed.
And this is a big reason why so many people still don't want to use linux. The "Why aren't you as smart as I am" attitude isn't as strong as it used to be, but it's still there. It turns off users. Nobody wants to use software when the other users are going to treat them like idiots. There are still many complicated problems to get linux up and running, and being a jerk ignores those problems, rather than helps solve them.
I'm not asking why people aren't as smart as I am. I'm not even trying to be controversial here. This is not a "user" we're talking about, this is not beginners help on reddit somewhere, this is HN.
If you can't install linux on some reasonably generic pc hardware in this day and age, and you consider yourself a hacker or coder, you probably ought to be a little embarrassed. Certainly not proud as some seem to be.
If I was interviewing someone for a role and asked them for their thougts on using linux and got "well I usually give up after a few hours trying to install it" I'd look elsewhere.
You are missing something very important here. I can setup almost every Linux I can think of for desktop / everyday usage. My relevant question is: "Is it worth my time?", and most of the time, it's not.
Look, it's fine if you breathe Linux in your free time and you don't watch movies or play games, or nail your girlfriend, or even watch the ceiling in half-meditative state with all lights out. That's not sarcasm. Everyone has the right to do what they want with their free time.
But many of us are hackers / coders AND have a happy personal life and don't want to bring their work to their free time. And want to make the best use of their work time, too.
Linux gets the job done and I very much like my XFCE 4.x on my virtual Debian, but it's still far away from a full-blown user OS on the desktop. Numerous articles agree with this fact.
Hey, I never said you have to like it, never claimed it would meet your needs, never claimed you should convert, never claimed it was better than windows, never claimed you need to spend your free time doing anything.
It's a simple comment - if you're working in tech in any capacity and you can't install linux satisfactorily, that really is nothing to be proud of in the modern day and age.
> I'm not asking why people aren't as smart as I am.
Not directly, but you are being condescending to someone who has legitimate issues. Just because someone is on HN doesn't mean they are great with linux, or that they are even a technical person. Maybe he's got a different skillset, and doesn't like dicking around with shitty config files, obscure error messages, and jerks that tell them they should be "embarrassed" for not being omniscient.
Can you tell me the best way to pull HE levels from a GE MRI's MM? If not, are you going to spend a lot of time looking it up just because? Why aren't you embarrassed that you don't know that?
> If I was interviewing someone for a role and asked them for their thougts on using linux and got "well I usually give up after a few hours trying to install it" I'd look elsewhere.
Depends on the role. If it was a linux-admin job, absolutely. If it was for a job in HR, then I wouldn't care at all.
Actually I would say that being on HN does mean that someone had a claim to be a technical person or in the tech industry, and that being able to install linux is now so easy and so common that it's more like the fizz-buzz test - if you can't do that we shouldn't even be talking.
We just want you to read the manual - we did. Summarizing knowledge to people that can't be bothered to read the manual is not a good use of my time. Please don't take that as mean spirited, it's just a fact.
"Giving back" to the open source world in my case is done by committing code and documentation, not free support. So when you hear "go read the manual" or RTFM - we aren't insulting you or being jerks. We genuinely want you to better yourself and acquire education, and are pointing you in the right direction to do so.
If that's the attitude you have (which I don't think is necessarily wrong), than that's fine. But people aren't going to adopt linux if that's the case. Frankly: The manual sucks. If mass adoption isn't your goal, then great, nothing more to discuss here. But the point of the conversation is comparing linux to win10. I've never used win10, but from what I understand it doesn't require you spend hours pouring over man pages, so there's a lot of friction to transition from win10 to linux.
The only thing embarrassing here is this comment. I can get linux "some sort of running" but that's frankly not good enough for me. I need both proper screen resolution AND audio working. I need logins without errors and networking that actually works. These are my experiences and no amount of an attempt to shame me will get rid of them.
Linux isn't for everyone. It's not that us happy-linux-users don't have any of the problems you describe, but rather that we are capable of fixing them and preventing their future appearance. If that doesn't sound like something you would enjoy or be willing to do - then it might not be for you.
I have had no issue with any of that for years now. Unless you have some really exotic hardware, it's hard to empathise here.
Linux is an industry standard, I think this sort of brag about how much trouble you have with just getting it running reflects much worse on the complainant than the system.
If anyone is bragging here, it's you, so no need to be a hero. If you're hearing this a lot, maybe it's not us and instead it's people like you. The whole point of a good UI is to make it easier for people that don't regularly use your system. So if you're hearing this a lot then maybe it really isn't good.
> Linux is an industry standard
On servers it is, on desktops Windows still rules and that's for a reason. Common users don't need to hit up a Reddit beginner forum or spend hours pouring over documentation to learn the command line and all the various utilities for their chosen distro.
On the server, yes it's rock solid and I've had a pretty good experience with it. On the desktop, no. Also, my experiences are not from 20 years back, try 2.
I see it on HN every so often, and I see it used as a way to say "linux is bad and complicated".
Installing it is not complicated, with a live CD it's arguably easier than windows these days. It reflects badly on you as a person working in tech if you can't do it.
Linux's greatest strength, freedom of choice, is also it's biggest weakness. Instead of a clear vision and focus, you have a million developers getting frustrated with half baked implementations and rolling their own half baked implementations. Because of this there's a constant churn and things can change radically from one version of a distro to the next.
A user can live on the same LTS version of a distro for years but eventually they'll need to upgrade. I have yet to see anyone successfully upgrade major releases of any distro without serious issues. The process is basically format and start over. That's unacceptable.
Sure Windows and macOS have their own set of problems but for the average user they just work. You don't have to dive into the bowels of the registry or modify ktext files to get most things to work and the majority of users don't even know about the CLI.
While I agree with you (for the most part), I do want to say that Debian is king when it comes to upgrading across major releases.
I've been using Debian since v1.3 and have upgraded I-don't-know-how-many machines across major versions. I've got production servers running Debian 8 right this moment that were running Debian 6 when they were installed.
With regard to RHEL, I always reinstall from scratch instead of attempting upgrades. I think they're a lot better about upgrades nowadays but I still remember, very clearly, the day when attempting an upgrade was just asking for trouble.
My feeling is that a lot of people around these parts use Linux as their desktop environment and think that just because it works for them, then it will work for everyone else. What they fail to realize is that if they're reading HN comments then they're probably above average when it comes to computer prowess and they're not reliant on someone else to support their computer needs.
I've been using Linux in various forms for over 15 years in all forms including daily driver desktop, daily driver laptop, HTPC, foisting it upon my children via netbooks, custom router firmware, Raspberry Pi development, custom ROMs on cellphones, home server hosting, cloud hosting, NAS appliances, development VMs, and production VMs. I tried use the desktop environments really hard for about a decade and constantly founding myself wondering why certain tasks had to be so hard. I always chalked it up to inexperience or unfamiliarity and pushed forward but then my job required me to use OSX. I quickly realized that it was the Linux desktop envs and not me. It was also around this time my kids started using computers more heavily for school and the Linux netbooks I provided them just couldn't cut it. Surprisingly they were able to use Nexus 7 tablets for their school work with relative ease.
I still use Linux daily but never on the desktop. CLI or nothing.
That's precisely my experience and preference -- CLI and servers (all the kinds you enumerated although I didn't do some of them) and please let's just stop there.
Linux desktop is just not there in terms of "set it up in 15 minutes and forget it unless you want to heavily modify it". Of all OS-es Linux has the biggest potential to be the ultimate desktop system but alas, not just yet.
they sure are: difference in e.g. desktop usability with like 15 years ago is huge. You can finally install something from cd and have it up and running including audio/network/automatically mounting pendrives/... without having to touch a config file. Usually. But the problem seems to be in the getting better: I'd rather just have it good enough already right now. Too often it does, for me, as a desktop, still not feel quite right yet and still too buggy. 15 years ago a lot of things Windows just worked and now they still do and a bunch of nice stuff was added (ok crap was added as well) and for what I do with it I have close to 0 problems. I just wish I could say that everytime I try any of the distros you mention.
It doesn’t really matter if off-switches exist, the problem is that the features are there to begin with.
In any piece of code, a switch is a point of added complexity. For ANY such toggle:
- The switch might not be saved correctly.
- There can be a regression where the switch stops working in the future.
- The switch, despite being “saved” correctly and displayed by a reassuring checkbox in a GUI panel, might not actually be CONSULTED in all the places it needs to be consulted (resulting in default-on, default-off or “whatever the developer of that component felt like” in various components across the system).
- There can be a regression in any one of the components consulting the switch in the future, leading to an inconsistent combination of things that may or may not check for this setting over time.
When they give over 100 options, I assume 50 of them don’t work and that they have no real incentive to make sure the other 50 keep working.
About the only thing you can trust is a single on/off switch for the whole thing, while simultaneously checking a bunch of low-level things (regularly blocking unwanted hosts, logging network activity, etc.).
TL;DR:
Fix Windows 10 Privacy can be used via a GUI or the
commandline. Right now it implements about 130 rules,
which keep Windows 10 in check, regarding data protection.
And:
OneDrive Cloud users or users of others of the above mentioned features should refrain from using this tool at this time, because the functionality of these services will be limited or disabled after running it.
This post begs the question, "how do these registry changes go above and beyond compared to the system settings?" I have a Windows 10 machine and the first thing I did was carefully pick through all of the menus and disable everything I didn't like the look of privacy-wise. I would be much more inclined to download and use this tool if I knew there was more to disable.
"Do you not want to disable spy feature X? Yes means no and no means yes. YES/NO"
The wording on these options is deliberately misleading and there are way too many of them in different places. Microsoft know full well that everyone would just click "NO" when asked whether to activate all these telemetry tools in simple language.
> Microsoft know full well that everyone would just click "NO" when asked whether to activate all these telemetry tools in simple language.
That's not true. I am more than happy to share anonymous usage and telemetry data with software providers, assuming that it's clear what I am sending, that it's anonymous, and that I can easily change my decision later.
Simple language makes me far more likely to give you as much anonymous data as you want. If I can't tell what I'll be sharing, I won't be sharing anything.
You're right about the anti-patterns and I don't like that they're referred to as spying.
In any case Microsoft has been addressing some of them. My father in law was reinstalling Windows 10 and I was guiding him through it because I remembered a couple of these trick questions(e.g. Cortana). I noticed a couple places where they fixed it.
It starts with the 'customize' button being almost illegibly small, hidden and not even looking like a button, compared to the 'use express settings' one.
I couldn't imagine having to fight paid software to not invade privacy like that some time ago. Things have gotten out of hand, it's a sad time. How did it come to this?
It's been like this for ages, it's like that stain in the ceiling you never noticed before.
Oh, and if you start digging you'll get even more angry. Things you trust without thinking twice (not really but you get the idea) is also compromised or doesn't have a practical alternative: hardware, TPM, certificates, your smartphone hardware and software, ∞
Side note:
The Author has invented a BSD 2-Clause license, making it a 2+1 non-BSD license, by adding:
NON-MILITARY-USAGE CLAUSE
Redistribution and use in source and binary form for military use and
military research is not permitted. Infringement of these clauses may
result in publishing the source code of the utilizing applications and
libraries to the public. As this software is developed, tested and
reviewed by international volunteers, this clause shall not be refused
due to the matter of national security concerns.
This is disappointing, because some of us do work as contractors, or a civilian employees for military agencies. This is the same issue with Mosh (which uses a patented library with a restriction on military use).
Microsoft officially ended OEM sales of Windows 7 with new Skylake hardware on Oct 31, but there are still some WIndows 7 OEM devices left in retail channels. Some effort is still needed to remove telemetry updates from Windows 7, but you get the benefit of compatibility with existing apps and very few global ongoing changes to your work environment. If you buy a new Skylake device now with Windows 7, you will receive security updates through 2019.
Recently (just within the last couple of days), I've been thinking about setting up a Windows 7 Enterprise VM for occasional use. Am I correct in understanding that this telemetry crap is going on in Windows 7 also!? I don't really keep up-to-date on it (I use only OS X and Linux on the desktop) but I was under the impression that only Windows 8 and Windows 10 were affected by this telemetry ("spying", IMO) issue?
Telemetry was aadded to Windows 7 around Aug 2015, https://news.ycombinator.com/item?id=10110155. With a bit of effort, most of the update KBs can be identified and manually removed. It's still a pain though, because there are many sites with conflicting claims about which updates can be removed safely. I also found at least one update that reinstalls itself after removal.
Since then, Microsoft has moved to monthly updates which combine security fixes + functional changes. If you want only security fixes, you need to disable automatic updates and download a separate security-only monthly rollup.
It's a shame, but we're approaching the point where we will need to whitelist all outbound traffic using a firewall that is external to the operating system. Anything that isn't whitelisted should be blocked, logged and audited. Someone should start a VPN service that blocks Microsoft, Google and Apple telemetry. If we had regulators, telemetry could be forced to use stable domain names to enable network filtering, and telemetry traffic would be unbundled from application traffic.
I thought maybe I wouldn't have to worry about the telemetry (read: "spying") too much if I stuck with Windows 7 Enterprise (although it wouldn't be on a domain), but apparently that's not entirely true.
I have media that has SP1 slipstreamed and saw that there was a "comprehensive"(?) update that Microsoft put out, reducing the need to downloads hundreds of updates. I'd have to double-check the date on it but I think it was recent enough to include some of the updates you mentioned.
I suppose I had gotten my hopes up that I could just install a volume licensed copy of Windows 7 Enterprise and "be okay". I should have known better. :/
That depends. Usually very little (single KB/s or less), but W10 is able to produce (write to disk) over 1GB/day of logs/traces/dumps that sits there and waits to be requested by the mothership.
Yet another "helper" that diminishes computer user's computing experience for the sake of supposed privacy, then just sets completely arbitrary policies that have nothing to do with transmitting information back to Microsoft.
This program will mess up most Windows 10 installations. In ways that may take you months or even years to ultimately notice.
Let's look at some of these rules:
- "[Disable] Let websites provide locally relevant content by accessing my language list." Meaning websites cannot provide a translated version that you'll want (e.g. if you visit a Chinese website, they may not provide an English translation since you aren't sending the language list).
- "[Disable] Let apps access/control my camera/location/contacts/microphone/etc." Breaks all apps that use the custom permissions (e.g. Skype). With it enabled they would still prompt you for per-app permission (camera, location, contacts, etc), with it disabled they're treated like you don't have a camera/microphone/etc at all... But only for modern apps, Win32 can still access the camera/contacts/etc. Effectively you're just breaking all modern apps on Win10.
- Turn back on outdated insecure Bitlocker encryption, turn on incompatible Bitlocker modes, and turn on Bitlocker modes that only exist in higher versions (e.g. enterprise edition). Uhh, k? Why are they dicking around with Bitlocker policies?
- Encrypt the page file (Even on Bitlocker enabled systems?).
- Disable auto-complete, password manager, and other useful browser functionality in Edge & IE. Also clear browser history upon exit (i.e. break browser history).
- Actually disable OneDrive via GPO (i.e. don't just limit it, kill it).
- Break automatic web proxy configuration (may kick certain people offline depending on network setup).
Just go look in the source code. This is amature hour. Someone's just gone through GPO, set a bunch of stuff without understanding what it did, noted down the corresponding registry changes and built this wrapper around it. But they never understood the GPO policies to begin with! Disabling the IPv6 helper may also kick people offline (even if it is very niche, what does that have to do with "privacy," what do 80% of these changes have to do with "privacy?").
Any amount of service toggling and hosts file stuffing will not suffice. It just screams ignorance. As a software developer you should understand that plugging holes in a black box is a futile effort. All these tools are doing is giving a false sense of privacy, that the next update will undo by flipping a switch or installing a new service.
If you think the OS is violating your privacy, stop using it or remove it from the Internet. Or both. It's the only way.
Edited to add: If you actually like Windows (I do), just switch to the Enterprise Edition and dial Telemetry down to "Security". Here is an explanation of what little is then shared, and how to even further minimize your footprint: https://technet.microsoft.com/en-us/itpro/windows/manage/con...
Edit to address the availability of the Enterprise Edition: If you are not able to get it via your $JOB, a valid key from MSDN surplus shouldn't be more than $50 if you look around. Of course you'd then be bending the EULA in your favor, but hey, since Microsoft is spying on everyone against their will I think it is fair game, right?