Principles for First-Use Tools

re: Principles for First-Use Tools

Andreas Johansson — Mon, 01 May 2006 21:20:29 GMT

Why not keep the concept count low and allow others to extend it using add-ins?

re: Principles for First-Use Tools

johnmont — Mon, 01 May 2006 23:40:57 GMT

Excellent point and there are cases where that will work very well and I'm planning on doing quite a bit of that. Sometimes the work involved in building these add-ins is high. For example, with ADO.NET and LINQ we have two rather different ways to access data. For upward compatibility with the higher-level products, we need to have runtime and language support for both in every Express product. If I have both, I'll be shipping two slightly different data access methods and that will definitely confuse many of the newcomers to .NET, Express, and programming. So to reduce concept count, I'd pick just one and ship it by default. I could ship one in the box and build an add-in for the other, but the cost is huge. So which one should I ship? Why? Also, which set of customers will be upset if I choose one over the other?

Now imagine that set of questions applied to about a bazillion features and you have a sense of why I need some set of guidelines to help guide the decision making.

Lessons in Visual Studio Express

notgartner.com: Mitch Denny's Blog — Tue, 02 May 2006 14:29:16 GMT

re: Principles for First-Use Tools

Dennis E. Hamilton — Fri, 05 May 2006 07:53:01 GMT

I'm liking this. I think contributors are getting to hands-on quickly and also finding steps that won't discourage someone, while leaving plenty of room for further exploration as well as moving to new experiments, experiences, etc.

In short, yes.