@filippo To begin the discussion not until the technical details of implementation is the major failure to begin with. This from the beginning is more a social fail, not a technical one. And failing to see that is concerning at a Code of Conduct level, way prior to anything technical. It looks like one might have begun to realize that.

@serpentroots @filippo It would appear so, that one fails to see beyond their own nose.

@filippo Sounds like an excellent way actively ignore anyone who is not already in the inside circle, and who had yet had a need to interact with Go, outside of using a fairly great language that otherwise "just worked" for them. I would be one of those. Also, no, I did not read the full proposal. Once I realized it's opt-out I could not care less of what you collect, opt-out is a hard no. If/when that change I may bother looking at it.
On the upside for me personally, I finally got around to check out OpenSnitch now, having been a LitteSnitch user previously.

@filippo It's frankly more worrying in general the one (that's Go/Google leader ship, not you as person) appears to have not anticipated the reaction, even more so when then continuing to argue that it must be opt-out. You might have seen slightly different reaction for opt-in. Either way, it would have been an uphill battle regardless.

@filippo @douglascamata it's already bad enough that module proxying is opt-out, now this.

@filippo @mariusor Not web (document) traffic, but HTTP. Craw by HTTP opt-out by SMTP, looks further away than robots.txt from here at least. At least somewhat standard, if imperfect. (it doesn't really help the case that it's "Google" that refuse robots.txt, no matter how little Google crawling is involved in Go)