Recall going back into Windows
arstechnica.com
arstechnica.com
Looks like this finally is the year of the Linux Desktop!
If you're technically minded, and are at least somewhat familiar with Linux, I can't recommend it enough. I wouldn't recommend it for a layman though; I did have to do some initial tinkering to get it spot on.
The flip side of this is that regular computer users don't actually have preferences nearly as strong as anyone browsing this site.
If one is technical enough to have an operating system preference, they're technical enough to manage Linux Mint. It may not be their preference, but they'll be able to manage
As always, the only groups that're really in trouble are "Knows just enough to be a danger to themselves and is entirely unwilling to learn something new", and those that depend on poorly supported, or unsupported, specialty hardware or software
I’m sure there’s a KDE fan writing up a reply right now, and while it’s also Windows-like, it’s considerably more indiosyncratic than Cinnamon is and has a bunch of bells and whistles that while great for power users have a decent chance of tripping up novices. Cinnamon doesn’t rock the boat at all which is exactly what makes it appealing here.
The dual ssd boot option for OS is also nice, as some games/applications in Win11 are inescapable.
However, for 99.8% of most day to day OS needs the Ubuntu repo desktop works fine (if you purchased a printer/webcam knowing it is fully supported etc.) =3
So, if someone wants a Linux to start with, installing Debian Stable, and selecting the Cinnamon desktop in the installer, is one great first option.
(The Debian default desktop is based on Gnome3, which does a few weird and annoying things. Fortunately, Gnome3 inspired alternatives, like Cinnamon.)
Similar situation, I'm thinking I'll have a two bootable drives and then a third data-only drive so that I can share some stuff between them. (Hey, I got a desktop for a reason, let's use that space!) AFAICT this should be safe-enough provided I avoid a special case where Windows has hibernated without unmounting an NTFS drive cleanly and I try to mount it on the Linux side.
Not sure how much work I want to put into the different flavors of TPM/secure-boot-y things. Less worried about evil-maid attacks as opposed to preventing a burglar rifle through my digital life.
Or in Win shell try:
winget install -h -e --id "WinFsp.WinFsp" && winget install -h -e --id "SSHFS-Win.SSHFS-Win"
A backup ssh server running F2FS on disk-encrypt for /home is good practice. =3
There is always a better way, but integrated backups are important. =3
If you're up for getting your hands dirty though, it's a gift that keeps giving.
Windows can't help screwing with users, and has been an IT security liability for years. When Microsoft abandoned backward compatibility for much of their installed base in Win11... they also removed 99.9% of the reason IT puts up with their BS. =3
Best to check your favorites are functional before nuking Windows, or do the dual ssd OS boot option. =)
It is a lot less of a pain for Linux users these days, but still YMMV =3
The only games that I do play regularly that work are Counter Strike 2 and DotA. Though I can't use Faceit for CS2 which would be ideal.
I think we need to accurately represent the shortcomings so people who switch aren’t surprised.
So far those are:
1. Laptop - Battery life is bad compared to windows. It’s about half.
2. Laptop - sleep doesn’t work.
3. All - multi-monitor setup with different pixel scaling doesn’t work for many applications.. unless you dig into all the Wayland options and issues and figure out how to launch all these apps under Wayland.
4. All - In general Wayland vs X issues. I can’t screen share with zoom.
5. All - Bluetooth driver issues - my Bluetooth headset won’t connect as an audio input and output device at the same time.
Now to be fair, I think all these are okay trade offs but they are a conscious choice. If you have anything outside a standard one monitor, wired peripherals setup you will probably hit issues you need to debug.I started paying for Ubuntu pro to put my money into it, so I’m hopeful for these kinds of things in the long term.
The fact it works for you while not working for someone else is actually worse than if it didn't work for anyone.
I absolutely agree, linux advocates must be honest about the shortcomings. In my case even on the thinkpad I experience the multi display scaling issue you mentioned, and bluetooth can be a little finnicky for my headphone (though this is much better than a couple years back! Usually simoly restarting the headphone solves everything).
I think it's very much worth it, and other than some of those minor issues I think current linux distributions are good enough to wholeheartedly recommend them over windows. That is if you're not held hostage by some windows only software.
E: about screensharing, I can't screenshare from teams on firefox, but from chrome it works fine, maybe that's the same for zoom?
That, and don’t expect brand new hardware to work well unless you’re willing to deal with a cutting edge distribution and all the trouble those can bring. One gen back from current is usually enough of a lead time for things to catch up.
Simply facts that are true.
There are problems on Windows too, but they are not these problems, and the problems I mostly have are only problems I have and not problems the usual Windows user has.
The normal windows user doesn't even try to login without a microsoft account, or even try to remove cortana/bing/copilot/whatever-this-week, remove edge, prevent the "HP Smart" driver bundle that installs for every HP printer or scanner these days and find the old style drivers without all the cloud shit, etc.
But I have not found scaling to be especially good on windows either, even with a simple single monitor. My mother in law can't run viber in her desktop because the app scales so bizarrely that some buttons are moved under other things or out of the window or even off the screen, but on top of that, the active areawhere a click is registered does not overlay where the buttons are displayed on screen. Maybe it's just an especially crappy app but she only uses like 3 things and two of those are firefox and libreoffice (which are because I set them up of coursae she never asked for that).
Fonts look ridiculously comically bad in browsers for some reason.
And of course the ads and notifications and onedrive nagging...
Thanks for understanding the spirit of my point about the shortcomings above and I really like the way you phrased the "Windows has its issues as well, they're just different ones" - and I completely agree there.
With Windows you need to navigate the Microsoft account, files getting stored in OneDrive, updates happening outside your control (arguably a good thing for most users), and more that I'm sure I'm not thinking of.
I do think the Windows issues are more abstract like security, privacy, and default on features - while the Linux ones tend to be more in my face usability ones. Again agreeing that choosing your hardware and desk/laptop setup can alleviate many of things. But that requires knowing ahead of time and people switching in reaction to something Windows is doing don't get that benefit.
I guess I'm writing all this because the idea of a Linux distribution working perfectly on most/all laptops really excites me and I think being candid about the shortcomings yet providing support to the distributions is how we can get ace these fit and finish issues.
Food for thought for anyone else reading this - the end goal of Linux for everyone is why I don't get too worked up about snaps. If they get to a point where I can tell my mom she can safely install apps X, Y, and Z by pointing and clicking in the app center it's a great computing future.
I wish X supported mixed DPI per monitor, ugh.
I will say one notable difference is that Linux issues as a rule at least are debuggable, whereas Windows issues can just be utterly intractible. It's not that rare for me to watch friends with computer science degrees frustratedly embark on the long misadventure that is "reinstalling Windows".
sudo cpupower frequency-info
# or
cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_available_governors
cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_available_frequencies
Also take a look at powertop you will probably have to install this. One can set any devices they are not using to optimal settings. I avoid touching USB used by keyboard/mice and network interfaces I am using to minimize lag. Powertop can output to a file and that can be used in a startup script to automate the optimizations one has chosen.There is also a sysctl setting called "vm.laptop_mode" which defaults to 0. On a laptop it can be set to 5 to combine writes and minimize storage wake-up. The caveat is that if the OS crashes one can lose up to 10 minutes of work. Most developers should avoid this setting unless their code editor autosaves frequently and syncs / flushes storage write caches. If unsure don't use it.
Another small gain is to ensure all daemons, desktop services and widgets not required are disabled or even removed. Some of them are power-hogs, some especially more than others. Powertop can sometimes expose this if left running for a while.
Another small gain can sometimes be installing "tlp" but different laptops and usage will see different amounts of power saving.
Oh and keeping the laptop off the lap can sometimes save power. More heat means more fan usage and thus more power usage. When at a dedicated desk using a laptop cooling stand multiple fans can extend battery life.
If one is feeling very adventurous they can install the latest bleeding edge kernel to net some small power savings but it may not be worth it if the laptop is used for anything critical.
long time Ubuntu user here, had been bulletproof on an i5 Panasonic toughbook until 24.04 and now it’s not so stable. Sleep also stopped working correctly on an i5 Lenovo yoga and I downgraded that one back to 22.04.
However that same distro runs smoothly (and the UI isn’t constantly glitching out) on an i7 thinkpad that I don’t enjoy using because it runs red hot and the fan is always going…. FWIW that’s also the only system I have that’s even capable of running win11 smoothly… but up until now, Linux was great on castaways that windows had forgotten.
I have acpi and charging issues on the stock 24.04 kernel tree with the Panasonic , which is a laptop that supports two batteries. If either battery gets pulled on that platform it stops charging on AC.
This issue isn’t present after putting Ubuntu packages for kernel 6.14 on it , which only came out two weeks ago.
It still wanders all over the place as far as whether I can get 8 hours on a charge (or two hours), swapping the batteries confuses the system still and I haven’t had the free time recently to nail down whether this is acpi, kernel, or Ubuntu specifically. I’ve mumbled a little bit about that one on launchpad and ordered a second battery for a different laptop that has that capability but don’t have answers yet.
Would need to know your Bluetooth chipset to speculate too much because some bleeding combo cards with wifi6 are also better supported by recent kernels. For example my Intel BE200 worked fine for WiFi but the Bluetooth didn’t work at all until either 24.04 or applying 6.14 to it. Not sure which, I just noticed it was there in the menu about a week ago.
with that said my laptop still has a resource conflict I haven’t pinned down where, when WiFi and wwan card are both powered on and active my WiFi speed is clipped down to about 2mb/s. I’m just powering the wwan off when I don’t need it and I’m inclined to think it’s still a driver issue or the two cards don’t get along or are conflicting for resources somehow… I don’t have a solid enough theory to report it as a “bug” or know for sure whether it’s just my hardware yet.
Ubuntu and Wayland were the first distro where I went “hey, using Linux on the desktop finally isn’t *ss” so I’ll give them that. But 24.04 has been the one that had me wondering if it’s time to get acquainted with another. Many are mentioned ITT, I just haven’t “distro hopped” and “tried them all” in almost two decades and it may be time again.
If only we'd had the foresight to say "client" instead of "desktop" we could have said we got there with Android. But, alas, in the 90s the mobile device in waiting was a wristwatch.
On Thursday, the company said it was reintroducing Recall. It currently is available only to insiders with access to the Windows 11 Build 26100.3902 preview version.
Some of those insiders with preview will also be active contributors (200+ contributors, 6 million users) to the Chris Titus Winutil project, an open source PowerShell GUI collection of open scripts and registry toggles to streamline common installs, sanity restrict updates, control windows features (such as Recall), access classic control panels, and custom build your own tweaked install ISOs and clean VM images.Good for gamers, developers, custom minimal window images, etc.
( "The best thing you can do for a gamer is strip everything non-essential from the OS" .. which is exactly what winutil streamlines and simplifies - along with a wealth of other user chooses options. The MicroWin tab allows a user to create their own minimal stripped out install ISO )
See: https://github.com/ChrisTitusTech/winutil
Improved for 2024 (18 mins): https://www.youtube.com/watch?v=5_AaHXrelTE
Windows Utility in 2025 (7 mins) (Addresses upcoming Recall V2 release): https://www.youtube.com/watch?v=IuaNw8Tpn7Q
The best thing you can do for a gamer is strip everything non-essential from the OS and let us enjoy the damn game. Windows used to be a great gaming platform but now I run nearly everything in ZorinOS except for Tarkov which has it's own hardware sandboxed stripped down Win11 drive.
2003, https://en.wikipedia.org/wiki/DARPA_LifeLog
>The objective of the LifeLog concept was "to be able to trace the 'threads' of an individual's life in terms of events, states, and relationships", and it has the ability to "take in all of a subject's experience, from phone numbers dialed and e-mail messages viewed to every breath taken, step made and place gone".
2007 Microsoft Research, https://www.microsoft.com/en-us/research/video/the-microsoft...
> The SenseCam is a personal, wearable camera developed by Microsoft Research in Cambridge, UK, and used as a lifelogging device in projects like MyLifeBits.. based on wearing the SenseCam for lifelogging of ‘events’ during your day, and generating a fast-forward movie of the event as the memory recall interface.
https://www.microsoft.com/en-us/research/project/mylifebits/ & https://en.wikipedia.org/wiki/MyLifeBits
> MyLifeBits is a life-logging experiment begun in 2001. It is a Microsoft Research project inspired by Vannevar Bush's hypothetical Memex computer system.. The "experimental subject" of the project is computer scientist Gordon Bell.. For this, Bell has digitized all documents he has read or produced, CDs, emails, and so on. He continues to do so, gathering web pages browsed, phone and instant messaging conversations and the like more or less automatically. The book Total Recall describes the vision and implications for a personal, lifetime e-memory for recall, work, health, education, and immortality.
Lifelogging has been referenced by 10,000+ academic papers over two decades, https://scholar.google.com/scholar?q=lifelogging
[1] - https://www.youtube.com/watch?v=a3Xxi0b9trY [video][44 mins][documentary]
Also - RIP Teddy K. (I'm a bastard that way)
Honestly it will probably help some people "hey where did I put that file?" but yes at a tremendous cost to privacy and security for those who use it.
I am also weary that it is opt in "for now".
Not only Microsoft's. I think Google were the first who introduced this crap. Now everybody is using it because users don't matter.
I think the UX of the future is: Please enter your card number and let us take what we want.
Keep in mind that anyone you email, chat, video conference, share files, or otherwise electronically interact with that has a Windows 11 machine with Recall will automatically opt you and your communications in as well, and you can not prevent it.
Not a bunch of independent, disorganized one-offs.
Consider this: your boss or another senior person at work has Recall and brings up information related to your employment, performance, and compensation. A friend has Recall, and reads some email from you in which you confide something sensitive. That customer support agent at the company you do frequent business with has Recall, because their employer wants to monitor their minute-to-minute activities, and brings up your account history.
Independently and uncoordinated, maybe not so bad. Together, tied back to Microsoft, it's easy to imagine constructing a profile of you from your interactions with these seemingly independent but in really collective third parties.
It would have been better if they'd done nothing. I've switched to mac and linux.
Additionally, they managed to refactor Outlook to Electron/webapp, which allows them to share the codebase between Windows/macOS/web, and deliver unified experience ... which might be a downgrade for Windows/macOS users.
A couple of months ago I opened the Clock app in an important meeting, to run a timer, and had to wait 10 minutes for it to update itself. For real. Same thing happened when I opened sound recorder.
It's probably not going to get better at this point. It's only going to have increasing amounts of online services integrated into the OS. The writing was on the wall when Windows 10 Enterprise shipped with Candy Crush
How is that better?
Or fixing bugs...
On the one hand, MS Research seems to have a long history with lifelogging. On the other, MS aren’t capable of implementing decent search in Outlook; I very much doubt their ability to do the same in this more complicated setting.
I don’t think Windows has any privacy-conscious users at this point (outside corp contracts where this will be limited), and they generally seem to be doing everything they can to get out of the home market, so the moment is chosen well.
Yes, it is useful, but not for the user, for the CIA.
I got a Windows 10 PC that takes 30 minutes to boot up, and bought a $2000 Windows 11 PC from Microcenter to replace it. I will install Debian Linux on the Windows 10 PC when Windows 10 expires in October 2025.
Hopefully the combined power of HN can figure out how to not enable this opt-in feature.
Feature toggles like this can be enabled on a single OS update, a single dial-home to the mother-ship. If the code is there I assume it is running. The only winning move is not to play.
So ... a bit off topic, but I want the opposite from OS vendors. The OS installation should take at most 3 minutes. Any features or geese that lay golden eggs that I desire should be a checkbox in a list of optional services. If I do not select anything the OS should not be bigger than 3GB for a desktop or 1GB for a server and even that is quite a bit bloated. If code is discovered on a machine that the administrator or owner did not intentionally install that needs to be treated as a security incident meaning "all hands on deck" and "all lawyers on a conference call with the OS vendor".
As a bit of a tangent, done right I should be able to click a button in Ansible and completely re-install the OS to a fleet of 100K bare metal servers in less than 10 minutes across 4 data-centers per region not counting POST time. Once the database OS is up their tamper-proof backups automatically restore, app servers grab their code and in 30 minutes every data-center is up and running. The first people in should be the customers and they might ask if we had a network blip.
That comment was for CISA. Make the above step 1 in chaos monkey tests for all corporations and governments and make live testing required by law to be publicly traded. This testing must be performed annually in a staging environment in every data-center to remain listed on the stock market or remain funded in government budgets. If code exists on a machine that need not be there it's a security incident. Let's not have another Sony/Tri-Star.
This is such a bizarre and dishonest mindset.
Windows also has code to allow automatic login without a password. Do you assume because the feature exists, it’s going to be enabled after an OS update?
What about the code to enable remote access for support purposes? That exists for a while, but people aren’t freaking out about it existing.
If it's not needed it's gotta go especially if it can talk to the network or modify data/code that can talk to the network. I will always assume it is or was running, period full stop.
I used to use it as a convenient way to find websites that I had read something about in the recent past but couldn't remember exactly where.
I mainly stopped using it because it was causing performance problems.
I have some improvements in the pipeline for lower capture intervals though.
You think shadowy governments are coercing Microsoft into adding an opt-in AI feature?
They could be, yes. Once the code base is there then it is just a feature toggle at that point. One dial home to the mothership means one could be toggled on, toggled off as desired. As for shadowy I think we've seen they need not hide in the shadows nor do they even try.
Several computers that ship from Amazon have malware pre-installed. I found a case where it was installed but not running. To nobody's surprise I format any disk I receive before I actually put it to real use but many people do not. Malware being preinstalled and Defender being pre-configured to not alert means the malware can simply be activated by a dropper for any C&C agents doubling as an application updater. "We pinky promise it's just to update the RGB LED's".
This code should be installed like any other optional code. Ask the person if they want it installed and if they click yes then install it in the background and let them know when it is ready to use.
Corporate customers demanded it for employee "monitoring and metrics".
https://www.theverge.com/2024/7/2/24190730/google-ai-pixel-p...
First off, the title is editorializing users' reaction to the feature. Apparently even the 99% of Windows users who have never heard of Windows Recall (a preview feature not even available yet and not even compatible with most people's existing PC hardware) are "groaning" over it. But I'm sure if I sit here and argue that people will like the feature it would be a lame argument, and I might even agree with you.
Still, the other problem with the article is that it brings up factually challenging issues with the feature:
> First, even if User A never opts in to Recall, they have no control over the setting on the machines of Users B through Z.
This isn't a problem unique to an AI feature. If you send intimate photos or passwords to someone else in the pre-AI world that was still a big security problem. Windows Recall does not meaningfully expand the risk of those actions.
> That level of detailed archival material will undoubtedly be subject to subpoena by lawyers and governments.
If a government or a subpoena gains access to your system, the presence of an AI categorization system is not going to make any meaningful difference to their ability to extract data stored on it. Those types of actors have their own analysis tools that probably won't even bother with Recall as they are almost certainly more sophisticated. Not only that, you can say the same things about standard backups: anyone taking a standard not-AI-enabled backup like Time Machine or Windows Backup would have the exact same problem of having previously deleted data available for search and seizure. The typical person's cloud data exposure is arguably much worse than Recall since it's physically stored on someone else's server, a company who may comply with subpoenas where you might choose not to, which Recall snapshots only exist on your local machine.
But, okay, let's say you're still not on board with my arguments. I think the last argument I have to make is the biggest problem of this article, which is that it's operating under the base assumption that Microsoft hasn't thought of any of this.
> Threat actors who manage to get their spyware installed on a device will no longer have to scour it for the most sensitive data stored there. Instead they will mine Recall just as they do browser databases storing passwords now.
The article is just assuming that Microsoft has done nothing to secure this especially since the last time they pulled the feature.
But Microsoft has a very detailed page that gets into how this all works:
https://support.microsoft.com/en-us/windows/privacy-and-cont...
This article points out:
1. Processing and data storage is local
2. The user has control full control over all the settings and they're secured (cannot be changed without authenticating with Windows Hello)
3. There is a huge list of exceptions where content is excluded from snapshots (e.g., incognito browsing, passwords, etc).
4. Saving, deleting, and all features related to the
> No internet or cloud connections are required or used to save and analyze snapshots. Snapshots and associated data are stored locally on the device. Recall does not share snapshots or associated data with Microsoft or third parties, nor is it shared between different Windows users on the same device. Windows will ask for your permission before saving snapshots. You are always in control, and you can delete snapshots, pause or turn them off at any time. Any future options for the user to share data will require fully informed explicit action by the user.
But this article is just interested in producing drama and calling a new feature that actually has unique functionality and zero dependence on cloud "enshittification." Okay, if you say so!
Because Microsoft has a stellar reputation for flawless security, we can trust that they did the job right!!1!
A question that comes to mind is this: can we give machines intelligence and memory, without extending to them at least the rights of their owners or operators? For example, if my computer starts recognising me and storing episodic memory of past interactions, these memories should be considered as private as those of a person.
Let's separate the two aspects, usefulness and risks, and argue:
1) why is this not a useful feature
2) do you really expect computers in the future to remain dumb tools without episodic memory and contextual awareness
3) if you answered no to 2), then how do you imagine the privacy and surveillance concerns can be solved? Or all you care about is sticking to your old dumb calculator, muttering "over my dead body", while the rest of society moves on?
Contextual awareness: integrating the information you have with your understanding of what is happening.
Absolutely trivial examples of interactions requiring them:
"where did I put that file I was working on yesterday?"
"Can you reopen that web page I was reading last week about Egyptians, the one with the red background?"
"What is this notification, is it something important?"
"Delete the intermediate versions and keep only the final one"
Etc.
But the whole point of recall is to store and remember past information. Plus you have a lot more control of what you send to OpenAI with memory enabled, and you can even temporarily disable it. With recall on it's just taking random screenshots and you don't really know what it was all given.
I dunno, it doesn't feel like the same thing at all to me.
This Windows feature is opt-in, off by default, and not necessary at all.
Are all of these angry comments from people assuming it just comes enabled?