[FoRK] The Future of Desktops, Re: Dunning-Kruger effect discriminatory environs

Stephen Williams sdw at lig.net
Fri Feb 28 22:31:44 PST 2014


On 2/28/14, 10:01 PM, J. Andrew Rogers wrote:
> On Feb 28, 2014, at 10:53 PM, Stephen Williams <sdw at lig.net> wrote:
>> However, OSX on top of Linux could rapidly take over the world, at least in terms of desktops.  I suspect they just keep this in their back pocket as a nuclear option.
>
> Why would this have any value at all to 99+% of all OSX users?  What does running on Linux provide? FWIW, in my experience Linux runs like crap on a lot of desktop hardware, hardly the kind of experience Apple would endorse. They would be better off adding hardware support to their existing OS.

OSX users have no need for this, except perhaps to have a wider choice 
of hardware to use.
This is for Windows and Linux users to get a reasonable desktop OS that 
has more complete than Linux commercial app support.

Apple would never attempt to add hardware support to their existing OS.  
That is an impossible task.  Even Microsoft didn't do it: manufacturers 
created drivers out of necessity, but old hardware will essentially 
never get supported.  Even Windows has this problem: Last year, we 
bought a top of the line Dell desktop to use as a quiet cube-stationed 
server.  It was intended that a Windows server version would be 
installed on it.  Couldn't because the device drivers only supported 
Windows 8 for desktop use.

Linux essentially supports every bit of PC hardware ever produced, and 
almost everything else too.  Clearly it should be the kernel, device 
drivers, network, etc.  OSX's top half could easily be placed on top of 
this with very little work.  Done properly, the Linux community might be 
helpful with this.

>
> Linux only has advantages over OSX on the server, and even then you would need to be a freakish outlier like me to really notice that suck and articulate it in specific terms. It is sort of like how VMs are godawful for certain types of software but only a few people actually notice.
>
>
>> The next 5 years will be a wild ride.  Even if Windows remains a popular base, it will soon be papered over by Android/ChromeOS/Web layers talking to apps that could run anywhere.
>
> The web remains a poor application environment. Dreadful lowest common denominator stuff, nothing sophisticated. Mobile environments (iOS, Android) are somewhat better but still pretty damn limited and intentionally so.

Most methods and practices are terrible.  The technology however is 
fairly powerful already.

>
> It seems more likely that the front-ends will remain limited and dreadful, because it fits a lot of future form factors better, and become dumber as more of the application heavy lifting is done on the server. The application will increasingly just steer the server, turning every application into an interactive video stream. The client can’t render what it lacks the data to construct due to bandwidth limitations but video frames have a fixed cost.

Parametric models.  Caching.  Advanced interfaces.  There are many 
possibilities beyond the status quo.

>
>
> However, the existing server infrastructure software was never designed to support this kind of use case so we are at that awkward phase where a big shift needs to occur but hasn’t happened yet.

http://aws.amazon.com/gpu/

> _______________________________________________

sdw



More information about the FoRK mailing list