WRONG! BIP-148, also known as the User Activated Soft Fork, actually showed that the miners do not vote. The community who run full nodes do. The miners simply followed, and gave the full nodes the type of blocks that full nodes demanded.
i hate it when people sugar coat BIP148 like this.
a true UASF is when there is a clause in the process about reaching >95% consensus. but when that is omitted and also the risks of splitting the bitcoin network into two chains due to lack of support and disrupting the billion dollar currency for even a short time is swept under the rug, it becomes more of a malicious attack instead of something positive!
during 2017 i don't remember the number of nodes signalling for BIP148 surpass 5k out of the 100k existing nodes by that time and the hashrate behind BIP148 was about 1-2%.
in my view there is absolutely no difference between BIP148 and MAHF (aka bcash) in that regard.
any change in bitcoin protocol must happen only after reaching more than 95% support from the entire network* or not happen at all.
* the entire network is
both miners and nodes. you can't exclude either one of these two.