To win in the cloud, Microsoft needs developers

A lot of people are asking if Microsoft will prevail in the cloud. While the jury is still out, it's becoming clear that Microsoft's strategy makes more sense than even it has explained.

In a recent post, InformationWeek's John Foley provided several points showing that Microsoft should be able to win, thanks to a massive user base and nearly unlimited resources. Most agree that Microsoft's cloud efforts are not so much a strategy issue as much as an execution problem.

I am starting to agree that "Software plus services" is the right approach--at least for Microsoft. One of the most overlooked Microsoft assets is Visual Studio, and if Microsoft adds a "cloud deploy" capability into the development environment, all of a sudden, things will start to get interesting.

Deployment and management are starting to be addressed for Amazon Web Services by companies like RightScale, but development is a much different story. There isn't an obvious cloud integrated development environment that does the heavy lifting typically necessary to build infinite-scale applications.

Microsoft has basically 100 percent of .Net developers using its tools. Once the cloud becomes an easy deployment option, it would seem obvious that more .Net developers will choose to deploy to a Microsoft cloud.

I definitely expect strong demand for Windows-based internal clouds, something the company hasn't yet mentioned as a deployment option. There are obvious enterprise reasons (including security and management) why an internal cloud would be very appealing to Microsoft shops, especially if they already have trained resources.

The big challenge to get adoption is that Amazon.com's totally open system profiles (i.e., run whatever you want) versus Microsoft's single-vendor platform won't likely cause people to switch to their cloud offerings. Azure won't be interesting for non-Microsoft developers for a long time.

Microsoft must give developers what they need in order for the Azure strategy to work. Without developers on board, Azure will be dead on arrival.

[Hardcode]

I've been tinkering with it for about a week now. Check the "Visual Studio Tools for Windows Azure" walkthrough for current deployment from VS2008. I agree that developer buy in is important, and as a rule, MS pays close attention to it.
Making the business case is another story. Integration as well as security and management will be key.
Wouldn't an "Internal Cloud" defeat the purpose?

[AppleSuxLeo]

Like this clown really knows what is going on inside the walls of MSFT.

[daverosenberg]

Oh hey, that's really helpful. This post is pro-MS you genius.

[Penguinisto]

This isn't as easy of a clinch as it appears... unless the company (or code-monkey) is specifically aiming for "the cloud" (most aren't), it doesn't matter how many copies of VS.NET are out there.

Also, if you're going to write apps for online use, guess what? You have to aim for more than one platform. .aspx isn't going to cut it in that respect, and no company wants to have to post a "you must use IE on Windows" notice when their competition can take all types of users. They'd pretty much have to in many cases because .NET is more often than not too proprietary for use by any other OS.

[shan_mcarthur]

Penguinisto, there is no requirement at all to use IE for a website hosted in Azure. The website can be built to support any browser and has nothing at all to do with the technology of Azure. You also should probably re-read the article. The premis of the article is that existing asp.net developers that are using Visual Studio can easily post their sites to Azure, moving them to the cloud. The argument you are making about developers having to write their web applications to support multiple server environments is way off base - the entire premis of all the cloud hosting vendors is that you write to their cloud infrastructure and they host your application and scale it into the cloud. You should really consider some more reading on cloud computing before making any more comments in this industry.

[Hardcode]

Penguinisto - Are you implying that ASP.NET applications require IE on Windows on the client? Are you also implying that using a web app developed in ASP.NET and hosted in IIS requires the .Net framework on the client? Am I reading this correctly?

[daverosenberg]

Internal clouds make a lot of sense if you already have hardware and the system admin abilities. The costs will be lower and so will the management functions.

A "Cloud" is a specific way to set up computing systems. Most Cloud stuff we read about are just systems running on the internet that may/may not adhere to Cloud principles.

[daverosenberg]

Internal clouds make a lot of sense if you already have hardware and the system admin abilities. The costs will be lower and so will the management functions.

A "Cloud" is a specific way to set up computing systems. Most Cloud stuff we read about are just systems running on the internet that may/may not adhere to Cloud principles.

[Hardcode]

Dave, I don't disagree that a "Cloud" in this sense defines a specific "way" to set up a computing system. I just don't see the benefit of the model deployed internally (Being careful not to confuse this with virtualization in the enterprise). Isn't the bottom line the bottom line. Just one of the major cost saving benefits envisioned is being able to scale smoothly to demand, on demand. That disappears internally, you still have to own/lease hardware in chunks. Not to mention maintaing and training employees to manage load balancing, IOS, and debug network problems.

[daverosenberg]

I hear ya. I just think that somewhere down the line people will realize that they have so much HW that they can build their own