（评论）

（评论）
(comments)

原始链接: https://news.ycombinator.com/item?id=40610435

本文讨论了围绕微软产品 Recall 的担忧，据报道该产品未经同意泄露了用户数据。作者认为，这一事件凸显了非技术产品经理的影响力日益增长以及技术的易于获取。他们认为，召回是一个伪装成隐私问题的安全问题，这一事件的发生表明微软内部多个级别的领导层对潜在后果的疏忽或漠不关心。作者进一步将当前情况与过去的做法进行了比较，过去的做法是消费者在产品注册期间故意提供个人详细信息。他们认为，微软承诺在需要时存储并可能利用所有用户数据，这引起了人们的严重担忧，即使数据仍未上传。拟议的更改包括多因素身份验证和加密，以降低风险。此外，作者还分享了与 Microsoft 软件相关的挫败感，特别是在保存文件和浏览文件夹结构方面。他们认为旧的方法（例如使用 API 来定位常用文件夹）可能会解决这些问题，但承认用户通常不会选择此类解决方案。文本最后批评了微软缺乏对用户需求的考虑以及对实施变革的偏好。

Heres something I just don't get. Microsoft just got their ass handed to them by the US Government because os (lapses in) security. Extremely more than coincidentally, Satya Nadella told the entire Microsoft org that if anyone had to choose between features and security, to choose security. I'm hearing from Microsoft people that all product roadmaps are deferred for a few months while security features are addressed. Their whole corporate spiel is "Microsoft runs on trust" (see the famous standards of business training on youtube).

And then someone goes and invents Recall. This is not the work of a lone engineer and a principal PM fishing for Impact or whatever they call success at Microsoft. This had to have gone through multiple levels of review. Microsoft PMs, CVPs, their corpo legal people, marketing approval. And yet no one stopped to say, "wait, this could blow up in our faces"?

I'll take a shot (not at MS, so I have no inside info).

GenAI is hotter than anything else right now. As Satya publicly stated, "we made them dance" -- which shows how high-priority it is to maximize "AI innovation" at MS.

If you disagree, think about how badly "Tay" blew up in MS's face and yet they still went ahead and bundled OpenAI LLM tech into all of Office365, just so they could have bragging rights about beating Google to it.

At this point, it's a race (to where? who knows) and no Big Tech Corp wants to be seen as "not at the forefront".

That's my $0.02, anyway :)

You are absolutely right. This AI race has been sort of funny to watch among the big tech incumbents.

Google kept on launching faked demos and hurriedly released an openly race biased image generator all in a bid to catch up with OpenAI.

Meanwhile Apple has been sort of lethargic. Recently annouced a deal with OpenAI to add GPT to their devices. They seem happy to continue playing catch up in this regard.

I think Meta is probably the only big tech giant that has kind of got their execution right straight from the jump. Can't point to any slip ups from their "AI announcements".

The user data is just too juicy a potential profit center to pass up on. So many models to create, to theoretically replace many of us.

We should assume that everyone is on their worst behavior going forward. Zoom, Slack, MS, Apple, etc.

> Heres something I just don't get.

> And then someone goes and invents Recall

Maybe you should read about the history of Microsoft, especially about its security.

But people forget easily.

It's interesting to compare this to the Chrome/Safari/Edge browsing history, which is stored in an unencrypted SQLite database, and tracks what you do for the last 90 days. It's just a bit less visual, Incognito/Private modes work, and some users clear it more often.

But a whole lot of the surveillance attacks people imagine about Recall apply just the same to the browser. I think it's the "little brother" casual attacks that are so well enabled by Recall - it makes it faster, easier, and way more visual.

Your browsing history is unlikely to contain personal information, secrets, porn images etc. And if you use Chrome, they get your full browsing history by default.

I get your point, but Microsoft's Recall can capture anything onscreen - emails, personal info, porn, passwords and the like. And it feels, bizarrely for 2024, that little thought has gone into privacy or security.

> that little thought has gone into privacy or security.

I think the thought is proportional to the amount of thought a non-tech customer will put into it. Nobody seems to care about or understands privacy these days. Everyone knows they're being tracked everywhere they go physically and on the web. People use their real names, address, etc for every junk service they sign up for, without seeing any reason not to. If you tell people that their TV is tracking and taking screenshots of what they watch [1], they say "yeah, Netflix knows too".

It's literally, "how it's always been" for any non tech person under 30.

[1] https://themarkup.org/privacy/2023/12/12/your-smart-tv-knows...

> I think the thought is proportional to the amount of thought a non-tech customer will put into it.

Part of me wonders if this is the consequence of how accessible tech has become, and the prevalence of increasingly non-technical product managers. I'm a former PM, and I'm not here to denigrate the PM role, but the fact that a product like Recall got shipped says a lot about the makeup of the product org that shipped it.

While I get that younger people tend to see privacy differently, I'd argue this isn't really a privacy issue, it's a security conversation, albeit with obvious privacy implications. Leaking what apps I use or what sites I visit is mostly a privacy issue. Leaking what I type into the boxes on those sites is a security issue. If the end result of leaking this info is the attacker can pwn all of my bank accounts, we're solidly into security territory.

The fact that this got shipped means that multiple levels of leadership either didn't think about the consequences or didn't care about the consequences. I hope it's the former, because that means they can learn from the backlash and hopefully recalibrate.

Microsoft is in a position of power that IMO requires a significant duty of care and responsibility to their customers, and lapses like this need to be judged through that lens, i.e. it is their entire business to make sure features like this are safe.

> but the fact that a product like Recall got shipped says a lot about the makeup of the product org that shipped it.

Microsoft is just tripping over themselves right now bringing AI to market because they don't want to miss the boat. Their copilot for office 365 stuff is hardly working, it's real beta quality. No normal company would have released it to market in this state. But they're just terrified that Google will eat their lunch.

I don't think security and privacy concerns are much on the radar there anymore. They just want to establish their name in this new market at all costs. And I think in their eyes it makes sense, they've always succeeded because they had the biggest installed base, not because they were the best. It makes sense they see value in being first mover at all costs.

It's just a bit frustrating as a customer. As usual with something they launch it's more promise than substance. I have to say that usually they do have the follow-through to really make it a success. But it does take time.

I think it’s a good point - these are still privacy issues, and being fatigued with the impossibility of defending privacy is indication of a power imbalance, not an acceptable default for humanity.

> The fact that this got shipped means that multiple levels of leadership either didn't think about the consequences or didn't care about the consequences. I hope it's the former, because that means they can learn from the backlash and hopefully recalibrate.

There was probably from lower decks, where they are closer to reality. However, people are scared for their jobs in this economy and likely didn’t take it farther.

It's not surprising once you consider that all the big tech firms hire MBAs for their PM roles. The ideal PM profile for these companies is someone with consulting experience who just finished an MBA.

As an engineer with an MBA and in an executive level twchnokogy role, an MBA is part of what I would consider a good background. That said, OP was talking about non-technical PMs. That is the issue. You can teach a tech person to understand business, vision, strategy, finance, etc. You cannot teach very well the business person who has all that the intricacies of technology.

> ... an MBA is part of what I would consider a good background.

Emphasis on "part." I'm totally guessing on this, but I think OP may have been talking about PMs where the MBA is their only notable feature rather than those where the MBA is just one part of a well balanced whole.

> You can teach a tech person to understand business, vision, strategy, finance, etc. You cannot teach very well the business person who has all that the intricacies of technology.

I'm inclined to agree with this. The thing about business degrees is that they are so minimal effort that actually understanding business isn't a prerequisite to being awarded one. I would know, I have 2 of them.

There's no guarantee a "business person" actually learned business, much less that they are capable of learning tech. Don't get me wrong, I'm not by any means implying that all business people are inept or that they are collectively unable to learn tech; it's often unrealistic, but not impossible.

When going to school for tech, such as an MS in engineering, CS, etc., typically requires enough effort that one will end up learning their respective field (I have met exceptions to this and interacting with them is infuriating) whereas going to school for an MBA is one of the easiest ways I know of to get government financing for a decade of partying.

It's how it's always been, always.

Many here may be too young to remember when many consumer products came with a "product registration" card. This was basically a postcard that asked for all sorts of information, such as your name, address, phone number, birthdate, sex, SSN, marital status, annual income, interests, other products owned, whether you own or rent your home, etc.

People willingly filled these out and sent them in. All the info went into databases that were merged with other sources and traded around various marketing agencies on 9-track tape reels. Advertisers could get mailing lists segmented by age, sex, income level, geographical region or specific zip codes, etc. for their campaigns.

It's all much more pervasive and invisible now, but it's basically what has always been done.

I can't recall a time where I or anyone I knew filled those out and sent them in. I realize this is anecdotal but it a lot easier to not mail in a form than to try to find the opt-out option in some computer OS.

> It's all much more pervasive and invisible now, but it's basically what has always been done.

Basically is doing a lot of work here, the level and degree of how much data is vacuumed, processed, and used for targeting nowadays is orders of magnitude of difference from these primitive ways.

A tent and a house are basically the same: a shelter.

Right, the pervasivenes today is much higher. But marketers/advertisers have always hoovered up and exploited as much information as was technically possible. That attitude isn't new.

Another thing that is absolutely new is the vast number of people using this data for things other than ads/marketing. The data on registration cards was never used to decide if you get a job or not, or how much you pay for a hotel room vs the next person, or how long you wait on hold when you call a customer service line, or how much your insurance rates go up, or whether or not you're a suspect in a criminal investigation, or if you get custody of your child in a divorce proceeding. The amount of things our data ends up being used for is on a scale that simply was not possible when people started filling out product registration forms.

> It's how it's always been, always.

I don't know, I don't think sending in product registration cards could/would often result in your bank account being drained...

> It's all much more pervasive and invisible now, but it's basically what has always been done.

So you admit it is far worse today than it was before? But the second half of your sentence seeks to disingenuously pretend that it has "always" been bad.

I can be sick with a cold or I can have stage-four brain cancer. People have "always" been sick but one is serious (terminal cancer) one is not (a non persistent cold).

Do you listen to music only with earbuds? Do you cover your face when going outside? Do you transform your voice for each person you’re talking to? Are you buying only with cash that you handled with gloves?

Privacy is not a binary concept. There are actions and information that some people are ok being public, and there are some they prefer to remain private.

What is not OK is spying and exploitation. I should know what data you’re collecting and preferably specify which I’m ok with. I also should know what is intended for and preferably for most of it to be anonymized.

Most people expect reasonable privacy policies from companies and they believe that there’s some regulation in place.

> Most people expect reasonable privacy policies from companies and they believe that there’s some regulation in place.

Absolutely, but if you ask/inform these people they will say "Well, guess I have nothing to hide." because they can't comprehend going without all their devices/services.

> Everyone knows they're being tracked everywhere they go physically and on the web

That sounds good to some people. But if I mentioned it to most people in my family they would probably be rather weirded out by it. They probably also would have no idea of the scope of the size of it and how it is being used against them.

> And it feels, bizarrely for 2024, that little thought has gone into privacy or security.

No, no. They thought about the privacy and security aspect. They decided that it's better for their bottom line if Windows users don't have privacy from the mother ship. Really, they already decided that way back when Windows Vista first came out and periodically asked Microsoft HQ if you should continue being allowed to use your computer.

I mean, you can't even install Windows 10 without it telling you several times that unless you opt out (again and again), it's going to send just about anything you do to Microsoft…

That was XP, and that was the beginning of my separation from Microsoft stuff in earnest. The Windows XP background and UI is a source of nostalgia for millennials just like Windows 3.x, old-school Mac, or Amiga would be for people my age... but I feel no nostalgia at all for it because I let that generation of Windows, and most subsequent ones, simply pass me by, fortune having smiled upon me and enabled me to work in Linux almost exclusively since around that time.

Even if they kept that promise and they never uploaded the data off our devices, someone else (a hacker, a cop, a stalker, a lawyer, a thief, an abuser) will. Their promise is that everything you ever do on your device will be stored anywhere and can be used against you at any time. It turns out that not many people actually want that.

I'm perplexed that anybody thinks Microsoft were being dumb. They know exactly what they are doing and putting the pieces in place to violate users' security is the point.

Theyre just boiling the frog slowly. It'll be turned on by default soon enough and then theyll start looking for excuses to upload it.

This can be used to make them a shedload of money one day.

I agree, these decisions in 2024 are thoroughly vetted. I think the only thing these companies don't know is when the news cycle will pick up on something they're doing and they get blowback, and they'll have to pretend this was some little oversight.

Honestly, it has the smell of an NSA pressure campaign which would also rake in the money like nobodies business.. an easy choice for Microsoft. They are as guilty as sin for turning everything they create into a surveillance product.

I was just remembering today what they did to the security/encryption as soon as they bought over Skype… they removed it. And who would that benefit - the spies.

No-one cared about that. No-one ever cares. Except for this rare occasion - tides are turning and people are starting to care a tad more.

Adobe figures that if MS can get away with it why can't they? Companies will just keep chipping away at our privacy every chance that they get because our data is extraordinarily valuable and we can't stay hypervigilant and outraged forever. Eventually, they'll get what they want, and then they'll push a little more.

Perhaps. A key difference though - history files can include the individual pages I requested from the same host. Right now I have like 50 entries for the various posts I read just from HackerNews, all as separate line items etc etc.

In the case of the phone, one simply sees recipient of call, duration etc, regardless of how much information was exchanged. The phone I'm calling is arguably analogous to the server I request a page from, in the metadata context.

I'd argue browser history is significantly richer in some regards due to this. It's not unheard of for user identifiers to appear in URL paths either - try visiting https://news.ycombinator.com/user?id= user name>... In my Chrome, that's instantly in the history file with my username.

At this point browser cache is known to everyone, and many people do clear it regularly. Browsers save passwords and form data but as far as I know they don't upload that data to Google. Still, chrome is very popular and it sends google people's browsing history and all their DNS traffic.

Yeah, but in theory Recall doesn't upload anything either, which is why it's analogous. And in fact Chrome does upload passwords, and they're not even E2E encrpyted in the default configuration.

As far as metadata versus data, the URL of a static image automatically discloses the image itself. The only way to claim that the history doesn't actually contain the image is if you assume that the site has gone defunct.

Unless, of course, you're willing to argue that a porn image stored on the local hard drive isn't contained in any folders on the same PC that soft-link it. You might have an interesting time trying to justify why it is contained in folders that hard-link it.

Am I confused about what browser history is or what? Unless you open a static image in a new tab to look at it or you download it (as opposed to simply looking at it on the page it's on), then how on earth would its URL show up in browser history, which by definition tracks user-visited webpages (i.e. top-level links) and not every single URL the browser makes a request to?

Sure, info about non-top-level links is extractable from e.g. request caches, but that's a different thing from the browser history SQLite DB.

It's a turn of phrase; it doesn't mean literally no though at all!

On a more relevant note, how can it know when a private browser window is open in anything other than Edge? Same question with the password manager - is there going to be some new API that apps have to "opt in" to to enable Windows to recognise them?

1. Browsing history doesn't show what the user is doing on the page. There is a big difference between logging "user visited his e-banking app", and logging his actual credentials as they are entered.

2. Browsing history watches one app. Screenshots watch everything across the entire OS.

Not just credentials - account balances, account numbers, etc. There's a big difference between your browser history recording that you opened your bank or healthcare provider's web site and Recall recording everything that appeared on the screen while you did.

People might use Incognito mode to browse porn, but I imagine it's a lot less common when looking at other sensitive sites.

The ickier parts are on the unintended capture side, like enabling "show password" on a site doesn't affect browser history but Recall may capture it in the clear.

Or from history you may see that you accessed a site, but not what you did on it (what comments you typed for example).

The browser history may not, the cache and other local storage may well.

The take-away is simple though: Modern desktop operating systems need a security model where individual applications are sand-boxed and protected from each other.

Legacy systems have security models that protect users from each other, but this isn't the personal computing world we live in anymore.

Continuing with the comparsion, Recall applies to the entire operating system not just one application. To avoid it, one has to avoid Windows.

Whereas to avoid browsing history, one only has to avoid the popular, graphical, advertising corporation browser. As I am not interesting in graphics, I do this everyday, with ease, because there are countless clients besides "Chrome/Safari/Edge" that work with the www for consuming information.

> It's interesting to compare this to the Chrome/Safari/Edge browsing history, which is stored in an unencrypted SQLite database ...

Recall seems to be storing its info locally in an unencrypted SQLite database as well.

At least, that's according to the instructions here on how to access and view the contents:

https://www.heise.de/en/news/First-experiences-with-Recall-9...

From the submitted article, it seems like Microsoft will change/secure the access (and maybe storage) in some way, though there's no details on the specifics.

At least on macOS, I can't navigate to the directory holding Safari's data with other apps (without special full-disk read permissions).

There's also always private browsing, which exists specifically because people are aware of the implications of a browsing history and a persistent cookie jar.

That awareness will be much harder to build for an always-on screen recorder.

One difference is that Web browser history has been there 30 years, since before most people at the time had even touched a Web browser.

At the time, it wasn't very thinkable that someone would have the audacity to take and abuse that information.

It dates from when Internet people overall were more savvy about privacy than users overall today are, but it was also when the Internet was closer to a trustworthy environment, and before Wall Street sociopath types took over the tech and the culture.

Lots of kinds of abuse that today are routine and almost universal, for even startup tech companies, (e.g., embedding third-party trackers into Web site, and getting even worse from there), I think would've gotten them ostracized, and outraged demands for criminal charges.

During the dotcom gold rush, there was such a flood of totally new, posturing people, and so much money being thrown wildly at everything, that any remaining outrage was lost in the noise.

And now virtually no one knows any different.

But if you're trying to push some new abuse today, I think ordinary people are starting to have some awareness of what vicious sociopathic buttholes tech companies have become, and so acceptance might not be a slam-dunk.

The vulnerability is that the first thing any malware that happens to run on the PC will do is upload the Recall database, giving the attacker your entire usage history since installation (and of any other user account on the same PC). This can then be analyzed for worthwhile targets for scams and blackmailing.

I expect it does, if you're using Chrome outside of Incognito Mode. Iirc, there is an opt-out about "web history" on the google account - which then disables some other things so that it annoys enough people into keeping it on.

no it isnt the same, you may know I went to my health care provider's website, maybe even to make an appointment depending on the url, but with recall, everything that is on the page will be stored, not just the url. It's totally different. So the message I sent my healthcare provider that is discussing some of my most sensitive medical issues will be available to read and a record is kept of it... not just the url. Do you not see the difference?

Yes, but one product cycle and there's metadata (like a background texture) that tells the OCR to skip this page. Or ask your local LLM if the user is talking about medical conditions? If you like the feature at all you can make these things work.

"If you like the feature at all you can make these things work."

It's not on the individual users to take steps to preserve their basic human dignity. It's not Microsoft to not take that dignity away by default as was their plan before this fiasco predictably blew up in their faces just like the Xbox One always-online Kinect requirement before it.

I hate that most browsers do not let you set them to keep history for longer than 90 days.

I want to be able to find things I've seen before. Recall would've been great if using it didn't require me to update to a version of Windows that contains "Copilot".

Their is a very different scope at the OS level.

Most of us know that the public Internet is based on surveillance capitalism, no matter if we hate it or are just complacent or ignorant.

OS wide is far more problematic and of low value to the user.

Browsing history doesn't contain what's displayed on the page, and what you input into the input boxes, or POST requests. It's sorta like telephone metadata.

On the other hand, I am always freaked out by Chrome extensions that "can read and change your data on all websites". Can't they have more granular permissions? You gotta have a lot of trust for those extensions LMAO. They can read your bank passwords, probably!! And if they are ever sold...

Exactly - knowing the content of each webpage is pretty easy if you're "big brother" surveilling millions of people, even more so if you have a Chrome extension to help.

It's "little brother" that benefits a lot here: bosses, spouses, parents, etc., who otherwise wouldn't click on 1000 links in your history.

Yes, they can change it, that's what Manifest V2 deprecation is about. It will break a lot of ad blockers, because they rely on being able to read anything and change anything on all websites. Many people feel that Google is doing it to make more people watch more ads, not to improve security.

To be fair for me the extensions that get that are uBO, Privacy Badger, and Tampermonkey.

I trust gorhill and the EFF to not fuck me over on my data, and Tampermonkey kinda needs those sorts of permissions to work. My password manager has read access to every website but I'm already trusting it with all of my passwords so...

Seems like a very juicy target.

These extensions should not store any data without a master password that you input every time.

What if someone stole the signing key, and submitted an update to Chrome store, even for a little? Oh wait that is only for Chrome Apps. For extensions, they can literally update themselves anytime. Someone would just have to steal the certificate.

If an extension that reads all data uses a CDN (like CloudFlare) that CDN can execute a MITM attack against it and download new code, that would he catastrophic even if it was caught 1 day later.

>Oh wait that is only for Chrome Apps. For extensions, they can literally update themselves anytime. Someone would just have to steal the certificate.

Mozilla reviews signed extension updates. Something tells me uBO is one of the most scrutinized given how very many users it has.

>If an extension that reads all data uses a CDN (like CloudFlare) that CDN can execute a MITM attack against it and download new code, that would he catastrophic even if it was caught 1 day later.

My threat model doesn't include state actors targeting me specifically. Not sure much of anything works against that threat model besides maybe iOS in Lockdown Mode as your only device.

Extensions can simply download and update their own code, eg by loading new stuff from localStorage.

I have seen Metamask update itself randomly, and it has access to read every website

This is a horrible comparison. Browsing history doesnt show the contents of the page. It doesnt show you what you were doing on that page. It doesn't reveal anything other than you went there and maybe how long.

Publicly available content, yes. But not the content entered by the user themselves (security question answers, for example), and not contents behind a login, which is usually the case for sensitive information. Screenshots capture them all.

Yeah, I think this entire debate is uninformed hysteria and manufactured outrage. "If an attacker has administrator access, they can see everything you have done on your computer!". OK? That has literally always been the case? "Attacker is root" is game over and always has been. The original writeup from DoublePulsar tried to justify that Recall is somehow different from other such scenarios, but I found it totally unconvincing.

I think it's the right move to have it off by default, but I'm just not convinced by the outrage here.

You’re missing the point. An attacker can only see the passwords in your Recall database if they have root, but if they have root there are (and always have been) a thousand other ways they can get your passwords. There is no new attack vector being introduced by Recall.

Another big, big difference, anybody, not just some black-hat pro with a long kill chain of zero-days, has a fantastic source of data to exfiltrate.

Perhaps you didn't note before, or are one yourself, but this includes e.g. abusive spouses. Sure, maybe the abusive spouse could hire a black hat, but this is very different to a drunk low-life wife-beater casually snooping through "recall".

It might not be a "new" attack vector, but its absolutely a complete degradation to any computer security.

I did read the article. The person I'm replying to claims the entire debate was "uninformed hysteria", which means they thought the previous security model already required admin.

If an attacker got root with recall they might not need to wait the user to type their password and risk detection. The information they want to know might be already in the recall database.

To be clear, I am not in favor of Recall or dismissing its intrusiveness. However, the correct comparison is not just "browser history". Google is also tracking your search history, passwords (built-in password manager), location history (Google Maps), ad clicks, and more. All-in, it's a LOT of data.

I'm with you -- I avoid Google products for the reasons you listed and am staunchly anti-surveillance capitalism. I just meant to say that even for a person with my very plugged-in perspective on these topics, Google's violations of my privacy still don't feel quite as invasive as Recall feels, even if on paper it's just as egregious and dangerous.

In a typical bigcorp environment, laptops are loaded with silently installed spyware. Certainly equivalent to taking a screenshot every second or an always-on keylogger.

The horse is out of the barn for many people during work hours. But in the OS and on by default is a different story!

If there's AI involved, everyone's panic level skyrockets.

No one retweets "Attacker gaining root access reveals all user information", but instead "Attacker gaining root access reveals all user information collected by AI program" will go viral for sure.

When Recall is enabled, it should have an overlay stating that it is active so that all users are aware. Something at least as obvious as the old Windows activation overlay.[0]

Otherwise, every creepy roommate, bad partner, bad friend, etc... will take advantage of this to do bad things.

[0] Ideally more obvious, like when Windows screen recording is running.

If you had screen recording on by default you’d run out of disk space pretty quick.

MS just did what every other micromanagement company did and took screenshots every second or so.

Will have to wait and see if the extra security measures actually improve anything or not.

However regarding it being opt out… what would prevent a virus from just enabling it on a bunch of machines silently. Sure it would be caught but the damage done and most won’t be bothered to go in and disable it after.

Or Microsoft just decides they need to really market the hell out of AI and it gets turned on my default anyways.

Without Recall, an attacker needs to get a program to stay resident in memory to log keystrokes, screen contents, etc. for an extended period of time without getting detected. With Recall, they can get the same end effect by exfiltrating the Recall database file whenever it's convenient (i.e. an infected version of a text editor could send it while pretending to check for updates). This significantly lowers the barrier to entry for getting a victim's data, while also making it much easier to avoid detection.

> Without Recall, an attacker needs to get a program to stay resident in memory to log keystrokes, screen contents, etc

Or it could just steal your cookies which are out there in the open.

Your auth cookie expires.

The username/password you type in next time it expires is far more valuable.

And it might not even be necessary to obtain cookies or credentials if I can just see whatever you could see when you’re logged into various sites.

This is all moot anyway because Microsoft has already said they are now going to encrypt everything behind Windows Hello making it as secure as my password manager.

That's old information. This is how Microsoft is intending to change Recall based on these criticisms:

Microsoft will also require Windows Hello to enable Recall, so you’ll either authenticate with your face, fingerprint, or using a PIN. “In addition, proof of presence is also required to view your timeline and search in Recall,” says Davuluri, so someone won’t be able to start searching through your timeline without authenticating first.

This authentication will also apply to the data protection around the snapshots that Recall creates. “We are adding additional layers of data protection including ‘just in time’ decryption protected by Windows Hello Enhanced Sign-in Security (ESS) so Recall snapshots will only be decrypted and accessible when the user authenticates,” explains Davuluri. “In addition, we encrypted the search index database.”

https://www.theverge.com/2024/6/7/24173499/microsoft-windows...

"Old" is a bit of a stretch here ;)

But I'm glad to hear they've committed to making changes. Given the misrepresentations they made regarding the initial rollout plan (the target of most criticism, mine included), Microsoft has to prove themselves here and I'll wait until qualified security folks get their hands on this before coming to any conclusions.

What we know is that the initial version was a non-starter, and this new info validates the concerns we've all been expressing.

I truly hope Microsoft does an acceptable job of addressing this. It remains baffling and worrisome that it took a public outcry for them to implement what sounds like a baseline level of acceptable protection.

Well it is "old" since the article is about Microsoft's blog post where they discuss all these changes!

https://blogs.windows.com/windowsexperience/2024/06/07/updat...

> It remains baffling and worrisome that it took a public outcry for them to implement what sounds like a baseline level of acceptable protection.

It's possible this was the intention all along but as a early-beta feature this was just the MVP. The reason it was rolled out to early testers at all was to get feedback.

> It's possible this was the intention all along ... to get feedback

If they're relying on public feedback to realize how completely unacceptable the initial rollout was, that again points to deep problems at Microsoft and is why I'm saying this is baffling.

Microsoft is a big organization with different teams. It wouldn't surprise me if this front-end AI team didn't consider the larger security implications -- having it stored in your profile probably seemed sufficient. It's the same security all your documents have, your browser cache, etc.

If so, that's a problem. It might explain why this happened, but that doesn't mean it's an acceptable practice, especially after recently claiming that security is a primary focus for all project teams.

Security requirements often completely change the architecture of a product. Things can be built without security that are significantly more challenging to accomplish when strict data security requirements are in place. Architectures that assume no security often completely break down when security is tacked on top.

If this is a matter of a product not yet getting "security added", that again raises major concerns about how Microsoft is building products.

It doesn't sound like they're going to have significant problems adding more security to this product. For advanced as it sounds, it's not that complicated of a technology. It's just plugging together a bunch of existing technologies. I could probably MVP this app in a week myself given what is available.

I think that exploratory development is, in general, a good thing. Bogging down all development with middle-management procedures might certainly have caught this early. But that doesn't necessarily make that a better way to build products.

The scary thing about Recall isn't actually Recall itself. It's that AI makes this kind of product possible and really easy. I'm sure we're going to see implementations of this idea everywhere and not just on PCs. Imagine AIs watching security cameras.

Virus turns on recall, user might not notice much. A real Microsoft service is running. It can then just wait and activate later. If the user notices recall on, they'll just blame Microsoft. You can then just turn it on again. You can already see that many users are suspect that it'll go back to being on by default sometime in the future too. It's not uncommon to see system updates change settings.

The virus doing the same things as recall will be much noiser and much more suspicious. Making it much more likely to be removed.

Not to mention that once recall has been running a virus only needs to extract the data. It records far more than what a password manager does and is far easier to search through. It just makes a very large attack surface.

Basically, why would anyone develop keyloggers anymore? Microsoft did it for you. And it'll never be tripped by antivirus software because it's an official and legitimately signed program. You don't see a problem with this?

> Or Microsoft just decides they need to really market the hell out of AI and it gets turned on my default anyways.

This is what will happen. And when you turn it off again, it'll be turned back on by the next update. Enjoy.

Please stop with these kinds of made up fantasy scenarios.

There's no such thing as "accidental enablement" for stuff like this, as if it's a switch every employee at Microsoft has access to, and one of them one day can end up flipping by accident with their elbow and it ends up in production without anyone else noticing.

Either they decide to intentionally enable it or not. There are no accidents , when stuff like this needs to go through a committee of people for approval before it makes it into production.

> Either they decide to intentionally enable it or not. There are no accidents , when stuff like this needs to go through a committee of people for approval before it makes it into production.

Absolutely. And all of them decided to screw largely defenseless non-technical consumer to make short-term profits. That's not a fantasy, that's our reality.

Few things that Recall can do: - Make sure Recall is per user and not something that can be installed system wide for all users.

- When a user enable Recall, it should ask to setup a "Recall password" and generate a private public key and use the password to encrypt the private key.

- Use the above public key to encrypt all the data it stores.

- When user wants to search Recall history, ask for the password, decrypt the private key and use the decrypted private key to decrypt the data and show the data to user.

- Show some sort of indicator on taskbar that Recall is running, not a tray icon (which can be hidden), but a proper big red circle kind of thing.

To me this seems like another case of MS top executives telling every team that they have to do something with AI. Typical approach used by many executives and managers - "Here is a new tech, figure out a product to build with it".

Corporate is never on the bleeding edge of Windows feature updates. They bring security updates first, but feature updates are at least one generation behind, maybe more waiting for Microsoft to fix bugs and doing their own regression testing, plus they get to choose wich features employees receive or are enabled by default via group policy. In other worlds, recall was never making it into any corporation anyway.

This is generally true, but Windows is the standard for far more SMBs than larger enterprise customers, and in that context it’s not nearly so straightforward. I have a client, a health insurance benefits broker for other local businesses. They do very well for themselves, but it’s just 2-3 full-time people, so there’s never been much cause for a full-on domain with GPO policies to maintain a strict, stable state across their equipment. Traditionally, off-the-shelf systems with SMB-targeted software had been more than sufficient.

When Microsoft decided to push a feature upgrade last year that automatically enabled OneDrive backups for their home directories, it technically violated HIPAA by moving electronic patient health information contained within their scanned files folder onto OneDrive servers without any prior consent or authorization. They literally called me when they were unable to find their files, Microsoft had (laughably, if it weren’t so serious) placed a text file on the desktop titled “Where Did My Files Go.txt”, and then directed them to the OneDrive folders where it had moved their desktops, documents, and pictures without their knowledge or approval.

I have since moved them to Microsoft 365 accounts where I can apply GPO, but my clients were understandably unhappy about having a new annual subscription that didn’t add any tangible benefit, rather they’re now on the hook for a couple hundred bucks a year for what’s essentially a shake down. Pay for the new service that adds nothing meaningful to their experience, or else face the consequences of Microsoft ruining your business on a whim.

I would have agreed with you until revently. And now, everyone is throwing email, chat, code, everything into cloud based AI tools at a highly regulated company. This happened 6 months after they just locked everything down for actual employees because of an IP leak. Very strange times…

maybe 50% of US business users have an admin of any kind who oversees their IT ops

everyone else just gets a laptop, unboxes it, turns it on, uses it, does whatever they want to it

see: any retail location in a strip mall, any mom/pop business, etc etc

I think on the product side it’s pretty straight forward. They saw RewindAI talking up a bunch of traction and people seemingly interested. Someone assumed customers wanted this because of that data, and it’s a pretty easy thing to build, so they went ahead. I am surprised it got past security reviews but I can understand how it came to be from the product side.

They’ll probably think twice before jumping into the fray again with the Microsoft branded Informant Wire (I mean AI wearable) ;)

I don't understand how Outlook isn't a compliance nightmare. Especially since it's moved to the cloud. The amount of very sensitive data Microsoft must have on just about every single business/industry thanks to outlook and excel is insane.

At one place I worked when the company replaced my old machine with a new windows 10 system it was configured to send every single keystroke back to Microsoft. There was zero concern over privacy or compliance, just an assumption that MS would never abuse that data for any reason. I did not have their faith and disabled that "feature" then changed a massive number of other policies to try and keep as much data out of Microsoft's hands as I could.

The corporate settings that care already do this to the employee screens ...

Compliance doesn't say "company can't watch employee" -- in many cases it mandates surveillance.

This just lets the employee leverage that too.

Depends on the compliance. If this monitoring sucks up any personal data (I don't mean employees' data here—personal data owned by anyone) there are erasure and data subject access requirements, for instance.

Corporate clients get whatever they want. I am certain that their Windows 10 support won't be pulled in Oct 2025 as MS has threatened for everyone else. And when they migrate to Win11, it will almost certainly be a separate OS image free of the garbage bloatware and ads that the consumer devices are plagued with.

Am I just imagining their saying that Windows 10 would be the last Windows? I had thought they would be moving to an Apple-esque model where OS updates would just become iterative and avoid the old EOL/upgrade cycle. It’s how I justified all of their tangential money-grabs on other fronts.

Not really a problem for enterprises.

Any company that has compliance requirements to keep devices supported with security updates, it's the same as Win 7 to Win 10; you either update everyone to Win 11 or you pay for the security updates for Win 10 (IIRC you have 3 years to update before you can't pay anymore). Many will likely already be on Win 11 as the upgrade path is easier/quicker than Win 7 to 10.

Also they will not have the gunk installed anyway as they will almost certainly have Windows Enterprise which has more policies that can be set, and then they will also be ordering devices from an OEM or distributor that doesn't have the junk included.

Heck, if they aren't doing Autopilot from the OEM or distributor, they will almost certainly be applying their own Windows image.

I bet there are a trillion companies and governments who want to know what all of their employees are doing every second of the workday. compliance won't stop them from trying.

On LinkedIn someone in my network pointed out that, apart from the security and privacy disaster, the name Recall was a bad choice because of negative events like product recall.

It would actually be a fantastic name if this were a real concern. Imagine, a well-known feature to mask any searches of a product recall. The only problem with this theory is that computer QA is so incredibly shit that the concept of a recall more or less doesn’t exist in the first place.

"Total Recall" in quotes makes me think you're trying to get your ass back to Mars and that you're trying to remember something because you had your memories wiped. It makes me think of nothing about a friendly service being offered forcefully upon you from your friendly and malevolent OS provider.

I have a really hard time understanding the use case for something like this. Stuff that I want to remember I just write down or reference something like my browser history or recently opened files. It's very low tech for sure but it works, is waaay more energy efficient, way easier to understand and audit, and doesnt have the same security concerns. I get that using "AI" has a Wow Factor that existing systems have but I cannot understand the thinking of folks that are OK with the trade-off. Ita just not even close for me.

I definitely get the use case. It's naive to ignore that there is utility.

But just because something has utility doesn't mean it comes at high costs. I mean it's a super powerful keylogger that is searchable without technical knowledge. Not to mention that it'll probably fail to LLM type of attacks, which even many non technical people are able to figure out.

But then again, I don't understand why people so passionately store all their chat logs (not just important/memorable messages) and take millions of photos. We kinda spy on ourselves

We used to use a similar tool in QA. Often when you accidentally reproduce something - especially something rare - you don't even realize it happened until it happened and then you can't remember what you did to produce it. Being able to look back even just a few minutes can save you hours of attempting to figure out what the magic was.

Now that I'm a developer, I often get into the flow and find myself knee deep in some work, but I forget to write notes about what I was doing. Coming back the next day, I often can't remember what issue I ran in to or how I fixed it. Having a quick way to review what I did the previous day is very helpful.

I can see a lot of potential uses for this technology, but I'm quite wary of any service that involves sending all of that data to some third party. Regardless of how much they might swear they won't use it for anything else, every company eventually sells your data for extra profit - it's just too tempting.

I think the product itself can be useful, but Microsoft is the second last organization that I would ever trust to implement it correctly, only after governments.

Giving your screen recordings to Microsoft is like giving a loaded gun to a toddler.

I agree, I think the current state of the AI is absolutely incredible technology, but I just don't see a 'product' yet.

If chat and co-pilots are all we get out of this wave of investment, then I'm not sure if it's been worth it.

I see a lot of cool little use cases (eg, LLMs are genuinely fantastic for creative brainstorming), but I'm absolutely not seeing the multi-trillion dollar AI industry that all the big companies are clearly banking on.

Too late. I've used Windows for over 30 years. Was never a "fan", it just got the work done without getting in the way. Good tool.

But now I'm moving all my computers, including work computers, to Linux. Will miss out on some hardware/software I use for music production (biggest loss will be TotalMix FX for my RME audio interface), but MSFT leadership has shown they don't get it.

Also, Fusion360 for Linux when?

What's funny about this and other "recalls" (pun intended) of "products" from so-called "tech" companies as a result of "feedback", or "backlash" as Wired calls it, is that the companies never asked for such input and AFAIK no one is contacting the companies to give it. AFAICT it is obtained through surveillance. To people born after the internet this might seem normal, but to me it is quite odd. The companies claim to be operating in service of "users" but there is generally no direct contact between these "users" and the company.^1 With some isolated exceptions that have increased over time, there is no customer service. And in most cases the "customer" has not paid the company for the so-called "product". Generally, no one is asking a refund on their purchase of Windows because generally no one pays Microsoft for it. Instead people just complain into the ether.

No need to pay for being the target of surveillance. The "products" are free.

1. Unless we count the telemetry and "auto-updates". Users never asked for this stuff though, it is not initiated by them. This "product" is broken on delivery hence the alleged need to keep "fixing" it by remotely installing more software, presumably that isn't broken and will not used for surveillance, on peoples' computers. All for free. There is no money to refund if the "product" does not work as expected.

Don't worry, even when you pay very high prices for products and service you're still being spied on. A company will always make more money by charging you as much as possible and then also collecting every scrap of data they can get their hands on. Every smart TV is spying on users and many are pushing ads. How many people do you think demanded their money back? Every game console sold still spies on you. Every car. Every cell phone. When there are no products that don't spy on us what will people do? Return them all and live in empty houses?

It's so weird to me that a company like Microsoft would care that much about "reputation". Everyone basically hates them already. Many of the most successful companies in the US are widely (if not universally) hated by the people who pay them. Nobody loves comcast, or exxonmobil, or centurylink, or EA, or equifax, or facebook. People feel trapped and unable to avoid paying some companies or using their products no matter how much they hate them. How many people have paid Microsoft for their OS at all? How much money would they really lose if they ignored the bad press? How many grandmas would start downloading linux?

I'm glad Microsoft is making changes, but I wonder how much is out of fear for their reputation and how much is just to try and comfort people and get the news to stop talking about it so that everyone doesn't just disable it as soon as it rolls out.

As someone who has tried to push back against what execs ask for many times, if they want it bad enough, it doesn't matter. They will push forward no matter what the objections are. And if the person objecting won't give in, they'll find someone else to do it.

Company-wide internal push to shoehorn AI into every product and service. All recognition and rewards are given to the sychophants, no matter how ludicrous their proposals. Even Principal and Senior developers are dragged into meetings with senior leadership to provide suggestions on how AI can be used in their microcosm. Whether it should be used is completely out of the question.

It’s a complete circus right now. Plenty of us just ignoring it and opting-out but it might reflect on our bonuses.

Non-sycophant employees are shut down and ignored once the whole corporate culture has bought in to the hype du jour. If you are the sole dissenter, it can even make you look like a “bad” employee for not recognizing the “opportunity” that the new hyped thing will supposedly bring.

Maybe a bit off-topic, but I sure wish they'd do this for OneDrive! I installed Windows for personal use for the first time recently (although I use it exclusively at work) and it drove me ABSOLUTELY BONKERS that my home drive was mapping to C:\Users\atribecalledqst\OneDrive.

What I hated the most was that the File Explorer just calls the folders in there e.g. "Documents" and "Pictures" without showing the full path. So it was hard to figure out just where in the file system you were looking -- a major annoyance if you do any work in the command-line!

Even after switching OneDrive off and doing as much as I can to try and get rid of the OneDrive folder structure, I haven't been completely successful. You can make some -- but not all -- home folders (like Downloads, Documents, etc.) point directly to their place in the local user folder, but others, particularly Pictures, don't seem to be movable. Additionally, some programs still seem to want to use the OneDrive folder by default, like I think Office programs still do their best to use them.

In the grand scheme of things it's a small annoyance but god it annoys the shit out of me! I didn't ask for cloud backup and it drives me nuts they tried to force it on me!

Dear OS writers:

Internet access is not always guaranteed or reliable. Please do not assume that the cloud is a viable solution for every user.

I ran into this on my phone awhile back. I knew I would be out of service for some time but had some PDFs I needed to reference. So I downloaded them to "files". Que surprise when I later go to look up a value and there's a little cloud with a down arrow button next to the PDF in the files app, which of course fails because I'm nowhere near any internet access. Even more fun: turning off the cloud integration in files just causes the files to disappear, even if you are currently connected. It's allergic to local storage.

This is the number one thing that annoys me about so many apps, especially apps with clear use-cases for offline use like listening to music, reading, and learning apps. I don't understand how so many app writers have never gone for a run through a canyon or flown on an airplane. I specifically pay money to SoundCloud for instance just for the "feature" to cache the music locally and somehow it regularly gets stuck clearly from lack of internet. It's probably some metric collection or some other spyware to make sure all the bean counters get their money at the huge expense to usability. Pimsleurs language learning app, and many book reading apps all suffer and all I want to do is not be bored to tears on flights that don't have internet.

> I don't understand how so many app writers have never gone for a run through a canyon or flown on an airplane.

In the UK, every time I got on a train, I'd experience that. And it was worse than not having internet; you had internet, but with extreme packet loss and instability, meaning that every app out there would simply stall, even if it already had the data to do whatever it is I wanted it to do, because it was waiting on some background request to complete. And because I had internet, the request didn't just fail, but it also wouldn't complete in any reasonable amount of time.

Very frustrating.

They just want live data on your activities and update without sync and stuff however expensive that is even for them, easier to be lazy too

Also every <35 years old person is a js/web dev, so that’s what they do on cloud

> learning apps.

So fun to spring for the paid duolingo only to realize you can only download the next lesson up, not like the entire course.

The lessons are like 5m long wtf am I supposed to do with that? I just want to spend my idle time on the plane or camping disconnected from distractions so I can learn, but app developers have made that effectively impossible

And this is why I don't pay for, or even use duolingo even though I'm actively learning a language

Seriously, it drives me absolutely out of my mind. I tell everybody who listen, "Remember your users aren't software engineers who are always connected over fast reliable pipes, and program accordingly" but it's a hard problem. No PM ever wants to hear that you're spending time optimizing for no/low internet scenarios.

I've gotten burned by that "isn't really downloaded" thing a few times before too, to the point where I don't trust apps to download anymore. I just adb push files from my laptop to my phone before I go. Can't always do that though, but I try to.

If this is the place to complain about broken patterns in Microsoft software, I wonder if anyone can fix this:

1. Create new office document (Word/PowerPoint/etc) and hit save.

2. No, the default location in OneDrive isn’t right so you click the down arrow to see more.

3. No, none of the other recent locations in the (short) list are right either, so you click “More locations”

4. Now you have to click Browse to see an actual Save As dialog that finally lets you navigate through folders. Even then the actual folders are right down at the bottom of the left hand “tree” pane, below a bunch of virtual folders, below OneDrive (aside: if you navigate “up” from here you get to “Desktop”, but it’s not the same “Desktop” that appears lower down in the list; that one is inside your OneDrive), below Music, Videos (you get no hint as to where these actually are), finally near the bottom there is This PC and Network which you can navigate sanely through. Oh, and right at the bottom there is “Microsoft PowerPoint”, as a save location. You can click on it and try to save a document in there, wherever “Microsoft PowerPoint” is. Just kidding, you are stopped by a dialog box telling you this isn’t a valid location.

JFC. No wonder people prefer the “everything is an app icon” approach. Windows is diabolical for managing files.

1,000% agree.

Saving files in Office has turned into a nightmare.

I don't understand what Microsoft is thinking with this behavior.

I'm fine with that being the default flow. But it can't even be turned off.

I imagine this design is better for non power users.

They no longer forget where they saved their files.

But for power users, this is terrible.

Yeah, my machine connects over WiFi on an external USB 3 adapter because I'm too lazy to finish my Ethernet project. The adapter requires drivers, which are handily included on the device itself as a mass storage device. But there's seemingly no way to get those drivers installed in the captive environment, I even tried using the "launch cmd" key shortcut and manually running the executable, but Windows wouldn't have it. And there's no option to install drivers so you can proceed with Microsoft Account sign in...

Literally my only option was to use the local account bypass. How long before they fully remove that, though, remains to be seen.

Yes, Onedrive started out as a pretty useful tool but has turned into a deceptive trojan that tries to force whatever growth metric MSFT managers are currently chasing through a combination of dark patterns (like hiding true file paths from view) and also simply refusing to operate in obviously useful ways which many users want and expect (like not having a built-in way to back up only specific sub-folders on different drives (forcing paying users to trick it by using junctions)).

There used to be no option to uninstall it - now there is.

You will still get it reinstalled during a major OS update, but at least it can easily be removed. Before it was a chore to clean up.

I would speculate there is even some way to prevent it from reinstalling during those major updates. That seems like the kind of capability they would build in because a huge Windows customer complained (i.e. realistically, the major check against dark patterns in Windows).

Not quite what you are describing, but you can prevent any specific executable from ever running by configuring a "debugger" for it in Image File Execution options (HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Image File Execution Options). You add a key with the executable name and then add a "debugger" value, then point that debugger at C:\Windows\system32\systray.exe. Every time the named executable tries to launch, Windows will try to "debug" it with systray, which immediately exits so the program never actually runs. After uninstalling OneDrive this can be set to prevent OneDriveSetup.exe from ever running for example.

You can also define software restriction policies to do similar things.

It id what SRP's are for (but yes, I would not put it past them to disable anything targeting OneDrive).

Hang around kids and even though they can be pretty good at using a computer, they have no clue how the thing actually works. They don't know what a file is anymore. Everything is a shiny little icon in a shiny little magic folder.

Not trying to make this sound like a value judgment, more an observation. But it makes you wonder, what do we lose by excessive abstraction.

This isn't excessive abstraction - this is just different abstraction. Files and folders are a human invention, and there's no law of nature forcing us to continue using them. It's like complaining about people forgetting how to use MS-DOS commands, when Windows (until PowerShell) was built on GUIs through and through and MS-DOS commands were only still there for compatibility. You don't have to learn MS-DOS command to copy files, you learn to use Explorer to copy files (which to a small extent is like using the MS-DOS command).

Or like complaining people forgot how to use teletypes. We didn't have to keep using teletypes, and we didn't keep using them. Our Linux terminals are still modeled after teletypes, but not in a way that has anything to do with using a real teletype. You don't learn teletypes, you learn terminals (which to a medium extent are like teletypes).

It isn't like when people don't learn to add numbers or how Quicksort works or assembly code. Those are still fundamental truths that help people understand things. It's more like not learning to write Roman numerals, or not learning ALGOL 60. Nothing is really lost except the ability to read old things. You don't learn Roman numerals, you learn western Arabic numerals, and they're better, not worse. You don't learn ALGOL 60, you learn C11, and some people would argue whether it's better, but it's not worse.

No, those shiny app icons are still using folders and files, that part is just being hidden from users to where they have less understanding of how things actually work.

Phones aren't secretly using Roman numerals or tiny embedded abacuses though. If they were for whatever reason, there would be plenty of value in learning those systems.

Nah. The files and folders still exist on all of these systems. So hiding them away is actually more abstraction, not “different” abstraction.

UX prognosticators have been preaching for decades that anything that computer users find confusing should simply be hidden. Not made more clear, or easier to use, but just papered over so users can no longer identify a specific thing to complain about. It’s just like the weirdos who try to get rid of the address bar on web browsers every few years, but the filesystem haters have been a lot more successful, and computers are more confusing as a result. You don’t solve confusion by hiding it behind a thin layer of paint. All the same problems still exist, but there’s no longer a way for experts to even try to help. There are so many better ways to simplify computing than pretending it’s magic.

> It’s just like the weirdos who try to get rid of the address bar on web browsers every few years, but the filesystem haters have been a lot more successful, and computers are more confusing as a result.

They figure that the less users know about how their device/software works the more dependent they are on developers who can then act as gatekeepers of what the user can and can't do even when the system is capable of much more. They don't want users doing things differently, or disabling things, or seeing what's going on under the hood. Keeping users ignorant, controlled, and dependent gives them a very secure feeling.

Storing things "in files" still writes a CHS-addressed sector on a disk, but it's less reliable and the user doesn't know where it is.

Files are currently used to implement apps, but that can be seen as a transitional measure, like an OS that supports both files and raw disk access. A fully app-based OS without files, though not existing currently, would be possible.

Another idea the industry discarded was to make the disk a big SQL database, again without files.

I recently tried to fully rid myself of OneDrive and it took me over 48 hours to accomplish. The only working method I found involved fully enabling OneDrive, signing in, and waiting for a full sync. Only then was I able to tell it to stop syncing and finally remap Documents, Downloads, Pictures, etc.

The fact that I needed to log in, wait 24 hours for my account to unlock due to inactivity (!!!), and enable sync in order to disable it was enough for me to finally decide that Windows 10 will be my last Microsoft product. It may be a small annoyance, but to me it was the straw that broke the camel's back.

That is truly insidious, but FWIW, you don't need to abandon Windows entirely because of this. There are ways of creating a custom Windows installation disk that removes OneDrive, along with other bloatware, spyware, and pretty much anything else you don't like. Look into tools such as Tiny11 Builder, MSMG Toolkit, NTLite, etc. This is a decent guide[1] for setting all of this up.

The process is quite tedious and takes a few hours, but in the end you end up with a personalized version of Windows, without any of the garbage. You still need to be vigilant of Windows Update undoing some of this, but you can also disable it altogether and manually cherry pick the updates you want to install.

It's insane that Microsoft is building such a user hostile OS that forces users to resort to this, but if you absolutely must use it, the experience after doing the above is not so bad. I've been running a custom install of Windows 11 for about a year now without any issues.

[1]: https://www.tomshardware.com/how-to/create-custom-windows-11...

And I can almost guarentee you it will magically all turn itself back on/reinstall itself eventually after the OS force updates/reboots itself in the not too distant future.

They are the house of dark patterns.

After a certain point anyone paying attention can see it's not accidental. Oops sorry! No. Their goal is your technological enslavement. Mis-features like that don't accidentally just always end up being evil and oops sorry when there is a real backlash. They wanted to see if they could get away with it, like they do.

I abandoned MS products in 1998 for good. Win98se pushed me over the edge.

Wim98SE was actually good though! Well, it was way way better than win98. Win2k and 7 (the last windows OS I ever had for personal use) were good too. The writing was on the wall back in the 98 days for sure though. MS decided that your computer was theirs.

If you think this is bad, there was a period last year that my documents folder would suddenly rename itself to "Documents" but in a different language. This would religiously change every few days. Other people have reported it as well.

I have disposed of my last PC now and have nothing to do with the infernal things, or onedrive, or any of that crap ever again!

Yes, my company just went through a merger and for quite a while we had two OneDrives showing up and it was difficult to tell where the default folders were in addition to being a huge mess any time a file dialog opened. I've actually reverted to creating folders in C:\ to store files so I know where they are.

This is especially obnoxious for Desktop and Remote Desktop Connection.

The former because my desktop is... where I want things just a certain way for THIS computer, not across the cloud. And because it's a PITA to undo and set it the correct way.

The latter because of course I use Remote Desktop on multiple computers, but it keeps saving a "default" file in the same place across computers, and throwing errors left and right because they conflict. So stupid.

Interesting, our experiences are different here. I suspect it's because I installed Windows 11 (23H2) using a local account using the OOBE bypass (not because I particularly hate the Microsoft account thing, but because this machine uses an external WiFi adapter and requires drivers in order to work, so I could not have done it even if I wanted to). The drivers are actually included on the device, but there's not a clear way to accomplish a driver installation while in the captive OOBE, even given the ability to launch a command line.

I did later connect my Microsoft account. In my installation the OneDrive folder is empty and the entries in Explorer map to the normal places (C:\Users\X\Pictures etc). If I open one of the default folders, it does show a "Start backup" entry in the address bar that is referring to OneDrive, though. If I open the OneDrive folder, it asks me to sign in (entering password) and set it up-- which is funny, because the Windows user is signed in using a Microsoft account already- so seems like they haven't connected those dots properly yet. In theory this might be their way of implementing a security check for uploading all your files, but if so it's an awkward way to do it.

> Additionally, some programs still seem to want to use the OneDrive folder by default, like I think Office programs still do their best to use them.

If I remember correctly, there is an API that programs can use to locate common folder locations for users (such as Documents, Pictures, etc). My guess is that your account still points to the C:\Users\X\OneDrive\Pictures instead of C:\Users\X\Pictures. If you could adjust those directly (maybe in the registry?), I would imagine that it will work correctly in these programs, especially since I doubt those programs would break on my setup, where there is no OneDrive subfolders (though I don't use Office so I can't check). And in case you wonder if there really are no subfolders in OneDrive since I can't open it in Explorer without signing into it- it shows nothing when viewed via PowerShell.

I just got a new PC and went through the same thing! Incredibly frustrating that in something Godot I have to manually traverse through folders to get to where I want to actually save a file (like.. Documents)

The proper way of doing it is to use the API calls that have existed for decades to get the paths of well-known folders. It is because they are known to move and in fact having a roaming profile on a server location dates to the mid 90s with WinNT.

If you're hard coding paths you're doing it wrong.

> The proper way of doing it is to use the API calls that have existed for decades

A user doesn't want to do this though.

I tried casually using a windows 11 machine for something the other day (I think I was fixing game folders for my girlfriend), using just explorer, and it was pretty obscenely bad how overly confusing it had gotten. I say this, and I fairly routinely debug old build systems with complex nesting file structures, I know my way around a file system.

This wasn't a case of "oh you're just a power user", this was a case of the system had broken, and the simple advice of "backing up your files" and "copy your files over here" wasn't working.

Telling everyone they need to use API calls is just ridiculous, the filesystem is just broken for the average user.

you just have to use dopus as a file explorer replacement and just use dropbox (with cryptomator of course...) to yield (in most respects) best in class file management and sync

Always set Windows up with a local account to avoid this nonsense. Used to be relatively straightforward in Windows 10, but MS made it a lot harder to dodge in 11.

It is infuriating when I open the file explorer and it takes many seconds to populate the side bar. This wasn't the case wit windows 10. Everyting in one drive really makes things take a long time. OneDrive is great, but I want a OneDrive folder where things are sync'd, not transparently transforming the file system into OneDrive.

The gateway to a monthly consumer subscription. Therefore important to Microsoft.

Apple also uses dark patterns to try and get a monthly income from customers. Apple has upsells and nag nag nag advertisements for iCloud.

The irony with Microsoft is that I would consider paying a monthly fee for a modern version of Windows 2000 without extra features. No adverts, no telemetry, no OneDrive, no cloud signin, no store, no games installed as part of the OS, no MS junkware, no bullshit. Aside: why is there no "Windows for developers" - even Balmer knew "developers developers developers" was worthwhile but Microsoft has deleted that from its DNA: even though Apple's competition is a mixed bag.

Add to these complaints that many folders are actually logical overlapping folders that pull from multiple places. I haven’t been able to bring myself to use Windows for years now, and I was a Windows sysadmin for over a decade! It’s basically impossible for someone like me who needs to feel in control of their computing environment to ever feel comfortable with.

Or better yet, make the OneDrive integration a public, documented API so we can plug in our own cloud storage and get all the same benefits (syncing settings, files, game saves, etc. but with the added benefit of choice). I'd love to get native integration with ownCloud / NextCloud and even other online competitors.

And for that matter, make Apple do the same for iCloud; I'd love to keep all my iPhone stuff in my own self hosted "cloud" and get 1st party integration.

One drive is awful. It keeps crashing and forgetting where it was. I have 3 copies now. I have to waste time to sort it out. And it messes up the dates. It is disgustingly f our workplace enforces this.

（评论） (comments)

（评论）
(comments)