Slashdot Mirror
How Microsoft Plans To Get Its Groove Back With Win7
shawnz tips a blog post up at thebetaguy that details Windows 7's huge departure from the past, and the bold strategy Microsoft will be employing to maintain backward compatibility. Hint: Apple did it seven years back. There are interesting anti-trust implications too. "Windows 7 takes a different approach to the componentization and backwards compatibility issues; in short, it doesn't think about them at all. Windows 7 will be a from-the-ground-up packaging of the Windows codebase; partially source, but not binary compatible with previous versions of Windows."
The thing is, the only reason most people run Windows is so they can run legacy Windows applications. A Windows that can't run Windows apps? Yeah, that'll sell like an iPod that can't play MP3s.
GCHQ Quantum Insert installed. If only our tongues were made of glass, how much more careful we would be when we speak
....shooting themselves in the foot. WIth Vista they screwed up half of the drivers and now with Windows 7 they screw up the entire lineup of software? WTF?!
No really... we'll get it right next time. The last five years were a mistake, but give us a few more years and we'll be more Mac-like. Honest!
I mean Cairo, I mean the next piece of vaporware that will be used to keep Microsoft in a dominant market position even though their current product is inferior to the competition in both the desktop and server space, because why migrate off when "Windows 7" is just a few years away and will be SO FAAARRR ahead of everyone else.
Same tune.
My Babylon
Oh, right - it's harder for force upgrades like that.
I'm gonna agree that this may not turn out how they want it to. Although I'm all for throwing out the old and starting new, the sheer fact that Windows has to support not just legacy software (which can be easy to emulate, sort of) but legacy hardware as well, probably means more people will have issues with this than not.
Wasn't this what Vista was supposed to do in the first place? It was supposed to be a dramatic departure from previous versions, but too much politics pressured developers into making backwards compatability a little too over-bearing on the system. This is clearly what they were trying to accomplish with Vista, but higher-ups were too afraid to do it, so they told them to half-ass everything to make it all work. After seeing what a disaster Vista has become, both on the development and user experience side of things, the Higher-ups have no choice but to listen to what their devs wanted in the first place; kill legacy. Not build it in and make it limp along half-working and hard to develop for, but just start with a clean slate and build a kickass base OS and worry about compatability with older applications and frameworks later. Basically, they tore a page out of OS X's plan of action.
...but not binary compatible with previous versions of Windows Sure Vista does that now.I seem to remember Vista was supposed to be a huge departure from what was done before - and then reality hit.
The mistake they are making (will make) is that that they think their software is what is broken - when in fact the software is just a representation of the business model they have chosen. Their system design is market driven not engineering driven - and whatever they produce from this point on will be the same as all the others. Windows, OSX, Linux, Unix etc are all products of the ethos in the organizations in which they are created.
If the mould is defective, there's no point is making a second one in the hope that it will turn out differently.
Genesis 1:32 And God typed
Unfortunately, the article itself is a work of fiction. The guy has lots of bad reasoning, poor memory and is desperately lacking in technical understanding.
;-)
For once, I'd say just read the article summary
Why can't they do what Apple has done about 3 times now?
Move to new technology, but provide a compatibility layer so legacy apps still work, even if they are in some sort of emulated environment?
The new hardware people will be using with the new system will be fast enough that even an emulated environment will be as fast (or faster) then their previous machine.
With the virtualization technologies available today this should be even easier to do then, say, Apple's transition from 68xxx chips to PowerPC chips, or PowerPC chips to Intel, or OS 9 to OS X.
Were they all seamless transitions? No. But they were arguably better then then the transition from XP -> Vista has been so far.
Microsoft seems to want to either take the course of backwards compatibility at the expense of progress, or progress at the expense of backwards compatibility.
Why not go for the best of both worlds through emulation/virtualization?
I couldn't get past the first paragraph.
"In the face of the mass-media criticism of Windows Vista, mainly with regards to the performance issues present when compared to Windows XP on hardware with similar specifications. However, very little information has been presented with regards to the performance of Windows 7, this article however shall change that."
Microsoft is always promising the next Windows will be built new from the ground up so not much is really new this time. The only difference here is the promise to break backward compatibility. Thebetaguy contradicts himself about that by having the balls to promise, "This should allow the majority of legacy applications to run perfectly," while Vista provided less than 60% of the same.
There are lots of other contradictions because thebetaguy does not really want to admit several things and he's angry about the few he's given in to. The Microsoft way of doing things was inadequate, but the change is blamed on legal challenges that competitors strangely don't have. He cites some of Vista's insane processes but fails to mention digital restrictions or the last minute elimination of XP drivers as reasons for poor performance. It's funny to watch a fanboy admit Microsoft is following Apple, but it would be nice for him to also admit that Apple followed free software and Unix practices.
Like I said, there's not much to this article. It's mostly a fanboy making excuses and casting blame for the failure of his favorite operating system. No real details have been announced and the game plan will, as usual, change before release - a sure sign that there's nothing really open about the "new" Microsoft. They are going to keep their secrets and continue to mess with anyone who's got any revenue potential.
Oh come on, do you really expect every article to specify the distributions it is referring to when the characteristic is something shared by almost all of the major ones. At best you'll get them saying Ubuntu instead (in place) of Linux, at worst they just won't bother mentioning it. You know full well what they meant, as did anyone else who knows what Linux is so why the attempt at criticism.
The first article tries to push the idea that all problems Microsoft is experiencing come from the antitrust wrist slapping they have got. This is stupid. Also takes some jabs at Apple and Linux.
The second part of the article is telling us the real problem Microsoft is facing. Code bloat. Dll hell. They have decided that they canÂt hold it any longer and they are going to start from scratch and run the old windows apps on a virtual machine for backwards compatibility.
There is a third part that is missing in the article. Most people around here suspects that some of VistaÂs performance problems, specifically on the the multimedia department are caused by the interference of DRM code. Is Microsoft removing all this code from Windows 7?
When his defense asked, "Which computer has Jon Johansen trespassed upon?" the answer was: "His own."
No single link to source - where did they get this info, just unfounded speculations.
Windows 7 early builds was already demoed and there's no evidence that it will be backward-compatible.
Also WinSxS (side-by-side dlls) is what windows xp uses to maintain different versions of runtimes from the start and obviously it has little to do with OS speed.
While reading this article the only thought prevailed - wtf author is smoking. Complete rubbish.
- Arwen, I'm your father, Agent Smith.
- Well, you're just Smith, but my father is Aerosmith!
No numbers. No estimations. Just some hand waving of "they are doing something different". The article doesn't change that fact at all.
Because OS X and Linux aren't de facto monopolies with 80%+ of the market.
Yes, because loading 1 MB of code as part of one executable is vastly faster than loading it as 1 MB of library. This is especially true when loading 10+ different executables that have the same code statically linked in. That is way faster than loading it once. More efficient too.
No, wait...
Besides, that code (such as MSHTML.DLL) was already an external library. Just about every operating system tends to get new libraries with major upgrades. Windows was not one monolithic executable before. Heck, it wasn't way back in the 3.11 days.
That has not always been the lure. The lure was it was pretty and not a DOS prompt. Then the lure was simply that there were more programs for it when it became dominant. But then again, Leopard runs programs designed for Tiger and before. OS 9 ran programs designed for OS 7. Just about every OS does that, including many UNIXes.
You've GOT to be kidding. "Proven" for OS 9? It didn't have memory protection. It didn't have preemptive multitasking. Heck, you still had to pre-allocate memory to programs at launch, didn't you? It was a fine OS design for 1992. It didn't work so well in 2000. It was a weight around Apple's neck and would have killed them if they didn't try to escape. It needed to updated, and previous projects had failed. A clean break was a very smart decision.
This is somewhat true, (quite on the laptop side later in life with the G4s), but it's also highly troll. "...in order to obtain the hardware-locked user experience of their new flagship operating system"? That's unnecessary.
It's not like anyone had ever thought of that before. If only Windows had a virtual environment in it. Maybe since 95. It could have run old DOS programs. Oh, wait, it did. Then there was WoW, Windows on Windows, that let 95 and up run old Win16 programs. Emulating older stuff is a common way of handling it.
Comment forecast: Bits of genius surrounded by a sea of mediocrity.
What Microsoft is doing here is a bold move. We all benefit if it pays off with an improved product.
We can sit and arm chair direct Microsoft in to all sorts of fun things, but why bother when we could just pick up some free software codebase and do better for ourselves? Hopefully hardware makers will start thinking like this rather than going down whatever SDK path Microsoft tries to sell them next.
With this announcement of total backwards break, Microsoft has declared complete defeat for their business model. It would be nicer if they would fly the white flag and be good sports about it. The free software community will welcome them if they just GPL their code and act nice. Hell, XP would survive longer than 2010 if they GPL'd it because the community could really make what they want. They don't seem ready to do that, so they can sink for all I care.
I read the internet for the articles.
The Kai's Semi-Updated Website Thingy
As reasonable as that sounds, it's not going to happen. Microsoft wants to *sell* you new OSs every few years, and letting you use the same OS for a decade, no matter how well it works, just doesn't make them enough (or any, really) money. Unless they do a subscription type service, which they have said that they're looking to do ...
Rather than making offensive rants and accusations, why not put your energy into learning how to write code, and let that do the talking?
How does a comment get tagged "Insightful" when it's completely obvious the person didn't read the article? It clearly states that there will be an emulation layer/virtual machine for supporting legacy applications. And to all the trolls that jumped on that bandwagon in response: You can have your opinion, but please make it an informed one!
With this announcement of total backwards break, Microsoft has declared complete defeat for their business model. It would be nicer if they would fly the white flag and be good sports about it. The free software community will welcome them if they just GPL their code and act nice.
Yeah they could just GPL the code and become a company for mostly just supporting their products. It's just that Microsoft seems to notoriously suck at support, so they won't be making any dime with that one. Seems it's really either fly or die for them.Power corrupts the few, while weakness corrupts the many.
With this announcement of total backwards break, Microsoft has declared complete defeat for their business model.
I don't think that announcing breaking backwards compatibility is declaring defeat for a business model. It is more a cleansing process. And I welcome that. A lot of the hardware and software we use could be a lot more efficient and, quite possibly faster, if backwards compatibility were dropped.We're to the point now where processors are fast enough now to handle VM's. Let VM's handle the backwards compatibility, translating old code for newer uP/uC code.
I, too, would like to see Microsoft's practices of messing with their user base to satisfy their customer base stopped. But for the sake of competition, I don't think Microsoft sinking is a good option, either.
(I would also like to say it's the year of the penguin, and signs are showing that people are fleeing MS Windows... they just also happen to be fleeing the WIntel world, too, towards Macintosh.
It's so funny how for over 20 years now every single time a new version of Windows comes out it's a huge disappointment relative to what it had supposedly going to be/have. Then a year after, like clockwork, we start hearing about how the next Windows is going to be so unbelievably awesome it's going to be an almost incomprehensible revolution in computing technology.
See also Sony Playstation for another example of the same "marketing strategy".
Apparently you didn't RTFA in its entirety. How does M$ plan to handle the backward-compatibility issue? by including a Virtual Machine to run all your legacy apps... exactly what Apple did with "Classic" for OSX.
This is exactly what I've been suggesting for some time now -- a modular version of Windows (consisting of core OS, drivers, networking, and a basic browser suitable for downloading a better browser with) where I can install as much or as little of it as I wish, and a VM to run my old shit that won't work with this new modular Windows.
Also, it's a great razor-and-blades marketing opportunity for M$: make the core OS cheap or even free, and charge for various levels of "Plus Packs" suitable for people who WANT a monolithic software experience.
The big OEMs can make hay from that too -- basic machines with the core OS only would be cheap, while "complete solutions" (with all the Plus Packs) would be proportionally more expensive. And I'm sure the OEMs could make a good enough deal with M$ for bulk licenses that they could make a hefty profit -- exactly as they do now with preinstalled software.
If M$ were to include VMs for both WinXP and Win98-atop-DOS, everything would be covered, including old games (maybe even DOS games!), old apps, old installers, old drivers...
Also, there is some security imposed by running potentially vulnerable OSs/apps in a VM, if only because it's harder for malware to reach. A few malicious apps can "jump across" into a VM, but most can't.
Also, at a guess the new core OS will be more UNIX-like or even *NIX-based, which ought to make y'all happy.... after all hasn't "*NIX is better" been the mantra around here since forever??
~REZ~ #43301. Who'd fake being me anyway?
This is just like Microsoft's ill fated Cairo OS. It will never happen. The only way I'll believe that MS will actually succeed in creating a successful OS is if they throw out their old OS completely and start again from scratch. This is exactly what Apple did, and it led to an extremely stable and secure new system. The legacy systems can be supported by some sort of VM, again, just like Apple did when it went from OS9 to OSX. The future increases in computing power will negate any drops in performance in legacy programs.
This and no other is the root from which a tyrant springs; when first he appears as a protector - Plato (423 to 327 BC)
Frankly, if they DO toss out all the old Win code and start over, that'll be the smartest thing they've ever done, but it'll be eight years too late and again, waaay behind Apple!
Since Apple isn't taking huge chunks of market share away from Microsoft, I don't believe it's too late to do anything. That's what's great about being a near monopoly, you can take your time and drag your feet.
In a few years, laptops in bubble-packs for $89.95 will be hanging on racks at the drugstore. Microsoft isn't ready for that.
Oh, it's worse than that. In a few years Apple could be selling a cheap iPhone for $150 that's more than twice as powerful as today's model. It'll probably support an external monitor and wireless keyboard, via a little docking cradle. It'll have 160+ GB of internal storage, and the ability to connect to your network storage, at home or at work.
So why buy a "PC" at all? If you're a company, just outfit your employees with iPhones and wireless headsets. They dock them when at their desks, use a keyboard and full-sized monitor, then take their relevant data (and their work) with them wherever they go. If you're a consumer, why buy a "PC" when you can just use your phone? Apple could sell portable docking cradles with a built-in monitor, keyboard and a big battery. Instant "laptop".
The future belongs to companies that control both the hardware and the software. It'll be the only way to have the kind of product control it'll take to lure consumers and business to your platform, and it's also gonna be the only way to earn the margins it'll take to survive. I don't want to say MS is doomed, but I think they're gonna be pushed into the server and application space by this development. Sorta like what happened to IBM over the course of the PC revolution.
Why throw out the old code? This is about the worst thing Microsoft could do. Rewriting code is just an invitation to introduce all the bugs that have already been fixed and to introduce a new set of bugs.
Far far better to refactor particularly bad code and restructure at higher levels. Takes less time, advances the product, and has a far better chance of actually being completed.
I am not a script!
It opens the door for alternatives like Linux to suddenly seem a lot more attractive. Remember, we're talking 2010 (2012 at best, realistically). That's 4 years. Ubuntu/Gnome/KDE/etc. will be a lot more polished by then. New technologies may conquer the desktop in the meantime (Adobe for instance is moving in strongly with AIR, Mozilla with Weave and XULRunner etc.)
I'm afraid that this move comes late. Vista is a detour that should have never been taken. It ate precious time at a very sensitive moment in IT evolution. It may come to be remembered as the second Windows ME (although arguably doesn't have the stability issues that plagued ME).
i ate crayons when i was a kid and now i have two braincells and the blue ones taste nicer
Linux is apparently immune from such criticism? Linux's total lack of an integrated media player, must be awfully subtle for it to merely be "apparent." A Toyota Corolla apparently doesn't have 7 wheels (but we're not quite sure, huh?).
Just how many thousands of libraries does the average application load? If you can actually perceive this load time on modern hardware, it must be an awful lot. And I guess they haven't learned the trick of .. oh, I don't know .. leaving libraries in memory until there's a memory crunch. Is this guy running Vista on 386SX with only 2 megabytes of RAM and a hopelessly fragmented 40ms drive?
Actually, I think the anti-Microsoft naysayers will say, "It's about time; you're only a decade or three behind the common everyday practices of every other computer programmer in the history of civilization."
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
The whole reason why I have stuck with Microsoft for this long (as well as many other people), is that apps I use aren't compatible with other OS's. If I could have iTunes for Linux, my wife would let me switch at home. Why doesn't Apple provide it? Because Linux doesn't have the marketshare. Why doesn't it have the marketshare? Because there aren't enough of everyone's favorite apps.
How much of the corporate reluctance to migrate to Vista is because of incompatibility with current apps? Some people are still running Windows 2000 to support old apps that were never updated to be compatible with XP, muchless Vista.
I understand that MS would have reasons to want to "cleanse" itself, but doing so would make them lose the one major advantage they have over Linux. If software companies have to re-write every app to work with Win7, why even bother with it? Who would use Win7, since all the apps are broken? Why not just write for Linux or Mac? The Apple market may always stay relatively small because of the price and the limited number of PC configurations, but Linux doesn't have either of those issues.
Linux has been in a tough spot for years because its marketshare is tiny next to Windows. But with no functional applications, Win7 would be starting over on marketshare, with no good reasons for anyone to buy into the new OS. Apple was able to start over with OS X because there was a relatively small number of users, who are fiercely loyal, and the change enabled them to get more users. I don't think MS can risk pissing off 90%+ of all computer users. Their biggest problem is that they could lose users, and breaking backwards compatibility can only increase the probability.
I'm sure they'll have some type of virtualization-enabled "Classic Mode", but you can do that from other operating systems as well, and if we have 2 years to prepare for it, Apple and the Linux community can have solutions that are just as elegant (or more so) than what Microsoft will cobble together, because whatever solution MS provides will most likely be an afterthought, since it's just a stop-gap solution until all the developers move over to Win7... if they ever do.
How come other every other OS vendor can build on a perfectly good codebase (Unix) and not wind up with a fragmented, unsupportable mess which requires tricks like WinSxS?
OK, it sucks to be an application vendor and have to recompile your application when the new OS comes out - but testing under the new OS has been necessary with every new OS release on every platform in the whole of history.
Will they really do what Apple did and help themselves to a new round of BSD injections or are they just going to shuffle their own cards into new piles?
God, I hope not. The world doesn't need another UNIX.
Please don't try to blame those "lazy" "third party" developers again.
What ? Upwards of 90% of Windows's "problems" are directly attributable to third party code.
The only explanations for Vista's lack of backward compatibility are incompetence or malice.
There are few products that have better backwards compatibility than Vista.
You have to be off your rocker if you think that Microsoft does not view the ability to run legacy applications as a competitive threat.
The level of delusion necessary to look at Microsoft's history and come to this conclusion is truly staggering.