The best solution would be to disable telemetry to the capacity you are capable or is possible without breaking something and block pings as backup to when it's enabled again with updates and repeat.

Yeah, that's what I was thinking..

Exponential backoff is baked in most libraries. So probably not.

Is there a service whit which you can automatially send fake answers to certain events?

Teams no linger does that. It just goes to an error page if it fails to load twice

To me, telemetry would be like a sofa company wanting to put some cameras in your home to see if you're using the sofa the way they thought you would. It just feels... off.

“90% of crashes happen right after the player uses a grenade”.

Imo, a simple opt-in crash report gets the job done. Technically it is telemetry, but a crash report is more justified than a "where have you clicked" report.

telemetry data for software from reputable companies does not get sold

There's just no trust in companies to not sell my data. I cannot trust Microsoft nor Google nor any other company to not sell my data, having seen the shenanigans every single company is willing to pull off to get a cent more a year.

In general I agree, but users should be able to make that decision themselves. I do not understand why you can't turn off telemetry, when it would be trivial to offer that option and so few users would bother to use it.

As a programmer: "your data is boring. I am not interested in leveraging this for anything besides getting the service you are using to work as well as possible"

Also me as a programmer: "yo, you don't need that data, stop asking for it. Ohh, your app is broken because it can't access permissions? Yeet."

I don't want you to know anything about me or my device. Simple as

How about shit breaking because everyone at some point is a bad programmer? Even Apple Music doesn’t work when I walk into the elevator until halfway through presumably because hitting play sets of a bunch of useless blocking network calls for music I have saved locally.

What those calls are, I can’t say for sure. Downloading artwork, license checks or telemetry. I’d venture to guess it’s the latter since music will play with placeholder artwork on a slow connection and license checks aren’t required if the subscription was recently validated (works offline for days).

But who really knows. I never bothered to inspect the traffic. The point is, if a company like Apple is creating such a crummy experience for a function so absurdly basic, you can imagine how easy and prevalent telemetry based user degradation is. Go browse the web with a tracker blocker and tell me it isn’t snappier.

PS: I’m also a programmer and collect error reports. So many developers will forego using connection pools, much less collect data with async api’s.

And let’s not even get into how telemetry is a shit tool that is misused 99.99% of the time and only used to surface popular features that aren’t necessarily good features only because we attach causation to every metric (x feature is highly used, therefore it must be good).

So why not ask, "hey you want to share some telemetry to help us improve the product" then?

It's what all reputable companies or projects, I am aware of, do.

Do you consider Microsoft a "reputable company?"

A few things (and I enable it usually, for the record):

• Not really the user's job to help you with anything especially related to your boss
• "from reputable companies does not get sold or used for marketing purposes. Our lawyers make sure of it," fuck man, this made me laugh. Good one

Policies get updated, companies are bought and sold, laws change, and most crucially of all, data gets leaked. It doesn't matter how airtight your asshole is puckered up or how many isolated networks are involved. It gets leaked. Leave the decision up to users about it and in particular maybe let them have full control of their networks.

Oh thank God, someone reasonable.

That would be good, considered the user could actually have a choose to opt out the telemetry. Windows don't ask you about gathering data for telemetry or "other" reasons.

Yup. You need more like a logger to get the actual measure of the data.

Which could in turn make their pc run worse because it'll constantly be retrying telemetry calls

...pay 9.99 to have us do nothing

I can concur. MS do not know the concept of retry limitation.

You can even know when I'm working from home just to the number of blocked MS tracking calls in my pihole log 🤣

The other side of that is that the telemetry data never gives you a "why" of something.

For example, users might spend a long time at a screen because they are thinking about what to do, or they are confused by the options and can't figure out which option they need.

This is why a QA team coupled with a large amount of beta testers is invaluable and necessary.

Telemetry, in the context of software development and UX design, is either a decision by the misinformed or just an excuse to save costs by axing the Windows QA department.

In reality it's likely the data is being sold off. But in either case, that's data Microsoft isn't entitled to (from a moral/privacy perspective).

I totally agree, but where I have a problem (and I imagine a lot of other users here) is that you can't fully opt out. You can only set "minimal" tracking but not none.

Also, Firefox, lemmys beloved browser, sends telemetry by default. You have to dig through menus you didn't know existed to even find out, and then disable.

Not only to Mozilla, but third parties as well.

Devil’s advocate: basically the only proper way to figure out how people are using your product

Focus groups and customer surveys work really well.

I think it's horrible and has no place at all in a operating system.

Conviva is "video streaming analytics." Any site with video content is trying to track who uses it, how long they watch, etc.

Wow, do you have a lot of Amazon devices? That's a lot of calls to Amazon.

Where are the Conviva calls coming from?

And who doesn't do it?

OSS operating systems. The more proprietary software you run, the less and less you actually own your computer and the more it becomes a tool to advance the interests of megacorps.

we just can’t avoid it anymore

Speak for yourself. Besides, all-or-nothing privacy is a false dichotomy. Giving out less personal data is still better than giving out everything, and you don't need 100% privacy to be unprofitable to advertisers.

If ET had been an android-phone he had been long called home before the intro started

That's a good one! But to be fair, Apple calls home just as much. They just don't sell that data (yet).

Desktop linux doesn't have any of this. And one day we'll get real linux on phones too (with full featured support).

It's always MS who has people on the internet defending them when they do it, to the point that it looks like MS social media presence is being carefully managed, like Apple. I saw it happen on Reddit too, and it happened less and less the more they got called out. The same "$70 a year for Office 365 is so worth it" is a talking point when I was on Reddit. Apparently, people here say the same thing.

All hail GrapheneOS

Get a device that is supported by degoogled android systems and get rid of all the garbage. My recommendation is always GrapheneOS on a pixel, just because you can reinstall google play services if you really need to and preserve usability, while massively boosting privacy and security.

I used blokada on my phone and it blocked around 10k calls per day. I've since moved to use Mullvad as an always on VPN and turned all of the blockers on. Haven't setup a pihole for my home yet because I didn't have access to the router.

My "smart tv" too.

Ubuntu lost me years ago. I still use the server version for.. servers. If you want something rawk sawlid for servers, go Ubuntu. Otherwise, go Pop. Or Elementary. Or Mint.

Don’t like your hand held? Fedora.

Hate yourself? Arch.

Draw your entire personality from knowing what a transistor is? Gentoo.

Despite your sarcasm, it really is.

I can recommend for that. It logs to RAM and only writes to the SD card once a day (or more/less, if you choose).

It's a must while using PiHole imo.

I've recently read about how to fix corrupted SD-cards in Raspberry Pis. Corruption can happen when the Pi is unexpectedly losing power during a write to the SD card.

To avoid corruption, you can change the (boot partition of the) SD to read-only, requiring either USB storage (flash/HDD/SSD) or a writable secondary partition on the SD card) if you need to save anything locally. The system itself will run fine without write access. Only your files could be at risk if you lose power mid-write.

You can also configure your system to boot from USB storage instead of SD card. Keeping the system partition read-only is probably still a good idea, if possible in your setup.

Modern versions of raspi-config offer a similar read-only overlayFS functionality out of the box! sudo raspi-config, go to Advanced Options, then enable the Overlay FS: Enable/Disable read-only filesystem feature.

Source: Shane S. @ https://raspberrypi.stackexchange.com/...

I had that issue too. Had to switch to running it on my NAS.

I'm wondering what could have caused that. Uneven power supply maybe? My pihole has been chugging away for years without much more than the occasional update when I remember.

High endurance rated cards are good for that. Don't risk using the cheap SD cards

Good try Microsoft

Unless windows update needs to check for something every few seconds... It's not that. This is behaving more like real time streaming of events, constantly sending to Microsoft, to a domain called events.

Probably sends what the user is doing all the time.

Probably right around zero.

Mine is a docker container on a Linux server in the hall closet.