GNU/Matt 
@blackard @[email protected] anyone can make a proposal if it meets the community guide lines. I could go propose #Fedora changes its name to BlueHat #Linux and the community would consider it if it met the work guidelines and they thought there was a benefit
john r red-horseYeah, this is a darn good example of what's happening. (Not really; that was sarcasm.)
This proposal was made by a Red Hat employee and appears to be fully supported by the Fedora Project project manager, who's also a Red Hat employee. Among other Red Hat employees.
Adam Williamson @jrredho
And it's opposed by several red hat employees, including me. This is how we do things in fedora. Also, matt has not actually stated any opinion on it yet. Matt considers things from every angle and tends to have a thoughtful opinion when he gives one.
We have rejected proposals from RH devs many times in the past, including ones from the desktop team. I wouldn't make any assumptions about how this one will turn out until the process is done.
@gnuplusmatt @blackard
I'm reading the same posts, Adam, so I see that there are plenty of Red Hat employees who are opposed.
But I can also tell which way the wind is blowing with people who are further up the food chain. If this decision was up to Matt? Throw in the towel, it'd be done.
I find this whole story line incredibly sad, to tell the truth. They aren't going to learn anything innovative from this telemetry data, and what they could learn they could get from RHEL users.
@jrredho
I think that's doing people a disservice. Michael and Matt are both good people who believe in Fedora and work hard for it. They deserve better than they're getting. We can disagree about things without going nuclear. And it'd be a mistake to think you know what Matt thinks just because he's trying to keep discussion civil and fair to all parties.
@gnuplusmatt @blackard
I think that's doing people a disservice. Michael and Matt are both good people who believe in Fedora and work hard for it. They deserve better than they're getting. We can disagree about things without going nuclear. And it'd be a mistake to think you know what Matt thinks just because he's trying to keep discussion civil and fair to all parties.
@gnuplusmatt @blackard
My take is that the bulk of the discussion is how best to do it, not if it's going to be done.
Do I think any less of the folks involved, including Matt and Michael? No. Do I think that from their perspective this is coming from a place that's insincere? No!
I still have hope it'll be voted down by FESco, but those two have made their decision on what they want; they are going about the business of socializing it.
@jrredho
Well, the discussion about whether to do it will mostly take place between fesco members, assuming Michael decides to continue with the proposal, since they're the ones who get to vote on it. I'm sure they're following the discussions and those will factor into the decision.
@gnuplusmatt @blackard
Well, the discussion about whether to do it will mostly take place between fesco members, assuming Michael decides to continue with the proposal, since they're the ones who get to vote on it. I'm sure they're following the discussions and those will factor into the decision.
@gnuplusmatt @blackard
@jrredho you wanted a personal opinion from Matt, btw? Here's one!
"my suggestion would be to implement the must-choose yes-no option as several have suggested (that is, “explicit choice”)."
F40 Change Request: Privacy-preserving Telemetry for Fedora Workstation (System-Wide)
For what it’s worth, Canonical says that about 66% of users leave the switch alone in Ubuntu; the other 33% change it. I would expect the number for Fedora Workstation to be roughly comparable (I can see reasons it might go up or down.) Stepping aside from “trying to facilitate discussion” to personal thoughts,[1], my suggestion would be to implement the must-choose yes-no option as several have suggested (that is, “explicit choice”). We can compare against DNF Countme estimates, and if the ra...
Yes, and notice how this is no longer a discussion on whether or not to implement telemetry from Matt?
It's migrated to a discussion on how to do it.
This is exactly what I'm saying.
@jrredho I mean, it can be both. Matt's implicit opinion is "we should implement it, but with an 'explicit choice' design". My opinion, which I expressed in the same discussion, is "we shouldn't implement it at all".
I think you're seeing more "detail" posts than "yes/no" posts because, well, there's only so much "yes/no discussion" you can have. Bob says yes. Alice says No. That's...kinda it. It's inherently possible to have much more discussion about the details of the 'yes' option.
@jrredho That doesn't mean that some form of the 'yes' option winning is a foregone conclusion, though. It just means that people like to get ahead of the details. It also obviously means that the details of the "yes" option are *important to making the yes/no decision*. There aren't really any details about the "no" option to discuss, because the "no" option is just...don't do anything.
I think Matt's opinion carries alot more weight with voting FESCo members than yours and, most definitely, mine. :(
Linking that telemetry code into the, say, gnome-shell, might well be a privacy exploit path that wouldn't be there without it.
And, finally, in my opinion, even working on it sends a message that the Fedora Project Mission Statement no longer truly guides the project.
Btw, I read your post there. I appreciated your take. So thanks for sharing those thoughts there!
@jrredho
Data from RHEL users would likely be very different to data from Fedora users, because they're very different sets of people who use the distros differently. I'm not sure it'd make folks happy if we made decisions about how to build Fedora based on how other folks use RHEL.
@gnuplusmatt @blackard
Data from RHEL users would likely be very different to data from Fedora users, because they're very different sets of people who use the distros differently. I'm not sure it'd make folks happy if we made decisions about how to build Fedora based on how other folks use RHEL.
@gnuplusmatt @blackard
All I've read is telemetry would be related to GUI/DEs, not innovative framework aspects. The bulk of the discussion early on was about down selecting Settings menus in Gnome for heaven's sake. Any class of user can give you that data.
That ain't deciding to push systemd or pipewire or btrfs or container frameworks, etc. The full history of those types of thing must be truly be vast. Respect. And so much of that germinated in Fedora without any telemetry.
@jrredho
Well, yes, it's a proposal from the desktop team, after all. GNOME is what they work on. (I think it's getting a bit lost in the discussion that this proposal is for *Workstation* only; other spins/editions wouldn't get the system at all). But this absolutely is an area where data would be useful! I see that myself.
@gnuplusmatt @blackard
Well, yes, it's a proposal from the desktop team, after all. GNOME is what they work on. (I think it's getting a bit lost in the discussion that this proposal is for *Workstation* only; other spins/editions wouldn't get the system at all). But this absolutely is an area where data would be useful! I see that myself.
@gnuplusmatt @blackard
@jrredho
So, real examples. Currently because they have no data, they work the way systemd or pipewire or btrfs were done: they build what they think needs building. So, they built a calendar app and a contacts app, because a desktop "ought to have" those things.
Then we (QA) find bugs in them, and then we have a debate about how important those bugs are. Should the desktop team drop other important work to fix them? Well how many people use those apps? We don't know.
@gnuplusmatt @blackard
I'm not saying the data aren't useful. I'm saying that that isn't the stated mission of the Fedora Project, as they've written it, and, until this, they've lived it.
None of that data would've given y'all an indication about pushing Pipewire or systemd, et al, imo.
Has the Project run out of ideas for new directions?
I also realize who the proposal is going from, and who they work for.
If this change gets approved it will move into all the editions and spins. It'll just be a matter of time.
And messing with Gnome, settings or otherwise, still ain't innovation. Everything about the goal is flawed, imo.