Month: February 2009

The Long Tail of the Information Glut

Lorin Olsen

 

Long Tail (via Mike Fruchter)
Long Tail (via Mike Fruchter)

Fifteen years ago, I coded my very first web site.  It was a Valentine’s day card to my wife and my kids.  The worldwide web was extraordinarily new.  Indeed, it was less that a few years old.  And I was running a web server on an OS/2 system running IBM’s GoHTTP (a combination gopher and WWW server).  I “composed” the site using a text editor (not notepad).  Yes, you really could make usable sites without fancy tools like Dreamweaver.  But the fancy tools are a huge time saver!  🙂
 
You may be asking yourself why I’m dragging you down memory lane.  Well, that’s a good question.  And I may not have a satisfactory answer.  But I was thinking about all the data I collect and process on a daily basis.  I use all sorts of tools to distill this data into some kind of information from which I draw practical conclusions for my employer.  

  • I use Google Reader as a means of aggregating lots of feed data.  Years ago, I used a tool called Pointcast.  And I moved from Pointcast to using a tool called Agent.  Since then, the volume of data has increased and the complexity of that data has also increased.  But it is extremely important to have a tool to aggregate all the miscellaneous source material.  Right now, I consume several hundred data feeds on a daily basis.  And Google Reader allows me to consolidate all of this information into a digestible form.
     
  • I use Yahoo! Pipes as a means of processing all of this data.  Right now, my use of Pipes is limited.  I mostly use it to aggregate and assemble related feeds together.  Like most geeks, I have many and diverse interests.  I must follow technology.  And my technology needs are broken down into a dozen different categories.  For example, I read social media feeds.  But I also read data center management feeds.  I am also keenly interested in politics.  In addtion, I am extremely interested in evangelical Christianity and the commissioning that all Christians share (i.e., like Peter, we are called to “feed” Jesus’ sheep).  With so many interests, it is imperative to consolidate all of my source feeds into clusters of related subjects.  I do this by joining sources that are related.  And I do this by parsing posts for key terms that can usually designate the subject matter.Why do I go through all this trouble?  Because I am interested in far too many things.  And I look at far too much data on a regular basis.  So I use Pipes as a means of filtering out what MAY be important.  In time, everyone will begin to tag their data.  When that happens, I will use a more canonical approach to assessing content.  But in the meantime, I use brute force dictionary and source data analysis.
     
  • I use other “trusted” sources as aggregators and filters for content.  Like most folks, I have a number of friends across the country.  And one set of friends are keenly interested in politics while another set of friends are keenly interested in IT technology.  A third set of friends are immersed in spirtual discovery, enlightenment and fulfilling the Great Commission.  In a way, I use these people as my editorial board.  I collect their feeds (including their Twitter streams and their Friendfeed streams) as a means of focusing my attention on the important things.  I can’t monitor everything.  And it is getting harder to do so with the passing of each day.  So when I identify a trusted source, I load their key feeds into Google Reader.  And I will sometimes put some pipes together as well.

So I use tools to process lots of data.  But it is important to note where we are in the evolution of these approaches and tools.  When I started processing data from the Internet, the tools were stone axe heads and flint.  Today, the tools are scalpels and lasers by comparison.  Indeed, the use of “intelligent agents” (promised in the late nineties) is finally coming to fruition.  We are finally in the long tail of these technologies.  The things that I did a decade ago are passé.  And the things I did two years ago are now mainstream.  More and more people will start using RSS as a source of data.  And they will use aggregators (like Google Reader) as a way of assembling and processing these massive feeds of data.  
So what’s next? If you were to put me on the spot, I would say that tagging is the next thing to really go mainstream.  Indeed, I am seeing political groups using Twitter hashtags as a means of subjectively categorizing their thoughts.  And while Blogger and WordPress have used tags and categories for a few years, I expect to see players like Facebook, MySpace and a host of other platforms start to really focus on tagging.  Is tagging new?  Nope.  Will it become mainstream?   Yes, just like blogging and micro-blogging are now mainstream.  
So what will I do on the fiteenth anniversary of my first web site?  I think I’ll just send flowers.  Poetry and photos on a website are just too “last decade” for me.
 
-Roo

Red Rover, Red Rover – Send Twitter On Over

Lorin Olsen

redroverStrategy in the information technology space is a lot like game theory – or at least it’s like kid’s game theory.  The normal inclination of companies is to acquire other companies in order to grow.  This is especially true when the target of an acquisition has a capability that you lack in your own service portfolio.
But during times of economic distress, it’s hard to cost-justify such an acquisition.  At the same time, it is very hard to justify a “buy” when you aren’t quite sure how federal regulators will perceive such an acquisition.  Enter kids games… If you remember Red Rover, you know that sometimes you call a weak player just to get numbers on your side.  And then sometimes you call a strong and pivotal player with the hopes of building your own line.  During economic situations like we are facing today, a courageous company may just try and call another company to come play on their team.
Last night, Twitter announced that they will be playing in the OAuth games.  For those not intimately familiar with OAuth, it is an open authentication model that promotes application authentiaction and data sharing.  OAuth is not the same thing as OpenID (which is for user/personal authentication).  At the same time, they DID NOT annouce support for OpenID.
Why is this important?  Well, user/personal authentication goes straight to the heart of your customer base.  If you maintain your own user database, then you can leverage that base directly.  If you rely upon someone else’s user database (e.g., Google or Yahoo’s users), then you can’t directly tap that base – at least, not as easily.  So I’m not suprised that Twitter is signaling OAuth.  They want to have a broader reach of interoperatiblity.  Unlike Facebook, Twitter doesn’t have a “platform” to exploit.  So it is in their interests to play on a field where other company’s can agree to work in a friendly (coopetition-based) model.
It also makes sense that they aren’t announcing any plans for OpenID.  Twitter just said “no” to becoming an appendage on someone else’s platform.  But they aren’t big enough to do the same thing that Facebook wanted to do (i.e., buy others and build themselves). This is especially true since there isn’t a LOT of spare VC funding these days.  So Twitter will keep its user base and “play nice” with others – for now.
By announcing OAuth support, Twitter is making itself an even bigger target.  But this time, they may be the target of a different corporation – and it smells like an infrastructure player might be in the mix.  Specifically, I see either Google or Yahoo! making a play (hostile or otherwise) for Twitter.  Personally, I think that Google is trying to bring Twitter into their universe.  It would make Google an even more compelling services “platform” if they had a “partner” who could pull the micro-blogging crowd into orbit.
Is Google thinking about it?  Who knows?  But they are certainly a big player in OAuth.  Their entire OpenSocial initiative hinges on the success of OAuth.  And recent posts by Google seem to indicate that they want to “play nice” with other companies and other services.  If nothing else, maybe they can drive more services to rely upon Google infrastructure (i.e., not their search infrastructure).
At the same time, Yahoo! might surprise us.  Their new CEO wants to make a big splash.  And there could be no bigger splash than an outright pruchase of Twitter by Yahoo!  I hope that this does not happen.  But if I were the CEO at Yahoo!, I’d really be tempted.
-Roo

Come Out, Come Out Wherever You Are

Lorin Olsen

cell-towersCell phone coverage is quite interesting.  I use a Sprint-based phone and my wife uses and AT&T phone (both provided by our employers).  So both of us have different qualities of service in various points throuhout the house.  I have reasonable coverage on the perimeter of the house.  And my wife has excellent coverage on the east side of the house.  Thanks to Google Latitude, I now know why Cindy has different coverage on different sides of the house.
Cindy has a BlackBerry Pearl.  And I have helped her to set up Google Latitude on her phone.  BTW, the BlackBerry version of Latitude is quite good.  And it seems quite different than the Windows Mobile version that runs on my phone.  But that is grist for a different post.  I had a few minutes tonight so I figured that I would finalize the setup so that her phone was associated with Google Latitude and her Google account.  I also went ahead and set her account to update her location automatically.
Since her phone doesn’t have GPS, Google got the location information from the cell tower that Cindy’s phone was associated with.  So Cindy and I learned that her phone was communicating with a cell tower across the highway about a mile south of us.   I thought this was kind of cool.  So I called her into the classroom to show her the Latitude application from within iGoogle.  And she walked into the classroom on the east side of the house.  Within two minutes, I noticed that her location shifteed to someplace to the east of our house.  In fact, she had moved between towers just by moving thirty feet to the east.   And when she walked back to the living room, her location changed in Latitude – indicating that she had associated with the original tower.
Both of us thought that this was realy cool – and quite startling.  Google has done a good job making sure that the customer must provide approvals before anyone can see/consume location information about your account.  But most folks won’t be paying detailed attention.  Therefore, I wonder how many people will unkowingly give all sorts of location data to all sorts of people – including Google.  Yes, tracking kids may become easier.  But being tracked by Google (or other entities) will also be much easier.
-Roo

Center of the Universe? Meh, Bleh or Ewww?

Lorin Olsen

google-cloudGoogle is taking an increasingly central role in all of my computing. [Note: Please take a look at a few of my recent posts.]  Here is the quick list:

  • All of my personal email is processed via GMail.  This includes my ISP email that I pull via GMail.  And the introduction of offline capabilities only locks me further into the Google camp.
  • 99% of my searches use Google.
  • I consume 95% of all of my RSS feeds using GReader. The other 5% is processed using Yahoo! Pipes.
  • I use Chrome over 80% of the time. I still use Firefox (for some of the extensions I love). And I still play with both Opera and Safari. But these will dwindle, not grow.
  • I’m beginning  to use Latitude for my location-based activities. I’m not sold on it yet. But Brightkite is going to have a tough time keeping me.
  • I use GTalk as one of several IM ecospheres I routinely frequent for personal communications.  This is even more important as my company will (in all likelihood) interconnect its enterprise IM solution to the public Jabber infrastructure via the GTalk servers.  Personally, I believe that XMPP will soon become the modern-day equivalent to SNMP (the protocol that unified all email systems globally).
  • I’m using iGoogle as a portal to all of these services. The current iGoogle actually has some very nice features.  It has certainly improved greatly in the last two years.  And the integration of GReader and Latitude make it far more compelling than it was a year ago.
  • I used to run this blog on Blogger (a Google asset).  But when I joined Microsoft three years ago, I had to drop Blogger.  Of course, squatters came in and took my old blog address.  So when I left Microsoft, I started to run this blog on WordPress.  Actually, I like WordPress a lot more.  When I transition to a site on my own domain, I will still be using WordPress.  But Blogger (Google) has a huge number of active blog sites.  And the fact that it is a free service will bring people to Blogger and get them hooked on Google tools and Google advertising.
  • Over the past year, Google has been a pivotal player in the Open Social movement.  And they are achieving an even great role in open/federated authentication.  With their work on OpenID, they will be one of the three top players in any federated authentication solution.

It is clear that Google has a huge postition in my universe.  And that position is growing, not shrinking.  With more and more Google assets in common use, there is a defniitve gravitational “pull” associated with these computational bodies.  Indeed, the gravitational well of Google is getting larger and larger.  For example, if I had the spare change for a new phone (and no time left on my current contract), then I think I might pick up a GPhone rather than an iPhone.
This is becoming quite reminiscent of the place Microsoft began to take in the early and mid-nineties.  Is Google becoming the next Microsoft? Gosh, I don’t want to be the millionth person to discuss that hypothesis.  But it might be worth looking at Google as the next IBM.  IBM had a lock on an entire platform (the mainframe).  Every decision that was made within the corporate data center had to factor in the current and future blueprints from IBM.  And it sure looks like Google is more and more capable of exerting this kind of influence – especially as cloud-computing evolves into a reality.
At this time, Google is a key vendor of some important client technologies.  And Google is a key player as a vendor of infrastructure services and capabilities.  Google is also a new and increasingly important player in communications. And Google is practially the only player in search-based advertising.
So what’s next for them? Is it hardware? I don’t think so.  They don’t mind specifying standards – like their handling of the GPhone.  They won’t build it – but they’ll design and direct it.  Is it software?  Well, I’d have to answer yes and no.  They will build the software and then distribute it.  But they will do this as a means of providing a platform for ad-based revenue.  And right now, Google is able to monetize all of their software and standards investment via a simple advertising tax they impose on nearly every platform.  Will they focus on services?  Hmmm. That might be interesting.  But I doubt that they have the stomach for that – especially in a market that is shrinking – not growing.
Is Google a Microsoft or an IBM?  Who really knows the answer to that question.  But should Google be watched by federal regulators?  I would have to say yes.  While I don’t think that Google is exercising monopoly power to the disadvatnage of others (either competitors or customers), I do think that they have the means to do so.  But will the current governement (as controlled by the Democratic party) have the stomach to play watchdog over a company that is known to fund many a Deomocratic pet project?  Now there is an interesting question.
-Roo

Google Is The Center Of The Universe – Part II

Lorin Olsen

google-latitude
I am a casual user of Brightkite.  I use it to log into four or five different locations on a regular basis.  And I use it to do a couple of location-specific tasks – such as location posts and location photos.  But as a general rule, I use it only a handful of times each day.
But there are a couple of things I love about it:

  • Entering location data is easy.  I have a few dozen Placemarks set.  And I can enter my location by referencing one of my preset Placemarks.
  • Phone integration is quite mature. SMS data entry is a snap.  I can send an SMS message to a single phone number (after associating my phone with my Brightkite account).  Checking in is easy. And sending pictures and text are a breeze.

But I am not altogether thrilled with everything that Brightkite does.  Indeed, I have a couple of problems with it.  

  • I have to build a completely separate friend/follower base.  I hate this.  I want to have one friend/follower/contact base for all my social apps.  It’s not hard to use Brightkite’s functionality.  But it is yet another user database.
  • I have to use a separate user/password to signon to the service.  I really want a single signon or federated signon.
  • I really want my location solution to integrate with my mapping solution.  In this way, I can do dirctions and distance calculations simply.
  • I would love for my social systems and my location systems to work together.  Specifcally, I want to be able to IM or send tweets to folks that are around me.  And I’d love to have the ability to use XMPP as the infrastructure for these communications.  In that way, I can leverage the immense automation potential of Jabber-based systems.

Do I need another location-based system?  Probably not.  Do I want another location system?  M0st definitely.
Enter Google Latitude.
 This morning, Google told the world that Latitude was available for testing and use.  So I downloaded itand gave it a try.  Google Latitude does some things extremely well.

  • Google gets the basics right. I can check-in. And I love the fact that I can use Google Maps to put a pin where I am located.
  • Google does get the GPS fundamentals right.
  • Google has mobile clients for all the major phone platforms (although I am hearing that the iPhone client really needs a little help).
  • Google uses the single signon system that my other Google apps use.  More importantly, it can leverage OpenID in a federated model.
  • Google integrates with Google Talk.  Hence, it is XMPP-enabled.
  • Google integrates with Google Maps and Google Earth.

While it would be fair to say that Google has the framework and fundamentals right, it still needs to polish some of the rough spots.  In particular, the check-in process is particularly clunky – especially for phones w/o GPS.  Since GPS is avaialble on all “new” phones, Latitude reaches the tech savvy folks who like new toys.  But the reach is limited for those phones w/o GPS.  And the SMS functionality appears very immature.  There is no simple way to check in (via a service like Placemarks).  
Bottom Line: Google Latitude is a great first step.  It gets the fundamentals and infrastructure right.  But it really needs some work on the UI.  With a few more polishes to the UI, this WILL replace my Brightkite use.  But for now, I will use both and incur the overhead of doing the same thing twice.  This offering will force Brightkite to up its game or risk losing everything it has worked so hard to earn.
 
-Roo
P.S. Louis Gray has a great write up on Latitude and its impact upon Brightkite and Loopt.  His thoughs are available on his site.