Mike Davidson, Newsvine.com, and over a dozen other sources have all been blogging today about the acquisition of Newsvine.com by MSNBC.com.

Unfortunately, I’m sick today and won’t be writing much about this, although I do have lots of opinions on the acquisition. Also, I highly doubt my opinions will be so unique that the other sources won’t have already covered them by the time I am repaired.

Update: Ok, so I’m feeling better today and it’s time I weigh in with my meager opinion.

As someone that works with “news” (quotes are sometimes important there, aren’t they?), I’m often faced with an argument of “should we allow user generated content in our product(s)?”

For me, the argument isn’t whether or not you should ‘allow,’ it’s at what level. For me, the argument is user generated content vs. user contributed content.  You see, I’ve always felt that the Internet was a grea way to get instant feedback on what works in a news organization and what doesn’t (for instance, what stories online gather the largest audience; this information may help build tomorrows leader story). This theory doesn’t always work, of course (considering America’s apparent obsession with reality TV and celebrity disasters), but when it does, instant feedback is a godsend. Further, a news organization already accepts contributions and suggestions (what do you think news wires are for?), so why not glean content from your core audience and feed it back to them?

Ok, that last line was a bit OSNAP.net-ish, so I’ll back off a little. What I’m trying to get at is my hope’s for Newsvine.com in the future are the following:

• MSNBC.com will use Newsvine.com as-is; meaning, they’ll just watch how Newsvine.com authors, contributors and visitors interact with the Web site and what ‘they’ find important (they being the audience, not MSNBC.com)
• MSNBC.com will integrate the user contributed portions of Newsvine.com into their own Web network, leaving the user generated portions alone.

There’s a lot more to be said about this deal (such as, hey Mike! mind giving me some pointers on MagStand?), which will have to wait until later. I can only handle so much on the computer today.

Many Seattleites now know the “Seattle Streetcar” by another name, the South Lake Union Trolly, and Kapow coffee knows just how to capitalize on that.

I spoke with Jerry Johnson at Kapow, who told me he was informed by city employees that

“they expect 100,000 or more riders simply because of the t-shirt campaign.”

That expectation could undersell the weight of the t-shirts and Web site, as tourists and journalists have blogged and reported on the name Worldwide.

I’m looking to mark my calendar for the inagural rides and will be sure my wife and I are in attendance, wearing our S.L.U.T t-shirts proudly.

The smart folks over at Curious Office have a good post today about user needs and how it relates to marketing.

I cannot overstate how important it is to remember the “fail fast, fail cheap” idea embedded in many Agile/XP programming methodologies and how well that mantra can be applied to new ventures. Feature bloat has been a problem with my projects in the past, and I’m finding much better success in breaking down a project into “what’s good enough” type of packages. If launching an application with features 1-10 would take 1 year to develop, could the market be better served in 3 months with an application that has only features 1-3? If so, wouldn’t it be better to launch a shorter-featured application and allow the userbase to provide feedback on what new features should come next? Perhaps the users don’t want feature #4, and without launching fast (and cheaply), I would have wasted time and resources on an unwanted feature.

I recommend reading the Curious Office post. You can never learn too much about marketing and timeliness. I certainly am a shining example of that need to learn.

So I’m working on this application, titled “ImageServer” that allows users to search hundreds of thousands of stock images. The application works many times faster than the previous software my employer was using, and returns results in a much better format (drag+drop for lightbox and download, flagging of inappropriate results, etc).

I am, however, facing a problem with my application that any search-related software developer must deal: How to return more relevant results to a user.

You see, I’m trying to figure out what determines “dog-ness” in a given set of images. For instance, in the search above, I’ve queried for “dog” images. Now, of the 300+ results, how does one image gain higher ranking of “dog-ness” than another? Currently, any image with a “dog” keyword is returned, in no particular order. Will an image tagged with both “dog” and “dog house” rank higher than an image simply tagged with “dog” and” puppy?” (considering there are two keywords with ‘dog’ in their text for the first). What about the condition where the user is looking for a single dog in a photo?

Relevance is key, especially with predictive (read: recommended) results.

Now, I know that search algorithms are proprietary, but there must be some literature on the topic. Perhaps I’ll perform a Google search for one.