4mo ago

Manjaro distribution is introducing a system for sending telemetry about the system

The developers of the Manjaro Linux distribution, built on the basis of Arch Linux and aimed at beginners, announced the beginning of testing a new service MDD (Manjaro Data Donor), designed to collect statistics about the system and send it to the external server of the project. The author of the MDD intended to enable telemetry by default (opt-out), but the decision has not yet been approved and, judging by the objections of some developers and users, it is likely that telemetry will be offered as an option requiring prior consent of the user (a request to enable telemetry is proposed to be added to the greeting interface after the first download).

The report includes data such as host name, kernel version, desktop component versions, detailed information about hardware and drivers involved, screen size and resolution information, network device MAC addresses, disk serial numbers, disk partition data, information about the number of running processes and installed packages, versions of basic packages such as systemd, gcc, bash and PipeWire.

The sent data is stored on the project server in the ClickHouse database and visualized using the Grafana platform. The IP addresses of users are not stored, and the hash from the /etc/machine-id file is used as the system identifier.

Аccording to the code https://github.com/manjaro/mdd/blob/master/mdd.py#L40 sends everything.

88 comments

Opt-out? Seriously? What are the Manjaro devs smoking?
- Whatever they can get their hands on, including your unique hardware identifiers
- Ad firm money.
  Maybe I'm just cynical, but my first instinct when I see stuff like this is they have a secret contract with an advertiser and are selling this information.

enable telemetry by default ... MAC addresses, disk serial numbers
Another reason to not use Manjaro. Just use Endeavour instead.
Edit: I'm not against telemetry pre se. I have the KDE feedback enabled for example but that was opt in and sends no unique data.
- It's all about trust. Manjaro has given me reasons to distrust them.
  
  When?
  Edit: I misread, though it said "trust" instead of "distrust"
- Another reason to not use Manjaro. Just use Endeavour instead.
  Endeavour could be useful if it's your first time running an Arch-based distro and you're looking for software/configuration suggestions. Otherwise, Arch Linux is fine by itself and it doesn't have telemetry
  
  I don't think anybody would say otherwise. Both Manjaro and Endeavour mean to make Arch more appealing to users who aren't comfortable with command line configuration.
  Endeavour has arguably done better than Manjaro, but yeah. They're just some configs on top of a system that does very well on its own.
  
  endeavor keeps pushing updates with cringe AI desktop backgrounds and its making me want to install vanilla arch
- Why?
  Let me put the question back to you. How do think the uniquely identifiable information will help them improve Manjaro?
  Do you think they’ve got a Russian satellite and will track down your HDD serial number from space?
  No.
  There’s lots of benefits to telemetry.
  As I basically said, if you bothered to read my comment.

network device MAC addresses, disk serial numbers
That's enough. I'm calling it evil from now on.
- Thought it's probably fine after reading the title, but this shit isn't fine. What the fuck.
- we seriously need to get the reporting domain added to popular blocklists
  
  What?
- The MAC address is anonymized with sha256, and IP adresses aren't stored.
  So this seems to me to be perfectly anonymous.
  
  Why collect such data though? And you can call some Big Tech telemetry completely anonymous too if you trust their explanations.
  
  MAC addresses are 48 bit, and half of that is just the manufacturer. So 24 bits really, and those bits aren't random, I think manufacturers just assign these based on some scheme, like a serial number. Point is you could easily reverse the SHA by brute force.
  You can't calculate any useful statistic from a hash so literally the only use this would have is some sort of tracking.
  Edit: I just looked up some data and I found someone using hashcat on an RTX 3090, which looks like it can do almost 10000 million SHA256 hashes per second of salted passwords (which are longer than 48 bit MACs, so MACs should be faster). 2²⁴ is 16.8 million, so it'll take about 1.7 ms per vendor. I found a database with (all?) 53011 vendor ids:
  
  >>> 2**24 * 53011 / 10000 / 1000 / 1000 88.93769973759998
  Yup, 89 seconds. You can calculate the SHA256 of every single MAC ever potentially issued in 89 seconds on a bog-standard 3090.

Opt-out? I see it's time for the seasonal Manjaro fuck up.
- They'll find some way to make this change break the AUR again

Why do they need information about the hostname? Is it really valuable for them to know how many systems are named daves-pc?

The report includes data such as host name, kernel version, desktop component versions, detailed information about hardware and drivers involved, screen size and resolution information, network device MAC addresses, disk serial numbers, disk partition data, information about the number of running processes and installed packages, versions of basic packages such as systemd, gcc, bash and PipeWire.
That's insane

Why on earth do they need to know hostname? MAC addresses?
- And disk serial numbers 😟

I get the usefulness of technical telemetry such as kernel version, RAM, disk space, processor type, etc... but NIC MAC? HDD serial? WTF?
- Those are absolutely ways of covertly identifying your device while technically not counting as "personal information" under privacy laws.
  
  Serial numbers are hardly covert though... but yeah.
- Yeah that makes no sense lol. Who needs MAC addresses to debug and fix bugs? No one.
  
  I said elsewhere, I hope this is just some way to track changes over time per user.
  But they need to take an anonymous hash of some non changing data or create an install id that is used for this and nothing else (e.g it identifies a unique user but not the person or hardware behind the user).
  Too much identifying info is just pushed around like we shouldn't care, it's become a real problem.
  
  The first three octets of a MAC specify the manufacturer of a NIC chipset. That could come in handy for driver debugging.
  Manufacturers and firmware versions of storage devices? You can make the argument; perhaps it would have helped figure out the SSD firmware bugs years ago.
  But stuff like whether or not you have video capture card or your current system temperature stats? Nah.. that's getting into "identifiable information as toxic waste" territory.

I just don't see a good reason to use Manjaro and many reasons not to.
- Friends don't let friends use Manjaro
  
  Like if you're going to use Arch btw, go all the way and use actual Arch.

It amazes me it's still as popular as it is and still own goaling at least once a year.

data such as host name,
Okay why do they need to know that? Why do they need to know if the computer is called "Melissa's Laptop" or "Workstation 15, Internal security division"? Seems like this kind of data could if stolen be misused and it has minimal legitimate purpose IMO as anyone can put anything as host name and while in organizations it often corresponds to use it doesn't have to for individuals. Someone could call their machine "Mack's Porn Rig" and they only use it for doing banking and a little coding.
kernel version, desktop component versions, detailed information about hardware and drivers involved, screen size and resolution information,
This all seems legitimate enough, this would be helpful for understanding the hardware their users run on and targeting features or bug fixes.
network device MAC addresses,
Not great but there is an argument for it, they could just grab and send the first 3-4 octets which would give them the info they need on manufacturers without getting uniquely identifiable data that along with some of this other stuff is concerning for fingerprinting.
disk serial numbers,
Okay, what the fuck. Why do they need disk serial numbers? What possible use is there for that. Those are used for warranty claims and could be used as part of uniquely fingerprinting a computer and person. Not cool.
disk partition data,
This is vague enough. I guess one could choose to see this as just info about partitions in use say if there's also an NTFS partition that looks like a Windows install that would be useful but on the other hand data encompassed within a partition could also nefariously be read as allowing them access to all your data. Partition layout, partition labels, and file systems used on disks available to the system would be a clearer way to put this and erase any doubt.
information about the number of running processes and installed packages, versions of basic packages such as systemd, gcc, bash and PipeWire.
All this is also fine just technical data stuff.

Manjaro is already less stable than arch, now it collects your data involuntarily? Fucking wild how anyone can use it.
- clown distro makes clown decision

That list about which data they're collecting is longer than my highschool essay

I've defended Manjaro many a time, despite the mistakes they've made. The main reason for this, Manjaro is the most stable Linux distro I've used.
However, the main reason I ditched Windows as my primary OS was telemetry (and bloat). If Manjaro introduce this, it absolutely must be opt-in.
I actually contribute to the Steam hardware survey as I want to ensure Valve, but more so hardware manufacturers, are aware desktop Linux systems for gaming and creative work are viable. But it's my choice to contribute.
If Manjaro don't implement this as an opt-in then I'll be installing Arch. It will be a pain to configure my software again but needs must.
- If manjaro is the most stable distro you’ve used you can’t have used a lot
  
  I mostly used Ubuntu based desktop distros and frequently had issues with the 6 monthly update cycle. Problems with Fedora too. I have not had a single update issue with Manjaro. I often have different distros running in VM's and whilst Arch has been the most reliable, most are not.
  I also setup loads of Linux servers in my I.T. job that I used to have, so I have plenty experience.
  The bottom line is Manjaro desktop has been ridiculously reliable for me. Therefore other peoples hate of it washes over me and is meaningless.
  
  Yeah the Manjaro devs have a long history of gaffes not to mention the infamous one with PGP keys requiring users to reset their system clock

This may be illegal in EU if they don't use opt in. ~~Even then it may be illegal for under 18 year olds to collect MAC addresses and disk serial numbers, as those can potentially be used for identification.~~
The data is anonymized, and the IP is NOT stored. So I'm not sure this violates GDPR?
From the code we can see the machine ID is anonymized, sending only a SHA256 checksum.

def get_hashed_device_id(): # Read the machine ID with open("/etc/machine-id", "r") as f: machine_id = f.read().strip() # Hash the machine ID using SHA-256 to anonymize it hashed_id = hashlib.sha256(machine_id.encode()).digest() # Convert the first 16 bytes of the hash to a UUID (version 5 UUID format) return str(uuid.UUID(bytes=hashed_id[:16], version=5))
This makes it somewhat a nothingburger IMO.
- That's not anonymous, that's pseudonymous.
  What is the point of this? The machine-id already looks to be some unique random number, so you're calculating another unique random-looking number from that, might as well use the original number.
  You can't glean any useful information from a unique random-looking number that would help with developing Manjaro. You can't calculate any statistics from that. The only use is tracking.
  Edit: And as mentioned in my other comment, reversing the MAC SHA by brute force is trivial, so that one at least (and possibly the other hardware serial numbers they collect) shouldn't even be considered pseudonymous.
- Nah, it's still considered Personal Data under GDPR, because it's possible to connect to natural persons. So GDPR applies. And this is illegal, there is no legal basis for processing this data.
  
  because it’s possible to connect to natural persons.
  That's debatable, and is only based on the claim that it's just a 24bit decoding that can be brute forced. I don't know for a fact that it's true that it can be boiled down to 24bit.
  I checked my own /etc/machine-id, and the folder doesn't even exist, so what exactly is supposed to be in it IDK. And yes I use Manjaro.

Why do they need half that data for a derivative of a distro? Fuck off. I don’t care if someone collects the model number of my GPU or whatever but that sounds like personally identifiable tracking data, not basic “telemetry” data to set development priorities or whatever.

Once again proven right that EndeavourOS is the superior downstream Arch distro
- I moved one of my computers to endeavor, but one is still on manjaro and the contrast is kinda hilarious. Manjaro machine always gets funky after updates, it struggles to deal with sleep and hibernation, and it feels slow even when its like 4x as powerful as my EndeavourOS machine.

Glad i said fuck it and went straight to actual arch when i wanted to try arch based. Literally like 9/10 times i hear manjaro brought up its not going to be in praise. Ffs lol

users can be identified
probably Opt-out (still in discussion)
Two nogos combined makes nonogogos. Why do they need host name, MAC address and disk serial numbers? Why can't people set how much they want to send in, like KDE Plasma does? Will the data be shown to the user before its send in? Steam does that perfectly (show data and its opt-in) and that is even a proprietary application. Telemetry is okay if its done right, without user identification, opt-in and not hiding whats sent, preferably in multiple levels of what is being send.
I used Manjaro before and switched to EndeavorOS because I was not happy. Now I am. Manjaro can't stop being stupid (not the users, I'm not attacking any user here, only the maintainers or developers of Manjaro).
- The way I read it, the developer wanted opt-out but it's likely it will be opt-in. I'm find with opt-in and vehemently against opt-out for telemetry.
  I would prefer the information was statistical only. Rather than hostname (making the assumption they only want hostname to be able to somehow separate the data to follow changes over time), a much better idea would be some kind of hash based on information unlikely to change, but enough information that it would be unlikely possible to brute-force the original data out of the hash. So all they know is, this data came from the same machine, but cannot ID the machine. Maybe some kind of unique but otherwise untrackable unique ID is created at install time and ONLY used for this purpose and no other.

I don't get why someone would use Manjaro after so many fuckups.. If you don't know what I'm talking about, you're either too new to Linux or don't care. Just look for "manjaro certificates" or "manjaro drama" and you'll find out for yourself.

With archinstall, anybody can install Arch in 10 minutes nowadays. Why use Manjaro ?
- To many options? A new user might be confused, by for example choosing a the correct disk layout.
  
  NGL on pretty much any install, I'd end up looking up pros and cons of every filesystem AGAIN...
  ... It's BTRFS now. Simple. Easy. Lol
  But it was a lotta research to reach that conclusion. So yeah I get that newbie apprehension!

hostname? MAC address? serial numbers? does "partitionx data also include names and GUIDs?
why would they need these? what is wrong with them??

Another reason to hate manjaro.

Dammit, Manjaro. Why you gotta be WEIRD?! I used to love their branding, but they keep doing crazy things that would clearly alienate the userbase that's left...

I tried Manjaro last year and I hated it.
Something about the distro would lock up my PC, it would freeze from time to time.
I disabled the standby/sleep function, but allowed my monitors to go into standby. But if I left my PC for an hour or two my screens would not wake up, different types and brands. I had so many issues with Manjaro and while speaking with a friend I told him I had moved over to Nobara but he was still on Manjaro. But then a few weeks later he mentioned he was running Nobara. Seems he also ditched it.

My brother's been injecting himself with Linux?

Don't like it, don't opt in
Even Debian has popcon
There are lots of benefits for developers to gather telemetry.
Don't like that? Fork and do your own distro (presumably though you don't contribute anything to open source, so id expect such people to simply whine and get angry at contributors)
- Debian popcon is opt-in, first of all.
  https://popcon.debian.org/FAQ
  Q) What information is reported by popularity-contest ?
  A) popularity-contest reports the system vendor [1], the system architecture you use, the version of popularity-contest you use and the list of packages installed on your system. For each package, popularity-contest looks at the most recently used (based on atime) files, and reports the filename, its last access time (atime) and last change time (ctime). However, some files are not considered, because they have unreliable atime. For privacy reasons, the times are truncated to multiple of twelve hours.
  [1] i.e. the dpkg Vendor field, see dpkg-vendor(1).
  So no fucking MAC addresses and machine-ids and harddrive serial numbers and stuff.
  They only want package statistics, the point being to have statistics about the popularity of packages, mainly so they can be prioritized for the CD/DVD isos. You know, information that actually has a use, not hardware identifiers that can only be used for tracking purposes.
  
  Each popularity-contest host is identified by a random 128bit uuid (MY_HOSTID in /etc/popularity-contest.conf). This uuid is used to track submissions issued by the same host. It should be kept secret.
  Oh, and by default, IP, unless usetor is enabled
  A machine I'd is just a hash too
  Can you explain to me how you track Mac address, serial numbers over the internet.
  Just fyi, the backend project I made 20 years ago was hardware related. There's potential reasons to grab this info..
  But, if it is a concern, I'm sure they'd welcome submissions to improve the parsing and allow things to be filtered.
  In fact, popcon could be used for digital fingerprinting technically
  In all likelihood, op never spoke to the manjaro developers either
- Yeah, my only concern here was if it was opt-out. That'd be bad.
  Now I completely understand the developer on this. This is useful info to have to help decide future changes/features and general direction, but balancing the right to privacy means this kind of data provision should ALWAYS be opt-in. Microsoft, you hearing me here?

88 comments