Windows 10 - Deeper Impressions

[NoelC]

Yes.  It's a full OS "in-place upgrade" install, not just a "cumulative update", and it takes a looong time even on a fast computer.

 

 

 

Files, yeah, but not settings.

 

Thing is, many settings ARE preserved, which implies it's technically possible - and that means the settings reverted are purposefully being changed by Microsoft.

 

I wonder if it's the same in countries with laws that take privacy very seriously.

 

-Noel

[JorgeA]

When you install the "upgrade" from 10240 to 10586 do you get the setup screens like you would have gotten during the initial "upgrade"?

 

If you mean the screens that give you vague messages in chatty condescending language (such as "we're getting some things ready"), the answer is yes.

 

TBH, in my case the move from 10240 to 10586 took so long that I walked away from the computer, so I can't say what all got displayed on the screen during the process.

 

IIRC, during the original Win10 installation there were several points where you needed to make a decision as to settings, and the process would halt until you made a selection. There wasn't anything like that when moving to 10586.

 

--JorgeA

[Tripredacus]

I did a Windows 10 upgrade test a couple days ago from Windows 8.1 and it upgraded to 10586 version of Windows 10. One thing different I noticed was that the context menu for the Start button is skinned differently now.

[JorgeA]

 

Sounds like in theory they could continue to support XP (because it doesn't require these certificates), but not Vista?

 

Sounds like in theory - and hypothetically (of course) - the good AV guys are giving you a full load of bullsh*t .

JFYI:

https://www.globalsign.com/en/blog/microsoft-announces-updates-sha-1-code-signing-policy/

http://social.technet.microsoft.com/wiki/contents/articles/32288.windows-enforcement-of-authenticode-code-signing-and-timestamping.aspx

 

jaclaz

 

 

Thanks jaclaz, that was helpful.

 

If I read it correctly, the TechNet article would suggest that the AV vendor could continue to sign an SHA-1 (instead of SHA-2) certificate for Vista systems, but only if that's the "only" OS for which it's developing software:

 

Code Signing Certificates: Windows will no longer trust files with the Mark of the Web attribute that are signed with a SHA-1 code signing certificate and are timestamped after 1/1/2016. With the exception of issuing certificates to developers who intend to develop only applications for Windows Vista, Windows Server 2008, CAs may not issue new SHA-1 code signing certificates after January 1, 2016.

 

Then again, the article is not exactly a model of textual clarity, as several of the commenters at the bottom of the page indicated.

 

Presumably, the AV company's current version is signed with an SHA-1 certificate and so should be fine for Vista, but they've announced a new version and will not only stop selling the current one, but will even stop sending virus definitions updates for it early next year.

 

--JorgeA

[JorgeA]

 

Huh, that's interesting. Vista, though, does require these code signing certificates, is that right?

Sounds like in theory they could continue to support XP (because it doesn't require these certificates), but not Vista?

I don't know, I've skipped Vista altogether, but I doubt it. Anyhow, the best person to consult about that would be MagicAndre1981... If he doesn't chime in soon, it might be a good idea to PM him about it.

 

 

Andre seems to have dropped out ever since the tussle over the value of UAC. I wonder if he'll reply if I send him a PM. I'll think about it.

 

--JorgeA

[JorgeA]

Check out this statement I found from another forum. Speaking of build 10586, a security software developer says:

 

Thats why I hate MS for giving devs just 2 weeks time to test their software on the biggest kernel change since Vista.

 

Biggest kernel change? Two weeks? Nice!

 

What do you think? I'm especially curious about the magnitude of the kernel change.

 

--JorgeA

[NoelC]

I haven't seen any reason to think there's been a significant kernel change.  I wonder what they're talking about.

 

• The file system implementation seems as slow as in build 10240.
• An idle system seems to need about 1 GB and roughly the same processes.
• Picky programs all still run, implying no change to Win32.

 

Anti-malware authors, who may use undocumented interfaces, might be especcially sensitive to system changes, but that's nothing new.  Even our own Big Muscle has a version of his software working fine with build 10586.

 

Frankly all I see they've changed is some window dressing, and only a little at that (ooh, colored title bars!)

 

I did see one improvement so far that stood out.  The small dialog you used to get with a one line text field for building the system path is now a multi-line dialog so you can see the individual folder paths more easily.  Nice feature, but did it really need an entire OS replacement?

 

 

-Noel

Edited November 17, 2015 by NoelC

[vinifera]

he wrote biggest kernel change SINCE VISTA

 

6.0 to 10.0

 

to us ordinary users thats nothing

to devs, its new driver type (since number changed as it didn't follow up 6.x numbering), and probably some API hooks beneath

[jaclaz]

he wrote biggest kernel change SINCE VISTA

 

6.0 to 10.0

 

to us ordinary users thats nothing

to devs, its new driver type (since number changed as it didn't follow up 6.x numbering), and probably some API hooks beneath

Yes and no, meaning that it is likely that there is still some "lying mechanism" about OS versions

JFYI:

http://blogs.msdn.com/b/cjacks/archive/2009/05/06/why-custom-actions-get-a-windows-vista-version-lie-on-windows-7.aspx

and:

http://blogs.msdn.com/b/ieinternals/archive/2014/02/19/internet-explorer-and-everybody-else-version-lies-for-compatibility.aspx

 

The “Windows NT 6.2” token is another interesting one—it lies in numerous ways. First, Windows hasn’t used the “NT” moniker since Windows 2000 was released. Next, my application is running on Windows 8.1, and yet the UA string claims that Windows is v6.2. That’s a lie stacked on top of a lie. With Windows Vista’s release, the Major version returned by GetVersionEx has been frozen at 6; Windows Vista was 6.0, Windows 7 was 6.1, Windows 8 was 6.2, and Windows 8.1 was 6.3. So, even considering the 6.x lie, you’d expect that my test application would be showing 6.3, but it’s not.

That’s because GetVersionEx now directly lies to applications that call it when run on Windows 8.1; it only returns the 6.3 value if the application specifically indicates compatibility with 8.1 inside its manifest. To prevent applications from lying (and, say, claiming compatibility with Windows 9 before it exists), the manifest requires that callers list their supported operating systems using a GUID; each OS version’s GUID is only documented when the new OS version is reaches its public preview stages.

 

 

jaclaz

[rn10950]

I haven't seen any reason to think there's been a significant kernel change.  I wonder what they're talking about.

 

• The file system implementation seems as slow as in build 10240.
• An idle system seems to need about 1 GB and roughly the same processes.
• Picky programs all still run, implying no change to Win32.

 

Anti-malware authors, who may use undocumented interfaces, might be especcially sensitive to system changes, but that's nothing new.  Even our own Big Muscle has a version of his software working fine with build 10586.

 

Frankly all I see they've changed is some window dressing, and only a little at that (ooh, colored title bars!)

 

I did see one improvement so far that stood out.  The small dialog you used to get with a one line text field for building the system path is now a multi-line dialog so you can see the individual folder paths more easily.  Nice feature, but did it really need an entire OS replacement?

 

 

-Noel

 

I find it hilarious, yet satisfying, that MS didn't try to replace this with a Metro "app". They seem to really be pushing that lately (as we all know) and they also seem to want to kill win32 at their earliest convenience.

 

he wrote biggest kernel change SINCE VISTA

 

6.0 to 10.0

 

to us ordinary users thats nothing

to devs, its new driver type (since number changed as it didn't follow up 6.x numbering), and probably some API hooks beneath

 

Yes, but the new driver model and the NT major version change have been in Windows 10 since RTM. What I got out of this is that the "biggest kernel change since Vista" happened between 10240 and 10586, and if that is the case, I am really curious into seeing what it is. (they probably integrated telemetry into the OS at a kernel level)

Edited November 19, 2015 by rn10950

[JorgeA]

Windows Phone market share drops to 1.7%

 

According to Gartner, in Q3 of 2014, Windows Phones represented 3% of the worldwide smartphone market, and still saw an established user base despite the low figures in comparison. Today's updated figures now show that market share has continued to plummet, and has reached 1.7% one year later in Q3 of 2015.

 

And so Microsoft's response to the universal (pun intended) failure of its mobile projects, is to twist and turn Windows itself into a mobile platform. Brilliant.

 

BTW, in 2011 as Windows 8 was nearing the light of day, Gartner was predicting a 19.5% mobile OS share for Microsoft by 2015. The share has actually fallen since then (from 4.2%).

 

Way to go, Steve Satya Ballmer Nadella.

 

--JorgeA

 

 

 

[jaclaz]

BTW, in 2011 as Windows 8 was nearing the light of day, Gartner was predicting a 19.5% mobile OS share for Microsoft by 2015. The share has actually fallen since then (from 4.2%).

 

Well, to be fair to the good Gartner guys , this particular prediction seems as accurate as most of the others they make.

 

jaclaz

[JorgeA]

Yup! 

 

--JorgeA

[JorgeA]

Left behind by the Windows 10 bandwagon

 

So, on paper at least, the stars all seemed aligned for a happy resolution to the very many debilitating video driver issues that plagued my Envy x2 (and the many, many other systems based on the Z2760 chipset) at the Windows 10 launch. I could most certainly expect a resolution soon after its release or, barring that, at least by the time Threshold 2 dropped. But now that the Windows 10 November Update is out, and with it no sign of a fix for my sad little 2-in-1, reality has begun to sink in.

 

Microsoft and Intel really did decide to screw-over millions of PC buyers with Windows 10.

 

And not just any buyers. We early Windows 8, 2-in-1 adopters were some of the company’s biggest fans. Stalwart supporters. True believers, indeed if not always in word (hey, it’s my job). When I bought that Envy x2 in 2012, I did so firm in my convictions that it was a good long-term investment -- that like so many PCs before it, I would get to leverage numerous Windows generations before finally putting the scrappy little beast out to pasture.

 

So in addition to "legacy" Windows users, Microsoft has managed to alienate users of "modern" Windows versions too.

 

What's not to like about the strategy?

 

--JorgeA

[NoelC]

Repeat after Microsoft:

 

All things are possible by spending even more money on useless technology.

 

-Noel