[lewisl9029]

I assume you mean the Enterprise version? Last time I checked that wasn't something individual consumers can just go out and buy.

But regardless of that, from Microsoft's perspective, does it make any sense to not offer the option to disable telemetry on all versions? From what I see, all it's doing is giving them bad publicity while driving potential users away to other platforms. And for what? Probably a few basis points in telemetry data that won't even make a dent to the overall trends in their telemetry data.

[caskance]

Both the enterprise and educational versions offer you full control. Consumer versions offer what I consider to be enough control, but obviously that's subjective.

If you offer the option to disable telemetry, some people will disable it. If there is any pattern to which people do this, and we all know there is, it introduces sampling bias and casts aspersions on any conclusions you draw from that data.

[khedoros]

It seems like they're trading one bias for another. I suspect that the population of users likely to disable telemetry have a lot in common with the population of users likely to just stick with Windows 7 and remove the updates that added telemetry collection.

Personally, I'd probably be satisfied if Microsoft gave me the tools to examine the telemetry that my computer wants to send. Not making that available to users makes me feel like they have something to hide.

[lewisl9029]

Yes they're definitely trading one bias for another in this case. Users who value privacy simply won't use Windows 10 to begin with. And you can't collect telemetry data from non-users.

Users who turn off telemetry can still give you one crucial data point: the percentage of users who do care enough about privacy to turn off telemetry. But this data point is something they simply can't capture because of their insistence on making telemetry mandatory. And any future decisions they make with regards to privacy and telemetry usage will now have to be based on speculation instead of hard data.

In addition to biasing their telemetry data to the subset of users who don't care strongly about privacy, this decision also has a real cost in terms of adoption rates, market share, and damage to the goodwill they've slowly built up over the years in other areas. I just can't see why they'd think this is a worthwhile tradeoff.

[caskance]

They already made that trade by allowing enterprise customers to turn it off.

They also tell you exactly what types of data they collect, along with where and how. That's not consistent with wanting to hide the extent of their telemetry.

[khedoros]

What they do for enterprise customers is a moot point to me, an individual, when I can't buy an enterprise-edition license from Microsoft.

I've seen hand-wavy explanations, in general terms, of the data collected. I don't want that. I want to see the actual data collected from my specific machine. Even a bullet-point list including items like "total time executing binaries tagged as 'game'" would be preferable.

[magicalist]

> They also tell you exactly what types of data they collect, along with where and how

where?

[caskance]

In the official documentation for the feature that critics never want to read.

[khedoros]

In the interests of providing a better answer than "In the official documentation":

This is the Windows 10 retail EULA: https://www.microsoft.com/en-us/Useterms/Retail/Windows/10/U...

It links to "aka.ms/privacy", which takes you to: https://www.microsoft.com/en-us/privacystatement/default.asp...

There are some good examples of the categories of data that they collect and some information about which features, programs, and apps might collect that data. There are examples within the categories, but not an exhaustive list. Maybe there are more complete lists available for each separate feature. I'd be interested in seeing an explanation, collected into one place, of which pieces of data are influenced by which settings, which ones can't be controlled, etc.

[lewisl9029]

> If you offer the option to disable telemetry, some people will disable it. If there is any pattern to which people do this, and we all know there is, it introduces sampling bias and casts aspersions on any conclusions you draw from that data.

This is true enough, but consider this: Giving the option to disable telemetry will result in users users who would disable it. Not giving the option to disable telemetry will result in users who who simply won't use Windows 10 because of mandatory telemetry.

With the former option they still get a data point in terms of the percentage of their users who care enough about privacy to turn off telemetry. With the latter option they don't get this data point, but in return they get some telemetry data from users who would disable telemetry if it was available, but don't care enough about privacy to use a different OS because of it (i.e. the telemetry data they have available is still biased to the subset of users who don't care very strongly about their privacy).

Is the data from this latter group of users important enough to be a worthwhile tradeoff for getting data on the percentage of your users who do care strongly about privacy and the damage to their market share and the damage to the goodwill they've built up over the years? My answer would be no.

[Aleman360]

Because telemetry is a key tool used to improve quality.

What software dev doesn't think logs and crash dumps are useful?

[lewisl9029]

Offering the option to disable telemetry means Microsoft won't get telemetry data from the intersection of users who 1. don't just roll with the defaults for everything (many UX studies have shown that this applies to only a tiny percentage of users), and 2. care enough about privacy to want to turn telemetry off (this number is likely tiny relative to the overall Windows 10 userbase considering how great the adoption rates of Windows 10 has been so far despite mandatory telemetry), and 3. use Windows 10 (a significant percentage of users in 2. have already avoided Windows 10 specifically because of the mandatory telemetry).

That intersection of users seems absolutely miniscule relative to the overall Windows 10 userbase, and in most cases the telemetry data they contribute is not going to be any larger than a margin of error in Microsoft's Windows 10 overall telemetry trends.

Nobody is going to argue that telemetry isn't useful for developers, but at some point you have to decide if it's worth it to force users into giving you telemetry data if it means driving potential users away to competitors and giving fuel to a very vocal minority of users that write unfavorable articles and comments on news sites, blogs, and social media at every opportunity because of this decision.

[Aleman360]

Agreed that there should be a way to opt out, and I think there is on the install screen. Whether or not it's a complete kill switch probably depends on a particular definition of "telemetry," and in today's cloud connected world it's probably hard to draw a line.

> many UX studies have shown that this applies to only a tiny percentage of users

Ironic because that's exactly the kind of thing telemetry is excellent for proving. It's quite useful for informing and providing quantitative feedback on UX.

[magicalist]

> Whether or not it's a complete kill switch probably depends on a particular definition of "telemetry," and in today's cloud connected world it's probably hard to draw a line.

Uh, no, it's really not.

[mattmanser]

Recording how long someone's had their browser open, how many hours of games they've played or how many photos they've viewed is not a log or a crash dump or in any way related to quality.

It's just creepy and invasive.

[us0r]

How about copying/pasting code, inserting tabs into code, inserting break points into code, how many projects you have, references, when you save, why your code wont compile.

Visual Studio is probably the primary reason I'm even bothering with Microsoft but fuck me they collect a lot of unnecessary shit. They claim its "anonymous"... Some of this doesn't seem so anonymous to me:

"Context.Default.VS.Core.User.Location.GeoId":x, "Context.Default.VS.Core.BuildNumber":x, "Context.Solution.LastSolutionBuildID":x, "Context.DebugSession.VS.Diagnostics.Debugger.DebugSession.StartupProject.UniqueGuid":"{x}", "Context.DebugSession.VS.Diagnostics.Debugger.DebugSession.UniqueGuid":"{x}", "Context.Default.VS.Core.User.IsInternal":"False", "Context.Default.VS.Core.User.IsOptedIn":"True", "Context.Default.VS.Core.User.IsMicrosoftInternal":"False", "Context.Default.VS.Core.User.Type":"External", "Context.Default.VS.Core.User.Id":"x", "Context.Default.VS.Core.Machine.Id":"x", "Context.Solution.ActiveProjectGuid":"{x}", "Context.Solution.SolutionSessionID":"{x}", "Context.Solution.SolutionID":"{x}"

Short of actual source they send everything (and details) you do in VS in real time.

[AlexandrB]

> ...how many hours of games they've played or how many photos they've viewed...

That part jumped out at me too. This probably means they're recording the name (or even hash?) of every executable you run and comparing to a list of known game .exe files. In addition they're recoding how long you have been using each executable (if not start/stop timestamps).

I'm not a security expert but it seems like this kind of information could be correlated with (for example) Tor exit node traffic to unmask a Tor user or other fun surveillance uses.

[WorldMaker]

Hashing .exe names is way too much work. Since Vista Microsoft has provided an opt-in way for a game installer to optionally flag "Hey, I'm a game". This was used for the mostly useless "Games Explorer", but also little bits of other functionality.

Also, the Microsoft Store has an entire category called "Game" and can just use that.

Worst case though, by "game" it might just use the Window's Xbox app's determinant for shortcuts like Win+G which is essentially, from what I gather, "Does it use DirectX? y/n".