[OT] Windows dying

[Joakim]

On Thursday, 2 November 2017 at 05:13:42 UTC, H. S. Teoh wrote:
> One thing is clear, though: claiming that Windows is "dead" is, 
> frankly, ridiculous.  Even a non-Windows person like me who 
> rarely has any reason to notice things Windows-related, can see 
> enough circumstantial evidence around me that Windows is still 
> very much alive and kicking.  (Even if in my ideal world there 
> would be no Windows... but then, if the world were my ideal, 
> 90% of computer users out there would probably be very angry 
> about being forced to use obscure text-only interfaces that I'm 
> completely comfortable in.  So it's probably not a bad thing 
> the real world doesn't match my ideal one. :-D)

Congratulations, you find a claim that literally nobody has made 
in this thread to be ridiculous.  Next you'll say that Walter's 
claim that Java will replace COBOL is ridiculous or Adam's claim 
that we should write a full crypto stack ourselves in D is a bad 
idea, both of which neither ever said.

On Friday, 3 November 2017 at 06:20:25 UTC, Tony wrote:
> On Wednesday, 1 November 2017 at 08:49:05 UTC, Joakim wrote:
>> On Wednesday, 1 November 2017 at 00:16:19 UTC, Mengu wrote:
>>> On Monday, 30 October 2017 at 13:32:23 UTC, Joakim wrote:
>>>>
>>>> I don't know how intense your data analysis is, but I 
>>>> replaced a Win7 ultrabook that had a dual-core i5 and 4 GBs 
>>>> of RAM with an Android tablet that has a quad-core ARMv7 and 
>>>> 3 GBs of RAM as my daily driver a couple years ago, without 
>>>> skipping a beat.
>>>>  I built large mixed C++/D codebases on my ultrabook, now I 
>>>> do that on my Android/ARM tablet, which has a slightly 
>>>> weaker chip than my smartphone.
>
> How does the performance compare between an i5 laptop and an 
> Android tablet?

My core i5 ultrabook died in late 2015, so I never ran any 
performance comparisons.  I'd say that its 2012 Sandy Bridge 
dual-core i5 was likely a little faster to compile the same code 
than the 2014 quad-core Cortex-A15 I'm using in my tablet now.  
I've recently been trying out AArch64 support for D on a 2017 
Android tablet which has one of the fastest quad-core ARMv8 chips 
from 2016, I'd guess that's faster than the i5.  But this is all 
perception, I don't have measurements.

>>> Why do predictions about the future matter when at the 
>>> present Windows dominates the desktop and is also strong in 
>>> the server space?
>>
>> Because that desktop market matters much less than it did 
>> before, see the current mobile dominance, yet the D core team 
>> still focuses only on that dying x86 market.  As for the 
>> future, why spend time getting D great Windows IDE support if 
>> you don't think Windows has much of a future?
>>
>
> The concept that you are proposing, that people will get rid of 
> ALL their desktops and laptops for phones or tablets, doesn't 
> seem to be happening right now.

To begin with, I never said they'd "ALL" be replaced in the 
paragraph you're quoting above, but yes, that's essentially what 
will eventually happen.  And of course it's happening right now, 
why do you think PC sales are down 25% over the last six years, 
after rising for decades?  For many people, a PC was overkill but 
they didn't have a choice of another easier form factor and OS.  
Now they do.

> At this point, were they do to that, they would end up with a 
> machine that has less power in most cases (there are Atom and 
> Celeron laptops), and probably less memory and disk storage. 
> That solution would be most attractive to Chromebook type users 
> and very low end laptop users. And while people buy low spec 
> laptops and desktops, there are still many laptops and desktops 
> sold with chips that aren't named Atom and Celeron or arm. If 
> phones and tablets try to get chips as powerful as those for 
> the desktop and laptops they run into the chip maker's problem 
> - the more processing power, the more the electricity the chip 
> uses. Phones and tablets don't plug into the wall and they are 
> smaller than the batteries in laptops. And in order to use a 
> phone/tablet as a "lean forward" device (as opposed to "lean 
> back") and do work, they will have to spend money on a "laptop 
> shell" that will have a screen and keyboard and probably an 
> SSD/HD which will cancel most of the cost savings from not 
> buying a laptop.

You seem wholly ignorant of this market and the various points 
I've made in this thread.  Do you know what the median Windows PC 
sold costs?  Around $400.  Now shop around, are you finding great 
high-spec devices at that price?  The high-spec market that you 
focus on is a tiny niche, the bulk of the PC market is easily 
eclipsed by mobile performance, which is why people are already 
turning in their PCs for mobile.

Battery life on mobile is already much better than laptops, for a 
variety of reasons including the greater efficiency of mobile ARM 
chips.  And the Sentio laptop shell I already linked in this 
thread has a screen, keyboard, and battery but no SSD/HD, which 
is why it only costs $150, much less than a laptop.

> In the case of trying to court Android development, I read that 
> 95% of Android is done on Java (and maybe other JVM languages 
> like the now "officially supported" Kotlin) and 5% in C or C++. 
> But that 5% is for applications that have a need for high 
> performance, which is mostly games. Good luck selling game 
> developers on using D to develop for Android, when you can't 
> supply those same game developers a top-notch development 
> environment for the premier platform for performance critical 
> games - Windows 64-bit.

I don't think the numbers favor Java quite so much, especially if 
you look at the top mobile apps, which are mostly games.  I don't 
know what connection you think there is between the AAA Windows 
gaming market and mobile games, nobody runs Halo on their mobile 
device.

btw, the mobile gaming market is now larger than the PC gaming 
market, so to think that they're sitting around using tools and 
IDEs optimized for that outdated PC platform is silly:

https://www.digitaltrends.com/gaming/pc-market-grew-in-2016-led-by-mobile-and-pc-gaming/

>>> I have seen conflicting reports about what OS is bigger in 
>>> the server market, but Windows is substantial and the more 
>>> frequent winner.
>>>
>>> https://community.spiceworks.com/networking/articles/2462-server-virtualization-and-os-trends
>>>
>>> https://www.1and1.com/digitalguide/server/know-how/linux-vs-windows-the-big-server-check/
>>
>> I have never seen any report that Windows is "bigger in the 
>> server market."
>
> I linked one that said:
>
> "And what OSes are running in virtual machines and on physical 
> servers around the world? It turns out like with client OSes, 
> Microsoft is dominant. Fully 87.7% of the physical servers and 
> VMs in the Spiceworks network (which are mostly on-premises) 
> run Microsoft Windows Server."
>
>> Last month's Netcraft survey notes,
>>
>> "which underlying operating systems are used by the world's 
>> web facing computers?
>>
>> By far the most commonly used operating system is Linux, which 
>> runs on more than two-thirds of all web-facing computers. This 
>> month alone, the number of Linux computers increased by more 
>> than 91,000; and again, this strong growth can largely be 
>> attributed to cloud hosting providers, where Linux-based 
>> instances are typically the cheapest and most commonly 
>> available."
>> https://news.netcraft.com/archives/2017/09/11/september-2017-web-server-survey.html
>
> Web-facing server is a subset of servers. Shared web hosting 
> services are probably a harder target for native-code 
> applications than internal IT servers.

Web servers are a subset but by far the largest one, so any 
accounting of market share is going to be determined by them.  
Native code has been dying on the server regardless of web or 
internal servers, but the real distinction is performance.  
Facebook writes their backend in C++, the same for any server 
service that really needs to scale out, which is not likely to be 
internal IT.

> But regardless of whether Windows is dominant, or just widely 
> used, you haven't made predictions that Windows servers are 
> going to die.

I don't think about niche platforms that hardly anybody uses.

>> Your first link is actually a bad sign for Windows, as it's 
>> likely just because companies are trying to save money by 
>> having their employees run Windows apps off a virtualized 
>> Windows Server, rather than buying a ton more Windows PCs.
>
> I would say that is an unlikely scenario. Companies use virtual 
> machines for servers because it allows for the email server 
> and/or http server and/or database server and/or application 
> server to be on one physical machine, and allow for the system 
> administrator to reboot the OS or take the server offline when 
> making an upgrade/bug fix, and not affect the applications 
> running on the other servers.

I see, so your claim is that process or software isolation is so 
weak on Windows Server that they run multiple virtualized 
instances of Windows Server just to provide it.  Or maybe that 
Windows Server needs to be patched for security so often, that 
this helps a little with downtime.  I doubt they are running many 
WinServer instances like you say, given how resource-heavy each 
Windows Server instance is going to be.  But regardless of how 
you slice it, this isn't a good sign for Windows.

>> Meanwhile, your second link sees "Linux maintaining a 
>> noticeable lead" in the web-hosting market.
>
> Don't know why I linked that as it doesn't even have a 
> percentage breakdown. My intent was to show a web server 
> breakdown but I will concede that Linux is bigger for web 
> servers. However, Windows is still big and you aren't 
> predicting it will die.

I've actually said elsewhere in this forum that the cloud server 
market is way overblown and will greatly diminish in the coming 
years because of greater p2p usage, so yeah, I think both linux 
and Windows on the server will largely die off.

>>> And if desktop OSes were going to go away, the MacOS would go 
>>> before Windows.
>>
>> Oh, Apple wants that to happen, one less legacy OS to support, 
>> which is why all the Mac-heads are crying, because macOS 
>> doesn't get much attention nowadays.  Do you know the last 
>> time Apple released a standalone desktop computer?  2014, when 
>> they last updated the Mac Mini.  They haven't updated the Mac 
>> Pro since 2013.
>
> Why do you think it is that they haven't come out with an iOS 
> Mac Mini or iOS MacBook?

The Mac Mini is easy, they're just winding down that legacy form 
factor, like they did with the iPod for years.  Their only entry 
in that market is Apple TV running tvOS, which is more iOS than 
macOS.

As for the iOS Macbook, it's out, it's called the iPad Pro.  
Their CEO, Tim Cook, is always boasting about how it's all he 
uses these days:

https://9to5mac.com/2012/02/14/tim-cook-ipad-80-90-of-tim-cooks-work-is-on-ipad-work-and-consumption/
http://appleinsider.com/articles/15/11/09/apple-ceo-tim-cook-says-he-travels-with-just-an-ipad-pro-and-iphone

>> They see the writing on the wall, which is why they're 
>> lengthening their release cycles for such legacy products.
>>
>
> Do they want them to go away, or do they see the handwriting on 
> the wall? The fact that they still make them, it appears that 
> they don't want them to go away. They can stop making them at 
> any time. And by them, I mean their entire macOS (i.e. their 
> non-mobile) line. I think that the Mac Mini/Mac Pro pale in 
> sales to the iMacs as far as Apple desktop sales go.

Simple, they see the writing on the wall, ie much smaller sales 
than mobile, so they want the legacy product to go away, which 
means they can focus on the much bigger mobile market.  The only 
reason they still make them is to milk that market and support 
their legacy userbase, the same reason they were still selling 
the iPod Touch all these years after the iPhone came out.

They don't break out iMac sales but given that it's much more 
expensive than the Mac Mini, it's doubtful that it sells better.  
I was only talking about Apple's standalone desktops because 
they're most comparable to the PC market, but it's true that 
PC/Mac all-in-ones like the iMac have done better lately, one of 
the few growing segments.  But when the entire desktop/laptop 
market is shrinking and the much more expensive all-in-one sales 
are so small, that doesn't mean much.

> If you look at the graph in this article, the iPad has declined 
> more as a percentage of Apple revenue than the macOS line has 
> in the last five years.
>
> https://www.statista.com/statistics/382260/segments-share-revenue-of-apple/

I don't have access to that chart, but yes, the iPad and tablet 
markets have been shrinking.  It's possible that more people 
would rather use their smartphone, which usually has a more 
powerful chip than Android tablets, with the Dex dock or a 
Sentio-like laptop shell than a tablet.  But neither group is 
using a PC: both are mobile, smartphone even more so.

> There is a case to be made for supporting  Android/iOS 
> cross-compilation. But it doesn't have to come at the expense 
> of Windows 64-bit integration. Not sure they even involve the 
> same skillsets. Embarcadero and Remobjects both now support 
> Android/iOS development from their Windows (and macOS in the 
> case of Remobjects) IDEs.

You're right that some of the skills are different and D devs 
could develop for mobile from a Windows IDE.  But my point was 
more about general investment and focus, the currently dominant 
platform, Android, needs it, while the fading platform, Windows, 
shouldn't get much more.

Frankly, I find it tiresome that some Windows devs in this thread 
think the reason IDE support isn't better is because somebody is 
listening to me.  More likely, Rainer or whoever would do that 
work is already invested in Windows, but doesn't have the time or 
interest to do much more.

You'd be much better off finding that person and helping or 
sponsoring them rather than debating me, as I likely have no 
effect on that person's thinking.  I wish it were otherwise, but 
I doubt it.