Recommended Posts

I'm pretty happy with IE10 on the Win8 DP, but it's good to see others are already planning a metro version of their browser

Also, desktop Firefox 10 (every beta so far, and the final) works just fine in the WDP.

The blog post earlier today on WOA (Windows On ARM) clears up a few misperceptions (including mine) on WinRT. A true WinRT version of FF will run without changes on any platform that supports WinRT(x86/x64/ARM) - true CPU neutrality. (Naturally, that hasn't been the case for any browser before - regardless of platform.) It also means that MetroIE is the *same browser* regardless of the OS underneath - and why plug-ins got banished. (Would you really want an ARM CPU sandbagged by some of the heavier plug-ins - or ActiveX controls, for that matter - that desktop IE and desktop FF have to deal with?)

Also to note. Chrome on ICS doesnt support Flash. If i remember correctly.

So flash is going out anyway. Till windows 8 hits the shelves everybody should move on to html5

And i dont think so other browsers can run on WOA (in desktop mode i mean).

And browsers will require native code to offer comparable performance to IE10 (immersive).

And browsers will require native code to offer comparable performance to IE10

Well, the C++ code in WinRT runs at generally native speed (albiet sandboxed within the runtime) - and they can directly access DirectX in WinRT. They already have a DirectX hardware acceleration layer working in Firefox desktop, and most of their code is C++, so there's not too much reason why they couldn't get the performance they want.

If they tried to be ridiculous and write it in C# & XAML, then they wouldn't have a chance of getting decent performance :p

Well, the C++ code in WinRT runs at generally native speed (albiet sandboxed within the runtime) - and they can directly access DirectX in WinRT. They already have a DirectX hardware acceleration layer working in Firefox desktop, and most of their code is C++, so there's not too much reason why they couldn't get the performance they want.

If they tried to be ridiculous and write it in C# & XAML, then they wouldn't have a chance of getting decent performance :p

Looking through Mozilla's notes on Windows 8 support, it seems that still isn't an ideal situation for them and possibly other browser vendors. The main issue it seems is that metro doesn't support native code, which most of Firefox is written in. They've noticed however that IE10 works differently to other metro apps, in that it is the same exe as the desktop app, running with native code outside the sandbox, just using the metro interface instead of its normal desktop UI. Therefore it's possible that Microsoft might permit this later on, removing the issues they have.

They've noticed however that IE10 works differently to other metro apps, in that it is the same exe as the desktop app, running with native code outside the sandbox, just using the metro interface instead of its normal desktop UI. Therefore it's possible that Microsoft might permit this later on, removing the issues they have.

It'd be unlikely Microsoft ever would. Allowing a WinRT Firefox app to access native code outside the sandbox paints Firefox as a big security risk, and a potentional attack vector for virus / malware to get into the system - considering here that Windows On ARM tablets will not allow any other native apps apart from Office, and the Metro apps are sandboxed. With IE, this isn't *as much* of a concern, as they can push a patch down Windows Update as soon as they can for any issue - it's their program in their hands, and they can make sure they take action right away. But I don't think they're in any mood to let a third party have that kind of responsibility.

Also to note. Chrome on ICS doesnt support Flash. If i remember correctly.

So flash is going out anyway. Till windows 8 hits the shelves everybody should move on to html5

And i dont think so other browsers can run on WOA (in desktop mode i mean).

And browsers will require native code to offer comparable performance to IE10 (immersive).

Actually, no.

MetroIE (the Immersive version) is straight WinRT code (going forward - there may be some native code remaining in the version in the WDP, as it does support ActiveX controls). The *desktop* version (which supports ActiveX controls) is native code (and likely, on x86/x64) Win32/Win64. Yes - that means *three* versions of IE on the x64 versions.

Looking through Mozilla's notes on Windows 8 support, it seems that still isn't an ideal situation for them and possibly other browser vendors. The main issue it seems is that metro doesn't support native code, which most of Firefox is written in. They've noticed however that IE10 works differently to other metro apps, in that it is the same exe as the desktop app, running with native code outside the sandbox, just using the metro interface instead of its normal desktop UI. Therefore it's possible that Microsoft might permit this later on, removing the issues they have.

Metro style apps can be written in 100% native code.

Metro style apps can be written in 100% native code.

As far as I can see that's not the case, you can only use managed code. The closest you can get as far as I can see is using managed C++ allowing you to reuse code from native C++, but of course with the lack of Win32 and many other APIs often used in native code.

Just a quick question

Are there any known plans for other companies to make internet browsers for metro? and can it even be done because from what i understand metro apps are made in HTML 5?

not worth it maybe!! who knows metro might be gone in Windows 9.

As far as I can see that's not the case, you can only use managed code. The closest you can get as far as I can see is using managed C++ allowing you to reuse code from native C++, but of course with the lack of Win32 and many other APIs often used in native code.

Brandon is a member of the Windows Shell team. He knows. :)

As far as I can see that's not the case, you can only use managed code. The closest you can get as far as I can see is using managed C++ allowing you to reuse code from native C++, but of course with the lack of Win32 and many other APIs often used in native code.

We actually do not really support (or at least emphasize) managed C++ (aka C++/CLI) for WinRT development.

The primary targets are via the API "projections." They are:

Native C++ with "Component Extensions" - also called "high level" C++

C# and VB.Net (managed)

JavaScript

C++ apps can use DirectX (D3D/D2D) to render directly (or via a custom UI framework built on top of it), or use the new (fully native) XAML system.

.NET apps can use the same XAML system (the implementation of which is all native).

JS apps use HTML / canvas / SVG

You can also forego the C++/CX extensions and the niceties of the high-level projection, and instead write purely standard C++. Here you'd interact with the WinRT in its raw form, what we call the ABI (Application Binary Interface). That level would be most familiar to experienced COM developers (HRESULTs instead of exceptions, etc). For those developers we also include the new WRL (Windows Runtime Library) helpers, which is analogous to a modernized subset of ATL. But it's all new and designed for the WinRT ABI (and new base platform constructs like HSTRING and such). That said, I don't know why you would :-) The ABI layer is... verbose, compared to the elegant and concise CX goodness (and the result the compiler spits out is the same).

Just a quick question

Are there any known plans for other companies to make internet browsers for metro? and can it even be done because from what i understand metro apps are made in HTML 5?

The UI is declared in HTML5 but then compiled into code just as a UI can be declared in XAML then compiled into binary.

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Posts

    • Huh? I have an Aeron Remastered and the backrest cannot go up or down (only the lumbar is adjustable). Yes, the Doro C300 Pro V2's backrest can be locked, but the locking mechanism is so weak that when you recline and lean back the backrest moves with your body. I think the whole purpose of this chair is that all of the moving parts (except for the rocking, which can be locked in place) are supposed to move with your body, as you shift your weight, that's why I couldn't get used to it. Some people might like it though. Its redeeming factor is the price, I mean it costs like a quarter of what a new Herman MIller (Aeron) costs.
    • Well done Massie. This is how to treat the Fox News filth.   
    • I notice how you dodged the questions I had about the racism shown by ignorant, gullible, cowardly people when the Poles, like your partner, were the immigrants. Ahem. I wonder how you'd feel if native born Brits suddenly treated you as "dirty crooked immigrant" for being half Trump-American? If they ordered you to leave and "go back to your corrupt country" (on the other side of the Atlantic), would you go? The truth is based on facts as supported by evidence. As requested in your previous posts, I have used the facts in your own post to show everyone the truth.
    • US citizens are paying to their government, who could use that to fund healthcare and tuition and relieve the costs of these for citizens instead of making tax breaks that overwhelmingly favor the rich. I'm not saying that tariffs are the correct solution, but what else would they be used for? What else could Trump have in mind for wanting them, if he hasn't figured out that labor costs are higher in the US?
  • Recent Achievements

    • Conversation Starter
      jessse3334 earned a badge
      Conversation Starter
    • Reacting Well
      JuvenileDelinquent earned a badge
      Reacting Well
    • One Month Later
      Excellence2025 earned a badge
      One Month Later
    • Week One Done
      Excellence2025 earned a badge
      Week One Done
    • Week One Done
      flexorcist earned a badge
      Week One Done
  • Popular Contributors

    1. 1
      +primortal
      506
    2. 2
      +Edouard
      207
    3. 3
      PsYcHoKiLLa
      151
    4. 4
      Steven P.
      73
    5. 5
      macoman
      62
  • Tell a friend

    Love Neowin? Tell a friend!