With this example, I wasn’t trying to insinuate that Fedora is doing the same, but rather to present an example that would better address those that fear and what they fear.
There’s also the aspect that a lot of people feel that the decision was already made that this was going to happen regardless of what everyone feels about it… and this discussion is nothing more than PR spin to make people feel that they’re being heard. Some changes will no doubt be made to get some people on side… but does anyone actually believe this change was ever going to be entirely rejected/denied based on feedback from this thread? I don’t. I 100% believe the decision was already made and now it’s just a matter of getting the community ‘on-side’ with whatever adjustments are needed to get people to go along with it. Maybe I’m wrong, but that’s how I feel.
Due to that, it heightens the worry of some (like muself) that future changes will take place that we also dont agree with will be implemented and that others who do have malicious intent will take advantage of it. Afterall if there was never a chance of this being denied from the start… how are we to believe that a later line being crossed won’t also be pre-decided.
I feel like the comparison was intended to be less Steam to Fedora and more Steam hardware survey to this metrics proposal since both are intended to help guide developers towards the features, toolkits, hardware, and platforms that are used by the most users so they can develop the best experience for the greatest number of people.
I doubt Valve is making any money on the steam hardware survey directly, and I imagine both users expect the respective groups of developers to be working on stuff that actually matters to them (or at least a lot of people) rather than, say, working on optimizing the platform to run on an iPod, or add driver support for a platform that hasn’t been relevant in decades
yeah, that makes a lot of sense too. I definitely agree that it would benefit the long term health of Fedora to work towards becoming more diverse in where resources come from.
From what I’ve heard, Red Hat’s policy is more “upstream first” though - which sounds like “hey, we want to do this, lets ask the upstream community first, and try and work with them, and if not we can do it ourselves”. I suspect if Red Hat wanted to quietly add metrics to a ton of Linux systems, they could easily add that to CentOS Stream or RHEL without going through the process of asking permission from anyone.
Fedora certainly has deviated from Red Hat’s wishes before (i.e. with filesystem choice)
That’s fair. But do we have a full detailed set of the data collected by Steam?
Because that’s required in order to do a fair comparison.
I’m not too much of a gamer, so I am not already familiar with them.
But from Fedora’s side, I was not too fond of the idea of collecting cameras and peripherals information for example… I feel it’s a bit overreaching.
I don’t think that’s a fair assesment, given how successful the pushback of the original proposal was.
The fact that the “stricly red-line opt-in as an absolute requirement” and the abundance of dark patterns that were introduced one after the other in the original discussion thread have been completely withrawn. I rather think that the community recognizes that the data can be valuable for development, the new proposal better respects both privacy-by-default AND user choice, and therefore members of the community have become more accepting.
I dont use steam either, but they have the aggregate data publicly available here
i previously talked about how i think this could be a cool pattern to follow because then users get access to the aggregate stats over time (thus getting something useful - to them personally - in return for their data), and maybe it could become an independently managed system (by a group like CHAOSS) to ensure things are neutral and not able to be steered by any one distro (and maybe it could be a cool additional way to measure Linux desktop market share, or strike some fear into big tech companies when they see the graphs spike as people switch to Linux immediately after they announce some invasive change)
I don’t think that’s a fair assesment, given how successful the pushback of the original proposal was.
The fact that the “stricly red-line opt-in as an absolute requirement” and the abundance of dark patterns that were introduced one after the other in the original discussion thread have been completely withrawn. I rather think that the community recognizes that the data can be valuable for development, the new proposal better respects both privacy-by-default AND user choice, and therefore members of the community have become more accepting.
I hear what you’re saying, but personally, that always felt like a Motte-Bailey tactic for PR purposes. Start out with a proposal (opt-out) that you know will get pushback, so you can then retreat to the more defensible position of “opt-in”, seem like the reasonable person and that you’re listening… so that you still get what you want. If the proposal started with Opt-In, and there would still have been pushback but there wouldnt have any position to fall back to in “negotiation” that gets them what they want… telemetry.
The original Opt-Out proposal was crossing so many lines I can’t believe for a moment that anyone in the libre and open source community would genuinely try to push such a thing. But hey… maybe I’m out of touch… and that sort of thing is accepted by more people that I’d imagine these days.
That’s what I initially thought as well… until the original proposal was (as far as I know) rejected even with the Fedora Project Leader (@mattdm) plus others in decision-making places being on board. This might seem bad, but there are people that come from the community (not from RH) to the top and make decisions too (from what I understood anyway). From the Fedora Docs:
The elected positions cover all Fedora subprojects that are not under the engineering or mindshare banners, and the community at large. One specific responsibility is to represent the voice of individual contributors to the Fedora Project. Each representative will also work on specific goals which they bring to the Fedora Council as highlighted during the election process.
[ … ]
No person who currently holds another Fedora Council seat can be elected. If a seat becomes vacant, the Fedora Council will arrange for a temporary replacement.
Even the FPL (Fedora Project Leader) needs consent from the Council to be the FPL:
The Fedora Project Leader is hired by Red Hat with the advice and consent of the Fedora Council.
Not to mention that, if this was already a done decision, why go through the hassle of announcing and gathering useless community feedback on Fedora Discourse of all places? Twice even?
I am just calmly following how this telemetry saga ends now
I was heavily involved in original discussion, in the new one I just voted against in the poll and that’s it.
I already made my decision too - I am not gonna use distro that comes with telemetry.
Good thing we have alternatives.
That was what I tried to convey earlier (likely entirely poorly). If an open community would spearhead a bad decision, the project (or parts of it) will simply be forked and the original will slowly die. However, if such changes are done slowly and without (bigger) repercussions, others might join in and do the same. Once it becomes mainstream, more ‘features’ can be introduced one step at a time with little choice other than to go along with it since there is no alternative anymore (unless you want to disappear in a niche).
Also, pressing through with the metrics feels pretty much like there is an ulterior motive behind it, which may start to unfold its true color in another 25 or so years when choices like this has long been forgotten by the mainstream and accepted as ‘unchangeable’ and/or ‘the way it is’.
As a comparison, cloud solutions were attempted as early as 1999/2000, but failed miserably. However, something changed over the years and something once viewed negatively is now mainstream with a lot of providers.
I am inclined to be against such metrics, in particular since there were mentions of alternatives, which (compared to the current opt-in method) has likely the same effect and, as others stated, are safer on the OS side and/or for the user.
I mean, I would not mind getting offered a survey once in a while and would likely fill that out, if done right, be it by visiting Fedora’s various sites or even an option by e-mail (unless one is getting bombarded) is totally acceptable.
I feel the users using the OS should be personally reporting their issues. It encourages cooperation from both sides (users using the OS want it to work good, devs want their OS to work good for users), and is probably less-ambiguity as to what to work on as QA (if 20 forum threads appear with NVIDIA drivers being busted, distro QA should probably rush to figure out what’s going on).
But I’m also not entirely sure what metrics would do for my benefit, which kind-of has two sides:
I document everything I do and talk about it publicly, so my suggestions are pretty visible to people at QA
I see apps/methods done popularly that I knowingly don’t want, but the metrics showing a majority is going to sway dev towards that vs stuff I actually use
The latter I somewhat feel about Wayland that I couldn’t find as a viable alternative for 8 years up until F41/GNOME 47. Others raved about it for years, distros defaulted to it years ago, but my real-world use had me prefer Xorg up until a month ago. Metrics from me wouldn’t have done anything about this, and while I’ve been plenty loud about it, I’m pretty sure someone just randomly decided to fix it
Basically, I feel metrics could be used to sway dev at less of a benefit to end-users, vs distro devs curating improvements from its users personally; QA seeking things to QA for their users, vs QA’ing against a mechanical list that reduces users to percentages. But I’m sure a mix of both is what most devs do and that’s cool.
I’d have enabled metrics yesterday (would have been nice for F41 ) in hopes it’d be useful to someone, and don’t have any reason to believe my metrics to be used by Fedora/parties nefariously. I like the idea of it being opt-in!
Hi, status update time! I haven’t actually done any work on this since the proposal was approved and will be working on other projects for the foreseeable future. Metrics for Fedora is still very important for Red Hat, but realistically I won’t be getting to it anytime soon due to other responsibilities. Red Hat will try finding somebody else to replace me on this project.
