4 ms·
It was activated, but not activated-activated (in that the addon gracefully decided to not do anything). But obviously we can't delegate that decision to the ad
by revelation 9y ago
It was activated, but not activated-activated (in that the addon gracefully decided to not do anything). But obviously we can't delegate that decision to the addons, that's why the addons pane exists after all.
Which begs the question: if you had to do something to activate it anyway, why abuse the "studies" sidechannel to install something that is deactivated, when you could just as well prompt to install it when it's meant to be activated?
- DonHopkins 9y agoThe "studies" sidechannel is the ideal place to distribute a deactivated plugin if you want curious people to activate because they want to test out new features. People subscribe to the studies sidechannel BECAUSE they want to try out new features, and those people tend to know about and use about:config, so it's likely many people will enable it, to see what it does. And here's a bug report about something it does: https://github.com/mozilla/addon-wr/issues/39 https://github.com/mozilla/addon-wr/issues/39
- stordoff 9y ago> People subscribe to the studies sidechannel BECAUSE they want to try out new features AFAIK, you are opted-in by default - I recently had to disable it on a clean installation.
- Crespyl 9y agoYou're right that "studies" is designed for distributing/testing experimental features, among other development and debugging uses. I'd dispute any characterization of LookingGlass as a FireFox feature though, and feel very strongly that using Studies to distribute marketing software (inert or not) was an entirely inappropriate use of the tool. Mozilla being willing to use Studies (or having a process that allows for it to be used) in this way means that, although I was fine with the extra telemetry, debugging, a/b testing, etc., I cannot trust Studies to be used strictly for development related things.