I had a wacky thought the other day, one that I think is actually quite viable.  Sorry, this article is not a solution but a suggestion to the Microsoft design teams.

Problem:
Media Center Extender video playback is limited by hardware to MPEG2 and WMV basic formats. 

Solution:
Transcode files into a format that is useful for Media Center Extender.  Windows Media Player 10 has a new synchronization engine for portable devices that already does this, the basic operation goes like this:

• Device publishes its capabilities (screen size, playback formats, etc)
• WMP10 determines what files should be copied over on next syc
• Files that are not already in a compatible format are transcoded to a compatible format, and are placed in a cache
• Upon sync, it copies over the transcoded files and not the original files

This all happens without changes to your current library, but each file has potentially a transcoded equivalent to be sent to the portable device.  Now, I certainly think that the current implementation of this has its flaws, but the basic idea is a good one.  I’d like to leverage this concept of “portable device synchronization” into more of a “remote device service”. 

Specifically with MCE and MCX, here is how it might work:

• MCX publishes its capabilities to the MCE machine
• MCE determines files that do not match those capabilities and places them into a transcode queue
• MCE transcodes these files into WMV9 format with a max bitrate of 4 MB/s (per the MCX’s capabilities)
• When MCX attempts to access a file, it is redirected to the transcoded version
• If the file is not yet available in a compatible format, it is added to the transcode queue and the user is notified

To abstract the idea beyond MCX’s, this works for multiple remote device types including:

• Windows Media Connect devices
• Media Center Extender devices
• Other PCs on the LAN (maybe without MPEG-2 decoders)

This suggestion will of course eat up lots of disk space and CPU power, but to be honest those things are becoming cheap commodities these days. 

By applying existing technology, it suddenly becomes possible to access my ripped DVD collection, my home videos taken from the Sony Digital8 camera, music stored in Ogg format, etc.

Please, please, please someone come up with a standard way to manage a library of digital photos!  The problem with the current state of the market is that everyone is looking for ways to lock in their own way of cataloging metadata rather than working with what we have.

I have several thousand digital images, and dealing with them is a royal pain in the butt.  Status quo is to organize images into a folder/subfolder hierarchy and deal with things that way, but there are plenty of companies trying to work with images in a more intelligent way.  Unfortunately these point solutions only attempt to solve the library management issue for their proprietary interface.  However for any given image, it needs to be viewable in the following ways in order to be useful for me:

• Windows XP native UI
• Windows XP Media Center Edition
• Web photo gallery
• Portable media center
• Email

So, right off the bat it is obvious that any metadata the gets applied to an image needs to be portable and reusable.  We have EXIF metadata embedded in images already, I would propose that we should just swallow our pride and use that as the standard.  In almost all cases that I’ve seen, image library applications try to be the “one and only” system where images are viewed and organized.  Case-in-point is Picasa, which by most accounts does some very cool things.  I can browse images by date, subject, etc but only from within Picasa.  The native PC interface is quickly becoming the least-used way that I view and share photographs, so that is not useful to me, and all of the cool context that picasa applies is completely lost when I leave that interface.

The best effort I’ve seen so far at organizing a photo library is a hidden feature in Windows Media Player 10.  It takes the same approach to digital photos as it does to your media collection – read the files, understand the metadata embedded in the files, and organizes virtual hierarchies based on that.  Any edits you perform to that metadata are re-embedded in the files themselves.  Huzzah!  The functionality is really only there to allow synchronization with portable media devices, so it doesn’t offer much in the way of slideshows or publishing (yet). 

If we can get agreement that metadata lives with the file, then all sorts of other cool things can begin to happen.  Web photo galleries can automatically be populated with information that comes from your camera.  Emailed photos can have context, and I can put together some very cool slideshows using Windows Media Center or Photo Story. 

Information without context is useless.  I’m not trying to solve the world’s problems here, but think about what we could do if we just used the information we already have:

• Browse through all of the christmas pictures over the years (based on date ranges)
• Navigate a timeline of the best images of the year, based on date taken and viewing stats (auto ratings) on the images
• Take a look at your night shots (based on exposure settings)
• Rotate images properly based on EXIF orientation settings (shocking, I know)

Next, let’s define some standard fields to track in EXIF for other interesting information such as place, subject, etc and insert that into the same architecture.  Now wouldn’t it be cool (and very feasible) if digital cameras could use GPS to track their location as well and automatically use that information to further catalog images.  Obviously I’m getting ahead of myself here, but the point is that this should be very simple stuff.

Stop trying to solve massive problems like facial recognition and work on getting the basics right.