I can't help but see the similarities in their design to concepts from Zanzibar/SpiceDB, and that's great! It also means if they ever do encounter a situation where the existing system is no longer viable (engineers leave or a new design constraint requires rethinking), they _can_ migrate to a dedicated solution easily.
One trade-off I didn't see acknowledged in this post is the sacrifice in terms of tooling. Dedicated solutions have toolchains for debugging and proving correctness that you can integrate into CI/CD and homegrown solutions can't typically spare the time to also build themselves a nice toolchain.
edit: fixed now
This reminded me of the tine i said something similar to teammate whom asked why our work always felt so bad. They Lost thier mind went off on some 20m rant and refused to talk to me for weeks and we were never on good terms again. I broke something in them i guess. They ultimately left thr tech industry.
It was the first of a number of incidents in my life that I learned that there really are a certian set of people that cannot be told the truth.
not really relevant but whatever it was a weird incident maybe that will entertain someobe
Make telemetry strictly opt-in then, that's really the only respectful way to do it.
Everything else is sophistry.
Not singling you out either, too many projects do it by default, yours at least makes it obvious here that you do and how to turn it off https://authzed.com/docs/spicedb/ops/observability#telemetry so kudos for not hiding it in the basement behind a "Beware of the Leopard" sign.
For open source projects I just find it more offensive, I'd expect closed source projects to do it by default (but I expect less of them to start with) but for open source it rankles much more.
You didn't explain why, in your opinion, opt-out is disrespectful; why respect requires opt-in. For most Western users, the bandwidth is hardly a drop in the bucket, and while there are certainly kinds of telemetry that are disrespectful of a user's privacy, telemetry in and of itself is not (in my opinion) abusive of user privacy, and therefore neither is opt-out abusive.
Indeed, if you were to argue that telemetry violates user privacy, then opt-in (i.e. to help send bug reports) doesn't fix that. If you add a "send bug report" button, but the data collected in the bug report abuses the user's privacy, why is it ethically any better? The data collected is far more important than the frequency or control over how and when it's sent. Users who desire taking that control should anyway run their own restrictive, allowlist-based firewall, which no opt-in/opt-out setting can overcome anyway.
With opt-in on an open-source project I get to choose to help contributing something back (telemetry data) to help the project.
Opt-out telemetry is not compatible with respect. It also adds an obstacle when I try to advocate the software to friends and coworkers. I have to warn them about it in advance, which is often where it ends because we’ve been burned in the past with other software adding extra things to their telemetry that weren’t there before, or changing the name of the environment variable that disables telemetry, or someone on the team does something that accidentally causes data to be transmitted unintentionally.