Improving the Modern Application Lifecycle

This week in Redmond, Microsoft is hosting the 3rd annual ALM Summit, a gathering of Application Lifecycle Management (ALM) practitioners interested in learning more about the craft and sharing their own experiences with others.   Over the course of three packed days and across four tracks of discussions focused on DevOps, testing, agile development, and ALM leadership, attendees are discussing and collaborating with others in the field, all with the goal of improving how our industry delivers software and services.

A key set of themes during this summit focuses on real change happening in the industry.  In a world of devices and services, we’re seeing that feedback and iteration are the name of the game, with multi-year release trains replaced by faster and thoughtful build/measure/learn cycles, with a need for friction-free paths to production yielding advances in quality enablement and continuous deployment, with the blurring of development team roles, and with teams becoming more and more distributed.

All of this has led us to shift our approach in how we improve and release Visual Studio, while at the same time ensuring that new value includes capabilities to propel this “new normal.” In November, we shipped Visual Studio 2012 Update 1 (VS2012.1), an update to Visual Studio 2012 that provided not only fixes for bugs in the RTM release of Visual Studio 2012, but also a wealth of new features, spanning improved support for agile teams and continuous quality in addition to improved support for Windows and SharePoint development.  Today, I’m happy to share that we’ve released our first preview of Visual Studio 2012 Update 2 (VS2012.2).  This preview includes all of the improvements from VS2012.1 while also introducing web-based support for Test Case Management (TCM), improved support for work item tagging, unit testing features for Windows Phone 8, and more.  You can now download this preview, and you can expect subsequent previews and the eventual release of VS2012.2 to contain many more exciting capabilities.

However, while I’m excited by this VS2012.2 preview release, I’m even more excited by another of today’s announcements.  As Brian Harry just announced in his keynote this morning at the ALM Summit, we’ve added Git source code management to Team Foundation Service, with Git repositories hosted in Team Foundation Service available today for use seamlessly from any Git tool on any operating system.  Developers can now benefit from Microsoft’s fully-integrated ALM suite, while at the same time having a choice of using Git or TFVC (Team Foundation Version Control) for their source control repositories.  We’ll continue to invest in both Git and TFVC (Team Foundation Version Control) throughout future releases, as we see both centralized version control and distributed version control systems as being optimized for different types of projects and development workflows.

As part of this effort, today we’re also releasing a preview of an extension for VS2012.2 that enables connecting Visual Studio to Git repositories hosted in any Git host, including Team Foundation Service, CodePlex, GitHub, and any number of other 3rd-party services.  To create this extension, we utilized the open source library libgit2, and in the process, several of our full-time engineers worked as committers to the libgit2 project.  These capabilities will be built into a future release of Visual Studio, enabling it to serve as an incredibly robust Git client, one that provides seamless integration with the rest of the simplicity and power provided by Visual Studio.  For more detail on these announcements, see Brian Harry’s blog.


Follow me on Twitter at

Comments (2)

  1. vinod shalgar says:

    Hi Som,

    This is a great news that there is a lots of improvement in Visual Studio 2012 Update 2.

    One of the feature or goal of Visual Studio 2012 was a quick installation , which is very true. But what I noticed, updates such as Visual Studio 2012 Update 2 taking more time than installation of Visual Studio 2012 which is killing actual goal of quick installation.



  2. Marcello says:

    This business has become something like building bridges: tons of steel and concrete to achieve the simplest goal, like joining two distant points over a canyon, being the weight of the infrastructure several times more than the weight of the road that effectively joins the two points.

    In the last years, I had experienced an increasing lack of confidence and reluctance to use *the software* as the number of "service packs" and *updates* populating the downloads section of your website grows exponentially: there are services packs, and services packs for service packs, for nearly all the products your company ever released.

    Your post sounds too good to be true. The programmer will become another piece of the puzzle: as more and more refinements and tools are incorporated into the software, the programmer will someday, I guess, be "upgradeable" or "discard able", just as the software, with some notable "exceptions". After a number of iterations, all the bugs will just disappear; the end result will be the perfect software, under a controlled "ecosystem", and ideal conditions, that is.

    The blunders in your reasoning are quite ostensible: the real world is far from your premises: your hypotheses are too idealistic and promising: the end result must be faultless, immaculate ie: the conditions on which the software will be executed: you are taking for granted ideal conditions about everything: the hardware, the software, the teammates. In your hypothesis, the software has no bugs, the hardware doesn’t fail, all the members of the team are ideal, perfect humans, in good conditions: they are joyful, laughing and delighted with their jobs, consenting to cooperate, to share their knowledge, to talk,etc: your hypothesis are fictitious in each of these areas, as we all know that hardware can fail, software has bugs, and many programmers these days suffer, in variable degrees, all the vices and problems common to all humans, ie: narcissism, selfishness, arrogance … Some may be phony imposters and pretenders or peripatetics willing to impress their workmates.

    Therefore, not one of your conjectures is satisfied in the real world, and the result will not be the one you predicted or supposed, but inconsistent.

    Ergo, it is increasingly difficult to find a justification to continue using *the software*, to lay down the groundwork of any mission critical project upon it, in the light of the serious flaws in its conception as I concluded in my previous reasoning: it is evident and crystal clear to me how incredibly weak are its foundations, being the principles of rationalism disregarded in every new release.