Funding the .NET Development Ecosystem

The more concerning trend, IMHO, is the tendency for popular .NET packages starting as open source or free and eventually turning into closed projects or "sponsored" by 3rd party vendors. Consider Reflector (now with Red Gate), ServiceStack, Fiddler (Telerik), Extended WPF Toolkit (Xceed), and others.

As a developer who worked on Visual Studio (Visual C++, C# and the debugger teams), I have to disagree as to why it takes Microsoft so long to add new features. Sadly, it has almost nothing to do with external politics and everything to do with the slow moving nature of the Redmond behemoth. (Internal politics, yes. External, not so much). Dev teams were never discouraged from stealing features from other products, as long as it was management's idea. But the managers aren't developers.

I worked with the team that produced the IDE (editor, project/build system, etc) enhancements for C++ and C#, and I can tell you, we were all eager to add improvements for developers. But unless those features supported whatever the newest Microsoft agenda was, they never got off the ground. (Regardless of the fact that these high level agenda items - think 'WPF' or 'Metro' were completely orthogonal to basic improvements for actual developer productivity).

I spent five (FIVE) painful years working to get VS 2002 ("VS.NET") out the door and I can tell you, the vast majority of that wasn't spent designing features or writing code. The next four versions weren't much better; their scope was smaller and large features were mostly verboten so that we could ship quickly ('quick' being a relative term). Massive management overhead meant that for every week that developers spent working on features, months would be spent navigating the "ship".

I've heard rumors that things have improved in the last five years or so, as Microsoft tries to force itself to become more agile. For the sake of the developers who still work there, I can only hope this is true.

Andrew, though i think you're right about pricing and ecosystems, a more interesting question is the viability of the ecosystem. Writing a application for the desktop isn't exactly forward thinking. That is, the main platforms now are mobile devices and web applications, both of which are more and more built around a service-oriented architecture where the front and back ends are decoupled. In that world, neither Microsoft nor Java represent the ecosystem as a whole, and in the mobile world at least, MS is not a significant player. I know of a few web applications that are using IIS on the back end, but have front ends built entirely on JavaScript/JQuery/Angular or equivalent. One of those shops just decided to toss IIS in favor of Node, which is to say that the notion of a Microsoft ecosystem is eroding.

Given that erosion, it's hard for them to justify a pricing model that supports an ecosystem that nobody will be using in five years :-).

The first paragraph of NickP131's response makes the same exact points I would make in response.

I believe you need to do some reasearch. JetBrains license is not per machine. A personal license is tied to a user, and then there are corporate licenses that are not tied to a specific user. CodeRush has been out for a very long time and is not young at all. You perhaps are thinking of Telerik's JustCode.

I hate the MSDN pricing, I want Ultimate Visual Studio but I do not need a access to everything Microsoft has ever written. I should be able to purchase VS Ultimate or Premium without MSDN.

A good article, but lacking some research (or intentionally anti-MS biased). The free versions of MS IDEs (Express) are sufficient for many, and the IDE enhancements deserve more criticism. JetBrains is expensive and has a horrible license (per machine, not per-developer), and CodeRush is good but young and is a still a bit buggy (I have owned both). Although I do pay for the MSDN version (about $1K per year) I believe I get value for my dollars. Sure, free is good, but not when the IDE is not supported, suddenly get bought and retired, can't get docs or support, etc..

I'm surprised this article doesn't mention SharpDevelop, the OSS IDE. It handles my needs quite well and doesn't have the bloat of Visual Studio.

Lost in the conversation is the fact that Java made early and lasting inroads into Corporate computing, where cost of tools is critical and limiting, by working cheaply, correctly and cleanly, and improving in a regular fashion over reasonable time. Microsoft and the 'other tools' venders have only slowly been incorporated into that market, and then only as their prices have come down and the MS networking and server options have improved into the range and capabilities of large corporate requirements.
MS tool documentation still drives me nuts. Read some of the SSRS documentation/books and try to figure out how the tool actually works (and is programmable), rather than how to push the magic buttons...argh!
Sorry. Pet peeve.

Great points and particularly relevant considering I renewed my MSDN membership last night. Painful compared to my costs for Apple, Intel (Fortran is getting pricey), Google and Xamarin. Every year I wonder if I can get by with the Express versions of the MS tools or if I should switch to Qt.