Windows RT + ARM support

[DoubleDutch]

Windows RT is going full steam ahead at Microsoft (rather than Win32).

It would be good if there was a WinRT compile option in PureBasic (Windows) otherwise PureBasic will be left behind.

This also may need ARM support though as RT is also available on the ARM version of RT. RT programs run on the 'metro' interface as well as the Windows 8 interface (soon win 7 too I think).

If ARM support is added then an Android + IOS destination would clearly be another next step? (PureBasic compiling for IOS/Android would be great) - Android is available in ARM and x86 flavours (dual binaries).

[fsw]

Just adding two links to make sure every one here talking about Windows RT knows what a mess this new Windows RT and WinRT really is.

Some people may say: "You can't trust Paul Thurrott..." which might be true, but if even a fanboy like Paul says: "Redmond, We Have a Problem" it's getting serious.

http://www.winsupersite.com/article/win ... lem-144554

Here more info about WinRT (and the differences to Windows RT):
http://arstechnica.com/features/2012/10 ... -new-again

I can understand why some PureBasic users might have a request in this regard, but this RT stuff is a mess. Don't know if it's worth it...
Well done Redmond.

[TI-994A]

Well done Redmond.

And as if things weren't confusing enough as it is, they've now changed the name METRO to MODERN UI. Way to go, Microsoft!

[DoubleDutch]

The Metro name was owned by someone else plus it doesn't matter what anyone thinks about how good or bad RT is, it's here to stay and it's the Win32 replacement in the long term.

I've just 'What is Windows RT?' - it's nonsense, SkyDrive is available as a file system to any App in RT - so your docs are synced, WinRT programs can run on Metro and on the Desktop, there is a built-in mail app/calendar/contacts into RT - so why have outlook as it would only confuse everyone?

The whole point of RT is to make a new Api/system that works on all future Microsoft devices - you write one program and it works on Desktop computers, Tablets, Laptops, Phones, even dedicated game systems. When you look at it you will find the Api itself is better than Win32 already - no complicated old interface methods to deal with in order to use new things like touch, etc...

[TI-994A]

Clearly, there's a lot of confusion surrounding this topic, but none more than the differences between Windows RT and WinRT. Windows RT is actually an ARM-based OS, and WinRT, short for Windows Runtime, is a new API built into Windows 8. It is a sort of framework to run Windows RT apps. Microsoft is clearly targeting this as a convergent platform, which it had failed to do with the .NET framework.

However, if we're not targeting ARM-based devices or Metro/ModernUI-style apps, we needn't concern ourselves with this, for now.

[OrangeJuice]

@DoubleDutch:
Do you actually think that is a good idea? IMHO there is absolutely no reason to be able to run the same App on Desktop, Tablet and Phone. The usage principles are so different that I highly doubt that you could write an App that would work equally well on each without actually more effort compared to supporting two separate apps.

I read up on WinRT today and - my nonexisting gods - is it a mess. I looked forward to finally getting something better then Win32 Api but WinRT looks to me like a collection of all the worst of all worlds. I'll try to keep an open mind, but it doesn't look very compelling right now.

[DoubleDutch]

It doesn't matter if I think it's good or bad. It will be huge.

[IdeasVacuum]

.... I don't think WinRT API will be, erm, hugely huge. Depends how you measure it. It might start-off with the potential, it might open the doors for 'gadget' developers that are currently writing scripts. Look at the analogy the hardware side gives us: Netbooks, well received at first, but at the end of the day, not powerful enough in all aspects and so now they are really evolving full-circle to be notebooks/laptops.

In kind of the same way, WinRT has to compete with both .Net and Win32. The largest, most serious apps are written in C or C++, on Win32 .Net has made an impression but slowly because there has to be a compelling reason for a developer to move thousands of lines of code over to it (Nope, can't think of one lol). Initially WinRT simply cannot compete, it is still being developed, so then it's a question of how hard MS are going to try when WinRT is not the developer's first choice - not even the second choice? A lot of 'big' MS initiatives have fallen by the wayside.

Back to the original subject of the post, which is the hardware. Laptops and Desktops are chosen by various industries because serious work requires powerful apps, large screens, large drive capacity and large ram capacity. You can't design an intricate component or a 5000+ parts assembly in a CAD app with finger gestures on a touchscreen. Not only is it impractical, the ergonomics prevent it - in earlier days we used 'light pens' on the screen, very uncomfortable to use after a while. You certainly would not get far running a CAD-CAM app on a 10" tablet or a smart phone.

I think Microsoft's Surface will prove the case. Looks like the Windows RT/Arm version is selling well, but the 'Pro' version running Windows8/Intel and supporting legacy apps, plus just having much more grunt all round, is I think going to be more popular with businesses. Since the only way is up, you will out-grow your Windows RT/Arm device and what will you buy? A Windows8/Intel device!

[DoubleDutch]

People will start buying 'apps' from the MS store, you will sell a lot more programs through this than by trying to market it yourself. That is why it's going to be huge - just like the Apple app store.

[Kukulkan]

I think Fred and team should at least take a closer look. Maybe there are some tools or help available for conversion of existing code to ARM topology from Microsoft?

Many commercial PB developers are focussed on Windows and a lot of their customers surely will ask for a Windows RT version of their products. Currently, the only option is to port the program to something like .NET. And if this also runs on the "normal" Windows, the PB branch might get removed to support and maintain only one version. I see some serious danger in ignoring Windows RT.

Anyway, Windows RT might be a dead horse from the beginning. Investing into it also may be wasted time.

To repeat myself, Fred and team should at least take a closer look...

Kukulkan

[MachineCode]

That is why it's going to be huge - just like the Apple app store.

There was a tech report not long ago that said literally half the apps in the App Store haven't even had a single download.

[DoubleDutch]

You mean the Apple appstore report. I was suspicious about that because it cannot possibly be true - as at least the owner of the app would have downloaded it themselves to actually test the download?

[Polo]

Asking for Windows RT (Windows RT as in the "Metro" or modern UI apps, whatever platform it is on) support is like asking for iOS support - that would be nice, but it's so different you cannot expect the same product (Purebasic) to output to those different targets.

[DoubleDutch]

It's not really as IOS requires a mac to sign the code. RT is a different lib (on x86) and if ARM output support is added then that same lib (but the ARM version). Then a side benefit to ARM support is possible future Android support (by a nice coincidence it's also a lib for ARM and x86 too), then possibly IOS. I personally think ARM support will have to be added eventually anyhow as there is a distinct possibility of Apple eventually moving from x86 to ARM on it's mainstream machines (OSX).

[Tenaja]

... I personally think ARM support will have to be added eventually anyhow as there is a distinct possibility of Apple eventually moving from x86 to ARM on it's mainstream machines (OSX).

Much of this depends upon ARM's success with their 64-bit processor. However, the performance hit is so big at this time, that ARM will have to make enormous strides.

Intel has just recently started their focus on power savings, and while their first result may not be a home run, with about 20% savings right away, it was certainly a great way to start the game.