If just 1% of users opt-in to telemetry, you would have more than enough data for a representative sample set.
This is the whole point. You can do the right thing without any loss of data usefulness. So why persist in doing the wrong thing??
The team behind Brew is doing a great job and if analytics can help them do a better job, then be it. Nobody is forcing you to use it and opting out is simple enough.
From initial discussion with Mike, there are two parts that would be needed for a (self hosted) setup:
a) The hosting itself (this is something I can likely help with)
b) Setup and ongoing management of Piwik. This isn't a skill set I have (nor am interested in). However I'm generally thinking the Homebrew Community is big enough to rustle up at least a few people with the right skill set for getting this done (and keeping it done :>).
Trying to get both hosting + Piwik setup/mgmt from the same set of people may be a tall order, so doing them separately is probably more achievable.
As a data point, this kind of approach - sponsored hosting/hw/similar + volunteers to look after Community infra - is used successfully in other projects. PostgreSQL is a good example.
Can't really think of any obvious reasons it wouldn't work for Homebrew too. ;)
Additionally, this (kind) offer is pointing to other organisations that may be able to host the servers and discussing potential solutions to the other options. It's not a slam-dunk by any means.
This is all done in my free time so I don't need an excuse to do or not do any of this.
Homebrew lead maintainer here. This is exactly why we do it. To repeat myself from below: we're chronically understaffed and underfunded.
> Furthermore, now that you know, for brew specifically, will you opt out?
It's worth noting: if everyone who uses a niche piece of software opt-out when it breaks we'll look at our analytics and likely remove it from Homebrew instead of fixing it. This may be unfair but given the amount of work: we need ways to prioritise. Analytics have helped us do that so we can prioritise fixes on things that are critical for many people.