As many of you know, late last week we sent mail out to Whidbey alpha participants alerting to them to the announcement this week. I was particularly excited to send that mail because we had some real news in the mail about the Express skus - not tons of detail and not lots of advance notice but at least we made an effort to give the early participants a heads up. Lastly we asked people to keep it quiet until the announcement.
As Robert McLaws talks about, the news was very quickly leaked. I want to thank the people who honored our request - I know almost everyone on the list kept it quiet, and we will to continue to push to be open with this sort of information. For the one or two of you who weren't able to honor it, well, it has definitely prompted some discussion and “management visibility” here. It has also given us an experiment - did the leak matter? We'll look at this in a couple weeks, trying to assess whether it had a negative impact on our ability to communicate the information in a way that was most helpful to customers. If the mail-and-leak caused a lot of confusion then we'll probably think very hard before doing that again. With respect to the leakers themselves, I'm actually interested in feedback on what to do. On the off-chance we find out who did it, we'll likely reconsider their involvement in the alpha. For Whidbey this is moot - everything about Whidbey is pretty much public at this point. But we'll keep the names around and look for them as we're pulling together names for future alphas. I'm not 100% comfortable with this approach but I'm not sure there's an alternative and to every extent possible I want to ensure that we optimize for the 99.9% of people who will honor our requests to stay quiet for a few days. Another approach is to wait and send the mail the moment we're rolling it out to the press, but that's hardly more than sending the press release directly.
Robert raises another point that has always bugged me - we often take our most enthusiastic customers, tell them our plans, get them excited, and then put an NDA gag on them. We have some ideas around this that we'll pursue in the next version, ranging from more radical (do an early non-NDA design review and invite people based on their ability to be a conduit for information/feedback to/from the broader community) to less radical (assume news will trickle out during the NDA and provide clear guidance/process through which well-intentioned people can make appropriate judgement calls without feeling like they are at risk of getting in trouble).
Lastly, as I said in an earlier post, a lot of this is about the transition. When we give less information out and then give a little bit, it's news and someone can gain brownie points by leaking. If we consistently do a better job, my guess is that people will mostly ignore the articles based on incomplete leaks and either look for the news directly from us or read the articles based on more comprehensive research.