Live data from Hacker News

Improved JavaScript and WebAssembly performance in EdgeHTML 17

blogs.windows.com

51–60 of 112 posts

Re: Improved JavaScript and WebAssembly performance in EdgeHTML 17

#51
post #18

Two things on my wishlist as a developer: 1. Decoupling Edge updates from Windows updates for quicker adoption. 2. Custom Elements as the basic feature for Web Components.

Edge is a Windows component, why decouple its updates? Is there any problems with quick windows updates? Ordinary users can't even disable updates now, so it shouldn't be a problem.

Re: Improved JavaScript and WebAssembly performance in EdgeHTML 17

#52
post #46
post #18

Two things on my wishlist as a developer: 1. Decoupling Edge updates from Windows updates for quicker adoption. 2. Custom Elements as the basic feature for Web Components.

I want to add: 3. Use my hosts file so I can connect to my local dev environments

I just tested it and it uses my hosts file.

Re: Improved JavaScript and WebAssembly performance in EdgeHTML 17

#53
post #12

> our early experiments show an average memory savings of 7% from this and a few other memory improvements This seems little. I wonder if they knew it beforehand and decided it is worth the effort, or overestimated potential memory savings. > Chakra was able to reduce RegExp bytecode memory in some popular extensions in this release by up to 10% Same as above. My current total size of ad filter rules is less than 1MB…

Some optimizations compound. Skipping anything that doesn't pass some impressive threshold (20%? 30%?) is the path to 500MB about:blank tab processes.

Re: Improved JavaScript and WebAssembly performance in EdgeHTML 17

#54
post #40

Earlier quoted context omitted.

> If you primarily work with Mac OS, Safari makes some sense, otherwise it makes no sense. Why do I never hear that complaint? Yes, but I work on all three desktop OSes, and two mobile OSs; I picked Firefox, which I use everywhere except on iOS (where it does exist, but is not really Firefox, even if it does sync with the rest of the Firefoxen). > Also, Edge on iOS and Android will soon have a built-in adblocker. Coo…

> Cool. Firefox already does, and I can also use it on Linux and Mac. But Chrome doesn't and I don't use Linux or Mac, but if I were primarily using a Mac, I would be using Safari. And If I were primarily using ChromeOS, I would be using Chrome. > I have no way to verify exactly what Microsoft sends to their servers (and their description is incomplete and out of date, if you care to trust it). https://www.windowscen…

> How is this a dark pattern? Windows 7 and 8 also have telemetry, but they don't let you view it as conveniently as in Windows 10.

Seriously? Were you not a windows user three years ago when it started? E.g. from [0] "The most famous example of digital bait and switch was Microsoft’s misguided approach to getting people to upgrade their computers to Windows 10."

Windows 7 and 8 are now just as bad, I do not in any way recommend using them if you value your privacy -- but at the very least, it's possible to turn off all telemetry, or at least it used to be possible when I last allowed updates on my Win7 (now it's firewalled off the world), by avoiding/removing specific updates. And according to the link you provided, "Although you can't completely prevent Microsoft from collecting diagnostic data, ...". Thanks, microsoft! but no thanks, even if you let me view what you are sending 3 years after you started snooping on me.

How exactly does the OS support "privacy controls" against 3rd party apps? Do you remember that its default setup was sharing your WiFi passwords with all of your facebook acquaintances?

> I do all my work on a Surface Pro 2-in-1, more than 50% of the time in tablet mode. I also keep a close eye on ChromeOS and iPads, but unless they support most of my software and use-cases, I see no reason to switch from Windows anytime soon.

As I said, according to my obviously biased sample, you are a niche market. YMMV.

[0] https://darkpatterns.org/types-of-dark-pattern/bait-and-swit...

Re: Improved JavaScript and WebAssembly performance in EdgeHTML 17

#55
post #54

Earlier quoted context omitted.

> Cool. Firefox already does, and I can also use it on Linux and Mac. But Chrome doesn't and I don't use Linux or Mac, but if I were primarily using a Mac, I would be using Safari. And If I were primarily using ChromeOS, I would be using Chrome. > I have no way to verify exactly what Microsoft sends to their servers (and their description is incomplete and out of date, if you care to trust it). https://www.windowscen…

> How is this a dark pattern? Windows 7 and 8 also have telemetry, but they don't let you view it as conveniently as in Windows 10. Seriously? Were you not a windows user three years ago when it started? E.g. from [0] "The most famous example of digital bait and switch was Microsoft’s misguided approach to getting people to upgrade their computers to Windows 10." Windows 7 and 8 are now just as bad, I do not in any w…

> The most famous example of digital bait and switch was Microsoft’s misguided approach to getting people to upgrade their computers to Windows 10."

They got a major upgrade for free with major improvements accross the board, they should be happy.

> And according to the link you provided, "Although you can't completely prevent Microsoft from collecting diagnostic data, ...". Thanks, microsoft! but no thanks, even if you let me view what you are sending 3 years after you started snooping on me.

The basic telemetry is just harmless diagnostic data that's on every modern mainstream OS. MS isn't snooping on you.

> How exactly does the OS support "privacy controls" against 3rd party apps?

https://pixelprivacy.com/resources/windows-privacy-settings/

And even more controls have been added with the recent Spring Update.

> Do you remember that its default setup was sharing your WiFi passwords with all of your facebook acquaintances?

Nonsense, this wasn't the default.

Re: Improved JavaScript and WebAssembly performance in EdgeHTML 17

#56
The article I find weak. They state things like 'an average memory savings of 7%' and 'up to 2.5x faster' without providing details. I would like to see benchmarks. Then again, I don't really care. Chrome has won the game on Windows.

Beyond that, I feel people who have chosen Chrome as a browser on Windows will be very hard or impossible to switch because of their browser history, passwords and so forth is saved and carried over across devices.

Re: Improved JavaScript and WebAssembly performance in EdgeHTML 17

#57

Earlier quoted context omitted.

> Something called Edge from Microsoft is available for iOS (where it is a rebranded WebKit) and on Android (where it is a rebranded Blink, IIRC). Why is it a problem that it uses webkit on iOS and Android? Why reinvent the wheel, when those rendering engines are already optimized for those platforms? The thing I care about is that it supports syncing of settings. > If you primarily work with Windows 10, Edge makes s…

>If you primarily work with Mac OS, Safari makes some sense, otherwise it makes no sense. Why do I never hear that complaint? Edge being multiplatform was part of that post, so what the commenter said makes sense. There are better options on Android and iOS. >Have you ever looked at the privacy controls of Windows? Privacy and control over the device. Both of which are limited in Windows by comparison.

* https://www.windowscentral.com/how-view-and-manage-diagnosti...

* https://pixelprivacy.com/resources/windows-privacy-settings/

And even more controls have been added with the recent Spring Update.

Re: Improved JavaScript and WebAssembly performance in EdgeHTML 17

#58

The article I find weak. They state things like 'an average memory savings of 7%' and 'up to 2.5x faster' without providing details. I would like to see benchmarks. Then again, I don't really care. Chrome has won the game on Windows. Beyond that, I feel people who have chosen Chrome as a browser on Windows will be very hard or impossible to switch because of their browser history, passwords and so forth is saved and…

I'm a Chrome user so I don't really know, but doesn't Edge have an import function for all that? And isn't there a mobile Edge on Android as well?

If Edge truly did become better than Chrome (has a long way to go) then I don't feel like it'd be too difficult to switch.

Re: Improved JavaScript and WebAssembly performance in EdgeHTML 17

#59
post #23

Earlier quoted context omitted.

> Edge is on both iOS and Android, Something called Edge from Microsoft is available for iOS (where it is a rebranded WebKit) and on Android (where it is a rebranded Blink, IIRC). If you primarily work with Windows 10, Edge makes some sense, otherwise it makes no sense. On iPhone, by Apple decree, ALL web browsers (including Firefox and Chrome) are rebranded WebKit; However, on Android, Chrome is (mostly) the real ch…

> Something called Edge from Microsoft is available for iOS (where it is a rebranded WebKit) and on Android (where it is a rebranded Blink, IIRC). Why is it a problem that it uses webkit on iOS and Android? Why reinvent the wheel, when those rendering engines are already optimized for those platforms? The thing I care about is that it supports syncing of settings. > If you primarily work with Windows 10, Edge makes s…

The privacy controls of Windows 10 are insufficient. In order to make Windows 10 usable, at least by Windows 7 standards, for people who actually care (you sound content with MS's worldview, that's fine for you), this is the latest version of all the crap you have to do: https://github.com/adolfintel/Windows10-Privacy

Re: Improved JavaScript and WebAssembly performance in EdgeHTML 17

#60
post #54

Earlier quoted context omitted.

> How is this a dark pattern? Windows 7 and 8 also have telemetry, but they don't let you view it as conveniently as in Windows 10. Seriously? Were you not a windows user three years ago when it started? E.g. from [0] "The most famous example of digital bait and switch was Microsoft’s misguided approach to getting people to upgrade their computers to Windows 10." Windows 7 and 8 are now just as bad, I do not in any w…

> The most famous example of digital bait and switch was Microsoft’s misguided approach to getting people to upgrade their computers to Windows 10." They got a major upgrade for free with major improvements accross the board, they should be happy. > And according to the link you provided, "Although you can't completely prevent Microsoft from collecting diagnostic data, ...". Thanks, microsoft! but no thanks, even if…

I don't think this discussion is worth continuing as our basic definitions of decency, privacy, control and discourse are so far apart that it makes no sense.

You ask "how is this a dark pattern", I give citation and reference, and your response is "they should be happy?" Seriously?

> The basic telemetry is just harmless diagnostic data that's on every modern mainstream OS. MS isn't snooping on you.

The only modern mainstream OS in which there is no way to turn it off is Windows 10; you can on Android, you can on MacOS, it doesn't even exist on Linux.

Let me decide what's harmless and what isn't, and which updates I want and which I do not. MS is snooping on me, and its worth it a lot to them or they wouldn't be so adamant about doing it.

> https://pixelprivacy.com/resources/windows-privacy-settings/

Did you actually read what you link to? This does not apply to desktop apps. Which are, I guess, 99% of the apps people use out there? Maybe only 95% by now.

> what are you talking about?

https://gizmodo.com/why-the-hell-is-windows-10-sharing-my-wi...

Based on your previous replies, I anticipate an answer of "but they're your contacts, you probably wanted to do that, you should be happy". So, pre-emptively - no. I do not want my passwords shared by default with anyone.

GDPR encodes in law the fact that everything like that must be opt-in, including telemetry and stuff. Unfortunately, it only applies to websites and not to the operating system. But it should.

Post reply on HN