You’re the one misinterpreting the concept of out-in vs opt-out…and the vote.
I understand your idea of opt-in vs opt-out…it’s just not accurate because it’s only opt-in. Stop assuming things are what you want them to be and actually read the vote properly.
The vote is about data collection being enabled by default (opt-out) vs the user enabling it (opt-in).
If it’s enabled by default then the data will be sent without user interaction. The user can’t choose to enable it because it’s already enabled, they can only choose to disable it.
You could have it enabled by default but then prompt the user before any data is sent, but that’s just a different way of doing opt-in.
My initial basic concept had nothing to do with the vote. It was a simple way of describing an Opt-in/Opt-out scenario. It was sufficient for the purpose.
Then we moved on from that, and I described how the choice might be implemented; perhaps in an ideal world.
Again, this had nothing to do with the vote; it had everything to do with responding to your objections, and expanding on how such a choice might be implemented during installation.
Thank U for sharing this with us. Personally I use Linux because Windows using telemetry. This was my first reason to use Linux. Now I simply like it. But… So when Linux starting using telemetry, I must ask myself: “Why use Linux when its no difference this days?” I understand that U need (or “need”) this kind of information. But it’s would be nice if U understand me and others like me. And It seems like U are trying otherwise this post wouldn’t existed.
And this voting… There seems to be missing one answer
Thank U.
“It was a good run”
No, apparently the purpose was to incorrectly explain the concept of opt-in vs opt-out. We already know what it means, if we didn’t already then the vote told us.
No it wasn’t, unless your purpose was to suggest a totally irrelevant scenario that amounted to opt-in vs opt-in.
Here the question merely was:
Should the user have to say No to prevent this data transfer - else it happens or
Should the user have to explicitly say Yes to allow it - else it will not happen.
I think our data is ours, so our permission must be asked before any attempt to send them. No matter they are anonymous, I think nobody can extract data from my computer without my permission and knowledge. Knowing it, I had no problem using mdd and sending data last week. That’s why is a clear opt-in for me.
May be the easiest way can be a checkbox (unchecked by default) in Manjaro Hello with a text “Do you let Manjaro send anonymous data to help development?”. Easy to deploy, easy to manage permissions and everybody sees after installation.
I think that answers in the forums are going to be quite biased because of the type of user that actually visit the forums. The silent mass will probably not even be aware of this poll.
Also you are always going to have strong opponents of the telemetry whatever the method you choose to implement it.
A lot of projects include telemetry and are not so open about it. And nobody seems to care. And those who care can disable it if they want (or can avoid such product if they can’t disable it).
In my opinion you should go ahead with the telemetry, but be very open about it. Don’t ask in the forums if it should be done. Just do it. But when you do it, simply explain in layman’s terms why is important and how can it be disabled if the user does not want to participate. On new installations is easy. As someone has shown, give the user the two options with the consent pre-filled.
For existing installations is a little more challenging as I think it should not be enforced by default to the user without clearly informing him/her about the thing. And I’m not talking about a note in the update announcement in the forums, I mean, the question (with the consent pre-filled) or a notice should be shown in the user desktop somehow, weeks or months before the mdd is activated. Giving the users enough time to disable it if they want.
I would be more than comfortable sharing the contents of inxi -F (without su permissions) with the Manjaro team.
However, I do not think many users would be and certain users will be very vocally against it, given that this is linux afterall.
I can also see how opt out makes the data significantly more valuable to a developer than opt in.
IMO, a decent compromise would be to roll back the amount of data collected without user intervention
( Steam survey seems like an appropriate corpus? ) and to collect the full data only on an opt-in basis.
The difference here is it’s voluntary; not only can it be turned off or on by the user, we can also see what is being sent. Plus, in my personal opinion, I’m happy to send useful data to a Community project, but not so to a greedy corporate monolithic spyware distributor.
No that’s not the point. The point is what types of devices are being used, so they can build a better product. Take a look at the data that is being sent run mdd in dry run mode.
Because the telemetry is different. You get to choose, if it is to be sent, and it’s ONLY about the hardware.
You are free to turn the telemetry on or off, you also free to move to a Linux OS that doesn’t send telemetry.
Microsoft doesn’t give you the choice, they even make it hard to move away from a software ecosystem you may have become used to.
For myself, I’m happy to help Manjaro improve and grow.
I think that any reason should be rather self-evident.
As quite far from anything like a data broker … manjaro’s interest in telemetry should be most obviously illustrated by … the norm for other similar technologies/projects.
This is roughly summed in 2 points;
1st: To better understand the makeup of users and the specific hardware+software combinations in use. This is to better provide features or fixes that are most pertinent to manjaro users.
2nd: To better bargain from a known position of popularity. This is common for anything from politics to tech. In order for a group to better advocate for itself and its constituents (users) it needs to be able to point to what kinds of numbers it has. “We represent a userbase of 50,000 people”. All the better if it can be current and/or verified. This can benefit Manjaro GmbH in various business dealings, but can also have benefits for users with examples like hardware partnerships coming to mind.
Note: I havent verified any of this with anyone else… because its simply the basic norms for this kind of thing.
Please, never compare a proprietary product with a FOOS, it is always an apples versus pears comparison, because with a proprietary product you ALWAYS have terms of use that are different for each manufacturer. If you have FOOS, however, you have a GPL-3, for example, but it is the same regardless of the software, as long as it is same version, because it’s standardized
Not to mention that Microsoft Windows has zero, zilch, nada in common with GNU/Linux, because GNU/Linux is a de facto UNIX system, while Microsoft Windows is a mix-match of essentially incompatible technologies, held together with rivets and duct tape, in order to turn a graphical user interface for a non-networked single-user operating system into a standalone operating system with a superficial semblance of security and multi-user functionality.
Opt-IN means for me that no action is taken until the user is informed about what is going to happen and does indeed tik the appropriate checkbox for his choice and confirms.
Opt-OUT for me comes in two flavours:
a) malicious - a feature is activated and actions are taken without the user even knowing about it, because any description was horribly vage and/or burried in 300 pages of EULA that noone has ever read and/or vaguely mentioned in a little anoying popup.
b) honest - no action is taken until the user is informed and has approved the pre-selected option by clicking on the button “Next” / “Continue” or whatever.
I have to say that I trust the Manjaro Team NOT to intend a malicious opt-out. At least that’s what I can see.
But I have the sneaky suspicion that most people equate any “opt-out” to be inevitably opt-out variant a) - the malicious variant of opt-out. At least that’s the only reason I can think of why some have trouble keeping their blood pressure below boiling point…
Besides that, I can easily see why the collected data is valuable for the devs and for Manjaro as a distro:
How popular is Manjaro overall → good arguments for negotiating with sponsors
Which hard- and software components are most widespread → maybe there are some legacy parts that can be dropped from support?
Maybe someone could somehow in a non foreseeable future succeed in hacking the Manjaro server, getting hold of the data base, deciphering the device_id that mdd collected now…
And then what? Send a Dell sales guy to forage the whole of Germany in search for my machine? Seriously…
Whilst some may consider b) opt-out because you have to take an action to deselect it, that’s not really opt-out, it’s a form of opt-in with a default selection. Either way, it’s not the opt-out that is being discussed here.
The vote specifically says that the telemetry will be enabled by default for opt-out.
I want data sharing to be enabled by default, but I can easily disable it (opt-out)
Translation: It’s enabled by default and it will send data without the user’s express knowledge and consent. There’s a setting somewhere to disable it if you want (and presumably re-enable it should you desire).
vs
I want to manually enable data sharing via some checkbox in a UI app (opt-in)
Translation: Nothing will be sent unless you enable it. Perhaps there’s a prompt that asks you to enable it…it could have a check box which could be selected by default or not.
If it was b) then they’d need to either:
Prompt you for your decision before it sends anything, and enable it if you say yes (which is a form of opt-in).
Enable it by default but the first time it’s run (before it sends anything) prompt for permission and disable it if you say no (which is essentially the same as 1).
In any case, there would still need to be a setting somewhere to modify the behaviour. Though it could be a command instead.
Of course if it was b) then the vote options are very poorly phrased, and this thread is pretty pointless. Few people if any really care if they have to click a single checkbox and a button versus just the button. Anyway it could just be 2 buttons.
Also in the previous thread it was meant to send data by default without user consent, and apparently prompting for permission would be “hostile”.
Some people objected to that (ie being opt-out) so this thread was created to ask our opinions about it.
Perhaps @romangg could clarify what each option entails, to finally end this useless speculation about opt-out¹.
¹ The linked post and the phrasing of the vote clearly show what it is, so unless they’ve changed their mind…
