Scobleizer Weblog

Daily Permalink Monday, November 03, 2003

As an example of a great conversational weblogger who works for Alan Meckler: Jupiter Media's Joe Wilcox takes Microsoft (and me) to task "Microsoft’s biggest challenge coming out of its developer conference, which ended today, will be getting everyone focused back on Windows XP."

Good point. It's only human to want to focus on what's coming rather than what's already here.

One thing we've done is pull tons of people off of Longhorn development to get a major security fix done (code-named Springboard). Windows XP +is+ job #1. We're continuing to invest in that and bring out great new products, games, and services for XP.

But, how do you pull back the hype lever once you've pulled the trigger? That's an interesting problem and one that no marketing department has ever faced.

Hey, Alan Meckler, notice what I did in this post:

1) I linked somewhere else.

2) I responded to what that weblogger said.

That's a conversation. I sent people off to another weblog. In the "old school" of Websites, that's a no no. I once had a boss who said "never send anyone off to someplace we don't own."

In the weblog world you get paid back for sending people to some other weblog and/or website. Most people haven't figured that out yet.

Oh, yeah, I just committed another sin: I just taught the competition how to do a great weblog. That's part of weblogging culture too. Share and share and share until you can't share anymore. Why? Because people will remember the guy who shared the tips that made them rich and/or famous. They won't remember the guy who kept everything close to the vest. Not to mention that my readers teach me FAR FAR more than I teach them. But, they'll only do that to people who share first. It's the definition of "win win." I want my readers to win. That makes them want me to win too.

Alan Meckler announces yet another enterprise-focused cdXpo (conference and exhibition). I still am bored with Alan's weblog, which doesn't have a conversation, but rather is just a way to deliver press-releases. Hey, Alan, yeah, this is a good cheap way to replace those press-release announcement services. I hated using them too. But, Alan's missing the real power of weblogs. Why doesn't Alan talk to his Jupiter Research webloggers about best practices? They have tapped into something real powerful: conversational marketing.

Does conversational marketing mean bigger events? I had tons of people come up to me last week and said "I came to the PDC because of your weblog." So, yes, I think it can be shown that conversational marketing is very powerful indeed. The PDC was sold out. How about cdXpo? Can it hurt to try?

Yes, the rumors are true. Marc Canter stayed in my room during the PDC. Weird thing was I barely even got to see him because I'd come in at 4 a.m. and leave at 7 a.m.

What do we want you to do now that the PDC is over?

1) If you were at the PDC, please log into and evaluate the sessions you attended. That'll help us improve for next time.

2) Catch up on your post-PDC reading. Mike Gunderloy has an excellent article that'll get you caught up.

3) Join in the newsgroups.

Hey, what happens if you don't make all the PDC materials downloadable in one download? Two guys take the weekend and write a .NET hack to do it for you. Thanks Sean and Scott!

Heh, Scott Hanselman has some fun with me at dinner the other night. Seriously, I'd be honored if he really considered me his friend. Scott was doing some very cool blogging with his little Blackberry.

One problem is just getting anything done the week after the PDC. Most of the team I'm on is getting some well-deserved rest. Jeff even has a bit of time to catch up on his weblog (cool pic of Don Box and Chris Anderson -- they really are the rock stars of PDC 03 -- their banter with Jim Allchin was classic and will be a "must study" for keynote demonstrators for years).

I started writing out my memories of the PDC week, but it just got way too boring. Let's just say there were a few high points. Having Ray Ozzie ask to meet me at dinner. Walking into the weblogger BOF and having people cheering me. Seeing Aero on screen. Touring Universal with a bunch of webloggers and meeting Don Box. Having Chris Anderson come to dinner with me. Yes, the rock star of the PDC!!

Seriously, I'm getting boring again. Thanks to everyone who said hi and now let's start your aggregators again. Be back later for a tour through what's in my feeds.

As of yesterday Maryam and I have been married for a year. Wow, how fast it has gone. It's been quite a wild time in the Scoble household. I haven't been good to her during the leadup to the PDC, so will try to slow my blogging down a bit. This weekend, for instance, we'll be taking off. The community is really rocking, though. The SQL blogs, the .NET blogs, the PDC blogs, the Longhorn blogs, are all fantastic and are growing in numbers and in quality (don't count last week's many "moos" into the signal-to-noise ratio).

I love Maryam very much, and believe me, she's the reason my blog is halfway decent.

At the PDC there was this $700,000 Mercedes taunting me. It was saying "go ahead, take a picture!" So I did. Only one exists in the US right now. $450,000 list price, but I met a dealer there who had already sold two of them for $700,000 each (delivery starts next spring). $50,000 deposit required.

Maybe we can talk Bill into buying one for the Longhorn team. :-) Yeah, right.

Anyway, these images make for great desktop wallpaper for your new Longhorn systems (click on "original" at the bottom of the page to get the full 5megapixel images).

Alright, I've slept off the PDC, but it'll be a while before I really get back in the swing of things. I see that some of you are digging into Longhorn. A great rant was done by Ole Eichorn, who is taking Longhorn to task on performance and reliability.

Ole: you missed something MAJOR.

We expect every Longhorn system to have a GPU (translation: something like an NVidia or ATI graphics card like what all the video game-oriented systems have today).

Now, let's look at what that processor is doing today on Windows XP: absolutely freaking nothing.


Now, what would happen if we moved quite a bit of the OS to use that processor? Guess what? Better performance!

Did you see the video compositing engine's performance? Off the charts. Multi videos on the screen at the same time. Smooth performance. Including previews with videos.

That is simply impossible to do with XP today.

How about text? Did you notice that that was being rendered in the GPU now? Whoa! Text is way faster than on XP.

Also, he claims that WinFS is done for developers. Not true wasabi! WinFS is for end users. Why? We do user testing. Research. Our users are telling us they can't find their files. I know that I can't find my files. In just the past year, I've taken almost 6000 photos. You wanna see how difficult it is to find my wedding photos? It's getting to be darn difficult. I waste a lot of time.

I can't wait until WinFS is here. Yeah, if you used a stop watch and measured the time that a file came up in WinFS today, and the time that it would come up in XP's file system (NTFS) I'm sure that XP's would be faster. But that's assuming you know where the file is. The real measure is to watch the user and use a stop watch on the time it takes for him to BOTH find the file AND open it. I've done just that and in WinFS the time is far far far shorter.

So, when you're measuring performance, you can't just measure the computational time something takes.

Oh, and he also says that "important apps" aren't being done in .NET. That is absolutely NOT true. I can name several apps done in .NET, including my favorite news aggregator.

If we make app development time shorter, all of us get better applications (and more applications). That's a major goal of Longhorn.

A few more things on performance. The PDCLonghorn has tons of inefficient code in it -- it has not gone through performance tuning at all. WinFS is currently undergoing a major rework for the next few months just to get its performance up. The rest of the system is also being performance tuned. That's one major reason why we're still at least two years from release (and, probably more). We won't release this sucker until it really rocks.

Finally, one thing Microsoft has done throughout its history is to deliver major platforms that outstrip hardware a bit. When Windows 95 came out, it was slower than Windows 3.11 in a lot of ways (at least if you measured computational performance with a stopwatch -- end user productivity was way way faster, which is why it sold so well).

Anyway, I asked you all to hate Longhorn and Ole is just following directions. I'm actually happy to see the feedback! Keep on tearing it apart.

Oh, I didn't even see the paging feedback from Ole until now. Hey, we care a lot about that. I think you'll see more work on the "fundamentals" side of the fence soon. The PDC wasn't about that kind of stuff. We wanted to spend most of our time talking about APIs and architectures, not on what we're doing under the hood. Keep in touch, there'll be a LOT to say about things like paging and performance in the future.

November 2003
Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
Oct   Dec

Referer Page
Robert Scoble works at Microsoft. Everything here, though, is his personal opinion and is not read or approved before it is posted. No warranties or other guarantees will be offered as to the quality of the opinions or anything else offered here.

Click here to visit the Radio UserLand website.
Subscribe to "The Scobleizer Weblog" in Radio UserLand.
Click to see the XML version of this web page.
Click here to send an email to the editor of this weblog.
© Copyright 2004 Robert Scoble Last updated: 1/3/2004; 3:17:04 AM.