My prediction: Cshell at build

[Drael646464]

See, here the problem starts. The OS has no clue what the user wants. If the user started a video, the associated semantic would be to play the video until end. When moving a process to the background, the OS should never assume that the user wants to have that process suspended. Instead the user should just stop the video or close the tab.

I don't reckon neither the OS nor the developer should be able to make the assumption.

Its more convenient to me, for the video to just automatically pause when I switch tasks. Its smoother for me. Just like when a skype call comes in and it pauses the music. IS there also ever a reason why someone would listen to music AND youtube at the same time? Yet, it happens. Theoretically I could pause the music, but I don't see I should have to, any more than someone else should have to have it stop if they want to keep listening to their podcast whilst browsing.

I believe this thing should be user controlled, not developer or OS controlled. Ie settings controlled behaviour. Only the user knows what he wants an app to do when he shifts. If the devs can't put this in their app, perhaps the OS makers can. Either way, I'd prefer to have apps behave the way I'd personally like them to.

I was only talking suspension (e.g. app is not scheduled for execution anymore) not about RAM management. Hibernation to disk is another topic. Issue is, WinRT apps are just suspended even if there is no need for hibernation.

Meh, its related IMO. If you have a phone with 4gb of ram, I don't really see why tasks should be suspended at the whim of the OS. If there's something you don't use often, but want to have run quickly and be more accessible for example. Or at least I see them as related. It's not RT, its quasi-RT. Windows doesn't manage its ram that way, it caches it in and out, on demand. But perhaps if we discuss that its too much of a tangent

You always have the option to shut down a service (or to stop an app for that matter) manually.

Okay, so lets say I have facebook installed on an android phone. If I stop the facebook service, and I load FB, the service will reload. In general the android system does not like you messing with services, or force closing apps, nor do the apps themselves. All in the name of 'the best possible experience". And what about ios - can you even kill services manually? Windows phone? IDK.

Even in use cases where that works - this seems like a messy and complicated solution for something that could be very easy, just preventing the app from running, or only letting it run when the app is open. I know folks like google will never do this, but I could do this in desktop windows 10 by disabling the service.

In any case the vast majority of apps would just naturally suspend if there is no user input so they would be suspended anyway in the background, because there simply are no events coming in.

The always connected quality of mobile apps means they usually keep running services, and those services keep updating the app system. I think probably the overall load of most apps is quite small though.

Then there are application, which are supposed to do a certain job, which needs to finish independent of any events. As example if i want to compile 500 files it would take a while and my expectation is, that compilation completes in the background.
Then there are applications, which are dependent on external events (e.g. no user input). These external events could be timers or coming from the network stack. Example of these type would be music players, gps navigation apps etc.
Finally you have application, which would naturally suspend due to synchronization with some hardware mechanism. Example would be games, which are suspended while waiting for the GPU to finish the presentation.

In all these cases i do not see a single reason, why the OS should enforce suspension when pushed to background.

I agree with all these case examples. There will be absolutely people who want to do those things.

Although I'm not generally the sort of person to compile 500 files on my smartphone, or browse/text while following GPS nagivation instructions, and every mobile OS supports background music play, Ive had a 100% real-time OS in bb10 for awhile, and honestly I've never found any benefit in it. Years of use, and not a single instance comes to mind.

I'm reminded of someone who was moving from bb10, who was complaining about the limited realtime emulation (and complete lack of in ios), saying he'd like everything he's doing to be real-time, all the time. Most people were just stratching their head wondering why he would care, but he had loads of specific examples, he was a very busy worker on his phone, and he had his reasons.

Or the fellow on here that was complaining that windows 10 mobile didn't keep playing youtube when he was doing something else.

Or even my issue with loading multiple youtube tabs in edge on desktop, and having them all play at once.

There's a lot of different user behaviours and needs out there and there's a lot of different software playing together. So for me, my ideal system rests somewhere in between, and something that doesn't exist.

I don't know if I'll ever get exactly what I want, because mobile OSes tend to be hand holders, not bastions of freedom and settings, but perhaps if WoA ever comes to mobile way down the track I'll be able to get close (by disabling services, and shifting battery management settings)

 

[Sedp23]

Didn't see any c shell

Sent from Idol 4s

 

[Drael646464]

Didn't see any c shell

Sent from Idol 4s

Yup, I was wrong (on cshell @ build). We did however get Cortana skills api. I've yet to hear about anyone trying the preview much.

 

[Drael646464]

That's a very old quote and nowhere does it state that it is just temporary. WinBeta recent speculations point out that W10M will get the features of the Creators Update (night light and custom accent picker) with the "official feature2" update but that's it.

https://www.onmsft.com/news/windows-10-mobile-to-receive-new-features-this-summer-for-enterprise

Apparently mobile is getting some enterprise focused updates in the FCU. Remains to be seen what they are, or what else we will get, but at least its on the cards mobile will get some features.

 

[PerfectReign]

The odd thing is that we insiders would see these features but we don't.

Just need Astoria.

Sent from mTalk

 

[Drael646464]

The odd thing is that we insiders would see these features but we don't.

Just need Astoria.

Sent from mTalk

Usually feature2 is used to delay features whilst work is done elsewhere, or testing is completed. Possibly all features on 10m, are being stalled until onecore is complete, and mobile is merged back into the main development branch? But who really knows what's happening in the dev's offices.

They keep it tight lipped, I'll give them that.

 

[milkyway]

https://www.onmsft.com/news/windows-10-mobile-to-receive-new-features-this-summer-for-enterprise

Apparently mobile is getting some enterprise focused updates in the FCU. Remains to be seen what they are, or what else we will get, but at least its on the cards mobile will get some features.

Brandon said "in the late summer". That would be before the FCU hits. So I think this will be part of the feature2 update

 

[Drael646464]

Brandon said "in the late summer". That would be before the FCU hits. So I think this will be part of the feature2 update

I'm not American, so I don't really know when later summer is.

But if were part of feature2, that would seem to defeat the point of splitting the dev branches temporarily for work on one core. Ie, if they could still add features and continue work on mobile features, there would be no point in splitting the branches, and stalling and delaying feature updates that were already supposed to come to mobile.
Splitting branches like this is almost always to delay features for testing, or other code work.

When is late summer for the US, in a few months?

Guess perhaps the delayed features are already baked, and they are just waiting for the one core dev branch re-merger, to implement and test. Ie, they have the features baked, and are just waiting on the new one core to bake, to put the two together and merge the dev branches back.

If that's true, we should see some insiders action before release, indeed, if that's true we should see the work on one core complete in about a month or so.

That would be kinda nice because it would suck for mobile to get behind things like timeline, and Cortana skills, versus desktop. It would help if mobile was merged well before then, so there is a little time to get some of the FCU desktop features on mobile.

 

[tgp]

When is late summer for the US, in a few months?

FWIW, summer officially ends and autumn begins on September 22, 2017 in the Northern Hemisphere.

 

[milkyway]

So it's official now: W10M will end with feature 2
https://www.neowin.net/news/microsoft-confirms-that-it-has-no-plans-to-move-windows-10-mobile-off-of-the-feature2-branch

 

[Drael646464]

So it's official now: W10M will end with feature 2
https://www.neowin.net/news/microsoft-confirms-that-it-has-no-plans-to-move-windows-10-mobile-off-of-the-feature2-branch

Would be nice if someone could link to an online, watchable version of the webcast, rather than this Chinese whispers stuff. I'd like to see who said it, what was said, what was the context. Those articles are 99% speculation, even accepting the quote is accurate.

Although this is exactly what I expected since I got more familiar with the topic - a last huge update for current phones (Cortana skills, cshell, FDS, files on demand, timeline), and then the push to 64 bit, with those same features plus some more - like active stylus and god knows what.


We know both files on demand, and cshell both work on 32 bit. Which means cshell at a minimum is intended for 32 bit phones. If they were rolling out to 64 bit only, they have coded on some 74 bit reference hardware, with a 64 bit kernel and app platform, and it plain wouldn't run.


I'm not sure though if the FCU will be the last big update or not. Maybe, but I suspect some of those features they have promised will roll out early next year. After all, there's not actually going to be much to replace existing phones for awhile, and even when they are released they need some market testing, before the transition can be completed.

After all no one makes 32 bit arm chips anymore. So the current platform couldn't be supported forever. It's simply not possible.

Whether they call it feature2, or the main branch is pretty immaterial, its both something with a limited life of support, and the only consumer base that actually exists.

Obviously I disagree than cshell is some king piece that will sway all consumers in its new state, or that windows on arm will do the same. It's hard to say whether those things alone will have any real impact on the consumer market, or even the enterprise market. They definitely aren't engineered to only work on 64 bit chips as the article implies, as they have been successfully run on more than one 32 bit model.

In time device convergence will have real value, especially as a lower priced proposition. ATM things like DeX and continuum are mainly toys for enthusiasts.

Windows on arm run on a phone? Well its going to run 40-70% of native speed. On phone hardware which isn't as powerful as intel tablet/laptop hardware and nowhere near a desktop. So its not a gaming platform. The apps don't scale for phones, they work for tablets and laptops, but that makes them slow running and only useful in continuum, unless loads of people buy windows phones and make developers want UWPs. Which seems unlikely.

Useful for business people who happen to have monitors and keyboards everywhere they go. Pretty much useless to everyday people, until the UWP platform fills out a lot more.

I also don't think microsoft's next gen ideas rest on the software alone. I don't think WoA will be coming to phones soon. I don't think cshell is going to change the world.

Andromeda is a mobile device that doesn't use 10m, and doesn't have a keyboard attachment, and isn't geared as some kind of mini-surface. Its an actually new idea. It's not so much a "game changer", as a new idea that might chip a bit of the market away.

I think the software refresh of win10m is mainly about the chips. I'm not waiting for the windows mobile messiah aka 'surface phone' to come and convert all the heathens. All I am expecting is little moves of the pieces on the board, slowly over time towards a slightly better market share.

Which I suppose isn't as exciting, until years down the track when theirs loads of actual desktop software and games you can run on any architecture, in any form factor. But that ain't happening quickly. We'll probably see AR and VR start to mainstream by the time we see that.

 

[Johnny Tremaine]

How old is feature2? I thought it was only recent.

Dan, from this site, says his insiders have confirmed that feature2 is temporary. His sources have been right on many things, he seems very confident in it, and personally I trust his word more than speculation that seems to spread like old wives tales.

Ahem: https://www.neowin.net/news/microso...-windows-10-mobile-off-of-the-feature2-branch

It's official, from the mouths of Microsoft execs.
Apparently feature 2 is W10 Mobile's dead end, as it's, "feature complete". W10 Mobile devices won't recieve redstone 3.
It's basically WP7.8

 

[Drael646464]

Ahem: https://www.neowin.net/news/microso...-windows-10-mobile-off-of-the-feature2-branch

It's official, from the mouths of Microsoft execs.
Apparently feature 2 is W10 Mobile's dead end, as it's, "feature complete". W10 Mobile devices won't recieve redstone 3.
It's basically WP7.8

No what MSFT said was "windows 10 mobile would stay on feature two" if we are to believe whomever is reporting that the insiders webstream said that, not that " dead end, as it's, "feature complete". W10 Mobile devices won't recieve redstone 3" which no one at MSFT has said, and the writers of that article (and to some degree you) basically made up (they never said anything about win10m never getting rs3 for example).

MSFT has said there are features coming to win10m on several recent occasions (build, and the enterprise features announcement). Also cshell (UI tweak), that for some reason the authors think is "the great next step" (lol) runs on 32 bit chips, that no new phones could possess - something that would not be possible if it wasn't written and tested on 32 bit machines...

The real reason why current phones would _automatically_ get put out to pasture eventually is they are 32 bit, and current chips are all 64 bit. But for that to be a workable transition, msft will first need actual new phones to write new revisions of win10 mobile for. Cutting support for current phones, while there are no successful new phones will simply bleed customers, I find it implausible any corporation would be that stupid.

 

[milkyway]

Drael, your optimism is admirable, but please have your expectations in check. Here are the relevant quotes from the Insider chat:
Microsoft Insider team provides answers about the future

 

[Drael646464]

Drael, your optimism is admirable, but please have your expectations in check. Here are the relevant quotes from the Insider chat:
Microsoft Insider team provides answers about the future

I think it's the cshell/all in one/windows on arm on mobile crowd who are expecting the earth. All I am expecting is for Microsoft to transition on its way to 64 bit chips. You see me as optimistic, I see these blog writers as almost utopic, and my view as pragmatic and down to earth.

Those quotes are helpful.

Digging into what is factually actually said, explicitly and contextually:

*Currently win10m is staying on feature2, according to the teams making the builds (as opposed to management). But I don't see the big deal if it stays split. It means something to the coding team, and what they are doing, but we are not privy to what their process actually is there.

*No plans for nitelight right now. I don't care.

*Enterprise focused features as promised. That may or may not be good. I'm in business, we'll see.

*Fluent design, may or may not come to the OS as a whole (Answer was ambigious, but at least it'll come to all the apps anyway)

*You shouldn't worry about the branches and what they mean (read, people are reading too much into it, or the wrong thing, don't be so gloomy)

*Redstone 3 updates for mobile - ambigious/not answered

I look at these things from a business POV, and I do expect 64 bit to replace 32 bit, in phones. I don't expect current phones to be updated forever. I'd expect this year, maybe early next year, might be the last new features before moving over the 64 bit (assuming any OEMs make phones).

I don't see the real gain, consumerwise for WoA on a phone. That seems like a niche product, as UWP apps are the actually scaling, native speed ones, people will want for a fluid, transforming OS - and that's a long project, not one that will suddenly be complete next year. Win32 is the past, WoA is just a messy bridge/enabling technology.

I don't think they understand that things like WoA, on tablets and notebooks are actually tricky means of getting more UWP development, and are not really intended to get users using win32 at all - but rather finding the experience poor next to UWP (which will scale better, run faster, be more secure, auto-update, run across more devices etc) - UWP is native on WoA. Win32 is a gangly add-on. win32 is not the future, msft wants to phase it out.

The core, fast running, properly scaling apps, would be the exact same apps win10m can run already - UWP. Anyone that wants to game, or use a mobile app - UWP.

I mean just look at this. Its like one of those aweful fan renders of dream phones, that no one could ever make:



These people aren't saying win10m is "over" because they think is "baked" and nothing can be added, or its an outdated failure, like critics. Because its quite clear MSFT still wants in the mobile space.

They are saying this because they believe the second coming of jesus in phone form, is on its way in 2018.
I simply don't agree that 2018 will be the year of WoA on a phone taking the world by storm, the great and powerful surface phone, and that cshell will be so mind blowing, all those dusty nokias will have to be left in the dust. That's not level-headed from my perspective.

What is level headed, is that 32 bit chips will need to be slowly phased out, when the 64 bit ones actually come in.

The thing is, who IN msft to believe? Satya? Panos? The insider team? They are all saying, possibly related but quite different things. Its more like a wall of ambiguity than a series of concrete statements.

the change to 64 bit, isn't as exciting as "the ultimate ultramobile device" to some, whatever that actually means. But living in a grounded word of commerce, such a device, were it to come next year, wouldn't put apple on its knees would it? Nor would it steal significant market share from Samsung. At best, when it was baked, many years later, it might suit people who can't afford a real PC as well as a phone.

UWP itself is basically way more exciting than that - apps that can run on your phone, desktop, tablet, HoloLens, whatever - full desktop grade goodness. Making a phone that runs another app platform, isn't going to make that miraculously appear.

Rather windows on arm, on far more popular for windows users, tablets and laptops, will make win32 such a poor comparative experience, developers will be forced to move to UWP, whilst simultaneously creating demand for mobility apps on laptops - things like snapchat, and banking apps.

Outside of enterprise its just not much use on a phone, and that strategy for creating more UWP, doesn't work on a phone because it doesn't have the screen size for win32s.

 

[Drael646464]

A lot of the really cool features coming to win10 m are app based anyway. Timeline, files on demand, Cortana skills, fluent design, cloud clipboard. Two of those, as far as I am concerned are huge - and they are all part of Cortana.

The only one's that are UI based are cshell, and nitelite (and the new keyboard), and out of those the only thing that excites me is the large tiles, lol , and the features that come with swfitkey.

Would I mind split screen for when I occasionally use continuum? No, but I'd hardly spend money to get it. Nor windows on arm on a phone, neat trick but I'd see it as a bonus extra rather than something I'd want - I'd rather use win32s on native intel chips at full speed, with screen sizes that suit it. I've got a tablet for that nonsense. Honestly I would shed no tears whatsoever if win10m never got WoA or cshell, or if I never saw them on any phone. In fact I think it would probably amuse me more, if WoA never came to phones. Although it might have its use in business, its pretty useless for consumers anyway.

I've other devices that do it better. Anyone that wants those features could buy a windows on arm 8 inch mini-tablet anyway, and it would be way more functional without a screen.

I might stage a small funeral for large tiles though XD

It wouldn't surprise me at all if win10m got a few OS features here and there, until mid 2018, assuming someone actually released a 64 bit windows phone either (and if they don't there's no point in stopping development, MSFT needs its mobile testbed).

We have keyboard changed over to the swiftkey based one, confirmed (which presumably will also carry cloud clipboard and some other build mentioned features with it), and the productivity features. The new keyboard supports scripting as well. We don't know what these enterprise features actually are yet either.

That and all the app based features promised at build - I mean, what would people actually be missing by way of this "dreaded" feature2?

That's a truckload of cool stuff, even if we only got most of what has been promised and not all of it. Heck, if we got 30-50 percent of that, it would be more useful features than any android revision ever gives people.

Cshell is actually coded in 32 bit. Wouldn't blow me away if the HP, or ever XL and 4s got it. It might actually surprise me if it doesn't end up on some kind of HP device.

Cshell isn't the second coming of Christ in technology form though, its just a part of the evolving windows vision. A nifty trick, that has its niche useage (a phone doesn't have the power, the storage, even when docked to replace a PC yet, even running native UWP, its not like you can store your media collection on it). Most people consider 2tb pretty standard these days, and a phone runs generally 64-256 at the most (unlike a tablet, that can actually go up to 1 tb via msata cards, and also can have more oomph via intel and also thunderbolt three).

But I also suspect its a long way from being baked, cshell. They are probably picking a late 2018 release date or similar. Originally they planned the continuum changes for the CU. The leak we saw, was less than ready. It's not coming to anyone soon IMO.

 

[tgp]

You see me as optimistic, I see these blog writers as almost utopic, and my view as pragmatic and down to earth.

In real life, you need to take what seems like a logical, pragmatic view and water it down quite a bit. Then you might come closer to reality! :winktongue:

Around here we've been saying for the last several years what you've been saying the last couple months. These utopian expectations have not come to pass. That's why we're skeptical. By your own admission, you're new to Windows mobile. We have experience in this ride with Microsoft; you don't. Trust us! :amaze:

 

[Drael646464]

In real life, you need to take what seems like a logical, pragmatic view and water it down quite a bit. Then you might come closer to reality! :winktongue:

Around here we've been saying for the last several years what you've been saying the last couple months. These utopian expectations have not come to pass. That's why we're skeptical. By your own admission, you're new to Windows mobile. We have experience in this ride with Microsoft; you don't. Trust us! :amaze:

There's just little benefit to a windows on arm, phone configuration. Enterprise users might get _some_ user out of it, but in general it doesn't bring devs to UWP like it does on tablets and notebooks, and without a robust UWP platform, it's not much of a calling card to have cshell yet (ie apps that can run as both desktop-style and mobile style apps).
What people are imagining in this "ultra-mobile pc" is actually accomplished via the slow growing UWP platform, not by windows on arm.

And we'll have to wait for awhile for MSFTs various plays to build that up.

I see no reason however to doubt the primarily app based features that will trickle down from desktop mostly to win10m, like Cortana skills, timeline, cloud clipboard - nor the new keyboard or the soon enterprise features (whatever they are), as all official announced and not particularly ambitious.

And those features, even if msft bungled and only delivered a third of core promised features there, they would still add more value than any android revision. But I see no reason for that, cloud clipboard will likely be built into swiftkey, and thusly essentially they are all trickle down from desktop (timeline and Cortana skills via Cortana, files on demand via onedrive), outside of the enterprise features.

Although I guess onedrive files on demand does require some OS bake in, they have already demo's it on a Lumia 950. Guess its possible they decide to sleep in that day, but doesn't seem super ambitious or hard XD