When I lose my MVP status, we’ll all look back on this blog post and understand why. I just can’t help myself though – Microsoft has made some real head-shaking decisions over the last couple of weeks.
Sorta-Kinda-Announced-But-Not-Really – At UK Tech Days, Microsoft staff announced a release date of May 21st. No press releases came out for days, and then we got an invite to a press conference on 4/21. This led to a lot of guessing and second-guessing about whether maybe the date was incorrectly given (April 21 or May 21?), and then…
Surprise! The Bits Were Available – Vidas Matelis scooped Microsoft when he announced on Twitter that the R2 evaluation download page was live. He’d been reading the SQL Performance blog on MSDN and decided he’d try it. Further digging suggests that the R2 Express Edition is ready now.
Odd “Launch” Labeling – If you can’t download the software for production use, it’s not a launch.
Period. Full stop.
The SQL Server team has a particularly hard time understanding this. In 2008, they “launched” SQL Server 2008 in February – but they admitted you wouldn’t actually be able to download it for months. This time around, the “launch” is is 4/21, but all you can download is an evaluation version. They’re getting better, but still not good enough.
No MSDN or TechNet Downloads – The public gets to download an evaluation version free, but those of us who actually paid money for MSDN access or Enterprise Agreements can’t download it until May 3rd. I could understand if we’d all been given a Release Candidate version to bang on, but…
No Widely-Released RC – The last bits most people saw were a November Community Preview. Microsoft has historically delivered multiple CTPs, then a final Release Candidate, and then given the Release to Manufacturing go-ahead.
I work in marketing for a software company, and I know how hard it can be to coordinate a global launch. Software isn’t always ready when the marketing team’s campaign is ready, and the dates of events like PASS Europe aren’t going to change. But seriously, Microsoft, you can do better.
9 Comments. Leave new
I concur with our resident SQL Master.
Good grief.
Perhaps next time the launch of SQL Server 2011 will go along the lines of “RT #SQL2011 is RTM! W00t! To the bar everybody, and then 2moro we’re installing it on production everywhere! #sqlserver”.
Would have been more fun than the conference call 😉
Anywho, SQL 2008 Enterprise licenses purchased with SA last week will mean a nice shiny upgrade to 2008 R2 Data Center Edition for my client for freebies …
I can go into lots of reasons from a marketing/product management standpoint why these things happen, but I won’t. Probably best for MS to comment on that. However, it’s why I am a proponent of holding launches until software has been out for a while. It’s really tough to predict exactly when a dev cycle is going to end and go RTM in time to schedule all the marketing activities that need to correspond to a launch. This is a lesson I have learned the hard way, with much pain.
So true.
What is also missing is ANY information surrounding Madison. Tech details are few and far between. Microsoft Marketing has been pretty quite which is not typical of marketing. I predict Madison won’t ship until Denali.
Just in case you don’t get an official reply, I thought I’d add my 2 cents, on the historical comments.
Comment 1:
The SQL2008 Launch was not an oversight or something the SQL Team “didn’t get”. There were many at Microsoft that felt exactly as you did, these concerns were raised early & discussed at length, many times. Initially Visual Studio, Windows 2008 & SQL were all going to ship at around the same time. It made sense to have a bigger multi-track launch, demonstrating the added value each of those 3 products gave to the other. As Microsoft simultaneously ships to a global market this meant booking venues up to 1 year in advance, locking in speakers etc.(but even then we knew SQL wouldn’t meet their planned dates)
The ship dates changed. They considered splitting out the SQL content. But a) the Value prop of the 3 apps remained the same, b) Some Consultants / Customers told us it was expensive for them to leave the office so, liked the idea of “getting it all” in 1 day, & c) many preferred the idea of seeing what was coming, a few months in advance. d) Many Microsoft Subsidiaries, hadn’t budgeted for a separate launch, & had issues getting venues & people.
In short, those in favour of “pre-launching” SQL2008, had a stronger case than those arguing “Yer but we’ll look a bit dumb launching a product a few months before it is complete”.
Comment 2, 3, & 4:
Not sure, how you got May 21. The target hasn’t changed for ages. I suspect that someone fat fingered on the date, read 4 & said May or just added a few weeks “just in case”. Dyslexia Rules KO.
OR Perhaps they were talking about the date of their local Launch Event as distinct from the Global date of Availability. Which is different for different languages of the product. But usually they are referring to the English release. Of course, The date also varies depending on how you receive it. Web Downloads are much quicker to “Manufacture” than pressing a physical DVD & shipping it.
I’m pleased to see that MSFT is a little less anal about keeping their Web content secure till the actual USA Launch event is complete. Having the Free SQL version available from one site makes sense.
Of course it is also understandable that the more secure s/w delivery sites, ie: those that ship fully licensed product, they have much more rigorous staging to production release schedules that any product team has to fit with. This may account for the time delay between Free & Licensed release dates.
Comment 5:
There was a Release Candidate 0 (RC0) but I don’t know how widely distributed it was. Perhaps only those registered for the beta.
Comment 6:
MVP comments & feedback are taken very seriously. I have seen whole product feature sets change purely based on one MVP or Customer feedback. They (you) have a phenomenal amount of influence.
But like evey walk in life, timing is everything. Your feedback is often much more effective if you can (a) give it to the person who makes the decision, b) at of before the time they are making the decision.
ie: Screaming “Look out” to the driver of your car when you are 18 inches from the front bumber of an oncoming truck. Is similar to saying “Please make a major change to SQL feature X” 2 weeks prior to shipping.
I love your blogs, keep up the great work.
David – hi, thanks for the comment! A few notes.
Comment 1 – I agree, there’s a lot of synergy between multiple products, and it’s easier to do one launch instead of many. But following that same logic, why not launch SQL Azure at the same time as VS2010? There’s a lot of synergy there, right? The launch time span is about the same. The logic only spreads so far, and the several months between SQL 2008 and its “launch” stretches that logic too far. Otherwise, you might as well be launching the next year of products all at the same event.
Comment 2/3/4 – if you look at the tweets pictured above, that’s where the May 21st date came from. You can rule out the fat-fingering because 2 separate attendees heard the same thing. You can rule out different languages because this was a UK event.
Comment 5 – Historically, the SQL Server team released the RC to MSDN subscribers. No RC was made available through MSDN this time. (I’m dancing around this one.)
Comment 6 – Thanks, I’m glad to hear that, but the actions don’t seem to line up with that. In the MVP forums and at the MVP Summit, we’ve repeatedly questioned why we’re getting things like the DAC Pack. We gave plenty of feedback at the MVP Summit on several different topics, and I was left with the very distinct feeling that we were getting features shoved down our throat, not being asked for feedback. Like you pointed out, feedback is only effective when we can give it to the right person at the right time. Unveiling something to the MVPs months before RTM isn’t going to get you good feedback – but that’s not something the MVPs can fix, because we’re not involved early enough in the decisionmaking process for those particular features. There WERE features we caught early enough to help, but these features weren’t in that group.
Thanks again for the feedback!