- Telemetry is not bad
I canât continue after that one
It isnât. Itâs a matter of WHAT is sent.
Is âhow many people are downloading packages from our serversâ telemetry?
Technically wouldnât it be? itâs gathering data.
I donât want to start flame-war
But to me itâs like a logical fallacy, regardless of what is sent.
Itâs kinda like classical dilemma:
- Big government is ok!
- WellâŚwhat about dictator coming to power next and use all that big government?
Iâd say generally you donât want to give anyone any possible power and means to use your personal data with or without agreement, because even if you trust that someone now, it can quickly change hands silently, which happens all the time.
So letâs say if i really trust some distribution or program, and then it introduces some telemetry - i would think twice before keep using it, to me it is already a breach of trust.
For example i donât want to use KDE anymore because theyâve introduced user feedback which is cool because itâs optional and opt-inâŚBut what about some bugs which can accidentally enable it?
Can you guarantee bug-free experience in such big DE?
Years of experience tells me no. Even if we assume no malicious intent.
It happened to me before too, so no thx - personally i donât need any additional potential point of failure in my trust
More like statistics iâd say, usually telemetry is meant as anything personal identifiable, hardware, software info, ip etcâŚthen moreâŚand moreâŚand mooooore
No, thatâs not telemetry. Thatâs just server logging the requests.
Telemetry is when a program running on a userâs machine connects to some server in order to send it information it collected. Spyware is when it does it without the consent of the user.
I kind of agree with the video in the OP that telemetry is not necessarily bad. But I think it needs to be specified exactly under which conditions telemetry is not bad, and I can think of these two conditions that would make me not consider telemetry bad:
- It is disabled by default and has to be explicitly enabled by the user. This is to ensure consent by the user.
- It shows to the user exactly what info it is sending and every time it does it. If this is annoying, there could be an option to disable it, but it should be enabled the first time it happens.
What I vehemently disagree with the video is about snapd not being bad. Yes it is. Snapd is the devil. Iâd rather run windoze than have snapd installed. The only distro that treats snapd properly is Linux Mint, bless their hearts!
Also, he is completely wrong about Arch. I use it on all my production machines, never had any issues. It does not break, because Iâm not an idiot (at least not an exceptional one), and even if I were, Iâd still use timeshift
.
âArch Linux: The Not-an-Idiot Proof Linux Distro!â
0,01% is a pretty decent proof factor
Only Linux from scratch is better!
It should definitely always be opt-in, and always easy to understand the options. BUT I completely fail to get the idea of running say Firefox Nightly and NOT have it enabled (as a user) since you actually donât help Mozilla at all if you DONâT let them see your crash reportsâŚ
This. If youâre running Nightly then you really need to be allowing submission of bug reports and telemetry data because it helps them develop the browser.
And if you donât, they do things like remove the RSS reader and Panorama because âpeople donât use themââŚ
When it comes to Firefox Nightly and development versions like that, I completely agree, telemetry makes perfect sense! Itâs helpful to the developers to know when stuff goes wrong and under which conditions, and also which features are used the most and deserve the most focus.
But still, the two conditions from my post (which I clarified a bit better by editing my post) should apply, in my opinion. For example, there could be a popup dialogue when the program is run for the first time that says something like: âThank you for using our development version. We would like to collect such and such data from you, do you consent? For more info, please visit this webpage that tells you exactly what data we collect.â
The only way you deserve the trust of your users is by being 100% transparent and honest. One violation of that is enough to destroy all trust.
I wouldnât go QUITE that far - but half is the most I could give him. I wouldnât use Arch for a server situation, unless it was one I was personally âbaby-sittingâ, and it was the only one. The big issue is the unattended operation - and no changes are wanted in those conditions!
For any other âproductionâ use, I have found it be as stable as anything out there - and the issues it DOES come up with are minor and easily fixed - usually by the time you find out about it! Timeshift is one factor (update rollback on demand) and another is that generally machines are specâed with enough in the way of spare resources to enable instant amelioration - whether by swapping in a backup, or by switching to an âon machineâ alternate (from EnOS to Arch for instance) while working with the same data. No advantage to not being on âthe edgeâ to be found in that - and improvements that may enhance your production are more frequently found!
I can see how servers can be an exception to this. When it comes to servers, nothing beats Debian Stable, simply because it does not require a lot of maintenance.
But if you can maintain it, there is nothing preventing you from running a server on Arch. It will not be any more unstable than a Debian Stable server, it will just require more maintenance.
But even if we take servers as a complete exception (and unquestionably accept that Arch is universally bad on servers, which is not, but for the sake of the argument), thatâs just one exception. A production machine is any computer on which you get stuff done. And for personal computers, Arch is completely viable (and I count EndeavourOS as Arch), though not idiot-proof.
If you have a deadline for some project tomorrow and you havenât even started working on it yet, and you decide to update your DE and the software you need for that project, and you donât have a backup, and something breaks, thatâs not a problem with Arch. Thatâs a PEBCAK
I have to agree. This issue of telemetry is very wearing. Itâs always accompanied by a bad attitude at anyone who disagrees, trying to force agreement, which just confirms that telemetry isnât a good thing. Not possible to trust, when pressure and criticism is the response. Respect of data is a perfectly normal thing to want, and saying âyesâ to some apparently tiny amount of data collection does always leads to pressure to accept the next little bit of data collection. The value of privacy and peace of mind is far far above any need for telemetry, especially in this day and age.
Without doubt, I drop anything that starts to sneak in, or openly force, any telemetry.
I guess I somewhat agree with most of his points. I use Arch on my web servers but thatâs because theyâre not important (just personal sites) and I find it fun. If I was using them for business purposes I guess I would probably still run them, but do the updates far less frequently and probably have some redundancy in place in case something dies. Only once have I had an update screw me over and Iâve been using Arch for around 10 years so Iâm pretty confident with it.
I donât have an issue with telemetry. I understand some people are very paranoid about their privacy but for me the kind of things I wish to remain private are rather specific an minimal and itâs generally never an issue when it comes to the information most programs wish to gather.
As for snaps, I donât care what Ubuntu does. 5 years time snaps wonât exist and theyâll pick up Flatapks or AppImages, or even both. Never place your bet on Canonical technologies cough, Mir, cough cough Upstart, cough Unity, cough cough Ubuntu One
That does make firefox bad . Because it uses google
as search engine by bdefault . When google
is set as default search engine it will send some info to google on start up . So even if you choose DuckDuckGo
in settings some info are sent to google already . And I do hate that .
Because sending the info or not
is a small issue ( or âthingâ ) but I canât even decide that .
So itâs some serious issue . The only thing we can do is either create a company of our own or relay on community projects Like Endeavour
The video just felt like heâs reading off a list he thinks he agrees with. Thereâs also the âitâs proven people click on faces in thumbnailsâ thing for Youtube. Thereâs no passion or interest, the points are vague. âA fringe part of a community can be not niceâ and âyou shouldnât attack people for what they do with a computerâ â very unpopular opinions right there.
Snowdenâs words