 |
 |
| ActiveWin | Anonymous | Create a User | Reviews | News | Forums | Advertise | VBA in Excel | Users Online: 0 |
|
|
|
|
|
User Controls |
|
New User |
|
Login |
|
Edit/View My Profile |
|
|
|
|
|
|
|
Active Network |
|
ActiveMac |
|
ActiveWin |
|
ActiveXbox |
|
DirectX |
|
Downloads |
|
FAQs |
|
Interviews |
|
MS Games & Hardware |
|
Reviews |
|
Rocky Bytes |
|
Support Center |
|
TopTechTips |
|
Windows 2000 |
|
Windows Me |
|
Windows Server 2003 |
|
Windows Vista |
|
Windows XP |
|
|
|
|
|
|
|
News Centers |
|
Windows/Microsoft |
|
Apple/Mac |
|
Xbox/Xbox 360 |
|
News Search |
|
XML/RSS Newsfeeds |
|
Pocket PC Site |
|
|
|
|
|
|
|
FAQ's |
|
Windows Vista |
|
Windows 98/98 SE |
|
Windows 2000 |
|
Windows Me |
|
Windows Server 2003 |
|
Windows XP |
|
Windows 7 |
|
Windows 8 |
|
Internet Explorer 6 |
|
Internet Explorer 5 |
|
Xbox 360 |
|
Xbox |
|
DirectX |
|
DVD's |
|
|
|
|
|
|
|
Latest Reviews |
|
Xbox/Games |
|
Fable 2 |
|
|
|
Applications |
|
Windows Server 2008 R2 |
|
Windows 7 |
|
Adobe CS5 Master Collection |
|
|
|
Hardware |
|
Microsoft Express Mouse |
|
|
|
|
|
|
|
Latest Interviews |
|
Mike Swanson |
|
|
|
|
|
|
|
Site News/Info |
|
About This Site |
|
Advertise |
|
Affiliates |
|
Contact Us |
|
Default Home Page |
|
Link To Us |
 |
Might Microsoft’s Midori be ‘Cairo’ revisited? | |
|
||
| Time: 07:00 EST/12:00 GMT | News Source: | Posted By: Kenneth van Surksum | ||
|
Mary Jo Foley: My post about Microsoft’s post-Windows operating system, code-named “Midori,” elicited some interesting responses — and a few potential new clues over the past week.
|
||
|
Read Only Comments Return to News |
Displaying Comments 1 through 19 of 19 |
This is an archived static copy of ActiveWin.com. |
| #1 By 72426 (71.94.8.164) at Monday, July 07, 2008 11:20:19 PM |
|
#Mary Jo Foley
Mary is 'supposed' to have inside sources, but she herself doesn't 'STILL' get that MinWin that was demonstrated IS the NT kernel core, and was nothing more than a stripped version of a slightly newer build of the Vista NT kernel with an HTTP server used as the I/O subsystem. She continues to assert things like that MinWin won't be Windows 7, when in fact it has been in NT since 1992 and NT 3.1 til today and Vista, as the core of NT is very small and without the upper API levels or subsystems is a MINI KERNEL technology. Microsoft has used a stripped version of the NT kernel internally for years and years, and even educators can obtain it and a good chunk of the source for it. And yet, again in this article, she AGAIN STILL DOES NOT GET IT, and writes about it like it is a separate product. With that aside, the Singularity kernel that Microsoft has been working on in an academic fashion has been around for a few years now, and will more than likely bring technology BACK TO THE NT kernel, rather than be a NT kernel replacement. In kernel technology, theorists and engineers that even hate MS will argue that NT is a small solid kernel technology that still does things NO OTHER kernel out there does or can do... Mary, who are your sources? How do I have better sources than you? How do you continue to get paid to write this crap? Holy Cow... This post was edited by anthonyspt on Monday, July 07, 2008 at 23:21. |
| #2 By 15406 (216.191.227.68) at Tuesday, July 08, 2008 08:50:01 AM |
|
#1: I think you're the one who doesn't get it. Minwin has been in NT since the 90's huh? That's a new one. I think you're confusing a kernel in general with the specific minwin kernel that may or may not be in 7. Considering how smart you are, why don't you start your own tech blog and show MJ how it's done?
In kernel technology, theorists and engineers that even hate MS will argue that NT is a small solid kernel technology that still does things NO OTHER kernel out there does or can do... That's an interesting statement. Can you provide any examples of things the NT kernel does that no other kernel can do? Mary, who are your sources? How do I have better sources than you? How do you continue to get paid to write this crap? I looked and looked, but must admit that I failed to find any citations in your post to back up what you are saying, considering you're disparaging her for her supposed lousy sources. |
| #3 By 23275 (68.186.182.236) at Tuesday, July 08, 2008 10:36:42 AM |
|
#2, Latch, anthonyspt is absolutely and fundamentally right and he presents the proper perspective here as well.
Let's agree to take the politics out of this one and simply discuss what little we do know about this... and yes, MJ is wrong in the sense that there is little exploration offered and no real tie into what is actually taking place, or why... which compells me to offer that none of us really do and similarly, there several engineering groups within Microsoft that are working these areas and not all of them have the same goals - however related they may be. Let's source this next bit first - Microsoft Fellow Mark Russinovich and Microsoft distinguished engineer Eric Traut... and before we go further, YES, MinWin is in Vista and WS08 as described in 2003! That said, the work being performed on and around MinWin now is not the same work - different goals and different projects. (I'll quote directly where it helps, but the goal here is not for me to be right per se, but to help people perhaps understand this a bit more, and at least encourage those interested to look for more and beyond largely lay tech blogs - one can fairly assess that MJ is doing the same - so no hit on her at all). Cont... |
| #4 By 23275 (68.186.182.236) at Tuesday, July 08, 2008 10:48:40 AM |
|
Before I go further, I have to send "props" out to Paul Thurrott for having the sense to dig into these things in a way that is both helpful and candid and if you just want to brush on this stuff, read Paul's take on the matter here, http://www.winsupersite.com/showcase/win7_minwin_inside.asp
Paul nails the most important question: " The first, and most obvious, is why would Microsoft undertake such a project?" <I know... I know... (like some anxious kid in the back of the class...)> I assess that Microsoft is going to do two things with this (see, in Vista/WS08 you have catalogs, so all the hunting about for upward dependencies that existed in XP is a thing of the past - and in this statement is the key to understanding what MS is likely up to)... 1) forget the idea that "Windows is bloated" BS - making Windows smaller is not the goal here - making it more efficient yes, but that is a by-product. Back to those catalogs... MS already has the highly componentized OS model they need (Vista/WS08) - so forget bloat folks - it ain't there. Let's say you have an app, an old app and it needs x, y, and z components to run. Like Windows Updates, as an example, the catalogs are queried and the right mix of components are dished up to support the old app... so number one is an environment that dynamically shapes itself to the app being run, or the service, or whatever.. so much for backwards compatibility problems, right....? cont... |
| #5 By 23275 (68.186.182.236) at Tuesday, July 08, 2008 11:23:56 AM |
|
right there in front of us Vista users is an example of what to do with such work and the OS many of us love... Windows Update in Vista/WS08 - remember when I spoke of it being an app? now you know how it works... it uses MinWin and the catalogs Vista/WS08 maintain - you know... all those granular components the nitwits say aren't there...
I guess you could call that virtualization, but of a sort that is far more advanced than what we have come to understand in that context. Windows 7 will feature this use of the new work around MinWin. "application virtualization (running client applications without having to install them locally" http://channel9.msdn.com/shows/Going+Deep/Mark-Russinovich-On-Working-at-Microsoft-Windows-Server-2008-Kernel-MinWin-vs-ServerCore-HyperV/ Mark goes into a lot of detail and its worth viewing several times. 2) "is that a MinWin in your pocket...?" "Is that a MinWin in your cloud?" Win7/MinWin and related projects center on one thing (that I see): Where are the dependencies? There aren't any! Nothing in the kernel runs outside the kernel. Hmmmm.... ok.. so there's a means to call the components you need dynamically... very cool!... and there's the means to assemble what you need around the calls you are making and even a means to store those values for very rapid re-use and re-call later? Yes!!! Yes!!! Yes!!! - SO.... Win7 is going to be blisteringly fast and it sure is going to make great use of Multiple cores etc... and apps virtualization is what? Going to be the norm, right? Yep! Yes and Heck yeah! Oh... and no matter where, or on what either... phone, cloud, whatever... Cont... |
| #6 By 23275 (68.186.182.236) at Tuesday, July 08, 2008 11:36:15 AM |
|
by the way... Vista absolutely had to come first... the driver model... the componentization.. one day very soon, a great many people are going to be thanking Vista and the ground it had to break and the path it had to cut... "Hey, these Longhorn guys really did know what they were doing.... right down to the freaking codename...."
I could keep going all day... but one thing has stuck in my head for more than 30 years... and it is something Bill Gates himself often referred to - the sub-routine. I mention Mr. Gates in this context, because he really was the "architect" at Microsoft and he conditioned an industry to think in the context of subs. As I said, I could go on about all of this all day - it's really that much fun, but I can't (work calls for now). I hope that we can get some discussion rolling around this that gets back to the fun in it and leaves the politics and the marketing at the door. For now... the work around MinWin is all about subs. It extends that thinking in so many incredible ways... and of course, I can only imagine what they are from out here, but if what Mark is saying is true, and Sinofsky can work with him and others as well as Bill did, we're in for some really exciting stuff. There has never been a better time to be a developer - you will be able to build things that will change the world. |
| #7 By 15406 (216.191.227.68) at Tuesday, July 08, 2008 12:41:17 PM |
|
#3: I read all of that and I'm still looking for how any of this ties in to what anthonyspt & I were talking about. I'm also confused about how you endorsed him right from the get-go, and then either completely ignore what was said or refute it. If Anthonyspt is "absolutely and fundamentally right", as you claim he is, then I'm still very eager to find out more about:
- how the MinWin kernel, a supposedly new thing, has been in Windows NT for the past 16 years - the NT kernel does things no other kernel does or could ever do Those were the two main statements that I took issue with. |
| #8 By 1896 (68.153.171.248) at Tuesday, July 08, 2008 01:36:22 PM |
| i could be wrong but I think that in the next couple of years we will be talking, and miss, Bill Gates a lot. |
| #9 By 72426 (71.94.8.164) at Tuesday, July 08, 2008 02:09:29 PM |
|
#7 Latch...
Friend you truly need to do a little homework on NT or even go old school and read the first edition of Inside NT written before NT was originally released in 1993. NT is a hybrid, client/server kernel technology. This means that it is very different from standard micro and monolithic kernel designs. It gives the advantages of both ends of the spectrum of kernel design, since at the core it is very small and tiny, yet has very rich upper layer kernel APIs. Thus it doesn't have the locking multi-tasking problems monolithic kernels do, nor it is tied down by having everything in the core kernel as micro kernel designs do. So NT's core kernel is VERY TINY, the HAL is less 256K in Vista, was around 140K in XP and less than 64K in the original release. Considering what the HAL normally does, this is very tiny. NT uses and internal object based API system, that allows NT to layer rather elegantly through the NT kernel layers, going from the generic HAL, up several levels of APIs in the typical kernel area to 'add on' APIs for drivers etc at higher levels that can be abstracted at any point. This is doing stuff that no other kernel technology does or can do. NT also has other aspects it does that other kernels, specifically *nix kernels just don't or can't do. Since NT is object based, all devices and I/O is object based, and even the security model is a token/object based system. Because of this NT understands the difference between devices and files, instead of treating everything like file read/write using basic textual construsts as UNIX does. What this core technology gives Microsoft is a very extensible kernel design that is very configurable and dynmaic. (See Windows Embedded for how light it can be.) If you look at Vista and the WDDM, this is a new driver model that runs alongside the existing XPDM driver model, which both are kernel level drivers, with the WDDM sitting paritially at the kernel layer and shoving the rest of the driver back to user mode. The WDDM is more tha just a driver, as it about GPU scheduling (pre-emptive 3D GPU), GPU RAM virtualization, Multi-GPU handling at the OS level. If WDDM or a WDDM concept were to be added to any other kernel technology it would require a lot of work, where MS only had to 'add' it to NT, even at the lowest level of the OS for Vista. This is where the beauty of NT's extensibility begins, as MS left the legacy XPDM inplace as well, since even being competiing driver designs, they can exist together without incident. A more general area to look for NT design and how it works differently is the 'subsystem' concept that the OS is designed around. Win32 is an OS Subsystem that sits on NT as the primary OS interface. Thus NT can handle and always has hand multiple-OS subsystems that all run equally, without virtualization or emulation. This is how the original POSIX subsystem on NT ran, and is how the SUA (UNIX BSD) subsystem runs on Vista/XP. The SUA is a real OS running equally to Win32 and Win64 subsystems. This is the 'upper' end level of the NT kernel technology being a client/server design, as not only the kernel APIs are separated, but even what the users see as the OS (Win32 for example) is just a subsystem layer on top of the NT kernel. This is something ONLY NT does and currently ONLY NT can do. (This is about the OS subsystems, not virtualization, emulation, translation, don't confuse the other technologies that other OSes use to maintain or run other OS environments.) This post was edited by anthonyspt on Tuesday, July 08, 2008 at 14:12. |
| #10 By 72426 (71.94.8.164) at Tuesday, July 08, 2008 02:12:57 PM |
|
#7 - how the MinWin kernel, a supposedly new thing, has been in Windows NT for the past 16 years
Only the people that didn't know better said it was a 'new' thing. Splitting apart the NT kernel and removing upper level APIs, the OS subsystems gives you a very tiny kernel called MinWin, and is something Microsoft has developed with and used for years. Again go look at Windows Embedded even, it is just a few more APIs from MinWin. NT can and does break apart fairly easily, being a client/server kernel design with lots of layers and 'external' kernel APIs that are brought back to the kernel as needed. So when everyone saw an NT kernel under 32mb and then looked at Vista (which is large) didn't not 'GET' that this was just the lower NT kernel ripped out and compiled with an HTTP server interface. This still confuses people, sadly. |
| #11 By 23275 (68.186.182.236) at Tuesday, July 08, 2008 02:13:01 PM |
|
#7, Sorry... I read a good bit of what Anthonyspt has contributed around the web's many sites and he is correct in the context he has shared here - he wasn't trying to split hairs but point out that all this talk about new kernels isn't accurate.
ok, for starters where do you want to go in regard to kernel architecture? dynamic thread feedback? scheduler re-tasker that allows a thread to be elevated? associated dispatcher monitoring for available resources/separate cores/CPU's? (base plus burst priority)? We could spend a month of queuing alone and still not fully address how the NT Kernel is truly unique and better than anything out there. I mean... dang... out of the box when first shipped, there was support for 8 CPU's! A biggie would be how changes in a thread's priority level, from that of its primary process base priority, is done to facilitate not only CPU utilization, but also to enhance the response time of the application. A thread's priority is lowered if its time quantum expires, since it is not already at its base priority, and its not real time. Up to you, but I'd rather focus on what MS will do going forward as I shared my guesses above. |
| #12 By 72426 (71.94.8.164) at Tuesday, July 08, 2008 02:31:10 PM |
|
#11 lketchum...
The trick of NT is that it is very portable not only in the sense of code, but in the sense of kernel API layers. Some of things you mention are upper level kernel APIs and some are lower level kernel APIs, and I think this is confusing people, as they often look at a microkernel like Linux for a basis of kernel understanding, and it just doesn't apply to NT. So remember that everyone tends to to look at NT or other kernel technologies through Linux glasses. As for you referring back to Paul, he has good sources and some level of IT understanding, however when you get to the kernel or technical details Paul doesn't understand as much as he even admits he would like to understand. So even when he has spot on information, unless the technical details are spooned to him, there is a disconnect. For Paul, read with this disclaimer: Good Sources <> Understanding PS I could be wrong about Windows 8, as the singularity and other academic kernel projects in Microsoft could be the next big thing and they might move to drop NT and replace it with something new. However, it would take a lot for kernel technology to 'jump' past NT enough for Microsoft to make this commitment. This is why from what I have seen, most of the new kernel work has been around moving the technologies back into the NT architecture rather than replacing it. Take care... |
| #13 By 23275 (68.186.182.236) at Tuesday, July 08, 2008 02:33:57 PM |
|
#8, you are quite correct - a great many peope already do. To give you some idea of what I mean... when I learned electrical engineering, I was taught one on one by one of the men who invented the transistor as we know it today - there were actually three. As a discipline, that field of engineering did not exist yet. This is important, because as practically implemented software as we know it did not exist before Bill Gates, and just like the transistor I mentioned as an example, that discipline did not exist yet. In a round about way, I'm trying to say what it is like to work with people that have vision which is so far reaching and the very rare ability to advance things practically to a point that others can share that vision, or at least see it. Now the real brain buster is when you think that men like Bill Gates stopped well short of what he actually saw. I say that with a bit of a chuckle... as I remember one thing I heard all those years ago.... my teacher said: "Someday we'll change what these do in software, but first we'll simply influence how they are used" But you see, at that time, software were simply lists of instructions and nothing like we have today. We literally ticked them onto gummed tape by hand.
Now obviously, the guy was talking about FPGA's and dynamically structuring circuits based upon soft instructions and interim steps in oops to emulate that. We're not there yet, and that is my point... we have not seen a thing that guys like Bill Gates knew would one day be common. |
| #14 By 23275 (68.186.182.236) at Tuesday, July 08, 2008 02:46:16 PM |
|
#12, great point... and yes, that was my intent... to get some discussion going around how Windows/NT is truly different and frankly, radicaly different from the monolithic designs associated with the *nix
I think Singularity is about where to take it after 7 - I think also it is because where Intel/AMD will take the architecture over the next few years. The flexibility they will require has to address what might happen when multiple cores exist on one die and each is dynamically tasked to perform instructions/services - e.g., they are all CPU's and GPU's or 8 are GPU's one moment and at the next 12 are CPU's (to illustrate what I mean). It's a guess for sure... |
| #15 By 1896 (68.153.171.248) at Tuesday, July 08, 2008 08:20:11 PM |
|
#13: Interesting, you are an electric engineer Iketchum.
Out of curiosity what is your opinion of Nikola Tesla? I have always seen him as a true genius, way ahead of his, and for many things also our, time. As a classic genius he was ripped off of a lot of things and some of his ideas have been buried away. I am interested to know your opinion about him. |
| #16 By 23275 (68.186.182.236) at Tuesday, July 08, 2008 10:05:06 PM |
|
#15, Terrific question and topic... first... he was a genius in mnay ways - not just as an engineer. See, he applied physics to many practical models "before" committing them to practical applications. Before I go further, I have to state as strongly as an old man that well remembers life without electrification, that Tesla personally made life as we know it possible - and for me personally (more on that in a moment). You see, Edison was not nearly as much of a visionary and he was very much about control. He advocated for direct current and Tesla, along with every other physicist of note during that period knew that alternating current was the way to go if all people were to be able to access affordable electrification.
I personally appreciate Tesla, because his designs for induction mortors made the TVA possible and the TVA finally made affordable electrification possible in rural areas. It took a congress over a decade to finally make that a reality for millions of poor people - up until then, we washed in buckets, labored long days and needless to say, we sweated a great deal. Not to mention how much of an impact on general health all of it had. Sadly, we embraced the reality of premature death as many around the world still do. Cont... |
| #17 By 23275 (68.186.182.236) at Tuesday, July 08, 2008 10:18:16 PM |
|
So for me professionally, and personally, Tesla is a real hero. Many decades later when I had a chance to serve in his native Serbia, I felt happy to be able to give at least a little back. I am still angered at how he was treated, lived and eventually died - without much at all and not nearly the recognition he deserved. That model frankly still motivates me - that no matter how well one does something, or how well they mean to do it, they can still end up very bad off.
By the way, very early electrical comutators I would design for the processing of very early digital signals, were based upon his designs for generators. Such comutators were still in use and dead reliable up until just a few years ago. It always astonished me how many of us walked around on the shoulders of giants - which Tesla certainly was. Similarly, I used the idea of out of step AC to modulate signals - exactly has Tesla's work did the same to generate AC! That stepping, called OOP, or out of phase - managing that was very tough before precise digital timing mechanisms were developed - think about how important time is within Active Directory and you can see over and over again the application of the same principles!!! Cont... |
| #18 By 23275 (68.186.182.236) at Tuesday, July 08, 2008 10:20:38 PM |
|
Any case.... we owe Tesla just about everything we know. There is so much more to him than will fit here, but allow me to say that he insprired me and many others in about every way a man possibly could have. He even became an American citizen, which is something I have had the privilege of helping many people do.
As I said, he inspired me in a great many ways and he deserves significant attention in this country and around the world. Thanks for asking about him. :) |
| #19 By 1896 (65.9.187.245) at Wednesday, July 09, 2008 08:40:01 AM |
|
Thanks you for your comments. It is interesting as your opinion, as well as mine, about Edison differ from the widely common one: he was indeed a smart business man but not a real "inventor". All his plots to link Tesla and his alternate current to something dangerous and nefarious like the "electric chair" really damaged and delayed progress.
I also agree that there are still so many Tesla's ideas, projects and intuitions that have never been thorughly explored. The famous machine to create earthquakes, his theories about teltransportation just to mention some of his most, too ahead in futures, projects were boycotted and buried for whatever reasons. A real, real shame. Thanks again. This post was edited by Fritzly on Wednesday, July 09, 2008 at 08:40. |
