[mdlug] Interesting information about Windows "7"

[Aaron Kulkis]

Robert Adkins wrote:
>> -----Original Message-----
>> From: mdlug-bounces at mdlug.org 
>> [mailto:mdlug-bounces at mdlug.org] On Behalf Of Ingles, Raymond
>> Sent: Tuesday, January 22, 2008 1:31 PM
>> To: MDLUG's Main discussion list
>> Subject: Re: [mdlug] Interesting information about Windows "7"
>>
>>> From: Robert Adkins
>>> 	You haven't been reading much about Windows "7". They intend on 
>>> making it considerably more modular. As I understand it, the OS is 
>>> being designed to be quite tweakable and modularized 
>> specifically to 
>>> make it faster to put together as well as easier to tweak 
>> for specific purposes.
>>
>>  That'll be a major cultural shift, then. I'm not sure they 
>> can pull *that* off, either, in that kind of timeframe. If 
>> they do manage to do so, well, it'll be welcome competition - 
>> but I'm still minded of Henry Spencer's famous line, "hose 
>> who do not understand Unix are condemned to reinvent it, 
>> poorly." Sounds like they are taking pages from "The Art Of 
>> Unix Programming". It took Unix quite a while to get 
>> everything right, even by copying the playbook I'm not 
>> convinced MS can turn around that fast.
>>
> 
> 	From some of the stuff I have been reading, when they initially
> started talking about the existence of a Windows "7" design team and product
> plan, the whole modular thing has been an express element of the design from
> day one of the next Windows OS. I believe at that time, they had already had
> a good year, two or more in the bag on this next OS.
> 
> 	The article I linked said they already released a preliminary
> version (Milestone 1 or M1) and plan on releasing 3 more of these before
> hitting the presses for the "Gold" release. The articles suggests that this
> is well within their timeline. 

But exactly how much is IN milestone 1?

Probably not much more than a demonstration of an
ability to load and execute programs, against a
very very small library of functions, including
some basic GUI I/O, and maybe networking.

On the other hand, it only took me 4 weeks to write
a multi-tasking kernel on a lowly 6809.

Re-creating the code base for XP and standard
apps (like Office) is going to be quite the task --
the current office suite assumes a non-modular
kernel design with all of the crazy side-effects


> 
> 	If anything in this article comes really close to correct, I have
> some concern for the potential growth in desktop marketshare for Linux this
> next year. If any of it is true, this is a crucial year for Linux to make
> significant inroads, since the window of opportunity could shrink
> considerably.

Well, given that the embodiment of everything that's
wrong with relying on MS has all come to the forefront
simultaneously with Vista, the next 24 months are
probably the best chance that we'll ever have to get
Linux on the corporate desktop.