7 - The Blogger Lunch
It is my last day in Seattle and it is my chance to talk Longhorn. I was invited to Microsoft's WinHEC 2005 Blogger Lunch. Joe Peterson, Vice President of product development for the Windows division, invited me to join him for a casual lunch and open discussion with the minds behind the Windows products seen at this year's WinHEC. I would also be chatting with Shanen Boettcher, director of Longhorn development and Clyde Rodriguez, Group Product Manager of the Windows division responsible for X64.
The chance to break bread with Microsoft Windows executives and developers from the x64 and Longhorn groups was exciting.
How to describe it? Okay, picture this -- 3 tables. At one table, Clyde Rodriguez and about 10 blogger/news sites. I had already met him. At the 2nd table, Joe and another group of bloggers.
And my table -- to my right, Paul Thurrott, Steven Bink, Robert Scoble, Robert McLaws of Longhorn Blogs, and Tom Warren of Neowin. And Jed Rose at the other end, the Product Manager and Featured Community Sites leader who invited me.
First was a "cocktail hour." I talked Longhorn with Joe Peterson. There were just 3 of us chatting with him. He is the overall head of Windows clients, Internet Explorer, the GUI, and Graphics including Direct X and Avalon. He is the man just below Bill Gates in charge of Longhorn. After our conversation, we sat down and Joe made a more formal presentation to the group. He spoke for about 5 minutes.
Then he turned things over to Shanen Boettcher. Shanen is the Director of Product Management for Longhorn. He is Longhorn's lead developer. I was lucky enough to be sitting right next to him, literally. Shanen was the one who went onstage at the keynote to do the Longhorn presentation. We ate (I think), then we had some Q&A with folks directing questions to Shanen. That was the end of the formal activities.
Now the fun begins. Shanen and I went for a cup of coffee and talked one-on-one for awhile. Some of the info he gave me was a recap of what he had told the entire group, so that went out online. But some of what he told me wasn't mentioned to the group and I was able to give him some feedback from our staff and members.
Overall, there were some concerns we had that we pointed out to Joe and Shanen. They seemed intrigued and pleased with the feedback, and a couple of ideas they thought were good enough that they would take them back to the team and possibly incorporate them into Longhorn. So at the very least we know that we may have possibly affected the product and Longhorn may have some features (or drop some) that it would not otherwise have had.
Some major points (we will discuss some of these in more detail below):
* The schedule is pretty much set in stone -
* Summer Beta 1 (we think later than the June 30th date that has been mentioned)-
* Fall Beta 2 (maybe at the PDC? Or possibly later)-
* Summer 2006 ship date (that's internal) -
* Holiday 2006 availability -
* There will be legacy driver support for XP, so anyone with a rig now running XP should be able to run LH with the same drivers -
* No hardware upgrade will be necessary -
* According to Joe, Longhorn will "definitely" run well on a 256 MB RAM rig - 512 MB, of course, is better, but if someone has 256 MB and at least 1 GHz processor they should be able to take advantage of all the features to some extent -
* There are 2 new (internal) builds put out every day, a 32-bit and 64-bit -
* There will definitely be simultaneous 32-bit and 64-bit releases -
* They have no plans to go all 64 and they have absolutely no plans to move away from 32 -- they fully expect most users who get Longhorn in Holiday 2006 will be getting 32-bit -
* Skinning - is still being worked on, but definitely will be there - right now, you have to hack XP to expand on the theming - Longhorn will take what now requires a hack and will turn it into a feature -
* IE7 - yes, it will be out this Summer and yes it will be in Longhorn but it will not be the same -- that is, while IE7 will definitely improve XP, the architecture of Longhorn is such that IE7 will have many more capabilities in Longhorn that it won't have in XP, even though it is the "same" product - so anyone who gets IE7 for XP and thinks "well, I almost have Longhorn now" will not be correct at all -
* Antispyware - will be built in, with the allowance for anyone to use other antispywares in addition or instead, much like the current Security Center allows one to use the Windows Firewall or not -
* Multiple monitors - right now, if someone has 2 monitors and wants to run across both with one mouse, you need 3rd party software -- Longhorn will include that capability -
*Sidebar - bye bye. The sidebar is dead. At the beginning of the lunch, it was a maybe. After we pressed a bit, Joe hesitated and said, "okay, it's pulled."
The 2 big surprises from the Keynote, discussed in more detail here at the lunch:
* Metro - this is going to be integrated with printers - we saw a demo --what it will do is allow a printer to duplicate the transparency and 3D effects of what you see on screen - instead of the way it is now, where you get a flat approximation of the graphics you see onscreen, this creates copies that are almost like a screenshot - very impressive - lots of questions about this at the lunch, all were surprised (more on Metro follows) -
* All the search features that were part of WinFS that supposedly would not be in Longhorn will be in Longhorn after all - more on WinFS follows
-----------------------------------------
From the conversations I have had with all the MS folks I met over the last few days, here are some of my thoughts:
On WinFS
MS screwed up when it put a lot of stock in the "3 Pillars Of Longhorn" (WinFS, Avalon, Indigo). Because they are mainly terms that described features aimed at developers, not the end user. So when they "pulled WinFS," leaving Avalon and Indigo, the press reports (and public perception) were basically that "1/3 of LH" was gone, and that the search capabilities we had heard about were gone too.
That is not true at all. WinFS was a misnomer. All the search capabilities that we talked about are still there, still on track, and being fully implemented at the end-user level. When they said "WinFS" was out, they were not referring to the search capabilities we had all talked so much about. They were only referring to the ability of developers to extend the metadata and create their own variations of the search capabilities. The term WinFS was only aimed at developers and is totally separate from all the search features we have all talked about.
So nothing has really been removed. From the standpoint of our members, everything we thought would be there is still there. We saw the first full demonstration of the search features at the Keynote and it was very impressive. Files are tagged in such a way that you can create "virtual folders" using pretty much any parameters you can imagine. It is exactly what we had envisioned and more.
Bottom line - whatever we thought was being "removed" when WinFS was removed has, in fact, not been removed at all. Some of it will be in Beta 1 this Summer and virtually all of it will be in Beta 2 this Fall.
On drivers and requirements
One of the major points the developers said over lunch is that Longhorn will have full legacy support for today's XP drivers. They know that not everyone is going to be upgrading their computers within the next 18 months and don't want anyone to think that they have to. Longhorn will run "just fine" on today's PC's, even at 256 RAM, although 512 is better, of course. You just won't get all the bells & whistles.
On backporting to XP
What about the features being backported into XP? My understanding is that MS knows not everyone with XP will run out and buy Longhorn. If folks can get a taste of what Longhorn is about, although XP won't allow all the bells and whistles, they will at least not feel ignored. And maybe they'll want more.
On the Sidebar
Why was the Sidebar pulled? Joe Peterson said that MS had originally assumed that more people would have widescreen monitors by now, or dual monitor setups. The idea was that a sidebar on a widescreen or dual monitor wouldn't take up that much real estate. But it hasn't happened. And they don't see it happening by the end of next year. So the basic thinking behind it fell away.
On the delays
I would like to just throw in a couple of thoughts on this. There was a point when Longhorn was delayed. That is, planned dates kept getting pushed back. Then, at some point, new dates were set. Those dates haven't changed. MS has been saying Beta 1 this Summer and Beta 2 in the Fall, with release by the end of 2006, for awhile now. Those dates have not changed.
So the idea that Longhorn "keeps getting delayed" is old news. The dates for Longhorn that were announced awhile back have not changed. And I do not believe they will be. And I will tell you why.
I learned a lot last week at WinHEC. Talking both publicly and privately with the developers of Longhorn, x64, and the heads of the Windows division I found out a lot about the internal structure at MS.
Basically, you have product teams. Each team has a product to produce by a certain date. By product I don't necessarily mean Longhorn. For example, WinFS has a product team. Avalon has a product team. So the different parts of Longhorn have their own product teams.
This is my assessment and I will not attribute it to anyone from MS. What was happening before was that certain product teams were not hitting their dates and that is what was pushing back the whole program. MS didn't want to just fire product teams because they missed a deadline. So Longhorn kept getting pushed back.
Well, that's not happening anymore. Because of all the delays and negative press, MS made a decision that if a product team doesn't meet their deadline, they are fired. That's it. They are simply out of a job. That is how they are keeping Longhorn on schedule now. And that is why I can say that I believe the dates are real now. Because I know that no matter what shape it's in, they are going to release it in whatever form it is. If product teams don't meet their goals, they will be fired. The Longhorn project will not be held up anymore because individual teams aren't hitting their deadlines.
On why all the "other" sites dwell on the delays
The reason most articles since WinHEC reference the delays is because that is what sells. Whether they are TV, newspaper, website, radio, or blogger reporters/writers, they have to compete in the marketplace with each other and the juicier the story the better the bottom line. If everyone else is writing about the delays, you have to also. It's much easier to whine and complain and be negative as a journalist. It is much much more difficult to be positive and accept the company's view. It is the natural tendency of journalists to be skeptical. That is why they focus on the negative and the delays. It sells.
The whole reason for the delays has been because they had the attitude, "let's not rush this." That is why the marketplace turned on them. Once they realized how much damage had been done, they changed strategies. Now Longhorn is on track. It is not being rushed. If a team is not going to meet a deadline they are fired. They are simply let go if the product isn't on time and of the quality needed.
Did security issues hold it up? No, that is not the issue that had held up Longhornand will not be an issue going forward. Many of the security improvements for Longhorn are now called XP SP2. We have them now. Security is always an issue and patches are always issued. That will not change. MS is certainly not sacrificing security, since that is probably the one area that did meet their goals. It is all the other product teams that fell behind. Security was ahead of schedule.
My conclusions
Finally, on Longhorn being a "train wreck"
I spent the better part of a week with the developers of Longhorn. Yes, I attended the same Keynote that you can see
here. But I also spent quite a bit of time with Joe Peterson, Vice President of Product Development for the Windows division and Shanen Boettcher, director of Longhorn development. My conversations with them occurred over lunch and I spoke at length with Boettcher, both at a table with about 10 people as well as one-on-one over coffee.
It was a real eye-opening experience. I challenged them to respond to the criticisms being leveled at them by the press and public: it's taking too long, it's a copy of Tiger, they can't meet the dates, features have been stripped out, etc. I came away with a new understanding of what really goes on in Redmond and what is in store over the next year.
Much of what we heard has been or will be in the press. You may or may not come across these accounts, but they put to rest many of the arguments against Longhorn. But some of what I heard was in private conversation, not meant for dissemination to the public.
Like you, I was skeptical too. And, as a journalist, I certainly have no reason nor desire to suddenly praise everything that MS has done with Longhorn. But I am all about facts. And much of the "noise" about Longhorn is lacking in facts.
For one thing, there seems to be a huge misunderstanding about what 5048 actually is. It is not meant to be the state-of-the-art of Longhorn development. It is certainly not meant to be anyone's primary OS. It is just a sampling of some features that were either not present in 4074 or have been refined since then. Some bugs have been fixed. But anyone who expected this Build to be anything close to the final OS is sadly mistaken.
The reality is that Beta 1 will, indeed, be pretty much up-to-date in Longhorn development. There will be a big difference between 5048 and Beta 1. Many of the features that will ultimately make it into Longhorn will be in Beta 1. And Beta 2 will pretty much have them all. So there will be a huge leap from 5048 to Beta 1. What people were expecting to come out in 5048 will, in fact, be present in Beta 1. Expectations of 5048 were way too high.
It was mentioned earlier but bears repeating: Longhorn is still 18 months away, and for people to be looking at this Build and saying "gee, this isn't that impressive -- Longhorn will just be a glorified XP" is just goofy. That's like looking at a clay model of a car and saying, "this doesn't look like it will be the Mustang we hoped it would be."
The most impressive features of Longhorn aren't even in the Build we have now. Some won't even be in Beta 1. I just don't understand why people are expecting so much out of this Build. High expectations lead to disappointment, and unfortunately, most people's expectations were way too high.
I saw previews of some of Longhorn's capabilities, features not yet in the Build, which will put to rest any notion that Longhorn is just a "glorified XP."
People complain about the lack of themes. Well, Longhorn will have skinning support. What you do now with a hack will be built into the OS. People complained about the Sidebar. The Sidebar is dead. It has been pulled. People complained about WinFS being pulled, asuming that the search features we had heard so much about were gone. That couldn't be further from the truth. WinFS was a misnomer. MS unfortunately focused on developer terms and not end-user terms. "WinFS" really refers to the ability of developers to extend metadata, to customize search features for their own applications. That's what was pulled. But everything we had been told would be in Longhorn will be in Longhorn. The search features are there. The idea that "1/3 of Longhorn is gone," that the OS was being stripped down, was just plain wrong. It was an unfortunate choice of terms and unclear representation to the media which caused the press and public to believe that the search features were not going to be included. We saw them demonstrated and they are impressive indeed. And they will be there.
I could go on and on. As far as dates go, they are set in stone. I heard that over and over and over, both privately and publicly. Beta 1 will be out this summer, Beta 2 in the fall, and Longhorn will ship next Summer for public availability in the Holiday 2006 season. If Product teams do not meet their deadlines, they will be fired. plain and simple. MS has learned from the past and they know that if they don't meet those dates, they will lose a lot of stock in the marketplace. I would bet on it.
