OK, I’ve now processed all of the feedback and written one-off responses to many of the specific questions. I appologize if I missed any. I want to thank all of you who took the time to provide feedback. It is tremendously appreciated and the best way I know of for us to learn and get better.
Here I will summarize and play back what I heard. I will try to clarify and explain our current and future approaches to a few things.
Not enough time
Well, I can certainly relate to that. I have the same problem. I can only test drive a few things a year. I hate to admit it but I was one of the slackers who didn’t get around to installing Vista until after it shipped despite countless internal mails cajoling all of us to do so – head hung in shame 🙁
One suggestion I saw was to create some virtual labs to substantially reduce the upfront cost to try out Betas. I’ve already sent mail to our release team asking about the possibility. In the Orcas product cycle, we’ve gone to releasing VPCs and have gotten good feedback on how that has helped. It sounds like taking even one more step toward simplicity will help some people.
I also saw a few suggestions for better samples to get people started. I’ll poke into that a bit.
There’s nothing I can do to give you more time but at least maybe we can figure out how to take less of it.
First, I want to cover the degree to which VS is (or is not) coupled to the version of the .NET Framework. Over the last several versions, there has been a strong coupling. It is not, for example, possible to develop a .NET 1.1 application using VS2005. We’ve heard strong feedback that this coupling is very problematic for people due to the fact that apps don’t change very often and enterprises are cautious in rolling out new frameworks. We’ve heard customers want development tool improvements without being forced to deploy new frameworks. Because of this feedback, we have added the ability for Orcas to target multiple versions of the frameworks (2.0, 3.0 and 3.5). To see this you can, in a WinForms project for example, right click on the project in solution explorer and select Properties. You will see a “Target Framework” combo box. Now, that’s not going to help those of you who are still building 1.1 apps but it’s a start. The cross targeting experience isn’t perfect (intellisense isn’t fully aware of it, etc) but we expect it to get even better in future versions.
So, you actually can use Orcas to build .NET 2.0 applications.
There were some comments about compatibility with GAT/GAX, side by side problems with VS2005, etc. I don’t think we did enough to ensure side by side before Beta 1. We’ve done a ton of testing/work on that for Beta 2. It’s something we need to continue to move earlier in the cycle. As for things like GAT/GAX, Resharper, etc. it’s hard to make a Beta work with everything else already out in the ecosystem. We need to prioritize, so this kind of feedback is really valuable. Knowing what is important to you, we can work on getting that working earlier.
I saw several comments about TFS compatibility in different forms. First, I want to clear up one thing: the Orcas TFS client is compatible with a TFS 2005 server. The Orcas server is also compatible with the TFS 2005 client. I hear you about the go-live, upgrade and forward data migration support issues. Our plan for Beta 1 was to get test driving kind of feedback and to support upgrade from TFS 2005, forward migration of TFS data to RTM and a go live license for Beta 2. I’m hearing you say that we shouldn’t expect much feedback until those things are in place and either I should set my expectations accordingly or we should have those things available for Beta 1. Fair enough – we’ll rethink this for our next release. There are, however, some good TFS client features in Orcas that work just fine with a TFS 2005 server and I encourage you to try those out if possible.
Too much stuff coming from Microsoft
This is hard feedback for me to swallow so I’m going to try to twist it into what I want it to say and then let you beat me if I go too far. I have, for a long time, believed that we don’t deliver enough stuff often enough. However, I’ve coupled that with the belief that when we do deliver something it’s big and it’s incompatible with everything else we’ve done before (OK, an overstatement but you get the idea). So my dream has been to deliver stuff more frequently but to have it usually be compatible with what you are already using and make it a really easy upgrade experience. I’m hoping that some of the stuff I talked about in “Not enough time” and “Compatibility problems” will help here. Imagine new versions of VS came out once a year but the vast majority of them could be installed cleanly side by side with everything else you have. All file formats were compatible (so you could use version N and version N+1) on the same solutions and projects. They can still build applications against the same framework versions. They just have new refactoring features, new debugging features, performance improvements, new version control features, etc. Would that be good?
So, I’m postulating that the “Too much stuff coming from Microsoft” problem is that we don’t make it easy enough or provide a smooth enough path. Is this right, or did I miss the point?
I did see some feedback from people who have tried out the Beta. There were several comments on the WPF designer and a few on other features. I’ve captured all of those comments and will be sharing them with my peers in the division shortly.
I’m not sure I captured everything but the feedback seemed to fit roughly into the 3 buckets above. Again, thanks for the feedback. If I missed something or mischaracterized something, make sure to let me know.
One additional note. Someone asked about the remaining Orcas CTP plan. The plan is to have one more CTP between Beta 1 and Beta 2 that will be feature complete but not yet have all of the testing and bug fixing that will go into Beta 2. I think we are expecting it to be in June.