I’ve seen a number of comments here (and elsewhere) about VS 2010 being another “Vista”. We are very aware how weakness in a few key areas (like performance) can cast a pall over an entire product. As I said in the original post, we are committed to shipping a great release (in all respects).
David mentioned in a comment on my last post that we are taking “risky fixes” later in the product cycle than we usually would. David’s right but I’d like to add a little color to this. We normally think of Beta 2 as our “validation” pre-release. We are looking for customers to confirm or deny that we are about ready to ship. We have tons of feedback opportunities before Beta 2 (CTPs, Beta1, more CTPs, TAP customers, dogfooding, etc) to try to make sure we are in a position for customers to say “yes”.
- This time they did not. As a result, several things have happened:
We have spun up an effort to reach out to every customer (external and internal) who said no to understand why and try try out fixes.
- We have assessed (and will continue to) why our metrics were telling us one thing and our customers another.
- We have reset many of our scenarios, metrics and goals to better align with the feedback. As importantly we’ve communicated to the division that “this is not good enough”.
- We have redoubled our effort on performance investigations and fixes. We’ve made some pretty substantial improvements.
- We have adjusted the end game schedule to allow for more performance improvements while maintaining a high quality bar for the final release.
- We are working on adding another widely available pre-release after the first of the year to give us another round of readiness feedback from customers.
We’re committed to making it great. I’m not saying noone will flame us over something – of course they will. That happens pretty much no matter what we do. But we will make sure the vast majority of people are very happy with the release.