4 ms·
I read through the page and these features seem... not great? Some of them might be considered anti-features. The Browser Launch feature says that Firefox alw
by SquareWheel 4y ago
I read through the page and these features seem... not great? Some of them might be considered anti-features.
The Browser Launch feature says that Firefox always waits for extensions to load network requests whereas Chrome gives an option to wait via a setting. So in this case, you're trading performance (parallelizing the work) for privacy (may not catch a tracking request). It's a tradeoff I'm sure many are willing to make, but I like that Chrome offers the setting. I tend to err towards performance, myself.
WebAssembly says it isn't used - not because it isn't available - but because it would add an extra permission to the manifest. That is to say, Chrome informs you of this when installing extensions and Firefox doesn't. That's a minor implementation detail at best, and I think a case can be made that providing more granular scopes is a positive thing.
Then pre-fetching. Are they really disabling a major performance feature by default? I've used pre-fetching to good effect when the browser's resource scanner ran at the wrong times. The option should exist but this is a terrible default.
The other features are more reasonable. HTML filtering could be useful for adversarial sites like Twitch.tv that actively fight adblockers. Storage Compression seems like a useful optimization if you have a lot of extensions installed that populate IndexedDB. I'm assuming it's caching those values rather than doing a
"first time setup" every time a new private window is launched.
I'll readily admit there may be more intent behind these features than is expressed in these blurbs, but on initial glance I wouldn't say anything here is critical. The pre-fetching one in particular seems like a negative.
- Vinnl 4y agoHmm, I haven't looked at all of them in detail, but CNAME uncloaking in particular I believe is a very commonly used tracking vector.