Wednesday, May 28, 2008

CDNs On The Ropes

If you work for a Content Delivery Network (CDN) you might want to shop your resume at Google. Current CDN bandwidth rates are running between $1.60 and $.85 per Gig. That makes for a very tidy margin when buying commodity bandwidth at $.06-08 per Gig. With the introduction of Google Apps Engine pricing of $.09-.11 per Gig the game has just changed. Google is starting a bidding war with Amazon for the Platform as a Service (PaaS) market and the collateral damage will be the CDNs.

I recently reexamined the benefit of migrating our video storage and streaming from Amazon to one of the top CDNs but could not come up with a valid reason to increase my streaming costs by 500%. The classic argument from the CDNs is that with thousands of regional edge nodes, my content will be delivered quickly and locally. The reality is that quality of service in streaming services is often less about the number of hops to the server and more about the through-put of the last mile to the consumer. If there is a 250ms lag at the beginning of the connection but the stream stays connected during the live of the download it's transparent to the user but let that same video file start buffering and the customer is going to complain. Bottom line, the CDN brings one big value, the ability to scale to big loads if your content goes viral. We did the math, we can build out a nice scaling infrastructure and still have money left over on Amazon. Let's face it, most of our applications will never have 10,000 requests per hour.

The destructive cycle of technology always weeds out products with relatively high profit to value ratios. As the PaaS solutions come online, more developers will be able to access the low cost commodity pricing and it's going to drive down the value of CDNs significantly.

Saturday, May 24, 2008

Go to Michigan State for a Stone-Age Education

I was recently forwarded a link to a directive issued last November by Michigan State University telling all faculty that they must "Avoiding use of online software tools such as Google Apps, Gmail, and Microsoft Office Live". In their Q&A on the MSU faculty website they go further to say that all faculty must discontinue any use of online software of any sort, including personal email accounts and collaborative accounts on Google or Yahoo that may be used if the faculty member works at other universities. This edict effectively says you cannot use online software services in your professional or personal lives, period.

I've run into people who work in technology but don't understand technology progress and the creative destruction life cycle before and I certainly have seen when lawyers run a muck in an organization. This seems to be a case of the later, the main complaint is the End User License Agreement (EULA) presented by Google that states that they have the perpetual right to use and store what you post in Gmail or Google Docs. While this on the surface will cause most lawyers to spontaneously convulse, a bit of context should be considered before reacting. The reason Google wants to store and share information you have uploaded is to sell advertisement. Nobody at Google (as far as I know) is stupid enough to want to sell your personal or business documents to the highest bidder. On the contrary, Google has been one of the stanchest protectors of it's users data. Google is simply sharing keywords associated with your page to their ad serving engines and the third party involved is the advertiser who posts ads associated with those keywords. If I'm an advertiser and my ad shows up on your personal Gmail in box page, that doesn't mean I can read your email.

This sort of thinking is the 21st century equivalent of outlawing automobiles for fear of the impact on horse sales. Software as a Service is here to stay. Cloud computing is not going away. The MSU stance on using online software services is exactly the wrong approach. All software services are not equal, all EULAs are not the same. Educate your faculty (ironic to tell a university to educate isn't it?) on safe practices. Participating in the process at least assures you have a voice. The MSU policy sounds like Jim Jones headed to Guyana. Watch out for cool-aid everyone.

With 46,000 students and 4,800 faculty (gather from wikipedia, another "bad" website) MSU probably has tens of thousands of policy violators right now. Wake up MSU! The genie is out of the bottle already. As a parent of a high school student I can say that MSU just made the stone-age college list. My daughter and my money are definitely not going there.

Tuesday, May 6, 2008

DigitalChalk teams with IBM for Closed Captioning

We are very pleased to announce that DigitalChalk has jointly partnered with IBM and Hunter College to be the first online video training system to offer automatic closed captioning capabilities.

What is exceptional about his process is that it's fully automated. Prior to this technology being introduced, user-submitted video content online had to be manually transcribed and the closed captioning had to be loaded into the play script using a tool such as Magpie. By sending the video through a text-t0-speech translation engine and getting a full transcript with time coding, DigitalChalk customers will get (CC) automatically. We are also offering an online editor that let's the trainer tweak the timing and fix any errors before publishing.

Monday, May 5, 2008

Google wins on (any) Microhoo deal

Yes, I know that Microsoft has officially walked away from a possible hostile bid move on Yahoo. The market pressures that pushed MS to attempt the friendly offer in the first place have only continued to escalate however with the emergence of a Yahoo/Google search ad deal. It very well may be a poison pill for MS if Google can ink the deal to serve ads onto the Yahoo network but we aren't there yet. MS has a major hurdle to cross, examples of the trending to Platform as a Service (PaaS) solutions like the deal between Salesforce and Google underscore how quickly the tides can change in technology. Microsoft can either become a holding company of mature products (read IBM and Oracle) or innovators (something never before done in Redmond). They seem to have resolved themselves to being an investor, hence the Yahoo bid.

What Wall Street and most of main street doesn't understand is how utterly bad a Microhoo company will perform. Mixing oil and vinegar only makes a vinegrette when there is an emulsifying agent involved. I see no binding agents between these two organizations. Right now Microsoft and Yahoo are about as technologically and culturally opposite as I have ever seen. What platform wins? The dominant Web platform (Linux) or the dominant small intranet platform (Windows Server)? The mass defections at Yahoo will likely become a wholesale exodus. Microsoft may think it's buying market share but the social community market is as fickle as teen shopper. Google will benefit no matter what direction this deal goes. They have to be praying that the deal happens and the Microhoo goo gets thick. If not, they are dealing with two organizations that have proven they can't compete in the Web platform space. Google's got to be loving this.

Twitter Updates

    follow me on Twitter