Just read this article and had to respond/..
http://www.computerworld.com.au/article/255954/windows_midori_doomed_failure
On a number of facts its just wrong , particularly.
1. Microsoft wants to evolve.
2. Its unnecessary as windows Kernel is light and replacing the kernel will not solve the security and DLL problems.
3. The purpose behind a new OS
4. Backward competability
Note this is based on Midori being a Singularity style kernel.
1. Microsoft wants to evolve.
Dont know about MS desire to evolve ..
DOS - Windows - NT
NT - Windows 2000 driver model
Vista Video driver model
Windows - Bob
Word 5 - Word for Windows 1 - clean break WYSIWYG
VB - VB.NET
C#
WPF
Office 2007 - tool strips
Com- Com+ -DCOM/Ent Services - Remoting - Windows Communication Foundation
All the above are major clean breaks some opportunistic ( Word for Windows) , some because the platform was getting to hard to evolve ( NT). Its no coincidence MS dropped windows after ME ( for NT ) and Vista is showing similar signs. No company liked to do a clean break but in IT its pain thats for the long term good ( looks at OSX and NT) , Linux will suffer here as well as community projects find it hard to make such a break without fragmenting.
MS need a new Windows Mobile platform and my prediction is Midori will replace WIndows Azure and Windows Mobile 8 ( note CE needs a replacement) . These platforms run mainly or only managed apps so its an easy transition.
2. Its unnecessary as windows Kernel is light and replacing the kernel will not solve the security and DLL problems.
Most importantly Midori and Windows kernel are similar and the kernel is not that important. Whats different is how user application , device drivers and systems ( the fat and insecure bits in Windows) run and talk , so your comments on the windows Kernel ( while correct ) are meaningless. ie replacing WIndows Kernel with Midori is pointless whats important is replacing all ( non Java/.net/managed) apps , device drivers and services. This is why itsa big deal.
3. The purpose behind a new OS
I also don't think you understand the purpose of Midori or what its about .
- Security . This is the #1 driver and note how MS get smacked for security violations when its prob more secure than its competitors . This is not mentioned at all. Those Apple adds hurt .. prob enough for Bill and co to spend $5B to build a new OS. Ok the article mentions that it cant be achieved by the kernel , that is partially correct the Kernel can restrict applications . How many managed virus are there ?
- Reliability. Native apps and device drivers are reliable. ( Note not the kernel)
- Marketing. It is very important for a company to have the technological edge its why they can charge $$$. And Midori presents a marketing coup for MS , they have a new style OS and they can point at there competitors and say how old they are eg Linux 70-80 style kernel and command line interface. Midori only needs to be available and used by 1% of users for then to have this benefit and they can charge a premium. Remember Singularity kind of shocked the sleeping OS research community not for a managed OS ( as old OS sometimes only allowed managed languages , though revisting it was interesting) but by basically saying you can have security ,reliability. AND performance in a managed OS.
- MS needs to see the world in 10 years or 15 years time . They need a single OS that scales from devices to cloud servers and an OS thats .net aware . They also need an OS thats super secure so it can handle this eg are you going to install a firewall on your mobile phone ?
4. Backward competability
Note MS are proposing to support windows drivers in one proposal at first as well as native Midori. This is possible via the Midori HAL . In addition Windows apps could be support with the same Virtualization technology as Vista.
Prediction Midori will replace Windows...eventually but first it will be an OS for cloud and hand helds. Then it will be a new style OS similar to NT3.51 and 4 when Windows 95/98 were the desktop standards.
These are the timeliness we are looking at . Note Ms needs to look at where WIndows will be in 10 years.
WIndows Mobile 8 ~2011
Azure ~2011-2012
New Server ( Applications , files etc ) ~ 2013
Windows Mobile 8 light Desktop version 2013-2014?
Power workstation ?? 2014
Desktop replacement ?? 2015-2017 ?
The product however has to survive on its merits and WIndows Mobile 8 will be a key release.
Print | posted on Monday, May 18, 2009 11:59 AM