Coming from the digital video world, I'm entirely opposed to standardizing on using GB instead of GiB in any context. Because lots of other industries, like telecommunications and digital media, have long used the correct ^10 numbers. Come to think of it, Apple was the last company to use KB/sec for compression bitrates, and even they dropped it as of QuickTime 6 back in 2003.

When you get provisioned bandwith, you're getting ^10 numbers. And when you compress video, you're using ^10 values (so, 20 Mbps is really 20,000,000 bits per second, not 20,971,520)

It's a big pain to have to always convert between the real values and the erroneous ^2 values when figuring out how much video we can put on a disc/drive.

It will be a horrible thing to have GB mean different things in different contexts and to have to know when to do or not do the conversions.

You won't get a FormatLikeWord95 tag if you're not using Word 95, of course :).

It's important for lots of users that they can maintain functional compatibility and bidirectional conversion with older Office documents, so that was a core design goal of the .???x formats.

You may not need that yourself, but a lot of people do, and the new formats give them that in a much more interoperable, searchable, and efficient XML-with-.zip structure.

It seems like 90% of the compliants about the new formats don't even acknowledge the design goal, which makes any discussion of how good it is pretty irrelevant.

A clean-room new office file format would have had much worse interoperability and hence much less adoption, for only aesthetic gains at best.

If OpenOffice doesn't have Normal/Draft, it's dead to me. I don't want to see page breaks when I'm still writing the darn text! That's bugged me about other tools for getting on 20 years now.

That killed Pages 1.0 for me as well.

Well, yes, of course you'd use PDF if you want to get the exact fonts and layout on the other end. .docx is a content creation format. It can work somewhat for that scenario, but it's not its whole reason for existence.

But nor are you going to write an article in Acrobat. Note that the Office apps now have an excellent "Save as PDF" mode that's much faster than Distiller.

ValueCost.

What does the Student/Home version of Word cost? $80? If you use it for 10 hours a week for a year, that works out to $0.08 an hour. Total rounding error for anyone who makes money writing, and pays for itself many times over even if it only boosts productivity 5%.

As for Word, I'd say its deep strengths are in easy, productive composition of structured prose, plus great revision and collaboration features. And it's not just about feature-to-feature checklist, but about how all the features work together and are preseted. I've never seen anything that can easily defork two different revisions of the same document like Word, comparing and letting you pick change-by change with all the variants on screen at once.

While it's no layout powerhouse, it works very well for making structured documents if style sheets are used correctly, which can them be enhanced in LaTeX, InDesign or whatever.

Yes, that's correct, if you're running Aero Glass. That was true for Vista as well, to a lesser degree, but Win 7 expands GPU support to GDI+ compositing, improves memory management, adds concurrency for multiple applications, etcetera.

An easy test is to open Task Manager and watch the CPU meters as you shake a decent sized window around the screen quickly. WIthout Aero Glass on you can get near to saturating one core, while with Aero Glass on it won't have much CPU impact at all.

Here's some other detailed information:
http://blogs.msdn.com/e7/archive/2009/04/25/engineering-windows-7-for-graphics-performance.aspx

A Netbook is a system with a very low powered single-core CPU. Everything you can do to move things off the CPU makes everything else faster. Windows 7 can offload GDI, window compositing, and many other effects to the GPU (even one as relatively weak as in Netbooks), saving a ton of CPU performance. And thus making everything else faster, even if it's just looking at a web page that's running some Javascript or Flash.

I just upgraded my kids' Dell Mini 9 (1 GB RAM) to Win 7 RTM from its OEM XP config, and it's remarkably snappier even just doing web browsing, even with a GMA 945.

http://en.wikipedia.org/wiki/Features_new_to_Windows_7#Desktop_Window_Manager

Media formats are pretty orthgonal to the browser; most playback is via plugins, and there are WMV playback plugins available for all major browsers. Microsoft has a NSAPI implementation for Firefox, Distributes Flp4Mac for free. And of course Silverlight supports WMV (along with MP4 and MP3), and is supported in the codec pack for Moonlight.

WMV is quite widely used for premium content where the studios require DRM, as Windows Media DRM and PlayReady is the only widely deployed DRM available for license (Apple's FairPlay is only available to Apple as a publisher and Apple as a device vendor). So WMV is used for Netflix, Blockbuster, and other services in the USA, and it's used even more widely in Europe and Asia's video services.

But again, nothing to do with the browser.

With Silverlight supporting H.264 and AAC now, the actual codecs and media formats aren't the interesting point of competition. The big differences between Silverlight and Flash today are much more systems layer stuff like adaptive streaming and rich presentation layers. HTML5 is interesting, but even the proposals are well behind what Flash and Silverlight have already deployed for complex players.

I'm 38, and have been loathing cursive for a good 30 of those years myself :).

In the Portland Public Schools, we had some crazy oscillation between cursive and "italic" - I vividly remember a school or district-wide switch from the cursive we'd been learning for a few years to a new italic form of writing that I found much preferable. Probably 4th-5th grade? And then when I got to middle school everyone else was still doing cursive and it threw me for a loop.

Wikipedia may have solved the mystery for me:
http://en.wikipedia.org/wiki/Getty-Dubay
The Getty-Dubay handwriting method was developed in 1976 at Portland State. We must have been a test location or something.

Anyway, I was a bit of an outlier. In 5th grade I was simultaneous in TAG and advanced English and math classes while doing special ed for spelling and handwriting. I've also been prone to hand cramping trying to push the pencil down so hard. It was a 10th grade math teacher who came up with the idea of trying other writing implements. I found that a fine-point felt tipped pen was much more legible and much less tiring. And I was able to apply a calligraphy class I'd taken before, which made a huge difference. Erasers are one of those good-only-in-theory things anyway; just crossing out a mistake worked fine. And a calligraphic technique, since it felt more like art than a burden, slowed be down enough to not be so sloppy, which probably sped things up on net. "Make haste slowly" is an important lesson I've been trying to teach my own kids :).

My 9 year old son's in a similar boat, in a TAG magnet school, but well below grade level in handwriting. Since he's quite good at drawing, I'm hoping to get him interested in calligraphy as a way to apply those skills and thoughtfulness to handwriting, instead of treating it as a burden he's racing through to get over.

My parents had gone to Reed in the mid 60's when they had a famous calligraphy teacher, and my father has continued to use that as his formal handwriting style since. On a trip to Europe he bought me a nice Lamy italic nib fountain pen, which really did nicely for calligraphy, and I started using it for everything. It kind of wigged out my physics teachers when I started turning in exams and homework that looked like an illustrated manuscript, but it sure was much more legible than cursive pencil ever could have been.

For all the time I spent learning handwriting, and the many classes I actually enjoyed, I have to say the single most valuable class I took in all of school was typing in the 6th grade. I do a good 120-130 words a minute these days, which sometimes is an almost linear productivity gain.

Speed? That's the best argument to run Aero Glass. It offloads compositing from the CPU to the GPU, improving performance for apps.

Easy test to try with Glass On/Off. Open up Task mangager.

Open up a nice big JPEG image or something.

Grab the window and shake it like crazy. Watch your CPU meters

With Glass off, you can peg a whole core trying to render all that motion. But with Glass on, the GPU's doing the work and your CPU load hardly goes up.

Windows 7 improves this over Vista, as it pushes GDI-style 2D rendering to the GPU as well, and adds hardware YUV overlay support back.

http://blogs.msdn.com/e7/archive/2009/04/25/engineering-windows-7-for-graphics-performance.aspx

It's one of those offhand jokes that slowly reveals itself as multifaceted brilliance, no?

I'd say in life the big challenge is almost always getting people on board with your idea at all. Trying to titrate between getting them to believe in it enough to figure out the flaws but no so much that they run with it seem more difficult than useful. But if you have to, pitch it to the people you'd want to work with on the project. If they get psyched and want to run with it, you can run with them.

But as many have said, ideas aren't hard. It's implementation, revenue, and luck that are hard.

To that end, here's a million-dollar idea I'm not going to do anything with.

It's Bitter - Binary Twitter, for the geeks.

Because for lots of us, 140 characters are too imposing. It's like doing multiple Haikus at once. But binary we can do.So, with Bitter, you just get a bit. Your status is 0 or 1. You change it as appropriate. Eat a yummy peach: 1. Miss your bus: 0.

And it saves a ton of time for the author AND the audience: you can immediately see the status of a million friends in a 1000x1000 bitmap! Just try that with Facebook.

And just think of the data-mining and mashup possibilities. Track how and when people flip their bits. Tie it into Facebook and other profiles to see what coorelations there are between activities there and bitflipping. It would be glorious!

Oh, and the million bucks? Just sell $1M in advertising. An idea like this, it can't miss! What could possibly go wrong? :).

Anyway, it's only half a joke. It'd be an amusing little Facebook app or whatever. Could make some money maybe, but maybe wouldn't. But I've got a book to finish, and a third of another book after that, and a house remodel, and making Silverlight awesome. No way I'd ever get to it.

Ideas are cheap. Attention is expensive. And other people's sustained attention is REALLY expensive (=employee). That's the hard part.

Which patent issue? Is there something specific you're thinking of that hasn't been covered under the Community Promise etcetera?

It's more complicated than that; it's not just classic stream switching. The big differences compared to past approaches are that there's not buffering on stream switching, http is the only protocol required, and the indivdual chunks of video are small enough to get picked up by proxy caches. The latter delivers a lot of the scsalability value of multicast, but with the existing web infrastructure.

http://www.microsoft.com/downloads/details.aspx?displaylang=en&FamilyID=03d22583-3ed6-44da-8464-b1b4b5ca7520

WINE doesn't have the active support of Microosft and a clear license for implementation details.

This player is already doing stuff that was in Silverlght 2 that's not even being proposed as part of HTML5, like Smooth Streaming.

Are you running the current Moonlight Preview? That's required for managed code support, and hence Smooth Streaming.

This definitely works in Moonlight. If you can see it, you've got it installed correctly.
http://www.iis.net/media/experiencesmoothstreaming