go ahead, mac my day

a Macintosh girl in a Microsoft world

the life and death of Virtual PC for Mac

the life and death of Virtual PC for Mac

  • Comments 40

Most of the other MacBU folks have been talking about VBA while I've been gone, and I don't think that there's anything that I can really add to that discussion. VBA is much less of an impact on the apps that I focus on, and some of our other WWDC announcements were more near and dear to my heart anyway. Let's instead talk about my favourite team at Microsoft: the Virtual PC team. (Dear Entourage, PowerPoint, and Remote Desktop Connection: Yes, I still love you guys, and no parent is supposed to have a favourite child, but the VPC guys give me brandy.)

The future of Virtual PC on the Mac had been in question for a year. The VPC team was happily working along on v8 (then code-named Oxygen), and an anvil dropped from the sky at the last WWDC. That anvil, of course, was the announcement of the move to Intel chips. VPC is an application that sits quite close to the metal. Any change in the operating system or the chip architecture has a huge impact on VPC. Throughout the history of VPC, any change to the OS had the potential to break VPC in new and interesting ways. Remember the introduction of the G5? It took the VPC guys quite awhile to get VPC to run on the G5. It wasn’t a lack of effort. It was the combination of needing very specific expertise (which is to say, we couldn’t just throw more bodies at the problem) and needing some assistance from the folks over at Apple.

So here comes the Intel chip, and Leopard too. VPC v8 would need the same move to Xcode that every other major Mac application has needed to make. On top of that effort (which is a huge effort, as any Mac developer on a big project can tell you), VPC would require a re-architecture of the bits of VPC that were PPC-specific. We could re-architect VPC v7, we could port code from VPC:Win, we could re-code it from the ground up, or some combination therein.

We said that bringing VPC to the MacTels would be like doing a v1. That’s true, but it’s not the whole story. It’s not just that VPC v8 would be like doing a v1. It’s that VPC v9 could also be just like doing a v1, or maybe it would be VPC v10. There’s a huge engineering effort involved in making a v1 product. But when would we be able to focus our engineering efforts on improving performance or adding features instead of having to update the existing code to work on the latest OS release? What happens if there’s another major chip change?

’But what about Parallels? What about VMWare?’ I hear you ask. Parallels has got a v1 out there right now. VMWare is about to enter beta on their v1. One of the great things about a v1 is that you don’t have expectations. Your feature set is determined by what you can get working. It’s not determined by what you had working before. I think it’s v2 where life gets interesting. Can you build upon what you have? Can you get more people using it? It’s the early adopters who jump on a v1, and they’re not a big market (although they are a vocal one). In some respects, you get an easier job on v2: you get to add features, you get to fix bugs, you get to tweak performance. You get to make v2 a better product for your users.

But v2 is generally where you pick up the average user. VPC:Mac already has the average user as a part of our user base. For the average VPC user (who isn’t a Mac expert, and definitely isn’t a Windows expert), imagine buying VPC v8 and having very few new features over v7. A savvy user is more willing to let that slide because they’re aware of the enormous engineering effort behind moving to the MacTels. The average user, who doesn’t know or care about the change in chip, is going to be upset.

We made a hard decision. It wasn’t undertaken lightly. The team wasn’t happy about the decision. Ultimately, MacBU made the decision that Mac users would be better served if we focused our resources on making the next versions of our other offerings as strong as possible. The decision to move away from developing v8 made sense from a development and customer perspective, even if it was a hard decision to make. We spent months trying to come up with alternatives that made sense. While we were working through it, including working on the codebase, we gave the MacTel version of VPC its own code name: Lanai, for the place that we'd all like to go on holiday. (Roz wouldn't let us move our operations there, although we did ask.)

So where is the Virtual PC for Mac team? We got a lot of people from Connectix when we bought out VPC three years ago, after all. One of them moved to Redmond to become the General Program Manager of the MacBU team there. One of them is the Development Manager here in SVC. Another one is a Development Lead for Entourage. The PowerPoint tester in the office next to mine was a VPC tester in the Connectix days, and one of the other VPC testers just moved to the PPT team as well. Recently, with the death of VPC, the several remaining team members in dev, test, and program management have formed a new team at SVC to focus on some of the code that is shared across all of our apps. VPC:Mac might be dead, but it lives on in the great people that we have from that team who are contributing to the rest of our apps.

  • Thanks for the great post Nadyne.  There's no doubt that the loss of VPC has ruffled a few feathers with some of the Mac users out here.  Having been a VPC tech support lead, I'll miss not getting to see what was going to be new in v8, and wish v7 a "fair winds and following seas" as it sails off into the sunset.

  • Super write up Nadyne, did you get to meet his Steveness? Do tell us poor souls who couldn't make it to the WWDC how it was :)
  • Sorry, the WWDC NDA prevents me from answering any questions about anything that wasn't addressed at the keynote.  ;)
  • :P
  • What I don't understand is that apparently Apple can port their entire OS and application suite (maintaining backwards portability), create an translation layer for old apps do much much more while they are at it, whilst mighty Microsoft finds it too much work (and hasn't up till now) to port a handful off apps? Isn't there a huge gap in ambitioin and talent there?
  • It is sad to see the end of a great product, but given the choice between VPC and the rest of the products produced by the MacBU ...

    I'm glad of the decision made. Porting the other MacBU concerns to MacTel must be a significant effort in and of itself!
  • Microsoft's termination of VPC (for Mac) is actually one of the more reasonable decisions by Microsoft and I as a long time user have absolutely no problem with this. Of course my being sanguine about this is helped by the fact there are superior alternatives now available.

    However there are plenty of other areas where Microsoft have, do, and will I am sure continue to make really bad decisions (from a Mac perspective and in some cases even from a Microsoft perspective).

    1. MSN Messenger / Microsoft Messenger / Live Messenger / [Insert this weeks new name] on the Mac continues to lack Video and Voice capabilities. Microsoft have announced that they are working on Messenger 6.1 for Mac and I have 100% confidence that 6.1 will not change this. This is despite the fact that this feature has been the number one request by TENS OF THOUSANDS of people for years and years. Microsoft should realise the one of the reasons Skype has been so successful is that unlike Microsoft they fully support multiple platforms including of course the Mac. This means with Skype Mac users can now do Video (as well as voice) and even have a selection of USB handsets to choose from.

    2. Microsoft have discontinued WMP for Mac. Ok fine, good riddance it was awful anyway. However while they have given their blessing to Flip4Mac as a replacement they have not provided Flip4Mac the means to add support for WMV and WMA DRM (Microsoft's statement that Windows DRM is an open standard is shown as being the flasehood it is, at least Apple don't claim their Fairplay DRM is an open standard). This means Mac users cannot (even if they were foolish enough to want to) access Windows DRM protected media. This in turn harms Microsoft because there is then no chance on earth of a Mac user being converted to using a Zune player for example [Ha, ha!].

    3. The death of VBA in Office for Mac. VBA support is apart from the brand name, probably the single major reason for Mac business users buying Microsoft Office for Mac as it has historically been the only way to ensure a fairly good level of compatibility with Office for Windows. Without VBA support then you are limited to plain vanilla .doc and .xls files which can already be used in a large number of alternative (and often free) programs. Now I understand the difficulties that porting VBA to XCode and Intel Macs may represent (but unlike Microsoft I don't believe it would be impossible, one option would be to re-write the Mac VBA code from scratch rather than porting the current ancient code which I realise is itself not a simple task). Now just discontinuing VBA for Mac support is bad enough, but Microsoft have completely failed to indicate _IF_ (never mind when) they would replace it with the new VBA replacement that Office for Windows will also need to adopt in the future (apparently VBA for Office for Windows is equally hard to port to 64bit processors).

    On top of this is the fact that the only totally new software product from Microsoft for Mac for possibly the last ten years is Remote Desktop Client (Entourage is little more than a resurrected Outlook Express), while during the same period product after product after product has been discontinued. Halo does not count as a new software product as it was effectively an existing Bungie product, and I also disallow mere upgrades.

    Where is Flight Simulator for the Mac? There is a humungous pent up demand for that. (I would think easily in excess of 6 figures and quite possibly 7.)

    Encarta - RIP
    AutoRoute - RIP
    MS Flight Simulator - RIP, gone but not forgotten
    MS Money - MIA
    Windows Media Player - KIA
    VPC - Shot as a deserter (from PowerPC)
    Internet Explorer - Mercy Killing
    Office VBA - Committed Suicide

    One could add...

    Microsoft Messenger - Persistent Vegetive State

    ...due to Microsoft wilfully refusing to add Video and Voice capabilities.

    Microsoft's efforts can be summed up by paraphrasing Sir Winston Churchill

    ...Never in the field of software development, has so little been produced by so many for so long.
  • Ben - I think there are two things working in Apple's favour here: time and resources.  Apple ported a whole OS over to the Intel chip, but as they said when they announced the transition, they had been compiling OS X for on Intel since 10.0.  They had five years of work that they were putting into it without anyone knowing.  The rest of us had to scramble when the announcement was made to take this into account in our product plans.  Across all Mac apps (not just ours, think of Adobe and everyone else too), how many dev-years were added unexpectedly because Apple made this decision?

    Likewise, Apple has a few more people writing Mac apps than we do.  The whole of MacBU is ~180 employees.  How big is Apple?  A quick websearch puts that number at ~16,000.  Discounting the small corner of the iTunes team that produces the Windows version, they're all working on Mac apps and hardware.  
  • Great post Nadyne.  It's good to hear the story from someone directly involved in the product.
  • Excellent post.

    If I was someone developing on Macs I would have given up on them ages ago, isn't this the 3rd time they've change CPU architectures?

    I wonder though, if Connectix had not been acquired by MS, would *they* have ported it to MacTel?

    I think the Mac (mainly due to OSX and now running on intel cpus) is a platform where you guys  can't afford not to be present. Think about it.
  • wpSlider - That's an interesting question, and one that I've thought about.  I'm not a part of the original Connectix team, so I can't speculate.  But even without my speculation, one important thing to point out is that Connectix was a pretty small company with a pretty small portfolio of Mac products.  For them, the resource question would have been a different one.  The engineering effort involved with doing the MacTel port would have been the same, of course, but if there's not something else that really needs your attention at the same time, there's a much smaller opportunity cost involved in spending the time working on this instead of working on that.
  • I can't speak for the efforts of the Connectix team regarding VPC, but I sure do applaud their efforts with RAMDoubler. I used that routinely through the 90s with ZERO issues on all my PowerBooks -- and running Photoshop, Illustrator, etc. I think '8' was their finest effort. Gone, but certainly not forgotten.
  • Play Station Emulator wasn't bad either - very missed - sure new Intel Macs could do PS2 now or something else.... , maybe even classic emulator for Intel Mac's - not had time to  try that sheep saver though.

    RAMDoubler was invaluable in the early days, my Mac experience would not have been the same without it.

    Yah I know Sony would get in the way - early days we just didn't give a %^&& - it was war back then, the frontlines...
  • http://homepage.mac.com/asam/.Pictures/marvin/borg-desk.jpg


    A long time ago :/
  • VPC:Mac for Intel processors didn't make sense anyway. VPC is an emulator, "porting" it to Intel would have meant getting rid of the emulation part and replacing it with virtualization. Now, of course such "search and replace" isn't easy (or possible) in this case. VPC, with its drag and drop ease of use, was a wonderful product, and I don't doubt the new virtualization software for MacTel will equal it in time.

    The real question remains : if VPC only made sense for PPC Macs (which are still the vast majority out there, if I'm not mistaken), why was development on version 8 axed, and what will happen to the product anyway (I guess Microsoft is not going to give it away for free...) ?

    Thanks for posting insider information.
Page 1 of 3 (40 items) 123