4 ms·

>you need telemetry to find out how users react to abusive notifications You need telemetry to find out how users react to abusive notifications? Wow. I know t

>you need telemetry to find out how users react to abusive notifications You need telemetry to find out how users react to abusive notifications? Wow. I know that A/B testing is in vogue, but can’t they just use some common sense? People despise abusive notification prompts. People despise abusive location prompts. People despised abusive popups to the point that every browser blocks them and websites started emulating popups via CSS to keep abusing their users (hi Medium!).

lifthrasiir 8y ago
Unfortunately we are living in the world where the common sense is wrecked. People do complain if abusive notifications are blocked and websites are rendered unusable as a result, though it is completely the fault of websites! Keep in mind that browser vendors are trying to solve multivariate equations that involve users, web developers, companies, abusers and crackers...
michaelt 8y ago
You need telemetry to find out how users react to abusive notifications? I run an ad blocker, so my perception of notifications is they're used for things like breaking news notifications on news websites [1]. That's not a feature I personally need, but 24 hour news channels exist so presumably someone feels they need regular news updates. So no, it's not obvious to me everyone dislikes notification requests, even if I have them all blocked myself. If you let curmudgeon developers like me dictate products' features sets based on intuition, there'd be no HTML e-mail, no third-party cookies, no WebGL, no emojis, no WebUSB.... :) [1] https://www.wired.co.uk/article/push-notifications-breaking-news https://www.wired.co.uk/article/push-notifications-breaking-...
- freeone3000 8y ago
  And what a world we would live in!
zaarn 8y ago
When is a notification prompt abusive? What rule do you need to distinguish a non-abusive prompt (for example a chat program wanting the ability to notify you or a GPS-related map on a website) vs an abusive prompt (newswebsite wanting to spam you or a tracking tool using GPS). Common sense doesn't work here because everyone has a different perception of when these prompts become abusive or unwanted. Mozilla uses the data collected to determine how they can establish a rule to distinguish between abusive prompts and non-abusive prompts. How users react to them is a good indicator if the prompt was abusive or not, most people will decline abusive prompts and accept good ones. A simple "wait until first DOM interaction" will likely be not sufficient since a simple click on a text would then create the prompt. With more data you can determine a better rule. That people despise these prompts is fairly obvious and exactly why Mozilla is collecting the data; they want to help people by establishing a good common ground rule for these notifications to be automatically blocked.

lifthrasiir 8y ago

Unfortunately we are living in the world where the common sense is wrecked. People do complain if abusive notifications are blocked and websites are rendered unusable as a result, though it is completely the fault of websites! Keep in mind that browser vendors are trying to solve multivariate equations that involve users, web developers, companies, abusers and crackers...

michaelt 8y ago

You need telemetry to find out how users react to abusive notifications? I run an ad blocker, so my perception of notifications is they're used for things like breaking news notifications on news websites [1]. That's not a feature I personally need, but 24 hour news channels exist so presumably someone feels they need regular news updates. So no, it's not obvious to me everyone dislikes notification requests, even if I have them all blocked myself. If you let curmudgeon developers like me dictate products' features sets based on intuition, there'd be no HTML e-mail, no third-party cookies, no WebGL, no emojis, no WebUSB.... :) [1] https://www.wired.co.uk/article/push-notifications-breaking-news https://www.wired.co.uk/article/push-notifications-breaking-...

freeone3000 8y ago
And what a world we would live in!

zaarn 8y ago

When is a notification prompt abusive? What rule do you need to distinguish a non-abusive prompt (for example a chat program wanting the ability to notify you or a GPS-related map on a website) vs an abusive prompt (newswebsite wanting to spam you or a tracking tool using GPS). Common sense doesn't work here because everyone has a different perception of when these prompts become abusive or unwanted. Mozilla uses the data collected to determine how they can establish a rule to distinguish between abusive prompts and non-abusive prompts. How users react to them is a good indicator if the prompt was abusive or not, most people will decline abusive prompts and accept good ones. A simple "wait until first DOM interaction" will likely be not sufficient since a simple click on a text would then create the prompt. With more data you can determine a better rule. That people despise these prompts is fairly obvious and exactly why Mozilla is collecting the data; they want to help people by establishing a good common ground rule for these notifications to be automatically blocked.