April 2008 - OurView: The Opinion Blog

First of all, congratulations to Ray Ozzie and the Windows Live Core/Live Mesh team(s) for an impressive initial offering, and for opening up the Live Mesh platform this early in the game.  We’re just getting our head around Live Mesh as a platform, and not just another file sharing app.  BTW this video (edit: fixed link) gives a good introduction into some of the platform aspects of Live Mesh.  One of the problems with this kind of early admittance is that a number of questions around the strategy and direction have to remain unanswered until the platform fills out.  Of course, that doesn’t stop us from asking!

What about Foldershare?

Windows Live Foldershare, still in beta, seems to be totally redundant to what Live Mesh offers.  In fact, with the cloud storage piece, and “coming soon” functionality for the Apple Mac and for mobile devices, Live Mesh seems primed to leapfrog Foldershare quite handily.  This is a familiar Windows Live theme, offering redundant, competing, and confusing services.  We hoped that with the efforts to clean up the Windows Live debacle that sort of thing was behind us.  Until someone comes up with a plausible answer to why on earth anyone would want both Foldershare and Live Mesh installed on the same computer at the same time, we have to ask: what about Foldershare?

When will we see Windows Live Calendar integration?  Favorites? etc.?

As much as Live Mesh is a cool development environment, and as much potential as it holds, end users are still faced with the same old problems.  Windows Live Calendar, for as long as we waited for it, just simply doesn’t do the job we need it to do.  Users need to be able to create, sync, and share calendars easily and seamlessly.  We need the rich abilities of a client app to create calendars, not some watered down web app.  And we need to then seamlessly sync to other computers, to a web interface, and to other users on a granular level.  Live Mesh seems to be built to do exactly that, so where is the calendar?  Are we going to have to wait until PDC or beyond to get a working calendar?

Ditto the same questions for Windows Live Favorites, etc.

What is the backup story, and what about a business model?

Live Mesh is shipping now with 5gb of storage.  Yet I have nearly a TB (not all of it used, of course) on my Windows Home Server, and Live Mesh would seem to be a logical either alternative or adjunct to a backup service.  Will I be able to subscribe (in the pay as you go sense, not in the RSS sense) to enough storage to make cloud backup an option?  The same questions have been swirling around Sky Drive, but cloud storage as a consumer subscription service is still an unanswered question.

Do we have to go through this naming mess again?

Most of the stories on Techmeme today are about a file syncing product, called Live Mesh.  Yet we know that Live Mesh isn’t a product, its a platform.  What happens when new sets of functionality are added?  Will the name change?  Will confusion reign (again)?  Microsoft concentrates on platform, and we’re very impressed by this latest platform offering, Live Mesh.  We think its a game changer.  But Microsoft still doesn’t seem to understand how much flubbing the naming of products, of not having a clear story going in, hurts.  Now when you’re changing the game, having a clear story is tricky.  But Windows Live has less of an impact today than it should because people wrote it off, mostly because of the naming debacle.  Will we have to go through the whole mess again?

Live Mesh, at first glance, rocks.  The model is well thought out, the tools will soon be in place to extend the platform and use it in new and exciting ways.  But Windows Live users need answers to the basic questions first, before we look to the sky.  Hopefully, in the next weeks and months, we’ll be able to look up and see the potential Ray Ozzie is offering us.  For right now, though, our basic needs aren’t being met, and we don’t know that Live Mesh will be providing the answers we need.

When we told you to “pay attention to Feedsync”, we were of course talking about Live Mesh, but what Ray Ozzie is bringing to the table with the Live Mesh Technology Preview, enabled by Feedsync, or Simple Sharing Extensions to ATOM and RSS, is a lightweight, robust, performant, and scalable method for connecting devices, data, applications, and relationships.

In his Channel 9 appearance with former O’Reilly writer and current Microsoft evangelist Jon Udell, Ray Ozzie gets almost giddy when he gets to talk about Live Mesh.  And with good reason, Ozzie has been working on these problems for a long time.  What Feedsync enables for Live Mesh, when it is connected to a mesh of devices, the array of services available to Windows Live users (Live ID, contact store, data storage, etc.), user data, and the relationships between all of them, quickly becomes much more than a really cool way to sync up some files.

In another Channel 9 interview, Windows Live Platform Architect Abolade Gbadegesin describes treating Live Mesh devices in a similar way to how Live IDs are handled.  You have an ID, and now your device has an ID, too.  With it, that device can share what it knows across other devices and to other users.  It can create and store information about relationships, as well as files and file sync.  The Live Mesh Notifier (it was called Live Mesh News internally, and the initial exposure is a news feed not unlike the one on Facebook), can not only display news items about these relationships (Joe just added 3 files), but can potentially expose a lot of rich information about them (show all files added to these folders by Joe).

Our relationships online, and how we connect, protect, manipulate, and store them become much richer and deeper when we consider our devices as part of what we relate to.  Live Mesh will know what computers I use, what files I need, who to share them with, what I’ve modified, what others have modified, and what I’ve shared in a number of complex ways.  This platform (of which the current Technology Preview exposes just a tiny bit) through a simple set of RSS/ATOM extensions has bridged a gap between our online relationships to people, and our relationship to our devices.   When the mesh is connected to devices, Windows Live services, data, and available online or on a client, by managing the relationships between and among them all, it suddenly opens up a whole new world of possibilities.  We said it before and we’ll say it again, pay attention to Feedsync.

We’ve been telling you to get ready for this since last December, and this week, Ray Ozzie and the Windows Live Core team are preparing to unveil their first big project, Live Mesh.  We gave you the rundown on Live Mesh last week: sync devices to each other and the “cloud”, share files and folders, allow remote access to files and programs from another mesh device, or from a “web desktop”, all from one common interface.

So what makes Live Mesh such a big bet?  Isn’t it pretty much like Foldershare, or Groove, or one of the web 2.0 startup offerings?  Those, of course, are good questions.  While we’ve been able to ferret out some information about what Live Mesh is, how it works, and what it does, we haven’t yet seen it in action. Is it just another variation, another attempt at file sync and sharing?  In an email interview with Steven Hazel, Mary Jo Foley asked:

MJF: Foldershare is supposedly a precursor to the mesh architecture/synchronization services Ray Ozzie described at Mix. In your words, could you explain how Foldershare fits in with mesh? Will mesh supersede (and replace) Foldershare?

Hazel: Honestly, I can’t figure out what the heck those guys are talking about.

However we know there are some key differences.  As a Windows Live offering, Live Mesh should be a free download, or at least have a free component (we’re expecting Live Mesh to start out with a 5gb web desktop offering).  Groove, on the other hand, is a $229 install (or a $79/yr subscription) per user.  Foldershare, along with some reliability issues, and shaky functionality, doesn’t provide the cloud based “web desktop” piece.  Other services are either just getting started, haven’t proven reliable, or have been pulled off the marketplace.

Live Mesh is an opportunity for Ray Ozzie and the new guard at Microsoft.  The Live Mesh bet could pay off big if the Microsoft Sync Framework, and RSS shared extensions (or Feedsync) work reliably, quickly, and robustly.  If it works as Ina Fried is hearing, cross browser, and coming soon, cross platform, it could help to signal real commitment to a more open Microsoft.  Internally, there have been rumblings about Ozzie’s ability to ship software, and Live Mesh could help there, as well.

We’ll have a lot more on Live Mesh very soon, and will be bringing you interviews and demos from the Web 2.0 Expo later in the week.  Will Live Mesh prove to be as exciting as we’re thinking?  It won’t be long til we find out.

Yesterday's flurry of activity on the MSFT-YHOO front had so many plot turns that my head is still spinning, but at the end of the day, analysts are still leaning in favor of a MSFT-YHOO deal.  We could breathlessly tell you what we think is going to happen (and indeed I was working up an opinion post yesterday morning), but then Yahoo! announced a trial run of Adsense ads from Google, and more news "leaked" out about a Yahoo-AOL deal (what was it Fake Steve Jobs said about a three-legged race?), and to top it off word got out about a possible partnership with Microsoft and News Corp. (owners of MySpace).  Whew.

So truthfully we have no idea what's going to happen next.  We're going to continue to follow the developments closely (reality television has nothing on these antics!), enjoy the heck out of reading everyone scramble to try and make sense of it all (and feel a little sorry for them).  But generally, until the dust clears a bit, we'll probably try our best to ignore the elephant in the room, and continue to focus on telling you about things we do have a handle on, like Live Mesh ;), a new Live Maps update, lots more to come involving Live Search announcements, and all the best on Windows Live.

In the meantime, check out our MSFT-YHOO pages, where we will continue to tag stories of interest (using del.icio.us and NewsGator), track the latest stock news and prices (using a nice widget from Yahoo!), and compile all the official releases in our Document Timeline (ps. if you have IE8, you can even subscribe to the Document Timeline WebSlice!).

When we introduced you to Windows Live Core (now Windows Live Platform Services) just over a year ago, we mentioned James Hamilton and his work on commoditizing server installations, in particular by using containers.  Well here we are a year later, and Microsoft is implementing this idea in its new data center in Northlake, Illinois.  While many web 2.0 pundits talk about running services in the cloud, its going to be up to companies like Microsoft, and people like James Hamilton, to build out the infrastructure needed to provide cost effective, geo-located, and energy efficient services.

In his blog Perspectives, Hamilton has been writing about a number of ideas around data centers and building out the infrastructure.  In talking about a service free, fail in place model, using containers, he says:

Going to a service free-model can save service costs but, even more interesting, in this model the servers can be packaged in designs optimized for cooling efficiency without regard to human access requirements. If technicians need to go into a space, then the space needs be safe for humans and meet multiple security regulations, a growing concern, and there needs to be space for them.  I believe we will get to a model where servers are treated like flash memory blocks: you have a few thousand in a service-free module, over time some fail and are shut off, and the overall module capacity diminishes over time.  When server work done/watt improves sufficiently or when the module capacity degrades far enough, the entire module is replaced and returned to the manufacturer for recycling.

In today's post, he compares the efficiencies of a) building a data center, or b) placing 1125 racks of servers, one each in a condominium(!).  Taking the container idea to the extreme, perhaps, but it's a good read.  When I commented that a rack of servers might be one of the better room-mates I have had, Hamilton responded:

Our goal is to draw attention to what makes a "real" data center expensive. It's not the security, it's not the shell (the building), it's none of those things. Typically over 75%, and sometimes more, of the cost of a data center is power and cooling. In the example above, that's $150M spent on power distribution and cooling equipment. I smell opportunity.

I just happen to be reading a novel, "The Invention of Everything Else", by Samantha Hunt.  In it, Nicola Tesla spends his last days in the Hotel New Yorker (a fictional account with a basis in truth).  At one point the author describes the electrical system outside Tesla's hotel window:

"Years ago power lines would have stretched across the block in a mad cobweb, a net, because years ago, any company that wanted to provide New York with electricity simply strung its own decentralized power lines all about the city before promptly going out of business or getting forced out by J.P. Morgan.  But now there is no net.  The power lines have been hidden underground."

That's pretty much where we are right now; just getting to the point of stringing up the wires, of building the mad cobweb.  Microsoft knows this and is racing to become a J.P. Morgan.  It is one of the reasons why Microsoft is hot to acquire Yahoo!, although Steve Ballmer would love for you to think it is just a desperate move against big bad Google.  While a move to web services might mean the end of Office (as we know it now, anyway), the demand for secure, cost effective, reliable, and at least somewhat green cloud infrastructure is going to grow immensely.  Microsoft seems to be intent on being a leading provider of those services.

Perspectives is definitely my new favorite blog, a welcome alternative to the daily blathering about Yahoo! speculation, Windows 7, and blogger stress.  Keep up the good work James!

Last week saw the public beta launch of Windows Search 4.0. While it brought several welcome improvements for consumers to the standard Windows Vista search, namely speed and better searching of remote computers, I can’t help get the feeling that the Vista search still (for want of a better word) sucks.

For starters lets take a look at what Microsoft started off at the drawing board with. Bear with me as the branding mess on this is unrivaled.

In May 2006 an application known simply as Windows Live Search (internally OneView & Casino) was demoed at the Microsoft CEO Summit, during Bill Gates’ keynote no less.  Brandon Paddock wrote a great description of what this team were trying to achieve:

“The focus of this demonstration was on searching in the Enterprise and how Microsoft is developing cutting-edge solutions to problems that information workers face every day.

But we aren't only providing a solution for Enterprises!  We're building the next-generation search application for Windows, and we want to take search to the next level for all kinds of users".”

This desktop application combined the Windows Desktop Search technology we have now as Windows Search 4.0, a Windows Live UI, and searching across both local machines, intranets and the internet. This federation of search results is something we heard this week “was a research project” but is still something we hope to see Microsoft releasing for consumers in the future (it’ll inevitably come for enterprises).

In July 2006 Steve Ballmer gave a demo of Windows Live Search Center, as it was then named and shortly after that it was announced that the project would fall under the Windows umbrella.

If you compare the designs above to what we now have in Windows Vista, you’ll see that most of the basic features made it across. Preview pane, ability to filter results by file type (All/Email/Document… etc)

While it was a great decision to combine the Windows Search 4.0 technology with native OS search as we now have, the actually usability of it has actually worsened in Vista.

Over the weekend DownloadSquad did a great comparison of Windows Search 4.0 (on XP) to Google Desktop Search. As a Vista RTM user for now 18 months, I’d forgotten exactly how WDS was laid out compared to Vista search, and so it gave me some food for thought.

XP users have categories to refine their search results, by images, documents or music. In contrast, Vista users just get a jumbled mess. Yes it’s easy to filter search results by hitting the All/Email/Document.. buttons, but this adds in another hoop for users to jump through.

While it is much easier to launch searches in Vista (each Explorer window has a search bar) what use is that when results are indistinguishable. Windows Vista saw changes to the folder navigation UI in order to encourage users to adopt search, but instead we just have two poor solutions.

Then there is the fact that searching local machines on the same small home network is still unachievable for the average consumer, nothing has improved since XP. If I've added a mapped drive or network share, I don't want to have to navigate there in Explorer in order to search for my file, I may as well just carry on browsing directly there myself. This is where the federation of search results I talked about above comes in, and makes me wonder why Microsoft haven't yet shipped this in a basic form for consumers.

Obviously Enterprise is a much larger scale, so it's understandable that there are many more requirements needed to make it operate effectively. However this technology was demoed almost two years ago, so what's the holdup? Home networks are commonplace and Windows Home Server is only going to increase the prevalence of these.

Sure search on the OS has come a long way, but compared to WDS on XP, Vista has barely made any progress.