Tracking impressions (something advertisers seem to want) is facilitated by third-party servers. Self-served ads defeat this and raise fraud concerns.
Standardised advertising units (display sizes) mean that blocking elements strictly on dimensions is possible. One of my early userContent.css stylesheets, borrowing from online souces, did just this, and was highly effective, for a time.
Obfuscated content and JS can get around some of this, thou stylesheets whitelisting elements would be yet another workaround.
That is why the edge computing - changes this trust issue... It's just a matter of time before you have a advertising module you can install in cloud flare as a service worker implementation that both parties can trust... this with obfuscation could make it really hard to block and pretty viable for advertisers... let alone the fact that you can shift the analytics into the cloud edge servers... this both eliminates the argument of speed to access the content and removes your ability to block it. I see this as the future of adtech...
What seems to happen in practice is that infrastrucure domains and hosts used for advertisig are blocked by default. Amazon's aws & s3 domains come to mind, and they're rather horribly abused (Bezos's own WashPo have covered this). Which may be why generic buckets are going away.
Tracking impressions (something advertisers seem to want) is facilitated by third-party servers. Self-served ads defeat this and raise fraud concerns.
Standardised advertising units (display sizes) mean that blocking elements strictly on dimensions is possible. One of my early userContent.css stylesheets, borrowing from online souces, did just this, and was highly effective, for a time.
Obfuscated content and JS can get around some of this, thou stylesheets whitelisting elements would be yet another workaround.