Google Grapes Aren't Microsoft Grapefruit
|
News Analysis. It's long past time to set the record straight on Office streaming. |
I refrained from blogging this topic, because there's not much new there yet. But bloggers and journalists, and even some analysts, can't seem to let go of wildly, ridiculous speculation about so-called streamed Office. So, to be absolutely, unequivocally, clear:
- Microsoft is not going into the hosted Office business.
- So-called Office streaming is not a competitive response to Google Apps.
- There will be little different with so-called Office streaming than what's available now.
- So-called Office streaming will have insignificant impact on Google Apps adoption.
Bits Streamed Centrally, Virtualized Locally
On Tuesday, I spoke with Gavriella Schuster, Windows Product Group senior director, about so-called Office streaming. The interview was specifically about Windows management tools announced on Wednesday. But one of those tools, Microsoft Application Virtualization 4.5, will be used to deliver so-called Office streaming.
Based on the conversation, I would say that "streamed" or "streaming" are really misnomers. The terms don't accurately describe what Microsoft will deliver (and already does). Microsoft uses "application virtualization" in the product name for a reason. The client application is essentially virtualization locally, but it's managed (including licensing) centrally.
The capability isn't new, either. Enterprises already can manage Office from servers and send out bits to desktops; it's all a matter of licensing. Stepping back before Microsoft's acquisition of SoftGrid, there were thin clients (and, of course, still are).
The application client licensing terms, and subscription enforcement through MAV, are the bigger differentiators here. For example, Microsoft will allow third parties to deliver bits from servers to business client desktops. Enterprises can do that now using MDOP (Microsoft Desktop Optimization Pack), of which MAV is a part (Microsoft rebranded SoftGrid to MAV in November). There are plenty of businesses that want to centrally manage applications such as Office, which is the real reason Microsoft offers the capability, like it supported thin clients with earlier Windows releases. There are IT cost, client-management and licensing benefits to the approach.
Businesses adopting MAV-delivered Office won't be running a hosted service like Google Apps. The products don't really compete, because of where the client resides and the aforementioned management and licensing benefits. MAV-delivered Office may be centrally managed, but code is distributed to Windows desktops. Google's hosted productivity suite provides anytime, anywhere access on anything that has a Web browser. The business benefits dramatically differ: MAV-delivered Office is more about managing costs and infrastructure, while Google Apps delivers the aforementioned anytime, anywhere access on most anything.
If Google Apps were a grape, I'd call MAV-delivered Office a grapefruit. The shared "grape" in the names doesn't make them the same fruit, nor do they appeal to the same people. It's true that they don't go well together. How many fruit salads have grapes and grapefruit together? Yuck.
It's About Licensing Terms
Gavriella refrained from giving more details on what the licensing terms will be for MAV-delivered Office, and there isn't much information coming out of Microsoft's Business division. I presume there will be a subscription fee, which is totally consistent with Microsoft's "rent me" strategy.
The question: Will there be two licensing fees? Microsoft requires businesses to sign up for a Software Assurance contract to use MDOP. From a marketing perspective, Microsoft doesn't treat Software Assurance as a subscription service. But the company accounts for the annual payments like a subscription service. There is a contractual obligation, annual payments and set time period. If the company stops paying for Software Assurance, MDOP usage rights end. That sounds like a subscription to me.
MAV 4.5 also will require Software Assurance. The software is currently available for free during testing. Release Candidate 1 is due next month. When released, MAV would be available to MDOP Software Assurance subscribers. Presumably, centrally managed, MAV-delivered Office would have a second fee, either through volume-licensing (consistent with the current model) or through separate subscription fee through third-party SAAS (software-as-a-service) providers.
Interesting aside: Gavriella told me that Microsoft "had about 7,000 customers download MAV [4.5 beta]," but "only 3,000 cusomers on MDOP, representing about 5 million seats." That's quite a difference between MDOP and one of its components. Gavriella viewed the difference in downloads as customer interest in MAV and its application virtualization capabilities. I don't agree, and I told her so.
I'm concerned that Microsoft is misreading enterprise interests, because the downloads carry different restrictions. MDOP requires Software Assurance, and MAV beta doesn't. I still hear IT managers gripe about the Software Assurance requirement. Analysts have shared similar sentiments from enterprises they talk to. I see the disparity in downloads as being commentary on the Software Assurance requirement as much as, or perhaps more than, customer interest in MAV.
Too Rich, Too Many Calories
Circling back to the main topic: Microsoft does plan to deliver application streaming through Live Mesh. As I understand it: Client code will be required. Until Microsoft delivers code for all clients capable of running a Web browser, there wouldn't be real competition with Google Apps.
Client code is where Microsoft's application streaming strategy is flawed. The main reason for using hosted productivity applications is access anytime, anywhere and on anything. Microsoft's strategy is about richness, the idea that end users need richer—what I'll call fat—clients. If that theory were true, no one would use a Web browser, no one would buy a 128kbps song from the iTunes Music Store. What matters: The product is good enough to meet the end user's needs most of the time.
Software clients don't go away because of anytime, anywhere, on-anything access. The Web browser is a software client, as are widgets. The difference: Where that so-called richness—the true functionality—resides. Microsoft wants it be more at the desktop, even when there are supporting Web services, and Web 2.0 platform companies like Google put it on the server. Secondary difference: Google uses a software client already resident on virtually all PCs and many cell phones: the browser. Live Mesh will require additional client code.
Microsoft could, and frankly should, make Live Mesh available as a plug-in for any Web browser on any hardware client. But that wouldn't dispel Microsoft's broader philosophical belief: that end users demand richness and that they must have it in a fat client. I don't agree; looks to me like the Web browser's success, its broad ubiquity, demonstrates how 20th century is Microsoft's fat client development thinking. Do you agree? Disagree? Please share in the comments.
Comments (13)
Thanks for making my points for me, Joe.
Posted by portuno | May 2, 2008 3:36 PM
Google is in the "meal-replacement" market. MicroSoft selling "all-you-can-eat buffets" that require whole new "wardrobes" when you try to consume them. It would be a lot more realistic for MicroSoft to incubate new companies (as you have astutely suggested) than for it to transform itself into a "weight-loss" product.
Nope. This pig is not going to fly, whether it has 20th- or 21st-century ideas.
Posted by Pinball | May 2, 2008 4:29 PM
You're making an all-up case, essentially. Many are on this bandwagon - that pure cloud solutions are inevitable, one size fits all, and ideal and that soon "the browser" will be the only installed application.
I'm sure that pundits of like mind will keep beating this drum until it is completely and utterly worn out.
It is very unlikley that you will be correct and it is FAR more likely that a mix of rich clients, enhanced by cloud services and pure cloud services for certain use cases, with seemless integration and management tools wrapped around them, will win.
Your assertion that somehow the success of the web browser and itunes is proof that rich clients no longer have validity is demonstrably ridiculous.
But of course this is a far less alarmist and dramatic view and it also feeds less into the "MSFT IS DOOOOOOMED!!!!" set that drives clicks, right?
Posted by Mark Lambert | May 2, 2008 4:37 PM
I guess you're true on leaving fat clients behind and going for Web 2.0 applications. Most of our clients ask how much it would cost and how long it could take to convert their existing applications for Web 2.0. Their almost all want to clear fat client applications when possible. It is much easier to deploy applications on a server only than on 100, 1000 (or more) computers. And people in the company can switch computer without problem reinstalling a bunch of softs. However I don't think Microsoft is doomed but it will have to adapt.
Posted by Martin Simard | May 2, 2008 5:59 PM
""the browser" will be the only installed application."
Not so. There is no real need for a browser. The browser is a container protecting the application from interfering with the operating system supporting the browser.
Get rid of both. They are both frozen virtualizations of some designer team's idea of what you need a computer for. Better to put your own solution on your own bare metal. Much smaller package with even higher performance. Selected from a list of specifications and done when and where you need it.
A bare metal computer with an ability to boot to an internet connection and point to a URL can have precisely MYcomputer squirted into the new machine anywhere in the world.
You guys are thinking too dotcom. This is far beyond website technology.
This is putting the processing power we all think is confined to the desktop out on any number of computers including those of our own anywhere.
The "fat client" "thin client" debate is a discussion using artifacts of two obsolete paths.
A new client configuration will come out of recursive refinements of the way to build these things. If the software development paths are not blocked by some other monopoly (they do this kind of thing by nature - Microsoft is worth much more broken up than it does to shareholders now but will resist because the CEO will not be relevant to the pieces) that is.
Posted by portuno | May 2, 2008 11:19 PM
Working a bit on portuno's idea. Why doesn't Microsoft create its own little .NET Web Start. An equivalent of Java Web Start that lets you run your own little apps independent of the underlying OS.
Why are they so obsessed with Windows as the underlying platform? I understand the need to leverage the OS to maintain a monopoly status and keeps selling. It all worked fine up till Vista.
How long have we had .NET? How many years to create a platform independent environment? I'm not a .NET developer I do most of my stuff in Java and other open source languages. I'd love to hear from someone who does .NET development why are those enterprises having issues moving to Vista. Focused more on the economics of things not so much the technical aspects.
Because .NET sounded really nice back then when it was released, but doesn't seem to have lived up to its promise. Or maybe its acceptance wasn't so great and most kept with ol' VB. This leads to very interesting questionings about this "new" technology from Microsoft.
Will it be something new or just some bit of the same ol' with some RPCs added in between? To me it looks like the same ol' thing. Leveraging the OS now over the net. Still stuck with OS dependent implementations that are web distributed. Will these new web apps break when you update to the most recent Windows? or will they break if you don't ;) !
Posted by Gerardo Tasistro | May 3, 2008 11:21 AM
Gerardo Tasistro has talked about the elephant in the room and I can almost feel everybody cringe.
Microsoft has been working for years to get .Net to the point where they can use it
http://news.zdnet.com/2100-1009_22-814631.html?legacy=zdnn
This article from Jan 15, 2002 decries the state of .Net early on. It reads like a news article from this year.
The question here is, if .Net was intended to give Windows access to the web, why has that not happened?
Now, the Microsoft enthusiast can spout acronyms and market speak about all that .Net is supposed to be able to do on the web, but, the reality is, .Net is an also-ran and not at all a reliable or preferred method to build web applications.
Something appears to be missing from .Net. What could it be?
Anybody want to take a guess?
The Microsoft is facing a Markman Hearing July 10, 2008 that may provide answers what's inside .Net that Microsoft can't use.
Then we won't have to guess any more.
Posted by portuno | May 3, 2008 11:53 AM
I have seen one open source program, Handbrake that uses Net 2.0 or higher to run. Just one that I have run into. While Handrake itself is a very good free open source application, I have found Windows users, who very much support Microsoft, that refuse to install/run the Net framework from MS in XP. Even they, it would seem, do not trust MS to not screw up something.
Also, on a different take on the whole cloud, net, google docs, office online, take. In ten years time will everyone in the USA have at least a T1 connection? I think not. But this is what its going take to get rid of the fat client, or to run the OS from the "cloud server." My guess is that in ten years time, 10% will still be using dialup in the USA, as its about 1/3 now.
Posted by chips | May 3, 2008 1:09 PM
T1 for everyone is attainable.
http://www.magneticsponsoring.com/community-blog/t1-speed-wireless-completely-mobile-internet-service-is-coming-19.95-mo.html
Current telecommunications transport speeds are climbing steadily and we are far from a theoretical limit on attainable optical speeds.
Verizon is king of speed and I believe they will remain that way for the foreseeable future.
Posted by portuno | May 3, 2008 3:00 PM
criminal portuno still pumping his penny stock.
this one's a real hoot:
".Net is an also-ran and not at all a reliable or preferred method to build web applications."
what he won't say is that his penny stock company's only source of revenue is a HR product BUILT on .NET. that penny stock's home grown (that they bought) product, hasn't generated a dime of revenue since they bought it, according to the public filings. .NET mustn't be all that shabby.
wrt to T1 for everyone. that penny stock's fiber optic patent is nothing but theoretical fantasy. That isn't Al's opinion but fact as stated in the company 10-KSB.
Posted by Al | May 3, 2008 7:06 PM
Al is a frustrated banker who posts on stock message boards. It looks like he works for a hedge fund trying to play the emotions and information search against them and their money as relates to the stock Al chooses to post about.
This is the kind of harrassment you have to endure when you try simply to explain the state of the industry these days with operating system builders. Al here is anti-Microsoft. He's pro-Linux. Thus he's on the stock boards attempting to steer pro-microsoft shareholders into selling their stock so the hedge fund can buy those shares cheaper.
The Linux people are threatened like Microsoft with this new web-platforming and Yahoo scaring Microsoft away sends a shiver through Linux hearts tonight.
Al is facing a bleak future with no money rolling in from his Linux and open-source hedge hogs. He's angry and turbulent. He's also dishonest as he posts using multiple usernames on the Yahoo message board. He does this to make the debate front larger; one username can step in and redirect attention from the primary username (original_Al) and he can overwhelm a typical poster.
Let's see him overwhelm this board. LOL
Posted by portuno | May 3, 2008 8:51 PM
none of your recent posts work...
just blank pages
Posted by user | May 4, 2008 9:53 AM
What is .NET missing? How about a secure operating platform for starters? I don't know why this is so hard... protect the core o.s. kernel from the applications that run on it. Then you can be happy and connect everybody to everything and share it all with the rest of the world polyanna style if you want. If doing that means a major rewrite... no doubt it does... then get after it before Google or somebody else launches something which is open source based and the functional equivalent of windows 2000. Because if that happens, we won't all be playing the game of monopoly any more.
Posted by Soap | May 8, 2008 1:45 AM