Lazy loading is just another feature to improve performance, and browsers already used heuristics to load assets with prioritization by what would render first.
Also in the last 5 years, every browser has added more functionality to block ads and tracking. Lazy loading is not going to somehow make up for that.
How does it improve performance? It is literally more work for the machine to (mis)manage. And most browsers already have viewport heuristics, no? What makes this more likely to succeed?
By not loading images at all until you get to it instead of guessing if you will. It's already done by some websites using javascript. More work in milliseconds of CPU time is a worthwhile tradeoff for network bandwidth.
Guessing on if you have a fast enough connection to load is literally one of the features called out as hopefully possible with this spec. Such that it is expected that this flag will be controlled by scripts. (Or, turned to always eager if scripts are disabled...)
My question is what makes this better than the current script solution? Especially knowing this is intended to be built up with scripts.
I don't know. I was just commenting on how this is useless for adtech, but I don't see any negatives with giving developers more explicit controls over loading. Loading less data is a good thing regardless of internet connection speed.
My gut would be that this is going to win back speed lost to adtech. So, not useless. Just not useful to tracking. Which, quite frankly, is already extremely good.
My gut would also be a lot of folks think this will be good for tracking sure usage. On both sides of the fence.
Also in the last 5 years, every browser has added more functionality to block ads and tracking. Lazy loading is not going to somehow make up for that.