[OT] Windows dying

[jmh530]

On Monday, 30 October 2017 at 13:32:23 UTC, Joakim wrote:
>
> The decline itself doesn't imply a collapse, the collapse is 
> coming because the mobile market is looking for new growth 
> avenues and releasing mobile accessories like Samsung's DeX 
> dock or laptop replacements like the iPad Pro or this laptop 
> shell:
>
> https://sentio.com

I look at this and just wonder why people wouldn't just have a 
laptop.

>
> Mobile convergence killed off standalone mp3 players, 
> e-readers, GPS devices, point-and-shoot cameras, feature 
> phones, a whole host of former mobile single-purpose devices.  
> They're going after the PC now, with all the massive scale of 
> the mobile wave:
>
> https://twitter.com/lukew/status/842397687420923904
>
> Can the PC market withstand that tidal wave?  I'm betting not.
>

And what does this show, a huge increase in smart phone/tablet 
shipments and a modest decline in desktop sales. I don't dispute 
this. Smart phones are leading to a huge increase in the amount 
of people who use computers on a daily basis. A whole bunch of 
people who use PCs may switch to just using smartphones/tablets. 
However, some people do need and want them. And they will 
continue to use them.

> As for the average white collar worker in a decade, if they're 
> using Google Docs on their Samsung S18 connected to something 
> like that Sentio laptop shell, do you really imagine they won't 
> be able to get their work done?  I think it's more likely 
> they're using software completely different than Office or Docs 
> to get their work done, as those suites are already way 
> outdated by now, but that's a different tangent.

Okay, but Google Docs isn't supported at my company. Microsoft 
Office is. We have a huge number of Excel files that use a lot of 
features that probably can't be ported over to Google Docs 
without a bunch of work. They might be able to get us to make new 
stuff with Google Docs, but we're still gonna need Excel for all 
the old stuff (so why bother switching). There's a reason why 
banks still use Cobol.

Now I would love to move everything to R/Python. It could be 
done. But not everyone knows R/Python, but everyone knows Excel. 
If I get hit by a bus, then someone can figure out what I've done 
and get to work.

>
> 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.

I would have gobbled up 4GB using Matlab 5 years ago...Nowadays, 
I sometimes use a program called Stan. It does Hamiltonian Monte 
Carlo. Often it takes 10 minutes to run models on my home machine 
that's got a relatively new i7 processor on it. It's not unknown 
for the models to take hours with bigger data sets or more 
complicated models. I don't even like running the models at work 
because my work computer sucks compared to my home computer.

>
> The latest ARM-based iPad Pro is notorious for beating low to 
> mid-range Intel Macbooks on benchmarks.  It is not difficult to 
> pick up smartphones with 6-8 GBs of RAM nowadays.  Unless you 
> need monster machines for your data and aren't allowed to 
> crunch your data on online servers for security reasons, a very 
> niche case, you can very likely do it on a smartphone.

Doing everything on an AWS instance would be nice.

>
> You may be right that your particular workplace moves slowly 
> and they're not going mobile anytime soon.  But this is such a 
> big shift that you have to wonder if many such slow-moving 
> workplaces will be able to compete with places that don't: just 
> ask all the taxi companies phoning in rides to their drivers 
> who got put out of business by Lyft, Uber, and their 
> smartphone-wielding hordes of drivers.

We certainly have big competitive issues, but they aren't because 
our competitors are using Google Docs.

>
> There will always be a few Windows cockroaches that survive the 
> mobile nuclear blast, but we're talking about the majority who 
> won't.
>
> As for you particularly, I can't speak to your situation 
> without knowing more, but nobody's saying D should drop Windows 
> support.  I started off this OT thread by saying that investing 
> more time in getting D somewhere close to the level of C#/C++ 
> support in Visual Studio or some other IDE is a waste of time.  
> I stand by that.  If Rainer or someone else does it anyway, 
> that's up to them how they want to spend their time.

Look at the growth of Python. Among the many drivers of that, are 
people who use Numpy and its ecosystem (SciPy, Pandas, etc.). The 
work that Ilya et al are doing on Mir is a fantastic effort to 
provide similar functionality for D. More users using Mir will 
help build out the ecosystem and hopefully get it to a 
competitive place with Numpy one day. This requires more people 
using D. If efforts by Rainer or someone else to make the Windows 
experience better and leads to more D users and more Mir users, 
then I consider a positive. I don't consider it a waste of time.