> I've read scuttlebutt about M$ moving Windows to a SaaS model, but I fail to
> understand how this could possibly work.

Locally you'd have an OS that was stripped down to just run Remote 
Desktop (or equivalent) and interface with your hardware. When you 
logged on it would start a remote desktop session with an MS VM 
somewhere. *Assuming internet speeds were fast enough* you wouldn't 
notice the difference to running full windows locally.

The benefits are obvious (a machine that has all your files and looks 
the same no matter where you log on, an almost limitless supply of CPU 
power and RAM if you wanted to do CPU intenstive tasks, automatic 
backups for everyone, etc), I fail to understand how you fail to 
understand how this could possibly work :-)

> Leaving aside what I personally think of the whole Software as a Service model,
> I can understand, after a fashion, why it might be appealing for a development
> team to be able to issue mass updates to subscribers, rather than waiting for
> people to upgrade and continuing to offer backwards support and compatibility
> for legacy versions.

Their big problem will be the medium-large corporations that take 
months, if not years to test and roll out major software updates. There 
is no way they would accept the possibility of one day their entire 
company coming to a halt with millions of pounds lost due to an MS 
"update" that has broken something somewhere within their business. Also 
a lot of systems are not connected to the internet for various reasons, 
how would they work?

> That said: I personally believe that it would stifle innovation, since an
> independent developer couldn't test a program on their own machine, but would be
> forced to run it in a sandbox VM.

If the whole of windows and all apps are running on a remote VM, then I 
don't see why you wouldn't just fire up Visual Studio (or whatever) on 
the same remote machine. You'd be using exactly the same environment as 
everyone else to develop and test your app.

> This is pretty much the dilemma I'm having with iOS apps at the moment.  I
> despise what Apple has done to iTunes, and there are features that I would love
> to implement into a music player, but unless I jailbreak my iPod, I can't write
> a program solely for my own, personal edification, and use it.

I see no indication that MS would follow such a route, if they wanted to 
they could have done so already. That's just Apple being Apple, and 
their philosophy that if they control everything the user will have a 
better experience.

> In reality, I understand that there are probably music players already on the
> app store that do most, if not all, of the things I want to do; however, sifting
> through them is a pain, and I could potentially learn something by doing it
> myself.

Sounds to me like you should have bought an Android phone. One click in 
Eclipse (free) on my desktop PC and my app is compiled, sent to my phone 
and running within a few seconds.

> As I aspire to, one day, program for a living -- A topic not for discussion on a
> public forum -- I am interested in the various development and marketing models
> used in the industry, and often find myself at odds with various philosophies,
> but I try to remain open minded.

If you're interested you might like to try and do some financial 
modelling of the SaaS model for Windows. You can probably find out how 
many copies of Windows MS sells, how much income they've got from that 
over the years etc. Try and predict that if they continued with the same 
model how well would the company be doing in 5 years, 10 years, 20 years 
time.

Then assume they switched to a SaaS model at some point in the future. 
You'd need to make many assumptions about many different items. But you 
should be able to predict the profit they would make after such a 
switch. How does it compare to your previous predictions? How sensitive 
is it to changes in your assumptions? What is the optimum subscription 
price? How to handle business users? When it the optimum time to stop 
support for old versions of Windows?

No doubt MS has teams of people doing exactly the above (with access to 
far more information than publically available). *If* the calculations 
show more profit with low enough risk can be achieved with the SaaS 
model then the executives will make the decision.

Post a reply to this message