Internet based DVRs, are they the future? I think so.

I used to use Azureus and RSS feeds of Bittorents to download my favorite TV shows, this actually worked pretty well; the core issues for me were:

1. Content typically came in AVI/XViD format and thus could not be watched on my extenders, this is now a non-issue (once I get a second v2 Extender) since AVI/XViD is supported on the new extenders.
2. Occasionally I would miss shows, this was almost always attributed to folks not publishing he shows I wanted on the associated RSS feeds; I would have to dig around and find the content I wanted manually in these cases.
3. I have three “heavy” internet users in my home, my wife in-particular telecommutes from time to time and the almost constant drain on bandwidth was occasionally a problem (e.g. finding the right throttling given our expectations of content availability was hard).
4. The high definition content was a bit over compressed for my tastes, we almost only watch high-definition content these days and although the content we got was consistently of a higher quality than that of standard definition content it was not “high definition” (my wife would always ask why the content did not look as nice as the stuff we recorded).
5. Settings recordings was awkward, I can do Regular Expressions like they are going out of style but having to test the expressions against a number of RSS feeds to be sure I would get the right ones was a pain at best.

Despite those issues I actually thought this was a great solution, in-fact because of the DVB-S broadcasts of some US shows (Like Battlestar Galactica) I was able to get higher definition content for channels that I only get in standard definition.

This got me thinking, how could one commoditize this? The core issues I experienced were usability and availability of content; if we were to presume at some point codec support on devices was to improve and/or content would become available in other formats we could start with the problem of usability.

The core issues with usability was really that the tools I was using were not explicitly designed for the task at hand they were powerfull general platforms that I was using to address my needs, if we were to leverage the same techniques but build a targeted solution for this problem we could make things much “smoother”, this is what I thought such a solution would look like:

1. A Windows Service that implemented Bittorrent, Subscribed to n number of RSS feeds and managed downloads on a system by system basis.
2. A two foot (e.g. regular Windows Application) application for managing “recordings” this would need to merge the “guides” from regular MCE as well as the RSS based guides so that one can have a holistic view of their recordings in this application; this would probably end up looking allot like WebGuide.
3. You could also probably derive XMLTV based guide for the available content, and integrate that data in much of the same way the Australian MCE community has done with Bladerunner.
4. A ten foot (e.g. a media center add-in) application for managing the RSS based recordings written in MCML, again presented in such a way it looks much like the current TV experience. 

With the above approach one could probably get pretty close to a native MCE experience that “normal” folks could use, that takes you to the content problem; that is that if we want truly high definition content and/or content in a format that is playable by all the devices in my home we would have to get into the “publication” business. 

One way to approach this problem would be to leverage the Windows Live Messenger SDK and/or the FaceBook SDK (I wish more social networks exposed their data) to infer who your buddies are, using this data you could come up with a model that shares your show preferences (think new series recordings and/or favorite shows based on behavior) with your friends, this data could be represented as a RSS feed of Click-2-Record files.

You could also publish your own recordings by generating RSS feeds and torrents based on the content you have recorded and/or downloaded;  You could then use the same “social data” to give preferential settings for bandwidth usage for friends, one could also either give preference to or limit access to content to those on the same networks (e.g. I would only share content with folks on COMCAST since I am a COMCAST subscriber).

This could all work really well, and in the grand scheme of things it would not be too difficult to build; and to be honest I fully expect something like this to be very successful in the future (Microsoft itself is dabbling in this area with the new Internet TV feature they have built for Media Center); with that being said there are a few problems here:

• Unlimited Internet is not, well unlimited – As an example COMCAST is canceling users who use “too much” of their “Unlimited” Internet subscriptions.
• High speed internet in the home is just not high speed enough – Verizon is coming, and when it does it will force other broadband providers to compete more on speed but until it does the anemic upload speeds and mediocre download speeds (in this day and age) offered by folks like COMCAST ends up being a deal breaker if you have heavy users in the house unless you’re willing to accept delays in content availability (download while at work or at night while folks are sleeping).

I recently ran across a new project called Webisode, that I think is actually going to take a run at this space; I still think it might be a bit early but for there to be success here and I have yet to come up with a reasonable business model to sustain a project like this that would not be totally annoying but I do hope they or someone like them comes up with a way to make this work.

Print | posted on Monday, September 17, 2007 2:15 PM