[nailer]

> If you post (by way of example only, please observe the use-mention distinction here) a polite version of "ads are good and adblockers are stealing", and get a massive pile of downvotes...

Sure, I imagine the grandparent poster means arguing something like "limiting access for extensons is good because they're often used to steal financial assets". Old extensions are sold, or cracked and updated to inclue malware.

[JoshTriplett]

I want to avoid letting a meta-level conversation slip into object-level. But using the object-level as an example, I would expect a comment that acknowledges the types of things that are hard to build with Manifest V3, particularly more advanced adblocking, and acknowledges that there need to be solutions for those things, and makes the point that letting extensions be all-powerful does lead to problems and that also needs solving, would not get downvoted to oblivion. That's much more nuanced than, for instance, suggesting that Manifest V3 is an unalloyed good with zero problems, which I would expect to get downvoted.

[nailer]

> acknowledges that there need to be solutions for those things

Why is this required, in order not for the comment to be downvoted to oblivion? You may be confusing bias with nuance.

[JoshTriplett]

I'm giving an example, which to some degree was meant as an existence proof of a way to support an unpopular position without getting massive downvotes. "X is entirely good with no problems whatsoever" being replaced by "I think the benefits of X outweigh the costs, and here's some acknowledgement of the costs". I'm not trying to suggest only one possible way to do that, or only one pattern to follow. (This is one danger of using an object-level example.)

[nailer]

OK. Yes I agree there’s a cost to anything and acknowledging that is important.