Multiple sources have told me that Microsoft plans to overhaul the software and hardware before releasing the device. At this time, the software and hardware do not create a compelling solution that would move the needle for Microsoft and more importantly the Surface brand which is why when it came to the ‘go, no go’ decision earlier this year, it was not given the green light.
[…]
What you need to know about Andromeda is that the project is still alive inside of Microsoft but that it will not be released anytime soon. The company will re-work the hardware and software, see if it will move the needle, and if not, re-work again, until they find the right formula. Microsoft will not ship a project simply because the first phase is done, they are trying to get this right so that they don’t have another Lumia/Surface RT project on their hands.
There’s just not enough UWP applications at this point in time to support such a device.
And you know what? I don’t want UWP to succeed. I am sick and tired of OS vendors demanding that devs rewrite their applications for their special little snowflake API, because it wastes the dev’s time in an era dev time is scarce. Dear Microsoft, just create an API based on something that already exists and already has apps, like the Android API or whatever portion of it you can support. Basically do what a little OS called QDOS did, which was mostly compatible with the CP/M API and hence allowed devs to port their apps easily. Ask the company that bought the OS and evolved it into MS-DOS how they did it.
Edited 2018-07-13 23:20 UTC
Or, better yet, just sell the hardware and let us install the OS we want.
How’s that for a crazy idea?!RE: Comment by kurkosdr
This is not a tinkerbox like a Raspberry pi.
This is not a regular laptop with commonly available drivers.
Installing your own OS+Drivers on a device like this would be interesting for (starts extensive marketresearch) 0.01% of the market
Open the hardware then so drivers can be written.
Maybe all those driver-writers that know so much about hardware and are just waiting for some specs to appear online are all just waiting for some new hardware to appear to unleash the power of Microsofts hardware for the masses….or maybe we can wait 6 more months for Microsoft to try to get it right by themselves.
One of the above is a recipe to sell thousands of devices, the other to sell millions.
MS is largely a big operating systems company, not some crowdfunded startup; you aren’t seriously so delusional to think that releasing such incomplete product to market would be a good idea in regards to grabbing some mindshare, right?
So why aren’t you still running DOS apps? That’s the logical conclusion of never rewriting anything.
You didn’t understood what I said. Devs rewrite apps when there is a “gold rush” to a input method going on. Once the market settles there is no incentive for Devs to waste their time on special little snowflake APIs like UWP. And no the Surface is not a new input method.
Edited 2018-07-15 00:32 UTC
What? I have a few program that were written for an Amiga that I now run on a Haiku machine because the code was POSIX in design so I could port them very easily.
The problem is that in Microsoft OSes there often is a different API that you must use instead on a old API that does the same job but Microsoft want to force you design to run on their system only.
Problem with POSIX is that it is stuck in a world of command line interfaces and server daemons.
http://www.opengroup.org/standards/unix
Which has lead to loose relevance even on modern platforms.
http://www.cs.columbia.edu/~vatlidak/resources/POSIXmagazine.pdf
Microsoft tends to support older APIs almost forever…
Then you better get off Windows, because as of BUILD 2018 sessions, in a couple of more Windows releases there won’t be anything else, as Win32 and UWP will be finally merged into one store model.
There is no problem with this. Win32 can be packaged in many forms like .exe, .dll, .msi or “.store”. This doesn’t require any rewrite
Except MSIX sandboxes are the UWP and Win32 worlds merged as one, more to come next Fall.
MSIX is another way of packaging. You could package a Win32 app, but you could also package a .NET Core program to run on OSX or even some “compiled C-code to run on FreeBSD” or just package some documentation
MSIX has nothing to do with UWP.
MSIX has nothing to do win32.
You can package a win32 app as a sandboxed UWP with MSIX.
Edited 2018-07-17 10:02 UTC
“MSIX: Inside and Out”
https://channel9.msdn.com/events/Build/2018/BRK2432?term=msix%20…
“A closer look at MSIX”
https://blogs.msdn.microsoft.com/sgern/2018/06/18/a-closer-look-at-m…
Notice the nice MSIX Container description in both.
Anti-UWP crowd will get their beloved Win32 sandboxed, come hell or high water.
Apple, Google, Red-Hat, Canonical and others also appreciate paying customers.
It has been possible to sandbox Win32 for a long time, that is what App-V and Med-V are for. MSIX is basically the same technique for virtualizing the application and all the surrounding artifects (dll’s, registry)
None of this has anything to do with rewriting code for a different API/framework.
Store <> UWP
Win32+Bridge <> UWP
Win32+MSIX <> UWP
The point is that they are all getting merged under the UWP umbrella.
Let’s see who is right when Redstone 5 hits tbe streets.
Win32 is an API that you access by writing code.
UWP is a different API that you access by writing code.
Rewriting a Win32 app to UWP is a lot of work for a team of developers. So much work that Microsoft doesn’t do it themselves except for smaller programs like the calculator which is now an entirely different program compared to before.
MSIX is like a self-extracting zip-file. Unpack a few files to various directories, set a few registry keys, detect of tool A is already installed and change an app.config, etc
These two things have nothing to do with each other!
Yeah right, hence why all the new UWP controls on Redstone 5 were mostly contributed back from Office team as they disclosed at BUILD 2018 their ongoing migration to UWP.
Haters going to hate.
There is always Windows 7 and ReactOS for you guys.
Again, lets see who gets to be right when the next releases hit the street.
One big problem I have is with old apps running in the new high dpi resolutions of modern PCs… well, they just don’t work.
Even modern-ish Qt apps sucks at this (I’m looking at you, Perforce!
).
So what do we do? We have to port them to a modern UI framework that scales, right?
Plus, properly architected code should be easy to port over to new UI frameworks.
Those were different times, most really succesfull apps kinda bypassed CP/M or DOS, made their own calls to hardware.
I always find it dumb, why the he’ll do I need an app.
The world works fine without native crappie data stealing apps. Facebook, instagram, snapchat, YouTube, LinkedIn, and all those must have services can’t steal your location information, your call logs, access your mic, media or anything from inside the web browser if it’s not loaded. But you can’t even really stop that on an app. The only thing a phone needs is a proper browser that behaves the way it’s supposed to. Like kiwi browser on Android, play videos and keep them playing in background, handle downloads correctly, get actual desktop websites when necessary. That’s not hard, and windows mobile has that already. It has a very capable edge browser, but Microsoft was an idiot and kept doing stupid things with their mobile platform. This has nothing to do with apps, it has to do with Microsoft being stupid over and over again. This is another Microsoft kicking it’s own ass moment, release a proper flagship phone. The 950xl was almost a proper flagship, Andromeda has been rumored for years on end.
Edited 2018-07-13 23:24 UTC
WebApps can do most things apps can do, but not all. The mobile/responsive versions of my bank work quite well, but the app just worked a whole lot faster and easier.
That said, we are clearly moving in the direction that programs and apps get replaced with (progressive) webapps. This was the idea about the initial iPhone that launched with a great-for-that-time mobile browser and no app-store. 10 years later we have the ios-appstore.
Google does everything web-first and has the most used browser by far. Yet Android relies on the playstore. Also ChromeOS got a whole lot more useful now that it can run Android apps.
Conclusion, we are moving to web, but very slowly
Interesting
Thom here:
Hmmm…
Both statements are true. They are not in any way contradictory.
The two of them together indicate fanboyism.
But not Without(both).
—
I like the look and presumèd utility of the form factor and would very much consider buying a released device if it came with Win32 & “normal” desktop application capability alongside UWP/tablet-mode apps.
But not without.
I would LIKE such a device to operate as a PC compatible device and allow booting into alt-os’s. Linux, BSD etc. Even if that was via a MS written bootloader allowing boot into sanctioned subset of Linux/BSD distro builds.
(I’m aware the latter is perhaps slightly wishful thinking) but no desktop capability would be a deal breaker for me.