Archive for September, 2010

Would you bet on WPF, Silverlight or HTML5?

There’s nothing like someone saying some technology is “dead” to really create a storm. It’s even more of a controversial thing when it’s said by a former Microsoft Product Manager for Silverlight about what many consider a key technology for Microsoft i.e. WPF (Windows Presentation Foundation).

Scott Barnes, made some interesting statements on where he thought WPF, Silverlight and HTML5 were headed in the overall Microsoft Strategy. His overall assertion was that WPF was on a road to nowhere and that most of the internal love was for Silverlight and increasingly for HTML5. This even prompted Scott Guthrie of Microsoft who manages the teams that builds these products to state

scottgu

I have no special insight into whether Scott Barnes is right and I’ve no reason to believe that someone as respected and trusted as “The Gu” is not being honest.

My own insight and thoughts on this, with 20 years experience, in the technology/software industry is that I would be very surprised if there weren’t tensions between product groups at Microsoft. I think that is the nature of a large corporate business like Microsoft and I’m sure the Windows 7 team doesn’t like the Managed Code .NET teams and possibly views it with suspicion. It might explain the reason that we’ve not seen any WPF app built into Vista/Windows 7. I have previously blogged and stated at Microsoft Partner  meetings why Microsoft wasn’t porting more of it’s own apps to WPF? Something even as less controversial as Windows Live Messenger remains as a standard Windows app.

Betting on technologies is what we do and we do it for and with our customers everyday. I know that often the range of technologies/solutions from Microsoft are quite large but from my experience customers do need choice as to how they build solutions. Even if Microsoft had only one way to build apps then people are sill faced with a wide choice of building software whether that be Java, PHP, C/C++, Python, Ruby, Objective C, Adobe Air and running it on different platforms.

The fundamental principle behind WPF/Silverlight that we needed to bridge the gap between UI designers and developers was a sound one, and that by having a declarative language like XAML we could do valuable stuff easily/quickly without hardcore programming in C# or VB.NET was absolutely right. It’s true to say that we haven’t seen a huge rush for people to build WPF apps and that Silverlight has probably been more successful. I have just installed the new Seesmic Desktop 2 Beta which is built on the out of browser Silverlight capability.

There is now talk of HTML5 killing Silverlight and I guess by inference Adobe Flash as well and I saw an article today saying whether it will kill Javascript. It is with little doubt that HTML5 is going to be a very important move in web based applications. HTML is a funny old beast! It’s one of those technologies which was designed to give a very basic user experience because we are essentially working with the lowest common denominator to get a uniform experience, i.e uniformly bad!!. We then had more stuff pushed to the browser with Flash for more interactive experiences and then AJAX and more recently Silverlight. HTML5 is still developing as a specification whereas Silverlight is here and now and provides a great set of developer/designer tools and capabilities.

Obviously, we have some “big beasts” backing HTML5 such as Apple & Google. Microsoft cannot be seen to backing away on HTML5 because it’s been there and done that before. That’s why it’s been quick to release the public beta of IE9.

So, which do you back? Well, I think it does really depend but it’s still a very good bet to be backing Microsoft technologies!

SMBs Still Wanting to Buy Servers?

With all the talk about “Cloud Computing” at the moment you’d almost be forgiven that SMBs would be dumping physical server purchases by the bucket load to move to the new promise land of public cloud services!

This is one case of where the hype does not match the reality as highlighted by this article on the Windows Server Division Weblog, which points out recent IDC stats showing the biggest year-on-year server shipments in more than 5 years! As the article points out is that this is explained by the fact that businesses are in a different stages of their adoption of cloud services. A typical Microsoft Small Business Server is a multi-role solution within a single physical server and even shifting email services to Gmail or BPOS, still leaves you with file sharing, identity management/security, print sharing, remote access & possibly remote access, software deployment, policy management of desktops and possibly some line of business application (LOB). You could move your storage to the cloud as well and sync the data locally and there are solutions to do that but about your LOB? It might well be a bespoke solution which is not untypical in many businesses and so there might not be a cloud service doing quite what you want.

Sure, the cloud market is going to grow dramatically in SMB but there is going to be a process of organisations shifting to that gradually. There is still demand for physical servers as SMBs adopt virtualisation as they need new hardware capable of supporting virtualisation technologies from VMWare and Microsoft to do this.

Microsoft’s move to providing Azure as a platform for providing an environment to build cloud services is absolutely right because competitors like Google, Amazon Web Services, Salesforce.com and others are pushing hard on this.

Subscribe in a reader