Will Windows Home Server be Microsoft's next flop?

Although you don't hear about them too often, Microsoft has had its own share of flops in the market. Perhaps the most legendary of these was the ill-fated Microsoft Bob. But one of the other flops that comes to mind -- one that required a good deal of investment from the hardware community -- was the Smart Display. In principle, the Smart Display -- a wireless display that could remotely control a PC -- was a neat idea. But practically speaking the cost and complexity of it made it a non-starter when compared to just buying another notebook.

Today, I'm reading an announcement regarding Microsoft's Windows Home Server and it occurred to me that this is a product that could easily follow in the footsteps of projects like Bob and Smart Display. Like with Smart Display, Microsoft has hardware partners (HP is on the list that Mary Jo Foley posted here) that appear willing to follow the software giant where ever it goes. But could this be a case of the blind leading the blind?

Yesterday, in a post having to do with the many unrelated forces, that taken together, represent Microsoft's most significant challenge in its roughly 32 year history, I wrote:

But even as its technologies improve over time, its journey to one corner of the universe (where computing is mostly about running Microsoft shrink-wrapped software on Microsoft operating systems, accessing Microsoft formats) while so many others are headed in completely the opposite direction has produced divisions in the industry that have ultimately made Microsoft less exciting to watch.

This of course is relative. If you're a Microsoft devotee, it's probably exhilarating to watch.

When I returned from my last vacation, sitting in my inbox was an e-mail from Microsoft's Public Relations team regarding Windows Home Server (released to manufacturing this past July). I hardly got past the subject line when I thought to myself, "Windows Home Server? They've got to be kidding. What on earth are they thinking?"

The funny thing is that four or five years ago, I would have been in love with this idea of a server operating system that targeted the home. In fact, I was in love with the idea. With no server positioned at the home market, I decided to come up with my own. Based on Red Hat Linux, I built one that offered the basics to my family: shared directories to which photos could be copied (and through Linux's symlinks, those directories were accessible on the Web using Apache Server coupled with TZO's dynamic DNS service), shared printing, and a central hard drive to which the other PCs could be backed up. In fact, that Linux box still serves as our residential print server to this day.

But more often than not, I find myself hating the idea that I'm running a server in the house. Once you set up a server and people start using it, it isn't long before you've bitten off more than you can chew and suddenly, your role as the residential IT manager takes on an entirely new dimension. Four or five years ago when there weren't many options, having a server in your house made some sense. Today, if you ask me, given the alternatives (especially the online ones), it doesn't make any sense. Of all the different kinds of technology I'd like to have in my house, something as complicated as a computer running a server operating system is at the bottom of the list.

Microsoft is of course going out of its way to make sure that Windows Home Server is not complicated. The company apparently ran a beta that involved 100,000+ testers. The WHS Web site uses words like "easy" when talking about setting users' access rights and phrases like "unprecedented simplicity" in an effort to compare WHS favorably to another way of creating shared drive space on a residential network: attaching a network attached storage (NAS) device to it.

But let's be honest (and this has nothing to do with Microsoft, it's the same for any server operating system vendor). Words like "easy" and "simplicity" have always been a part of the server marketing lexicon because everyone knows that servers are inherently difficult to work with.

In many ways, WHS exemplifies the way Microsoft is swinging right while the rest of the computer industry seems to be swinging left.

The tagline on the WHS home page pretty much says it all: "It's all coming together." Presumably, we want the benefits of WHS ("Now, it's easy to keep and share documents, photos, video, and music" says the Web site) to be delivered to us from one central location in our house. But do we really want it to all come together? Or, judging by the success and usage of many online alternatives, is there evidence that we actually prefer things spread apart, but loosely coupled?

My first thought on the idea of a home server is that Apple would have done it already if the market wanted it (and done it better than anybody else could hope to do it). In fact, in many ways, Apple's offering of AppleTV serves as evidence that architecturally, the last thing consumers want to do in their houses is push data from one PC to another. Instead, in peer-to-peer fashion, pulling on existing data from an existing PC as AppleTV does or as iTunes can do (with the outcome essentially being the same as having an additional and central server) is diametrically the opposite approach to the direction Microsoft is going.

Quite honestly however, I think Apple has it wrong too. Big businesses are loathe to tightly couple their technologies because, despite everything their solution providers tell them, once a chain of integrated systems is created, the ripple effect of something breaking is often unbearable. It's not about anti-integration as much as it is going for something more loosely coupled that isn't nearly so brittle. In many ways, the IT politics in my house were like the IT politics in a business. When I first told my wife and teenage son they had this new shared infrastructure at their disposal, the first question from my wife was "What happens if the server's hard drive crashes?" I had the answer to the question in my head (I'll develop some sort of backup process) but the last thing I wanted to do was share it with my wife and my son because then I'd be really committed. Technology should be about fewer commitments. Not more.

The next thing I knew, there was zero adoption of my new centralized infrastructure (or at least the central storage part). What I was left with (the part that's still operational today) was a glorified print server: something I can't wait to replace with something far simpler, turnkey, and headache free -- a cheap wireless printer (which, for all intents and purposes, is an appliance).

If installing local hardware invites all sorts of support-related issues that you weren't really prepared to handle, then the direction that the rest of the industry seems to be taking (the industry swinging left part) where services on the Web take the place of hardware you might have installed in your house are a validation that your not alone. Particularly since those services are getting used.

Take something as simple as photosharing and ask yourself what makes more sense: buying a server, installing it in your house (connecting it to the network), showing family members how to share and/or backup photos, music, and videos with it, and then supporting it for the rest of your life? Or, using an online service like AOL's recently announced Bluestring where AOL is saddled with the responsibility of ongoing reliability and support? (Note, I'm not even considering the fact that long term, we'll all want access to that content on our mobile devices and how a home-based server running over a paltry DSL or cable modem connection is probably the least desirable approach to that problem). Or maybe, you'll go for some combination of services from multiple service providers like Flickr for photosharing and iBackup.com (or even Microsoft's own Live SkyDrive) for online backup/filesharing and loosely couple them together with your personal start page at iGoogle.com.

Longer term, for local hardware (even appliances like NAS devices) to stay relevant, they'll need to be bundled with online services. Let's say you do want to put a central NAS device on your network. Much the same way today's cheap wireless printers are configurable through their front panels, the NAS makers will find a way for you to programmatically keep your NAS(es) mirrored to an online service so that, if the NAS should ever fail (and it will), recovery will be almost as simple as getting a replacement NAS unit. Turnkey solutions like this are much closer to what consumers and small businesses want than solutions that require administrative expertise.

Full-blown servers? Well, they're an entirely different story. They have no choice but to violate the KISS principle and once you cross a certain threshold into the non-turnkey world, you end up with many of the same headaches that businesses are finally getting hip to, and trying to get rid off. For example, salesforce.com would be a complete failure if all businesses wanted to do was run their own software on their own servers. Can consumers learn something from businesses about what it takes to run successful IT? You betcha.

Still not buying off on my logic?

Earlier today, Microsoft's PR contacted me with an announcement that tomorrow, "Microsoft will release a first update to Windows Home Server via Windows Update." Updating servers (not to mention the all important upgrade that will one day be in the wings)? That sounds exactly like the sort of thing I, as a homeowner, want to be thinking about. Not. The e-mail points to a blog entry with details that couldn't better emphasize the complexity involved in running servers -- headaches homeowners can't possibly want. Here for example are some of that blog entry's bullet points:

• The out-of-the-box experience warns users not to reboot Windows Home Server while updates are being downloaded and applied. [Now] a text message has been added to the "Windows Home Server Update" step of the home server setup explicitly instructing users to not reboot their home servers.
• Windows Home Server Connector Firewall Blocking Issues: The installation of the Windows Home Server Connector software can fail due to firewall software installed on a home computer. The Windows Home Server Connector software and corresponding help files have been updated to help further identify and troubleshoot issues with firewall software or incorrect proxy server settings on home computers.
• Based on feedback from our customers and partners we have made changes to provide additional guidance on creating user accounts and passwords on Windows Home Server. The User Accounts tab in the Windows Home Server Console has been improved to provide additional information and user education about user accounts and passwords to first-time users of the product.
• The file system driver installation required to mount a backup on a user’s home computer incorrectly reports that it requires a reboot to complete installation. A text message has been added to the Opening Backup dialog to ignore the reboot request.

Translated (for the benefit of the consumers that WHS is targeted at): "Servers are complicated. More complicated that you want to know." After reading this (particularly the parts I boldfaced), I heard myself thinking "The people at Microsoft are clearly nuts." Again, WHS is barely out of the gate and already, the best Microsoft can do in terms of dealing with ease of use and integration problems is to offer more help and education where, what consumers really want is to just plug it in and have it start working.

There's a reason that salesforce and companies like it are hotter than anything else on the market. They just work. Updates and upgrades require little else but the refresh button on a browser. Worrying about system reliability is someone else's headache.

If you buy-off on the theory that the world seems to be heading in the opposite direction that Microsoft wants to lead it, then you can't help but wonder what the long term prospects for an offering like Windows Home Server are. Not good, if you ask me.