[freakboy2]

So what's the guess on this? Will virtual PC finally be really useable?

My bet is that it's going to be so fast that nobody would ever want to dual-boot their intel macs.. instead just run a VPC emulator.

is this the main reason why apple went to intel? will the argument of "not being able to run windows apps" finally be hard to make?

[osxrules]

We won't really have PC emulators any more. Emulator's are only called that because they make the software running under it think they are on different hardware. On X86 hardware, there is no need for emulation of the X86 CPU. The Virtual PC versions for Mac and Windows are vastly different. The Windows version is just a Virtual Machine like VMWare.

Essentially we would have something equivalent to what Java and X11 are in OS X. There is still a performance hit but the software should work much faster than under an emulator. I think the main difference will be native graphics card support. This allows Linux users to play some Windows games - sometimes faster than under Windows - natively under WINE.

I honestly think it could make Mac games firms like MacPlay or Aspyr obsolete if it works well enough. We'll have to wait and see how well the transition happens and it won't be for a while yet but I think this is one of the best moves on Apple's part.

[Maflynn]

I think with Apple going over to the intel chipset, there will no longer be a virtual pc. Since Apple already stated they won't stop people from putting windows on the apple hardware, there will be little need for VPC. The market for such an application is small as it is, the intel move just shrinks it to a degree that MS will stop development imho

[mduell]

Who needs emulators or dual booting when you have Vanderpool?

[wataru]

Originally Posted by Maflynn

I think with Apple going over to the intel chipset, there will no longer be a virtual pc. Since Apple already stated they won't stop people from putting windows on the apple hardware, there will be little need for VPC. The market for such an application is small as it is, the intel move just shrinks it to a degree that MS will stop development imho

Except that you're forgetting the fact that Microsoft also makes Virtual PC for Windows. Why? Because virtualization can be useful for software testing and other reasons. There will still probably be a Virtual PC, but it won't be emulating Intel hardware anymore.

[Maflynn]

Originally Posted by wataru

Except that you're forgetting the fact that Microsoft also makes Virtual PC for Windows.

I agree, its just that your not going to see VPC for the mac. They bought connectix for the virtualization software, not to provide mac users the ability to run windows. They'd sooner drop that one facet, now they have that chance.

[Catfish_Man]

Originally Posted by Maflynn

I agree, its just that your not going to see VPC for the mac. They bought connectix for the virtualization software, not to provide mac users the ability to run windows. They'd sooner drop that one facet, now they have that chance.

Why would they object to more people buying Windows?

[kupan787]

Originally Posted by Maflynn

I agree, its just that your not going to see VPC for the mac. They bought connectix for the virtualization software, not to provide mac users the ability to run windows. They'd sooner drop that one facet, now they have that chance.

Why? Microsoft is in the business of making money. If they can sell Windows to 100% of the market, why wouldn't they? Are they some how hurt when a copy of Windows is installed on a Mac vs on a PC? They would much rather have every mac dual booting into Windows than no mac booting into Windows.

[dazzla]

Perhaps you've not tried VPC on Windows but it's not exactly a speed freak. Admittedly it's slightly faster then running on OS X but it's nowhere near native speeds.

[Tomchu]

VMware on Windows XP feels more or less like 75% native speed, and that's only due to the fact that disk performance is poor (NTFS host drive hosting a disk image hosting a slave NTFS partition, two layers of file system to go through, one layer of virtualization).

[Maflynn]

Originally Posted by kupan787

Why? Microsoft is in the business of making money. If they can sell Windows to 100% of the market, why wouldn't they?

Indeed they are here to make money. VPC isn't free, they need to have a support structure. If they can convince people to buy windows for the intel macs instead of purchasing VPC don't you think they will? VPC is such a small segment, its probably not a very lucrative line.

Originally Posted by kupan787

They would much rather have every mac dual booting into Windows than no mac booting into Windows.

Exactly - that's my point. Why spend more resources maintaining VPC when they can have mac users dual boot.

Originally Posted by Catfish_Man

Why would they object to more people buying Windows?

They don't object but they also want to maximize their profit margins and if they can reduce the overhead (such as the support team needed for VPC) and still sell the windows to intel macs - all the better - Yes, No, Maybe?

Mike

[JLL]

Originally Posted by Maflynn

Indeed they are here to make money. VPC isn't free, they need to have a support structure. If they can convince people to buy windows for the intel macs instead of purchasing VPC don't you think they will?

That requires that Windows can run unmodified on Intel based Macs.

The Macs don't have to live up to some PC standard and perhaps it's not cheaper to add Mac support to Windows than it is to maintain VPC.

Originally Posted by Maflynn

Exactly - that's my point. Why spend more resources maintaining VPC when they can have mac users dual boot.

And you want to reboot every time you need to switch between Mac and Windows apps?

[chefpastry]

I would really like to see Windows apps running in OS X under WINE.

[himself]

Originally Posted by chefpastry

I would really like to see Windows apps running in OS X under WINE.

exactly. whether MS does it or not, someone will make a VM for windows on OS X.

[Maflynn]

Originally Posted by JLL

That requires that Windows can run unmodified on Intel based Macs.

The Macs don't have to live up to some PC standard and perhaps it's not cheaper to add Mac support to Windows than it is to maintain VPC.

Your right and at this point, product availability and performance is all conjecture. My "guess' is that microsoft wants to turn Mac users into windows users. It will be a lot easier for them to do that if they need to boot into windows rather then using VPC. If the architecture is different then that's a whole new ball of wax and since the mactel aren't out yet we will have to wait and see.

Originally Posted by JLL

And you want to reboot every time you need to switch between Mac and Windows apps?

Do you think microsoft really cares if you the mac owner has to reboot out of OSX and into windows? In fact they probably would welcome that because you may be more inclined to stay in windows after using the app that you needed windows for. There desire is not to satisfy you the Mac owner but sell more copies of windows, office, and other MS apps, most of which are not available on OSX. The more people using windows the more money they make. Emulations are typically slow so people are generally less inclined to get software for that unless absolutley necessary where as if its running natively they (MS) may sell more copies of project, or visio or what ever then if it was running under VPC.


Mike

[Person Man]

Originally Posted by himself

exactly. whether MS does it or not, someone will make a VM for windows on OS X.

Wine is not a VM.

[power142]

Many Mac users have Office for Mac. Those that buy Virtual PC usually need some kind of Windows to run on it. That's opening your wallet a sum total of three times. Why would Microsoft want to change that?

Also, from their point of view, the core emulation in Virtual PC becomes virtualization, which is a whole lot simpler when you already have a product that does the same thing

[JLL]

Originally Posted by Maflynn

Do you think microsoft really cares if you the mac owner has to reboot out of OSX and into windows?

Since VPC makes it much easier for the end user (no reboots), MS might sell more copies to Mac users, so yes, they care.

[goMac]

Originally Posted by Maflynn

Exactly - that's my point. Why spend more resources maintaining VPC when they can have mac users dual boot.

Why spend the time writing Virtual PC for Windows when Windows users can dual boot?

[Person Man]

Originally Posted by goMac

Why spend the time writing Virtual PC for Windows when Windows users can dual boot?

Would you please stop stating it as fact until the real machines ship? We don't know yet whether the machines can readily dual boot until they actually ship.

Besides, even if they can dual boot, there are plenty of advantages to not having to quit out of OS X to run Windows. Why spend time writing Virtual PC for Windows on Windows machines when you can boot Windows natively. Microsoft has a version of Virtual PC on Windows too, not just Mac OS X.

Concrete example: Web developers who want to make sure their page looks properly in both Safari and Windows XP with IE 6. You can do it from one machine with Virtual PC. Dual booting would be a PITA if they were making several changes and had to switch from one OS to the other.

[Tesseract]

WINE is a compatibility layer which allows Windows apps to run on a non-Windows machine (typically Linux), alongside native apps, as long as that machine has an x86 processor. WINE is already ported to Mac OS, the main limitation now it is that it can only run specially compiled (for PowerPC) Windows programs. It's very likely that WINE will be usable on x86 Macs.

WINE is not an emulator (hence the name), it's not dual-booting, and it's not a virtualizer like VMware. It's basically a recreation of the Windows API and thus allows Windows apps to run without any need to buy, install, or otherwise rely on Windows itself. Since it's not an emulator, it's really fast - probably approaching 100% of native speed. (On my [old, slow] Linux box, I used IE 6 under WINE as my primary browser for a while because it was faster than Firefox.)

I predict that WINE and (possible) dual-booting will make Virtual PC and its ilk obsolete on x86 Macs.

[Person Man]

Originally Posted by Tesseract

I predict that WINE and (possible) dual-booting will make Virtual PC and its ilk obsolete on x86 Macs.

Nope.

WINE will not make Virual PC (or other VM type software) obsolete because it may never be completely compatible with the Windows API. There will always be differences between it and Windows.

Dual-booting won't make Virtual PC obsolete for the reasons I mentioned above (i.e. in a production environment where you develop things to run on both platforms... you don't want WINE, and you don't want to boot into each environment separately, either).

[JLL]

Originally Posted by Person Man

Would you please stop stating it as fact until the real machines ship? We don't know yet whether the machines can readily dual boot until they actually ship.

Read the sentence again

[the_13th_saint]

So if we happened to give Windows complete access to our hardrives, and have it run on the same hardware using the same processor and ram, I'd be more worried about security than speed.

I'd go for a more virtual pc approach. give windows a certain amount of hardrive space, ram and processor cycles all regulated by the kernal.

if a virus affects windows, the integrity of the computer will not be affected. There could even be processes that monitor windows for unacceptable behavior like sending massive amounts of email, DoSing, crashes, etc. and restrict it if something isn't going correctly. The information on the macintosh side would not be able to be affected by virii by using only copies of the files that exist on the macintosh hardrive. The windows hardrive portion would also be stored in an expandable disc image, preventing actual boot-up of windows on the computer, thus preventing further compromise of the data security.
-13th

[cambro]

Is anybody else here a little worried about the very easy and fast running of Windows/Windows Apps on OS X?

Why would developers keep making OS X NATIVE applications when they can invest $$ in Windows development only and say to the OS X community, "hey, the windows version runs fine on OS X?"

[Person Man]

Originally Posted by cambro

Is anybody else here a little worried about the very easy and fast running of Windows/Windows Apps on OS X?

Why would developers keep making OS X NATIVE applications when they can invest $$ in Windows development only and say to the OS X community, "hey, the windows version runs fine on OS X?"

You'll have more people DEMANDING that developers make native OS X applications.

Not to mention the rumors that Apple is going to reintroduce "Yellow Box for Windows."

I'm sure Apple has thought of that before deciding. I really don't think it's going to be easy to run Windows on Apple hardware. They said they weren't going to actively prevent people from trying. That's all.

[Chuckit]

Originally Posted by cambro

Is anybody else here a little worried about the very easy and fast running of Windows/Windows Apps on OS X?

Why would developers keep making OS X NATIVE applications when they can invest $$ in Windows development only and say to the OS X community, "hey, the windows version runs fine on OS X?"

Those would be the same ones who tell people to run it in VPC now.

[besson3c]

Possibilities:

- Machine virtualization via Xen
- VMWare
- WINE
- Virtual PC
- Hardware based machine virtualization
- other OSS solutions (PearPC, etc.)

I've been noticing that the development of WINE has been extremely rapid lately, quickly approaching a version 1. My prediction is that Apple will retrofit one of these Open Source solutions to work as a Yellow Box (WINE seems like an ideal solution since no Windows purchase would be necessary, if they can get it so that it works reliably). They'll make it so it is about as good as Classic - good enough, but not good enough that customers don't stop complaining for a company to produce a native Mac experience with all the Mac bells and whistles.

[Angus_D]

Originally Posted by Person Man

Not to mention the rumors that Apple is going to reintroduce "Yellow Box for Windows."

Read Wil Shipley's take on that.

[besson3c]

Possibilities:

- Machine virtualization via Xen
- VMWare
- WINE
- Virtual PC
- Hardware based machine virtualization
- other OSS solutions (PearPC, etc.)

I've been noticing that the development of WINE has been extremely rapid lately, quickly approaching a version 1. My prediction is that Apple will retrofit one of these Open Source solutions to work as a Yellow Box (WINE seems like an ideal solution since no Windows purchase would be necessary, if they can get it so that it works reliably). They'll make it so it is about as good as Classic - good enough, but not good enough that customers don't stop complaining for a company to produce a native Mac experience with all the Mac bells and whistles.

[goMac]

Originally Posted by Person Man

Would you please stop stating it as fact until the real machines ship? We don't know yet whether the machines can readily dual boot until they actually ship.

Besides, even if they can dual boot, there are plenty of advantages to not having to quit out of OS X to run Windows. Why spend time writing Virtual PC for Windows on Windows machines when you can boot Windows natively. Microsoft has a version of Virtual PC on Windows too, not just Mac OS X.

Concrete example: Web developers who want to make sure their page looks properly in both Safari and Windows XP with IE 6. You can do it from one machine with Virtual PC. Dual booting would be a PITA if they were making several changes and had to switch from one OS to the other.

I don't think you read the comment you replied to very well.

The argument was if Mac users can dual boot, why would Microsoft continue to write Virtual PC for Mac? My reply was Windows users always have been able to dual boot, but Microsoft continues writing Virtual PC for Windows.

[goMac]

Originally Posted by Angus_D

Read Wil Shipley's take on that.

I don't think Wil Shipley gets the idea (not to mention like a lot of others he thinks Dharma is meant to be a Lost reference. If you know anything about world religions you're more likely to think the name at the bottom was a reference to Dharma as Dharma is a wonderfully good name for a Yellow Box type thing.) He talks about people converting from CFM to Cocoa in order to be compatible with Yellow Box. That wouldn't be the point of Yellow Box. The point of Yellow Box wouldn't to be convert Mac applications to run on Windows. It would be to convert Windows developers to Cocoa, and get Windows developers writing Windows and Mac compatible binaries. I don't care who you are, that's good no matter what viewpoint you look at it from. Windows developers get a great framework for creating applications, and get for-free Mac ports. The Mac users get more applications. People on the Windows side can't run anything not built in Yellow Box, such as iLife, Final Cut Pro, or any other big Mac apps. The Mac side doesn't loose any of it's unique apps.

Personally, I do know some very interesting things that are under NDA, nothing having to do specifically with Yellow Box, but certain things in the rumor do check out. The move to x86 is part of a larger change of direction on Apple's part. If Apple wanted a faster chip they could have gone with AMD or stuck with the PowerPC and used Motorola's coming mobile chips, but that wasn't the point of this move. Steve Jobs has a lot of personal interest in all of this, including personal interest in the Yellow Box. Don't forget, the Yellow Box was Steve's idea in the first place, I wouldn't think he'd be against it in any way.

Interestingly I learned a few things that lead me to predict the return of the yellow box probably 6 months ago. The Dharma article probably does have a few things wrong. I doubt we'll see any trace of the Yellow Box on Monday. But I would expect that with the move to Intel we're going to start seeing a lot of things that were part of the original Rhapsody plan that were dropped when Steve was convinced to stay with the PowerPC.

[msuper69]

Perhaps, there won't be any need to dual boot. The new Intel chips support hardware virtualization which should allow one to hot-key between concurrently running operating systems, for instance, OS X, Windows or Linux.

[paulc]

Well, running something like WINE may be problematic; as in if one has a "constantly running" winblowz API translator, then maybe the machine would be suspectable to all thew winblowz malware. That would be one thing I'd be concerned about. Don't get me wrong, I have no specific knowledge about this, it's just speculation on my part.

Can one turn WINE off and on dynamically? That would lessen the chance for malware.

Let's step back and look at exactly what we want... to me it's to run the very, very occasional winblowz app, but mostly gaming. I'm positive the whole process of "port certain, but not all, A-List titles to Mac OS (not to mention timing)" is going to eventually die away. I'm sure that at a minimum, we'll be able to re-boot into that environment to play XXX. That is the big barrier (IMO).

Lastly, nobody has mentioned the whole "virtualization in Intel CPUs thing. I've read lots of marketing palaver that never even hints at what is going on, so I speculate that it's keeping 2 complete OSes separated in memory with the CPU support chips providing the support. Kinda like fast user switching, but with a whole OS & apps instead of separate users. This seems very exciting... I'd love to know more details...

[goMac]

Originally Posted by paulc

Well, running something like WINE may be problematic; as in if one has a "constantly running" winblowz API translator, then maybe the machine would be suspectable to all thew winblowz malware. That would be one thing I'd be concerned about. Don't get me wrong, I have no specific knowledge about this, it's just speculation on my part.

Darwine x86 doesn't run all the time. One copy will be attached to each Windows application you run, but if you're not running any Windows applications there won't be any instances of it on your machine. HOWEVER, it is possible to run a Windows malware application through it, as shown by WINE's susceptibility to the Windows metafile flaw.

[besson3c]

Originally Posted by paulc

Well, running something like WINE may be problematic; as in if one has a "constantly running" winblowz API translator, then maybe the machine would be suspectable to all thew winblowz malware. That would be one thing I'd be concerned about. Don't get me wrong, I have no specific knowledge about this, it's just speculation on my part.

Can one turn WINE off and on dynamically? That would lessen the chance for malware.

Let's step back and look at exactly what we want... to me it's to run the very, very occasional winblowz app, but mostly gaming. I'm positive the whole process of "port certain, but not all, A-List titles to Mac OS (not to mention timing)" is going to eventually die away. I'm sure that at a minimum, we'll be able to re-boot into that environment to play XXX. That is the big barrier (IMO).

Lastly, nobody has mentioned the whole "virtualization in Intel CPUs thing. I've read lots of marketing palaver that never even hints at what is going on, so I speculate that it's keeping 2 complete OSes separated in memory with the CPU support chips providing the support. Kinda like fast user switching, but with a whole OS & apps instead of separate users. This seems very exciting... I'd love to know more details...

WINE is not a daemon that constantly sits in the background, it is only active upon demand. Mac stuff would not be at risk, because stuff that is installed through WINE is sandboxed, meaning that it only writes to a specific directory, and cannot affect any stuff outside of this directory. WINE stuff does not interact with Mac stuff at all.

The most compelling reason for running Windows apps is not only gaming, but for businesses to be able to run specialized custom Windows apps (of which there are many). If a business could migrate to the Mac without any loss of business, and if it seemed like a financially viable route for them requiring a minimal of retraining, what's stopping them?

[besson3c]

Originally Posted by goMac

Darwine x86 doesn't run all the time. One copy will be attached to each Windows application you run, but if you're not running any Windows applications there won't be any instances of it on your machine. HOWEVER, it is possible to run a Windows malware application through it, as shown by WINE's susceptibility to the Windows metafile flaw.

It's possible to run Malware in Virtual PC too, but the question is, can it take down your entire machine, or is it nicely sandboxed? I believe the latter is true.

[goMac]

Originally Posted by besson3c

It's possible to run Malware in Virtual PC too, but the question is, can it take down your entire machine, or is it nicely sandboxed? I believe the latter is true.

It entirely depends. The WMF bug didn't do much to WINE because iirc the exploit attached to the metafile had to be Linux specific, and how often are you going to run into a Windows Metafile with a Windows exploit built for targeting Linux? The same should be true of Mac OS X.

I mean, we're not going to suddenly get ActiveX files in Safari. The only risk would be if an exe was downloaded in the background, but fortunately (cough cough) Safari already secures your machine against exe's being downloaded in the background.

[goMac]

Originally Posted by besson3c

The most compelling reason for running Windows apps is not only gaming, but for businesses to be able to run specialized custom Windows apps (of which there are many). If a business could migrate to the Mac without any loss of business, and if it seemed like a financially viable route for them requiring a minimal of retraining, what's stopping them?

That would also be part of Apple's thinking. Currently Darwine has a lot of trouble running DirectX apps. I think they just got it displaying some 3D graphics, but it took a lot of hacking. Also remember it still runs under x11 for the time being, which is something your average user doesn't have installed.

I wouldn't expect WINE would make a huge splash on Mac OS X until you get a commercial distribution. Until then it will just be used by the power users and maybe deployed by some IT departments.

[besson3c]

Originally Posted by goMac

It entirely depends. The WMF bug didn't do much to WINE because iirc the exploit attached to the metafile had to be Linux specific, and how often are you going to run into a Windows Metafile with a Windows exploit built for targeting Linux? The same should be true of Mac OS X.

I mean, we're not going to suddenly get ActiveX files in Safari. The only risk would be if an exe was downloaded in the background, but fortunately (cough cough) Safari already secures your machine against exe's being downloaded in the background.

Yes, but I don't think that stuff that runs in WINE has r/w access outside of the directory used for installing Windows apps. IOW, WINE knows of nothing outside of the WINE directory. I don't think that attaching Linux code to a Windows exploit would be able to do anything, especially without root access (WINE stuff runs under normal non-root users).

WINE exploits can go nuts on the Windows registry, but the Mac ignores the WINE Windows registry. WINE exploits can go nuts writing to the WINE/Windows directory, but they don't have access to anything outside of this directory. Even if they did, they would still need root access to do any real harm.

[besson3c]

Originally Posted by goMac

That would also be part of Apple's thinking. Currently Darwine has a lot of trouble running DirectX apps. I think they just got it displaying some 3D graphics, but it took a lot of hacking. Also remember it still runs under x11 for the time being, which is something your average user doesn't have installed.

I wouldn't expect WINE would make a huge splash on Mac OS X until you get a commercial distribution. Until then it will just be used by the power users and maybe deployed by some IT departments.

WINE has a hard time running under x86 BSD too (such as FreeBSD), because it was written primarily for Linux. However, there has been little point in getting it to work outside of Linux because until recently, WINE was alpha software. It still may be a moving target. If WINE won't even run in the x86 BSDs, it's no wonder it has a hard time running in OS X.

However, even considering the immaturity of WINE, there *still* has been a successful commercial distribution called Crossover Office that pulls out snapshots of WINE, and sets up an environment to be compatible with specific apps (including iTunes and Microsoft Office). Apple could do the same.

My experience with WINE is actually outdated, because WINE is no longer alpha software. It has been improving at an astonishing rate (I've been on their announcement list for a while). I honestly think that Apple could be secretly contributing to this project, as it seems like every few weeks there is a new release that boasts a significant number of improvements.

WINE will break under Vista, but there will still be a need to run Windows XP apps for many years to come.


I really do believe that WINE is the way to go, and somebody will jump aboard this bandwagon. It is a very elegant solution. For those who have never heard of it, it basically mimics Windows APIs, allowing Windows apps to run in an included bastardized sort of Windows. Since an entire operating is not preloaded, but simply the APIs needed to run apps, the performance hit is minimal compared to solutions that require loading Windows off of a disk image. Additionally, a copy of Windows isn't even necessary to be installed, all you need to do is install WINE.

I think the runner up solution will be Xen:

http://xensource.com/

It's like VMWare, except open source. I've heard that performance is very nice.

[goMac]

Originally Posted by besson3c

Yes, but I don't think that stuff that runs in WINE has r/w access outside of the directory used for installing Windows apps. IOW, WINE knows of nothing outside of the WINE directory.

I don't think that's true. I can run programs in WINE that are over on my Windows drive and not even in WINE's virtual folder. It can also see documents outside it's WINE directories and read/write them. But then again, it's going to hit OS X's permissions as a way of sandboxing it.

[goMac]

Originally Posted by besson3c

I really do believe that WINE is the way to go, and somebody will jump aboard this bandwagon. It is a very elegant solution. For those who have never heard of it, it basically mimics Windows APIs, allowing Windows apps to run in an included bastardized sort of Windows. Since an entire operating is not preloaded, but simply the APIs needed to run apps, the performance hit is minimal compared to solutions that require loading Windows off of a disk image. Additionally, a copy of Windows isn't even necessary to be installed, all you need to do is install WINE.

I think the runner up solution will be Xen:

http://www.cl.cam.ac.uk/Research/SRG/netos/xen/

It's like VMWare, except open source. I've heard that performance is very nice.

The problem with XEN is it requires you to have a modified version of the guest operating system, which means you can't just install Windows to run under it.

I also talked to the VMWare folks, and they said as of now they have no plans to bring VMWare to Mac OS X. I think it would be nice as I'd like to be able to run multiple versions of Mac OS X at the same time for testing purposes. So, if you'd like to see VMWare on OS X x86 throw them an email.