Scali wrote:

I think that about covers it, don't you?

Heavy on corporate robotalk (special testing as opposed to normal testing?), but enough for the plebes, I guess. It's also good that they're not arbitrarily limiting OS support.

Scali wrote:

Modern hardware needs updates to drivers for legacy functionality on legacy OSes (to emulate legacy behaviour on new hardware). This may lead to issues. Hence, MS tests drivers with legacy OSes and publishes a list of hardware that will work properly with legacy OSes/drivers, and on which Microsoft will give continued support for the lifetime of these OSes.

Windows functionality has been largely abstracted for a while. I read it more as the opposite, where subsystems and functionality in the new OS is required by the new drivers or new firmware. Or more commonly, only the new OS will have tweaks for deficiencies and bugs in new releases of UEFI.

gdjacobs wrote:

Windows functionality has been largely abstracted for a while.

And how do you think this abstraction is done? Exactly, via drivers.
So I think you didn't get the point of the text.
Which is: new hardware needs new drivers for old OSes to get the proper level of abstraction.
If hardware doesn't have these drivers, or these drivers are of poor quality, these older versions of Windows will have issues. This is what Microsoft is trying to avoid, by specifically putting their stamp of approval on hardware that they have tested.

I'm not sure what some people think, but I recently pointed out in another thread that the latest nVidia and AMD GPUs do not have drivers for XP or Vista anymore. So although these legacy OSes would work on that hardware in theory, you're not going to get very far without proper drivers. You could only run in the default SVGA-compatible fallback mode, which means you won't get support for DirectX and OpenGL, and most applications will not work properly.

Scali wrote:

And how do you think this abstraction is done? Exactly, via drivers. So I think you didn't get the point of the text. Which is: […]
Show full quote

gdjacobs wrote:

Windows functionality has been largely abstracted for a while.

And how do you think this abstraction is done? Exactly, via drivers.
So I think you didn't get the point of the text.
Which is: new hardware needs new drivers for old OSes to get the proper level of abstraction.
If hardware doesn't have these drivers, or these drivers are of poor quality, these older versions of Windows will have issues. This is what Microsoft is trying to avoid, by specifically putting their stamp of approval on hardware that they have tested.

I'm not sure what some people think, but I recently pointed out in another thread that the latest nVidia and AMD GPUs do not have drivers for XP or Vista anymore. So although these legacy OSes would work on that hardware in theory, you're not going to get very far without proper drivers. You could only run in the default SVGA-compatible fallback mode, which means you won't get support for DirectX and OpenGL, and most applications will not work properly.

If that were true and their motivation then MS would not be charging companies huge fees to get their drivers MS certified - yet they do.
.

PCBONEZ wrote:

If that were true and their motivation then MS would not be charging companies huge fees to get their drivers MS certified - yet they do.

I don't see how one precludes the other.

gdjacobs wrote:

Heavy on corporate robotalk (special testing as opposed to normal testing?), but enough for the plebes, I guess. It's also good that they're not arbitrarily limiting OS support.

Plebes? As apposed to what? Pedantic, perpetually enraged "techies"?

I'm with the plebes - this is a non-event. Untwist your knickers.

Scali wrote:

PCBONEZ wrote:

If that were true and their motivation then MS would not be charging companies huge fees to get their drivers MS certified - yet they do.

I don't see how one precludes the other.

You make it sound like MS is benevolent and doing it's users and companies a favor by certifying drivers for free. Hardly the case.
One of the reasons there aren't drivers for newer hardware on older OS's is that MS charges extortionate fees to certify them and the hardware manufacturers don't want to pay for it.

PCBONEZ wrote:

You make it sound like MS is benevolent

I don't see how.

PCBONEZ wrote:

and doing it's users and companies a favor by certifying drivers for free.

I never used the word 'free' or any phrasing of that nature.

The point here is that Windows 7 is already out of regular support. Normally this means MS simply will not do any WHQL or other certification whatsoever. In this case they are making an exception.

PCBONEZ wrote:

One of the reasons there aren't drivers for newer hardware on older OS's is that MS charges extortionate fees and the hardware manufacturers don't want to pay them.

Hardly.
Firstly, your information is outdated. Microsoft indeed made WHQL testing free as of January 2014: http://download.microsoft.com/download/4/D/D/ … ses-hck2-1.docx
Secondly, even when Microsoft did charge fees, they were very modest, namely $250 per tested device per OS.
Certainly no issue for companies of the size we're dealing with here (Intel, nVidia, AMD, Qualcomm, Dell, Acer etc).

Next time, please check your facts before you go on a rant. It saves you from making yourself look like a tinfoil-hat fanatic.

I like my hat. Only it's Aluminum, not Tin. Aluminum has a much higher melting point.
I had it anodized this pretty deep blue and I have sprinkles embedded in the clear-coat.
.

"... If 10 was any good, it would be selling itself. It's free for home users of 7 and 8, after all! If we won't even take it fo […]
Show full quote

"... If 10 was any good, it would be selling itself. It's free for home users of 7 and 8, after all! If we won't even take it for FREE, what does that say about how good it is?

What a pickle (of their own creation) MS is in-- they can't give us an OS we'd actually WANT to use, because that won't sell any Windows mobile devices, and they can't get us to adopt 10, because it sucks. They're left trying to FUD their own prior products to death. If they can't sell us on it, they'll browbeat us, scare us, threaten us into taking their horrible product. After all, it's a make or break product; MS cannot afford to fail on this one.

It will be interesting to see how it plays out once the free upgrade period ends. If Win 7 users would not take it for free, what are the odds we'll take it for a hundred bucks?"

Ms doesn't support new hardware on old operating systems, wasn't it always like that? I don't see anything strange. Well maybe the problem is that they are discarding support for seven too fast (I'm not talking about 8 on purpose: there's really no reason on earth to stay on 8 when 8.1 is free and is better than 8 in every possible way).

TELVM wrote:

"... If 10 was any good, it would be selling itself. It's free for home users of 7 and 8, after all! If we won't even take it fo […]
Show full quote

"... If 10 was any good, it would be selling itself. It's free for home users of 7 and 8, after all! If we won't even take it for FREE, what does that say about how good it is?

What a pickle (of their own creation) MS is in-- they can't give us an OS we'd actually WANT to use, because that won't sell any Windows mobile devices, and they can't get us to adopt 10, because it sucks. They're left trying to FUD their own prior products to death. If they can't sell us on it, they'll browbeat us, scare us, threaten us into taking their horrible product. After all, it's a make or break product; MS cannot afford to fail on this one.

It will be interesting to see how it plays out once the free upgrade period ends. If Win 7 users would not take it for free, what are the odds we'll take it for a hundred bucks?"

That's some very nice emotional rhetoric, but for the umpteenth time, Mirosoft didn't do anything to any of their products. They didn't change anything, they didn't remove anything. They didn't "browbeat", "scare" or "threaten" anybody. It's all appeals to emotion, putting on a victim complex in lieu of any real argument. THEY are trying to GET US. WE won't let THEM GET US. GUBMINT. TRIANGLES ON THE MONEY. JET FUEL CAN'T MELT STEEL BEAMS.

Scali wrote:

And how do you think this abstraction is done? Exactly, via drivers. So I think you didn't get the point of the text. Which is: […]
Show full quote

And how do you think this abstraction is done? Exactly, via drivers.
So I think you didn't get the point of the text.
Which is: new hardware needs new drivers for old OSes to get the proper level of abstraction.
If hardware doesn't have these drivers, or these drivers are of poor quality, these older versions of Windows will have issues. This is what Microsoft is trying to avoid, by specifically putting their stamp of approval on hardware that they have tested.

The subtext in the article makes me think that UEFI compatibility might be more of an issue, especially considering that UEFI development is much more fragmented.

Myloch wrote:

Ms doesn't support new hardware on old operating systems, wasn't it always like that? I don't see anything strange. Well maybe the problem is that they are discarding support for seven too fast (I'm not talking about 8 on purpose: there's really no reason on earth to stay on 8 when 8.1 is free and is better than 8 in every possible way).

It's never really been the job of MS to support new hardware. They always punt that to the vendors. Then they run it through WHQL to provide some consistent quality control.

badmojo wrote:

Plebes? As apposed to what? Pedantic, perpetually enraged "techies"?

I'm with the plebes - this is a non-event. Untwist your knickers.

Believe it or not, I am for the "plebes" too (although that was a poor choice of words).

gdjacobs wrote:

The subtext in the article makes me think that UEFI compatibility might be more of an issue, especially considering that UEFI development is much more fragmented.

Windows 7 does not support UEFI at all.
Also, they are reasonably specific:
"For Windows 7 to run on any modern silicon, device drivers and firmware need to emulate Windows 7’s expectations for interrupt processing, bus support, and power states- which is challenging for WiFi, graphics, security, and more."

Most of these things have little to do with UEFI at all, and everything with driver support.

Scali wrote:

gdjacobs wrote:

The subtext in the article makes me think that UEFI compatibility might be more of an issue, especially considering that UEFI development is much more fragmented.

Windows 7 does not support UEFI at all.

https://technet.microsoft.com/en-us/library/h … v=ws.10%29.aspx
http://blogs.technet.com/b/askcore/archive/20 … d-computer.aspx
https://technet.microsoft.com/en-us/library/d … 28WS.10%29.aspx
http://www.sevenforums.com/tutorials/186875-u … indows-7-a.html
https://www.youtube.com/watch?v=WbyGTwa3UXY
https://www.youtube.com/watch?v=PSbQr4f2YA0

Scali wrote:

Also, they are reasonably specific:
"For Windows 7 to run on any modern silicon, device drivers and firmware need to emulate Windows 7’s expectations for interrupt processing, bus support, and power states- which is challenging for WiFi, graphics, security, and more."

I know Intel has been messing with power states to squeeze out additional energy savings, but I doubt they're changing too much with the way the APIC and PCIe bus operate. If they are, this is a much bigger issue than just Windows 7 support. We go back to my original question regarding the advantages presented by such platform changes and whether it's worth the loss in compatibility.

As for UEFI, the spec is so huge it leaves way too much room for board makers to screw up.

gdjacobs wrote:

We go back to my original question regarding the advantages presented by such platform changes and whether it's worth the loss in compatibility.

I think the question is irrelevant, since the hardware vendors have already decided that apparently it is worth the loss in compatibility. As they have done time and time again over the years (where is our ROM BASIC, where are our ISA slots, where are our RS232/LPT ports, where is our 9-pin keyboard connector, where are our PS/2 connectors... etc?)

Yet so we still have the LPC bus (mostly software-compatible with the old ISA bus), many motherboards have RS232/LPT port headers which are 100% software-compatible with the ones from the very first IBM PC, and you can easily find motherboards with PS/2 connectors (and, for keybaords, they still have their advantages) and laptops still prefer to use the PS/2 protocol for keyboards. And even when the motherboard does not offer PS/2 conectors, it will still emulate the hardware until the OS takes over and disables the legacy support. And even the latest motherboards will still emulate IDE controllers if set to do so.

In a way, the PC might very well be the ecosystem with the strongest resistance towards breaking backwards compatibility: you can still mostly run DOS and Windows 9x on the latest platforms. You will lack drivers and will not be able to use most of your hardware's capabilities, but they will run (default VGA, standard IDE controller, etc.). So I REALLY doubt we will reach a point that we won't be able to install and run Windows 7 on new retail hardware anytime soon, MS supporting it or not, but, as with WIndows XP before it, the process will get progressively more involved and eventually be unfeasible when hardware vendors start dropping Windows 7 support, and given its marketshare it might take a while.

By the way, it is kind of senseless to acuse Microsoft of trying to "scare off" people into upgrading. They are not the ones publishing click-baity articles about them dropping support as if it were something its not. They did not try to induce panic or anything like that. Do not blame MS for Forbes' (and other sites) screw ups. Their biggest fault here is not addressing the issue better by publishing more statements countering these crap articles.

Scali wrote:

I think the question is irrelevant, since the hardware vendors have already decided that apparently it is worth the loss in compatibility. As they have done time and time again over the years (where is our ROM BASIC, where are our ISA slots, where are our RS232/LPT ports, where is our 9-pin keyboard connector, where are our PS/2 connectors... etc?)

I think asking about the value of product changes is an important question, maybe the most important question in a market economy. In all your examples, the legacy and replacement technologies existed for quite some time. This allowed consumers to make direct comparisons and transition at a pace of their choosing.

The Ars article was quite good in attempting to flesh out this whole issue, but even they were left with many questions regarding the back end motivation and felt many of the stated reasons didn't make a lot of sense.
http://arstechnica.com/information-technology … -to-windows-10/

gdjacobs wrote:

I think asking about the value of product changes is an important question, maybe the most important question in a market economy.

My point was that the IHVs understand this as well, so given that they made the change, they figured it has added value.

Scali wrote:

My point was that the IHVs understand this as well, so given that they made the change, they figured it has added value.

No question, it has added value for them as they can potentially trim their driver development and support costs. Whether it has added value for customers is undetermined.