Confessions of a Longhorn User

A Computer Says 'Mooooooo!'

“Wronghorn” is just plain wrong – a rational response

As has been mentioned previously on LonghornBlogs, and on Scoble’s blog, Tony McCune published an article on ZDNet today that’s slightly unflattering for Longhorn.  Ok, that’s a little inaccurate.  It was more like he put Longhorn into a bag and proceeded to beat it with a stick.  I wanted to take a moment to respond to some of Tony’s comments, and set the record straight, where necessary.  So here I go.

“Myth #1: The Longhorn suite will be a worthwhile investment.
Microsoft typically pushes big software bundles that force customers to pay for much more functionality than they actually need, and Longhorn will continue that tradition. Microsoft’s own research shows that 30 percent of PC desktop users don't use the entire Office suite. They only use the word processor. This means that a huge percentage of businesses are made to pay for functionality that they don't use. It's like going for a fully loaded SUV when only one person will be driving the vehicle to the train station. A Ford Focus would do, but you're force to buy a Lincoln Navigator.”

I don’t think you’ve really made your point here, Tony.  First of all, I’d like to know what features of Longhorn you consider to be superfluous.  While I’m not claiming that there’s no such thing as “bloat” in Windows, I fail to see how Longhorn – even at the pre-alpha stage – is overkill.  Using your own numbers, 30% of people only use Microsoft Word, as opposed to the rest of the apps in Office.  What you’ve neglected to mention is the other 70% who use more than just the word processor. 

Assuming these numbers are a constant (which I doubt) across products, 30% of users will use Longhorn for nothing more than reading their e-mail or surfing the web.  But what about the other 70%?  Should they have possible features pulled from under them because a minority of people won’t use it?  I don’t think so.  Personally, I appreciate the fact that Microsoft isn’t just catering to the 70% or the 30% - they cater to both.  If you just want to check your mail or check out a website (like the beautiful J) there’s nothing stopping you from doing that – in fact, if the demos at PDC were any indication, doing just that is a much more pleasurable and encompassing experience with Longhorn than with previous version of Windows. 

“Myth #2: Longhorn will not drive customer dependence on Microsoft products.
Customers who chose to migrate to Longhorn will be faced with three alternatives: First, run an all-Windows shop and be saddled with Microsoft devices and licensing fees as your only option. Second, run a mixed-vendor environment standardized on Longhorn and accept the fact that you’ll have to rewrite all non-Windows applications. Third, go with Java and don’t worry about rewriting tons of code since you’ll have the freedom to use your choice of middleware and hardware, including Windows. It’s obvious which scenario Longhorn will perpetuate: locking customers into Microsoft software riddled with second-class interoperability and integration and, of course, security flaws and vulnerabilities.“

Once again, I don’t think you’ve made your point.  To put it plainly, Tony, ‘Where’s the beef?’ These are the options that all companies and consumers face already.  Run an entirely Microsoft shop, run a mixed shop, or run a non-Microsoft shop.  The fact that you recommend Java is, to me, a little disturbing to be quite honest.  Java is not the end-all-be-all that you claim it to be.  Without going into a huge amount of detail, its performance is simply repugnant on the desktop.  There’s a reason that you don’t see a great deal of Java-made software sitting on the shelf at Best Buy – performance is not so hot, at least on Windows – and that’s what customers care about.   Home users want applications that run fast and that have a familiar interface (Java’s interface looks totally foreign to Windows users).  Corporate users want programs that… well, run fast, and make sharing of information between systems easy.  Both Java and .NET support XML, so information interchange isn’t that big of a deal, and if you only have one platform to worry about… well, you don’t have to worry.

“Myth #3: Longhorn will provide a better alternative to Java.
With Longhorn, Microsoft claims that its new software is so easy to use that developers won't need or want the “complex, specialized” Java platform anymore. They fail to note that Java has made great strides in usability and--unlike Longhorn--is not tied to a specific operating system. Building applications on the proprietary Longhorn framework will continue to force developers to work within the narrow confines of Microsoft devices.”

The problem with your statement, Tony, is that Longhorn isn’t a programming language, Java is.  Longhorn is an Operating System.  A better match for your comparison is Java vs. the .NET Framework and WinFX, which is the new Windows API in Longhorn.  What you say has some truth to it… .NET programs only run on Windows.  There is an open-source implementation of .NET Framework for Linux (called Mono), but it’s still under development. 

I wouldn’t, however, call the “confines of Microsoft devices” “narrow”.  Microsoft has operating systems that run on PCs, Servers, Cell Phones, PDAs and watches.  That’s not too narrow, if you ask me.

“Myth #4: Longhorn will not require a multitude of customer upgrades to implement.
Microsoft has built its business on a model that forces customers to spend money on software upgrades every few years. Every successive upgrade restricts Microsoft's client base to fewer options and increased dependence on its platform. Even developers are not immune to this upgrade stranglehold. In the next two years, Microsoft is planning three new versions of Visual Studio .NET: Widby in 2004, Oracas in 2005 and the Longhorn version of Visual Studio in 2006. Each version has a new framework that cannot be used with previous versions. Developers must continually update their skills, which costs customers time and money.”

There are actually only two new versions of Visual Studio due out between now and Longhorn, according to the schedule I saw at PDC:  Whidbey, which is due out next year, and Orcas, which is due out in the Longhorn timeframe.   You’re also incorrect that the different versions of the framework don’t interoperate.  First of all, the different versions work as “Side by Side” assemblies.  They can all coexist on the same computer. 

Secondly, applications compiled with one version of the framework will more than likely work with a future version.  When the hosting provider for my website updated their servers to .NET 1.1, my blog engine automatically targeted the newest version (that’s a configurable setting), and, as a result, broke a feature or two.  Admittedly, the features that broke collected un-validated data from user forms, and submitted them to a database – a big no-no.  The fix was to either add one line of code to the web.config file, or to (GASP!) actually write secure code.  I chose the latter and had the feature working again in 10 minutes. 

I’m of the opinion that developers should be constantly updating their skills.  That doesn’t mean that every new technology should be embraced immediately upon release, but it’s a dangerous career move to be stationary in a moving wave.  Does that cost time and money – yes.  Is that a bad thing?  Not necessarily.

“Myth #5: Longhorn will support open standards.
Microsoft has never fully support open standards, and Longhorn will be no exception. Steve Ballmer says that Microsoft delivers the benefits of open standards through XML connectivity. However, the company creates new barriers to true interoperability by promoting Microsoft vendor lock-in. This defeats the purpose of open standards because Microsoft products are open only as long as you develop applications on the Windows platform – and the same hold true for Longhorn. This is not the same as truly open platforms, such as IBM's, which are agnostic and run on virtually every operating system and device, from hand-held computers to mainframes.”

“Open Standards” is a nice buzzword.  If you’re already using an all-Microsoft platform, it doesn’t really matter if you could interoperate with Linux, because you’re probably not going to.  If you do need to interoperate with another platform, you’ve got XML.  I’m not sure how you get much more cross-platform than well-formed, validated plain text.  If you’ve got a system that can’t read that, it’s probably time to pick a new platform.

Now, let’s face it, it is in Microsoft’s best interest to make its products interoperate well, because it helps them sell products.  It is also in Microsoft’s best interest to provide a way to interoperate with non-Microsoft platforms (like this award winning package).  It’s really not quite as closed as you make it seem.

“Microsoft promises the world with Longhorn, but customers may call it "Wronghorn" by the time it sees the light of day, several years from now. Even then, can Microsoft credibly claim that issues that have plagued its products in the past won't reappear in the Longhorn--security flaws, poor integration, limited scalability and lack of interoperability? History and logic tell us not to count on it.”

I think it’s time that we give Microsoft a little credit, don’t you?  Integration is not all that poor, scalability is not all that bad, and neither is interoperability.  In fact, I would hazard to say that Microsoft integration, scalability and interoperability is pretty damned good.  Security, on the other hand is absolutely something that they need to work on, and they know that.  I, personally, will give Microsoft the benefit of the doubt that they can get this down to a controllable pitch. 

And if they can’t, I’ll be right out there on the frontlines with you.