Posted By: Ray7 | Oct 1st @ 11:12 AM
page 1 of 1
Comments: 22 | Views: 925

According to eWeek.

 

The only one I disagree with, is open sourcing it. How would that help? The iPhone OS isn't open source and that doesn't seem to be having the same kind of problems.

ManipUni
ManipUni
Proving QQ for 5 years!

I would agree with that list.

Generally speaking Microsoft needs to throw out "Windows" and start from scratch. Stop trying to "carry the experience" or "leverage the IP" or other bull marketing speech and just write a made-for-mobile-device OS.

Windows CE/Mobile is just bad. It is poorly designed.

Bas
Bas
It finds lightbulbs.

Don't you know? Open Source is the magical buzzword that, according to everyone who isn't sure what it is, is automatically 'better'.

 

Anyway, they make some good points, but they all have caveats. Touch, obviously, but what about the gigantic amount of non-touch phones people are still buying today? Many people like buttons. An app store? The iPhone was a success long before the app store. Reliability is obvious, but is WM really that unreliable? And is the iPhone that rock solid, even with the dropped calls thing? Is the iPhone that enterprise friendly?

 

It seems like they're listing a bunch of stuff that any platform, not just phones, would benefit from, and are pointing to the iPhone's success where it applies.

It is open source though, you can get the Darwin sources.

I'm definitely in favour of Windows CE/Mobile being scrapped and re-written from the ground up. Not because I care particularly about the phone OS, though.

 

I'm in favour of any move which could result in the Windows CE variant of the Windows API documentation being removed from MSDN Library so that I only have to worry about the .Net, MFC and ATL variants/wrappers appearing incorrectly when I select something and push F1 in my Win32 source.

 

(Or someone could, you know, fix MSDN Library so it works again but it's been completely broken in this regard for 5-10 years now so fat chance of that ever happening.)

 

I do not like WM at all, it feels like a large hack after trying newer OSs like iPhone and Android, and I say that as end user

RLO
RLO

1. Commit to touch

Already there with 6.5 and currently with the Zune HD

 

2. Consider open source

Maybe for the browser.  It seems MS is losing the browser war, and it may be a nice "Pawn Sacrifice" to shake up the markets.  Right now most of the web developers on the sites I visit are going with design for firefox and patch for IE.  If MS was to embrace WebKit, it may be an opportunity to win back marketshare even though it would be a temporary win for the FOSS sector.

 

4. Ensure reliability

The best way to do this is maximize the ram available to key functions of the OS (email/phone/text/gui) and limit all other apps (especially browsing) to a seperate memory pool.


5. Remember intuition

It's easy to say that in one sentence, but when you have OEM's customizing your GUI to hades and back, kind of hard to bring it all together.


6. Forget everything you know about Windows Mobile

WinMo is still being designed with a Win98 GUI in a Win7 World.  6.5 appears to have taken on some design cues from Zune, but not enough to make it revolutionary.


7. Don't forget the enterprise

I don't think it was ever forgotten, which has been part of the problem. 


8. Focus on the keyboard

Works awesome with the Zune HD, so mark that as done.


9. Simplicity is OK

Agreed, but should never replace features, or make the device seem half finished.


10. Learn from Apple

Agreed, spend money into advertising and design.   Have an end to end solution that the carriers won't customize and mess up. And more importantly aim it at consumers, not enterprises. 

 

I am fine with WinCE, and think that it is an elegant solution, let it be the engineering base, and sell a custom version of that to OEMs to create their own GUIs with.  No sense in having the triple solution that is currently in the market.  WinCE with WinMo on top, with a tweaked GUI by the carriers and oem's. 

 

I really believe that MS would do well, to break out Office Mobile as a seperate app and sell it on more platforms.  Can you imagine the cash injection the company would get if they sold it for the iPhone?  Especially, if it was at a price point of $39.95.  Ride the money train like every other developer does until you can get your killer product on the scene.

 

WinCE is designed for limited-spec embedded hardware, now we have phones that are faster than some netbooks.  Should run NT instead, it scales and has an actual I/O model.

ManipUni
ManipUni
Proving QQ for 5 years!

You say that an intuition UI is "easy to say, hard to do" and you aren't wrong. But that's really the point, Microsoft should try and do something that is hard. I think the iPhone receives 9/10 on the intuitiveness scoreboard. Haven't tried Android long enough to have an opinion on it.

 

I think most of the problem with Windows CE/Mobile is the UI. The APIs are also a problem within their own right since they allow third party code to completely "own" your phone while they run.

 

And while people are still buying non-touch phones, you need to look at the price difference. The iPhone costs £800 to buy and they pass that entire cost on to the consumer (either through contract or price). A Nokia non-touch non-multimedia phone costs less than £100 which mobile operators can often give away for "free."

blowdart
blowdart
Peek-a-boo

I think most of the problem with Windows CE/Mobile is the UI. The APIs are also a problem within their own right since they allow third party code to completely "own" your phone while they run.

 

Well the APIs can be sandboxed, but as soon as that happens then people complain about operators stopping programs running. It's perfectly possible for an operator to lock down bits of functionality - but no-one likes that any more.

 

It is telling that the Windows Mobile phones that reviews like are the ones that hide the UI completly. But then you run into problems with 3rd party apps, which have to use the native UI so they run everywhere. So you have a nice HTC phone, then all the apps you load look so different and are unusable.

Good points.

The eWeek article was much too non-specific and so not very useful.

ScanIAm
ScanIAm
On a scale of 1 to 10, people are stupid.

I'm confused about why every WM device I've owned has had problems with ActiveSync at least once a week.  The error messages are useless and while sometimes, simply reconnecting the device solves the problem, other times, the entire device must be reset.

 

 

DCMonkey
DCMonkey
Monkey see, monkey do, monkey will destroy you!

I've been under the impression that Windows Mobile 7 *is* a rewrite, at least of the UI and related layers. Note how long it's taking them to ship it. I wonder if it will turn into another, quieter, Longhorn (Phonehorn?), and we'll see a more incremental "WM 6.9" release from the 6.5 team.

This is the very question I asked when I interviewed Loke Uei Tan from Windows Mobile. They're not rewriting the OS (from scratch) for WM7.

Good list. I hope Steve Balmer can read it and take it to his heart. Besides, open source, I think they are all important. (ths site is kind of annoying, so hard to find previous page)

 

I am going to add one more:

Easy Expression. What is it? Gives the manufactures an easier way to customize the GUI and so on. Not entirely flexible that un-Windows the entire GUI. But, it has to be an consistent unified customization, like Theme and a special pre-installed apps page. And pre-installed apps can be disabled or enabled in a simple manager.

 

Why this is important?

Because manufactures needs to have their own flavor, otherwise they look the same as other brand and I know big manufactures don't want that. If Balmer want to get that real big pie rather than "another nitch" market, they need to know what major manufactures want and deliever it. Otherwise WM will always just for some startup phone manufactures.

 

ManipUni
ManipUni
Proving QQ for 5 years!
They could just give Manufacturers a console that allows them to select the packages they want (e.g. Bluetooth, Web Browser, .Net APIs) and to place a few logos and strings on the phone to mark it as something they have produced. But ultimately the OS could be cross-phone compatible with the same packages installed (*and the same CPU for binaries).
 
I wouldn't give phone operators any control over the content since they have shown that they're all evil and only want to limit what the phone can do (e.g. Sell "unlimited internet" but block P2P, VoIP, Streaming Video, and other high usage apps off of their phones).
 
So throw the operators a few logos and strings too but keep them out of the phone's technical bits. If either the manufacturer or the operator wants to build themselves an application then let them but make them play by the same rules as Joe Whoever off of the street.

Yeah, totally. This is why a consistant app management has to used. Actually, IMO, all WM components are required, only to disable/enable. And pre-installed apps has a special place to disable/enable. Meaning even if those evil operators make a crapware to disable certain component/software, you can still disable it from that special place. And it is a dedicated place as MW components. Meaning you can't delete it, only update or disable/enable.

 

So, it will be like Locked Dedicated Space for MW Core, Locked Dedicated Space for MW build-in apps, Locked Dedicated Space for HW vendor, and finally an open dedicated space for user apps. Smiley

- rewrite the OS from scratch, fast operations/responsiveness required

- either write the OS in managed or expose wrappers for all the functionality in managed 

- stick to one programming model/framework only

- put touch at the ground zero

- invest $,$$$,$$$s in UX

- put strong hardware requirements in place for OEMs, both design and spec wise

- lock the OS from any 3rd party changes, only allow 3rd party apps to run

 

 

ManipUni
ManipUni
Proving QQ for 5 years!

You work there, make it happen! I hear throwing chairs makes things happen!

littleguru
littleguru
<3 Seattle

Something needs to be done. I am not sure if that is what Windows mobile needs but it needs a change. I still remember the days when it was cool to have a pocket pc that run Windows Mobile... that's a while ago. The problem is that the UI of these pocket pcs and the WM phones right now looks very much alike, even thought the hardware evolved...

 

I don't know if the APIs are so bad. I, personally, enjoyed coding for my old WM phone and didn't find it that hard. But it's true... a phone should expose a phone specific API set. I remember how hard it was to access some of the phone specific features.

A friend of mine is really into his WM phones and started a SMS conversation with me earlier. I paused my TV show to talk to him. He was saying how much of a pain it was to get the copy of TomTom he bought to work properly after upgrading handsets.

 

He said one of the problems was you had to hack their stupid installer to stop it overwriting the system bluetooth DLL with one which didn't work.

 

I texted back how crap it was to have to worry about system DLLs being modified when installing apps on a phone and how, to me as an outsider to WM, it gave a really bad impression of the idea of having Windows, in any form, on that kind of device.

 

Just before I hit send on the SMS, my HTPC (the one the video was paused on) spontaneously rebooted. lol. I guess there's a lesson there, too. (Then again, every other PVR I've seen has been even worse, so perhaps not. I just found it a funny coincidence.)

 

page 1 of 1
Comments: 22 | Views: 925
Microsoft Communities