Sources close to Microsoft have confirmed the veracity of last week's Windows Phone leaks – but say no decision has been taken to base the mobile platform on the Windows 8 kernel. The information that leaked last week concerns 'Apollo', the next-but-one release of Windows Phone. It's all genuine, but should be thought of as more …
Jam tommorow ?
What MS really, and URGENTLY need to do, if they want to stand the faintest chance of competeing in the mobile market, is get apps onto Windows phone - oh and some BASIC FUNCTIONALITY.
Ever tried to send a single contact to another phone via BT on WP7 ? Did you know you can't ?
Its the one thing I can't understand; Ballmer keeps shouting "developers, developers, developers...", maybe even to rock himself to sleep. But when it comes to introducing a new platform those developers are all of a sudden held at bay (to some extend).
Sure; the tools and information is all freely available. But if people want to use their apps on their own phone then they need to pull out their creditcards first.
You can say about Android what you want, but at least they managed to get this issue sorted out; stuff goes through the marketplace /but/ if users opt-in for "insecure" applications then you can also supply apps. through http or e-mail.
And that is much nicer on the home-developers. Which IMO should be key considering how they are most likely the ones who may eventually fill up the marketplace.
Neither can the iPhone - and it's hardly stopped the uptake of that device...
It was a big feature on my Nokia 6310i though.
Didn't know that ...
strangely, no one with an iPhone I know has told me that, when I whinge about it.
Got the Mrs an HTC Desire S for Xmas ... she's overjoyed with it. The ONLY reason we got a new phone is that she has (finally) discovered the power of an online calendar which can sync with a phone, and the Nokia she had wouldn't work with Windows Live.
So far, it pisses all over my Windows Phone (supplied by work). And it's getting boring, her discovering new apps that I can't have.
Blackberries can Bluetooth contacts
The phone that is, not the fruit.
My old Symbian Sony Ericsson could too, and the millions of Nokias that preceeded it.
I have a very high regard for WP and ill be blunt and quite happily say that 'almost' everything that people throw up against WP is utter crap, i honestly dont think anyone who thinks its terrible has spent more than a few moments with it, however, whilst i think your going a bit over the top and far to generalised with your "basic functionality" comment, I do agree with your point on BT vCards, sending vcards via bluetooth is a much missed feature and i would hope that it comes to WP sooner rather than later, it should have been included from day one.
i suspect its to do with the limited BT profiles that are supported at the moment, id love for them to provide the full BT profile package which not even the beloved Android Supports, but then thats what i keep my Pocket PC running WM to hand for :)
Bluetooth to send contacts...
Sorry, but I can't see much use for it these days when you can text or email them about as quick.
Personal choice, I guess, but not a feature I've used for years.
@GitMeMyShootinIrons: The good old days
Bluetooth could do it years ago, for free, using very little power, regardless of network coverage, was nigh on universal, and is going to the phone of the bloke you're talking to. Plus it defines the format of the contact data (VCard so far as I know), and is presented on the Bluetooth link as being a contact. This allows the receiving phone to add it to the address book with little room for misinterpreting the data fields and merely prompting the user as to whether they want this to happen. Problem solved well over a decade ago.
*If* a handset manufacturer has put the 'Bluetooth' label on the box then all of the Bluetooth stack should work properly. But Androids and iPhones were (still are?) definitely a bit dodgy in the whole Bluetooth area generally (don't know about WP7), mostly I suspect because they were in too much of a hurry to do the job properly to a high quality. You couldn't even use a hands free kit reliably, something that not even Steve Jobs would ever have been able to convince the world it doesn't need. But I digress. Solving manufacturer laziness by saying "Oh just email it" is a backward step, an acknowledgement that things are a little bit worse than they used to be, is making things far more complicated than they need to be for no good reason.
But you are right, it is a personal choice. Personally I'd far rather not have to type in someone's email address or phone number just so that I can swap contact details with them when I could just zap it directly into their phone with only a couple of button presses. I mean, once you've typed in their email address you've pretty much done the whole thing anyway. By way of analogy, when you're giving someone a business card you don't want to be writing it out long hand in front of them do you? Bluetooth is instant, so there's no awkward email/text delay and there's so little room for error.
That "15-year-old Windows CE kernel"
The rumour *I* heard was that the original CE "kernel" was a hatchet job on the Win3.1 "kernel", making it over 20 years old and devoid of any architectural features that one associates with the word "kernel". Nothing of my actual experience of programming for that platform made such an ugly rumour terribly implausible.
The linux kernel that android and a few other mobile phone platforms are based on is about 21 years old, you can trace the history of the kernel in iOS back to about 35 years ago and the NT kernel is at least 19 years old.
'VMS' + (1,1,1) = 'WNT'
Puts NT's lineage at about 34 years; or about 40 if you consider RSX as version1....
"RSX was a separate path at DEC and the progenitor more than anything of VMS that went to NT via Dave Cutler." — Gordon Bell, Vice President, Research and Development, Digital Equipment Corporation.
But being 40 is only a bad thing if your under 30. ...
"Original CE kernel" is something that's fairly hard to define. CE was rebuilt from scratch more than once. Maybe one of the early products with that names was indeed a cut-down Win3, but the CE that WP7 is based on has nothing to do with any of the desktop OSes.
MS needs to do what they do best
Offer a platform for diversity which neither Apple nor Android are currently offering.
Apple are doing their usual thing and will continue to make money precisely because they don't want to be all things to all people - it's not their game.
Android is just amateur rubbish with not a single player serious about the future of the platform. Seriously - it is the most insecure computing platform I have ever seen and no-one seems to care. I wouldn't touch anything in Android apps if I was wearing a BNC suit.
If MS can get their act together with a serious mobile OS and a proper apps environment they will push android aside as the mass market alternative to the iPhone without any problem.
Uh oh...you want this icon.
BNC connectors are too big for any smartphone. You got the wrong mental image there.
"If MS can get their act together with a serious mobile OS and a proper apps environment they will push android aside as the mass market alternative to the iPhone without any problem."
Well, quite. Also, if Linux could just get a few high-end businesses to develop business applications exclusively for Linux, it would easily topple the Windows behemoth, flooding the business world with Linux-based systems.
And if Opera could pull together some decent marketing and get a group of supporters as energetic (and numerous) as Firefox had 8 years ago, they could become the top browser!
Also in the line-up for obvious yet useless statements: if $BUSINESS would just start outselling $TOP_BUSINESS, they would become number one in no time! If $POLITICAL_CANDIDATE would just get a few more supporters than every else, they would get elected! Finally, if stupid people would stop communicating their stupidity, the world would be a smarter place!
Suddenly I get the urge to pick up a copy of NT4 somewhere (TechNet) and slap it onto a VM. Simply because I can, and its sometimes actually fun to see how those older OS's worked out.
Having worked with NT4 (client) I have to say that for the stuff it did it wasn't all that bad back then.
Anyone remember the 'server hack' ? With a simple registry key change (and copying a splash screen over) you could actually change Windows NT4 client into a server.
Hmm, makes you wonder if they'll port that one over as well :-)
Ah yes NT4
I was working for an international pharma company back then and all our lab data acquisition units were NT4 powered. We had a great couple of days rest when all the boxes were zapped with a worm.
Good old days
Never had the 'pleasure' myself, but remember reading a hilarious article about a company whose CEO had been sweet-talked into adopting NT4 server by some marketing chappy going on about modernising in line with the rest of the sector.
So the IT guys were forced to dump all their superbly reliable Netware servers, only to find that their clients were now being so overworked managing all the bloated traffic that they ended up replacing all 1500 client's NICs with expensive new ones that would manage the TCP/IP stuff themselves, just to get some CPU cycles back for real work.
A fortune spent, and they finally had file and print, just as before. The piece ended with the beautifully sarcastic "and they'll probably announce it as a streamlining exercise."!
What were your "lab data acquisition units"
doing talking to the Internet, then? Just sitting there waiting to get cornholed by Stuxnet 0.01-alpha?
Ahh, the server hack. That one worked all the way upto 2K if I remember right. There was even a handy utility to do it.
After that trying it makes things go very blue. :(
If you feel really nostalgic, slap NT 3.51 on a VM with a chunk of memory and a couple of virtual CPUs, it's weird seeing that. Runs like an absolute daemon too. :)
Actually works up to XP / 2003 Server. All you need is NTSwitch. Makes XP into Server 2003.
RE:What were your "lab data acquisition units"
No stupid, they were connected to the server and BOOM.
Computer security lesson 1:
This might come as a surprise but something doesn't have to be connected to the internet to get pwned.
Pah, for the true retro experience go for the original NT 3.1!
What they really need to do first is get some direction. They need to pick a kernal and stick with it (even if its a new one!) and then they need to pick a development platform and stick with it. I dont know how they think they can gain momentum and control over mobile/tablet/etc markets when developers have no clue what platforms will be supported in a years time.
you just nailed it on the head.
AAPL doesn't make fist loads of money and having done a complete round-end without the curtoursey of a reach around on the mobile industry by not having clearly declared their intentions
AAPL's Approach - love them or hate them, AAPL has a plan, its doing, *uck you all!
Don't see why not
When Microsoft eventually releases Windows 8 tablets they'll be running an NT kernel. What's the fundamental problem with running it on a phone? We're talking of a device with a similar CPU and performance characteristics, just a smaller screen.
Besides, Windows CE has some serious issues in its current Windows Phone incarnation. It doesn't support multicore processors whereas NT does. So why not put the same kernel under both and save some development & maintenance grief for everyone?
Sure, why not?
Worked out great for Apple, or has everyone here forgotten that "iOS" is just the short way of saying "a carved-down BSD with a lovely UI"?
I'm not sure why you got down-voted twice. What you posted is 100% accurate, and (if my memory serves me correctly) was one of the selling points of the original iPhone to the geek/developer crowd.
Also, fwiw, I remember when the common usage of IOS was in reference to Cisco gear.
Don't you mean a mashup of Mach 3.0 and FreeBSD, with a weird-ass OO device driver model ...and a lovely UI?
Not the craziest idea ever...
...it just doesn't seem to be particularly useful. Is the Wince kernel really holding back Windows Phone? There's an awful lot of application-level stuff that is either absent or very poor which I'd hope is a slightly higher priority.
Broader hardware support is definitely 'aspirational' given the dire standard of driver development and the need to port existing drivers to a new architecture (eg. ARM) which will probably almost as much work as writing Wince ARM drivers for most vendors, I'll bet. It isn't like there are any x86 smartphones on the horizon, after all.
The Windows Phone flop just gets funnier and funnier.
Both Nokia and Microsoft running around like headless chickens, not knowing what way to go, what platform to ditch, what to embrace, what to do next week etc etc...
Pretty bad new for anyone that was stupid enough to invest in any Microsoft or Nokrosoft mobile products. Here today, dumped tomorrow.
New Nokia WinPhone review
That was hilarious too; couldn't even keep its charge for a day.
Just imagine what the battery life will be like with a general purpose kernel with context switched drivers and application multi-tasking. Maybe they want to bring back the old Motorola brick for the battery, it's so retro it'll probably sell as a fashion accessory.
Until you realise that the failure of winpho means we have a choice between the malware ridden cess pool of android and the super-max lockdown of iOS.
Ok exaggeration in both cases but consumers *need* a strong third platform it will keep Apple and Google honest.
"but consumers *need* a strong third platform"
JUST like the desktop - Oh wait
but consumers *need* a strong third platform
No. They only need the _threat_ of a third platform.
It is the threat of 'the Year of the Linux Desktops' that keeps Apple and MS working, not the actuality.
"In particular, the decision to throw out the 15-year-old Windows CE kernel in favour of the even older Windows NT/XP/Vista/7 kernel has not been finalised."
So it's going to be buggy and crashy but they haven't decided just how buggy and crashy yet?
> It may seem odd that a mobile phone should be based on the same technology that underpins a powerful desktop operating system
Yes, that would be crazy! Imagine running Linux on a mobile!
Or worse, Mac OS X
Seriously, you'd think that as soon as somebody mentions the 'M' word around here, all sanity flies out the window (whatever there was of it, at least) and everybody's so busy stoking the bonfire and tossing the effigy on top that they forget how other people have done much the same thing and not gotten nearly this much stick about it.
Yes, imagine Nokia having phones that run Linux!
what about a wrist watch
didnt IBM get Linux running on a wrist watch about 10 years ago.
NT is a trademark of Northern Telecom
So in one way a good OS to pick for a phone.
is that the Northern Telecom that...
became Nortel Networks, which became Nortel and then went bust :)
NT3.51 vs NT4.0
In high reliability situations the NT3.51 lived on quite a while after NT4.0. They blew stability for a measly 10% increase in video & game performance. Yet it wasn't until XP that most games really used NT rather than the Win9x platform.
Since Printer Drivers use GDI also and sort of emulate a graphics surface on NT4.0 a bad printer driver could "BSOD" your workstation or Server. Impossible on NT3.1 to NT3.51.
Ironically the WHOLE NT3.1 was release about the same time as ONLY the kernel of GNU/Linux, I think in 1993.
Why NT3.51? Well, MS "invented" largely fake APIs so Win95 Apps wouldn't run on WFWG3.11 (some only really 16bit). In reality WFWG3.11 had 32bit TCP/IP, 32bit Diskmanager option and Win32s to run NT applications (no named pipe creation was only significant lack).
So they had to add these to NT3.5 and give many a free upgrade to NT3.51. Oh, no version of win9x could create named Pipes either. Win95a little more than all the 32bit options for WFWG3.11 with the Explorer shell.
NT4.0 could use Program Manager and Filemanager and there was a "Technology preview" Explorer shell for NT3.51.
I think Win CE was more a cut down NT with very limited thread/process space (a fixed number) done for a Sega console, and not based on WFWG3.11/Win9x. It was designed for very limited memory with no paging, hence small fixed resource limits rather than dynamic structures. The main stupidity of WinCE PDAs and Phones was using a miniature version of the Windows interface instead of something designed for a small screen. 320 x 240 was a LARGE win CE screen for a long time!
Well put. Indeed the aspect of pulling in the graphic driver to the kernel was not the greatest of moves, for server usage, let alone workstations which would BSOD enough to negate that performance gain.
But there again they did use to have NT3.5 available for non-intel platforms, so it's probably better written code with regards to comiling for new platforms. Though I'm not sure it's upto the real-time aspects needed. Maybe there better of buying RIM, selling on the email service and phone production aspects and going with QNX. As long as the API's map then it's just a compile instead of having to migrate from one set of API's to a set of API's that change direction every other update.
Either way it will get to a stage were feature set is not the factor and it's all back down to battery life and size/weight.
NT and kernel mode printer drivers
Yes, it took a while before games really started using NT, but Adobe Photoshop was one of the apps that benefitted from moving the GDI into the kernel. A dual CPU system would run PS slower than a single CPU system prior to NT4. (and keep in mind that XP kept GDI in the kernel for obvious reasons)
As for the printer drivers, they were moved back again to user mode starting with Windows 2000. (http://blogs.technet.com/b/askperf/archive/2007/09/14/kernel-mode-print-drivers-gone-the-way-of-the-dinosaur.aspx)
I quite liked NT4 back in the day, but I have never been big on printing. :P
"The main stupidity of WinCE PDAs and Phones was using a miniature version of the Windows interface "
Have a look at the HP Jornada 720. Released in 2000, nearly Psion size (but not weight or battery life), nearly sensibly sized colour touchscreen (640x240), nearly usable(ish) keyboard, 200MHz ARM.
Totally crippled by the useless WinCE-based HPC2000 OS which MS couldn't make interesting so they dropped it altogether fairly soon afterwards. ActiveSync wasn't real helpful either.
You can even get Linux for them these days; it'd be a laugh if there was an Android port (won't happen, only 32MB of RAM, though the near-identical 728 had a whole 64MB).
Ten years is a long time for the wheel to turn again, even in Redmond.
All of the above is exactly what I remember as well. In fact at the Microsoft WfWG 3.1 preview session (where I got my copy) the MS rep was publicly commenting they had more 32bit network code than OS/2. (IIRC that was still true until Warp3 or 4)
The Win32s API was single threaded, so threaded apps wouldn't run under WfWG3.11, but all other Win32 apps did. It was pretty trippy stuff running NT apps under WfWG. It was an add-on though, it wasn't included by default. I typically only did it for fun, like running OS/2 character mode apps under NT and surprising people who didn't know NT actually started as OS/2 NT.
WfWG3.1 itself was a backport from the semi-aborted Chicago project. Chicago was much more like OS/2 in nature, but the project ended up morphing into the less advanced and slightly disappointing (after Chicago) Win95. Wish I'd saved off the beta diskettes it came on. I'd like to fire that up for fun now in a VM and see how much my memory clashes with reality. :)
On nt3 and microkernels
This is what ast has to say: "Windows NT 3.1 was a half-hearted attempt at a microkernel system, but it wasn't done right and the performance wasn't good enough on the hardware of the early 1990s, so it gave up on the idea for a while. But recently, it tried again on modern hardware, resulting in Singularity."
They've had such influence that this wasn't a good signal for others tinkering with the same idea, though counterexamples like QNX show that it can be done with frightening efficiency.
Personally I've seen them botch too much, too often, too horribly, to trust they won't botch it on the (n+1)th attempt. Reason for harkening back to NT3 could be that it might be ported as it was to alpha, which may be less true of this apparently already abandoned research project.
A better approach --for us, not redmond-- might be to put andy's minix on those arm-based things in phones and fashion a programming environment and UI on top of that. At least there'll be someone with a few clues about building solid operating systems nearby. The porting to arm he's looking to pay a developer for, in fact, so if interested, go ahead and ask. I already did and he declined, worse luck.