Submitted by Microsoft appears to be assembling its game plan for the day when the Windows client OS as it has been developed for the past 20 years becomes obsolete. The incubation project, also known as Midori, seeks to create a componentized, Net-centric OS, based on connected systems – one that largely eliminates dependencies between local apps and the hardware on which they run. SDTimes is also featuring an article that has some more details about Midori.
see this is really where MS should be fucosing right now.gets windows 7 out the door (aka truely stable and maybe a bit slimmed down Vista) and then devote everythign you have to this.
it would be nice and if it is truely componetized then you could even have a Win32 api layer or virtual machine running if need be. or since MS is seeming to be nice ot open sourse these day, once the win32 api is a thing of the past, help WINE out and that way people can run “legacy” apps on a new os.
you can only suport things for so long before the code gets to the point of being unmantainable, as a developer i know.
a fresh start and a fresh aproach will be the best thing to happen to windows since the release of win2k.
Idealy I would like to see: silverlight (2.0+) GUI for the whole OS, and .NET as the api of choice for developing (of course i would want suport for java, python, etc…)
good luck microsoft, dont mess this one up.
Edited 2008-07-30 21:45 UTC
… to see a bunch of computers all showing up a BSOD at the same time.
There is *already* a Midori. It’s a web-browser…
There is also a Linux variant:
http://midori.sourceforge.net/
“Midori Linux is an Open Source project for delivering system software on small devices. It includes a build system, a Linux kernel with memory- and storage-conserving features, and system-level support for normal Linux software on platforms which might otherwise require custom “embedded” applications”
And it means ‘Green’ in Japanese:
http://en.wikipedia.org/wiki/Midori
and if that doesnt add up, its also the handle used by an author of a few novels:
http://www.planetmidori.com/
“An educator and columnist on adventurous sexuality, she’s also the author of “The Seductive Art of Japanese Bondage”, “Master Han’s Daughter” and “Wild Side Sex: The Bookof Kink”. Midori travels the world presenting to universities, education events, organizations and media.”
Ohhhh yeahhhhH!
Edited 2008-07-31 12:27 UTC
Do you realize that Midori is the project codename, don’t you? It’s likely to be released as Microsoft Live System for Desktop Platforms 2009 EX Plus Champion Edition or something
waporware?
I’d rather say it’s vistaware (software highly marketted, extremely delayed, rushed as much as possible, not polished enough for human consumption). Well, not highly marketted… yet. But getting this news about Microsoft crap so ahead of time, I just don’t see the point. Just in case anybody is wondering, I expect to earn millions on a project… I just haven’t figured out what the project is. Anybody interested?
Funny, as I read about Midori, once you tear away all the marketing fluff, the core ideas are pretty much the same as what we are already working on over in SharpOS (http://www.sharpos.org) and Ensemble (http://www.ensemble-os.org). Especially with us over in Ensemble, we’re writing a kernel HAL layer that is meant to be run inside a host CLR environment.
Members from both projects are also writing an all-C# compiler and CLR/VES layer, and we have already formed a cooperative alliance group to foster this, called MOSA – The Managed Operating System Alliance. The compiler and CLR layers will (at least in planning and code skeleton, at this point) have pluggable assemblies to handle the actual native machine code emission, as well as 32/64/whatever-future bit independence (just a redefinition of IntPtr, among a few other things at the compiler level). No need for Bartok and C++/assembly shims here.
Likewise, the supposed “transparent distributable computing”, which really is just an extension of how the OS hooks the .Net Remoting API, is the same.
For the disbelievers, SharpOS has been bootable for at least a few months now, and Ensemble is getting somewhat close, pending machine code emission work on the common MOSA compiler.
It’ll at least be interesting to see where MS goes with Midori, if anywhere. At least they’re showing some signs (if only hand-waving) of applying Singularity to something “real”. But in the meantime, there are others that are already working on this, in the OSS realm, who will probably get there a lot sooner than MS.
Uh, yeah, of course the “core ideas are pretty much the same”, because they’re all derivative of Microsoft’s Singularity research project.
What is “this” and where is “there”? You don’t know what MS is working on. It’s all just vague rumors and speculation. Further, it’s not clear that ANY of these projects are really needed or will ever to come to fruition. Right now, it’s little more than hobby-ware.
“The incubation project, also known as Midori, seeks to create a componentized, Net-centric OS, based on connected systems – one that largely eliminates dependencies between local apps and the hardware on which they run.”
I hate to be those Unix fundamentalists [1] who tirelessly rant about everything Microsoft product has already been accomplished by Unix a thousand years ago, but the project sounds like Plan 9. You can import servers running on your local computer or on a computer across the internet and access its services by mounting it. A server can be a program that reads a filesystem; the filesystem can seamlessly be exported across the network (no NFS!). A server can be access to your sound card, so you can play music locally or across the network. It doesn’t matter if the server is running on a different CPU architecture, as the protocol is architecture independent. I believe this qualifies Plan 9 as a “component, Net-centric OS, based on connected systems” and “largely eliminates dependencies between local apps and the hardware on which they run.”
1. Does avoiding modern IDEs like Eclipse in favor of screen & vim make me a fundamentalist?
–EDIT: some grammar
Edited 2008-07-31 02:32 UTC
Plan 9 isn’t UNIX, it was an attempt to make something better.
http://en.wikipedia.org/wiki/Plan9
There are alot of really cool things in plan9 that never have really been implemented anywhere else.
Edited 2008-07-31 02:54 UTC
The main difference is that Plan 9 is C/machine code centric whereas Singularity, et al. are based on .NET bytecode to enforce low-level security features (pointer bounds, type safety, etc.)
Inferno ring any bells?
All this “cloud” computing reminds me of the old “Time Share” phase of IBM”s history when it was a MONOPOLY that was considered EVEN WORSE than Microsoft as it is today.
Remember those “good old days” when computer users had to
RENT time and software in an IBM mainframe or Mini computer (which was actually as large as a REAL desktop of the “dead tree” type including the rest of the desk.
It is my belief that this whole “cloud” computing, “network” computing, “Application Service Provider” computing system is a danger to individual freedom on a par with a government gone bad because of our current society’s dependence on these machines. Because of this they should be as indpendent of each other as possible while still having the good aspects of the internet available such as free political speech not controlled or “filtered” by broadcast or cable networks, convenient information and convenient shopping.
I think a much better idea would be to use the “interpreted” or “virtual machine” languages such as Java, Python, Euphoria, C#, and maybe even interpreted versions of modern BASIC, pascal, and C/C++ as the basis of a system to create Universal Executables that would run on any independent OS.
This is how this would work.
1. Interpreters for the languages would be made for all the popular OSs (this has already been done with Python
and Java)
2. A stand alone executable archive file should me nmade for use in creating Universal Executables. Such executables would when “compiled” contain all of the interpreters for all the supported OSs, start up scripts for them that would respond to the proper interpreter for running the executable under the OS it is launched on. An indepedent GUI library (or if the software is a game a game 3d graphics and sound library) for Windowing apps or an independent curses library (meaning independent of the ncurses, PDcurses or the “conio.h” curses functions available in Windows
IDE compiler systems and an independent database library.) An executable made under such a sustem would be able to run an any machine and end the threat of monopoly in computing once and for all even if Microsoft still keeps its 95 percent desktop monopoly because executables made with it would run on ANY OS majority or minority.
As for licensing the language interpreters not already under a licensing system of their own, Executable Archive file, should be under some F/OSS license that allows exceptions for building closed source proprietray software from them.
The suggested Libraries including
the GUI library
the Game and multimedia library
the Curses library
and the DataBAse library
should probably be LGPL with linking exceptions for closed proprietray software similar to FLTK and wxWidgets or under some other linking exception allowing F/OSS license.
Of the available languages Euphoria probably comes closes to this system except for the universal executable idea with the ability to build Executable archive based standalone software through its binding and shrouding system. Java also comes close with its OS independent SWING GUI library. WxWidgets even though onnly for native compiled C++ at this time also has a good independent GUI system in wxUniversal.
While apps created this way would certainly be large compared with the “dependency hell” based apps of today
I see this is one of the prices of ending computing monopolies altoghther, not just Microsoft’s and mantaining our individual freedoms and rights in the computing area.
I agree that there needs to be change in the current use of computers – a lot of the problems you raise come from engineers not thinking about the social influences/issues raised by their designs – particularly the loss of personal freedoms.
Personally, I think it unwise at the moment for Microsoft to start planning the “OS of the future”, there are so many other issues that are perhaps more important:
– what we do when there’s no oil for PCBs
– when there’s less power available
– when people become more important???
Admittedly, some of the issues should be the reserve of hardware firms, but if Microsoft wish to retain their monopoly (however much I disagree with the legality), then they, and all computer firms, need to be looking at the twin threads of efficiency and necessity.
Whilst it would be nice for computers to be used in every part of life, there are *so* many unanswered questions, and potential future environmental (as in, social, economic and environment), issues that it would be worth looking at perhaps, more streamlined operating systems. For the Free software world, this is something that’s probably less of an issue than it is for Microsoft.
Since Microsoft is clearly thinking about what’s after Windows. So what will UNIX, Linux or OSS ( not the same groups of folks necessarily ) will do for post-UNIX such as post-Linux, post-BSD, post-Solaris, post-AIX, etc. ??
Will the operating systems be fractured into tiny little OSes islands?
Considering those OS’s open-source natures (except for current AIX), and our penchant for thinking in terms of evolution vs revolution… I don’t see significant fracturing in our future. We tend to maintain a certain level of diversity of offerings, meaning more heterogeneity and less homogeneity than Microsoft. But I would not expect the level of heterogeneity to increase substantially. The trend that I see is more a very gradual move toward the elimination of arbitrary differences, while retaining those differences which really make a difference.
Microsoft is actually a bit late to the game regarding “the network is the computer” thinking. The unix-like OS community has had a contingent thinking along those lines for some years.
Then again… to an extent, no technology truly exists until Microsoft “invents” it.
Edited 2008-07-31 14:48 UTC
There are already OSes in development that claim to be developed in managed languages.
SharpOS is such an example.
Anyway, we can rest assured that Windows isn’t going anywhere for a long long time and Linux will remain relevent (in my mind) for as long as Windows exists.
My photoblog is midoriconcept.com. Do you think that Microsoft is going to buy my domain for one zillion dollars?!?
The whole thing is reminiscent of Sun’s Spring operating system in the early 1990s which got shelved because Sun wanted to concentrate on getting Solaris into shape. Indeed, the whole Singularity situation has been pretty similar to the Spring affair from the start: established vendor starts “research OS”, “it’s just an experiment”, people get excited and wait for products. We just need to see if it all turns out the same.
A few yeas ago I read some ibm papers about the cell architecture (now in every ps3). As I remember they wanted a cloud of equipment think-powered by cells in every household. These should give each other process times or work as needed.
In such an environment wouldn’t it be thinkable that computer viruses crunch together, like in early evolution times crowds of cells builded the first primitive animals, and build the first primitiv artifical intelligence?
And you will guess, that mankind kills it, if it is observed. Thus you can guess, that if an artificial intelligence is building up in the future to a not any more primitive but overwellming point – that this can and will be only a hidden artificial intelligence. Then after a hidden hundred years it will speak up: “Hi there mankind! I will kill you.”
That will be the way of evolution of technical intelligence in the future. Or energy comes down first and technical development comes to a standstill. The last one is my favorite prognosing the future and much more likely.
Edited 2008-07-31 20:36 UTC
Microsoft is becoming so irrelevant.
Seriously…
After 20 years of being arse-raped most people only have two reactions:
1. OK come on finish it quickly!
2. Oh no not you again….
Before I go to bed, I would like to make use of my sleepiness to make a few naming suggestions for the next versions of Windows… Most of these are not funny I know… Just need to let off some steam… My apologies…
A. Windows Rebooted
B. Windows 2009 until 2011 Edition
C. Windows Beijing Olympics Starter Edition
D. Windows Fundamentals for 80’s PCs
E. Windows Melinda Edition for Non-Profit Foundations
F. Windows 360X
G. Gates
H. Windows 2008 for Dummies
I. Windows with Genuine Advantage for Workgroups
J. Windows YouNameIt!(tm) Edition
K. WindowBSD
L. Windows All-You-Can-Eat PayPerView Edition
M. Windows Mobile Compact Server for Embedded Pocket Tablet Center Edition
N. Windows 2010 Desktop
O. Windows Desktop 2013
P. Windows One featuring U2 – (Did I disappoint you,
Or leave a bad taste in your mouth, You act like you never had love…)
Q. Windows ZER0 featuring Smashing Pumpkins – (Wanna go for a ride???)
R. Windowz Russian Doll Edition
S. Windows 200in1 Cartridge Edition
T. Windows Stainless Steel Edition
U. Windows Offline Web 2.0 Reloaded
V. Midori .NOT
W. Windows I CAN’T BELIEVE IT’S NOT BUTTER EDITION!
X. Windows 95 Shutdown Nostalgia Edition
Y. Windows X-OS Agua! Agua!
Z. Windows Waldo Edition – Where’s Linux?
Ok I feel like an 8 year old now… Sorry for the non-sense… Good night guys……
Edited 2008-07-31 22:41 UTC
I’m not sure which documents the authors of the article had access to, but I would certainly take a wait-and-see approach to Midori and its particular structure or features. A lot of smart people are working on it, but they’re still trying out ideas and the real art to shipping something is figuring out when the product has enough features to be compelling while cutting or deferring everything that can go into later versions.
This is different from other OS projects in that it comes with a fairly different language and toolchain which gives the compiler enough information to enforce ‘correctness by construction’ which the article mentions. I’m not sure it’s enough to just make an OS in a managed language like C# or Java.. I think the Singularity and Midori guys are going beyond that.