If this was a conundrum that was at all resolvable, that had any hope of being tackled, I think it would be a valid concern perhaps.

But there's no suggestion or idea, no way to specify our way out of this one.

And the idea of providing too much guidance because one model has quirks? That seems so so very so-what. Oh no we over-prompted isn't zero impact. But it usually isn't that cataclysmic. I feel like this is such a 0.001% gripe, and to hold up all agency on the web because of this is unbecoming. Is ridiculous.

I know browsers wanted to get rid of websql but generally I thought the consensus was websql was disliked generally by vendors. It wasn't that they wanted a different websql: they didn't want a websql at all, period. The quirks weren't the problem.

Again though, fucking help! Fine, there are problems. I personally think this is some absurd fucking ridiculous mountains out of mole hills and ya'll are being absurd over this. Absurd. But if this was a discussion where there were interesting future directions to look to, that we could pursue and try to follow up on? Ok! Fine! Have your pound of flesh maybe or maybe it really does pay out! But it is some conservative Fear Uncertainty and Doubt soul rotting evil to obstruct & deny while making qualifications that are utterly unobtainable.

Meanwhile the rest of the software world is going to keep moving. The web is just going to be colossal negatively impacted by this impossible "it must be perfect" impasse created for no reason with no expectations of resolvability, in spite of the base premise of giving users access to their agents being an obvious direct and clear improvement over every other possibility by country miles.