You can find the source code for this entry at:  http://vallery.net/wp-content/uploads/2012/01/ValleryBlog.zip

So I decided to migrate my blog from WordPress to SharePoint.   The motivation was driven primarily out of wanting to “eat my own dogfood”.    Given that most of the work I do and the posts on my blog all have to do with SharePoint I thought it would be appropriate to actually host my blog with SharePoint.  

I decided to seek out a hosting provider which specializes in “SharePoint in the cloud” given that my blog is high enough traffic and important enough to me that hosting at my house isn’t an option.   I first turned to Microsoft’s Office 365 service given that I’m already hosting my personal email with them.    I was disappointed to learn that you currently cannot host a public facing website with anonymous access.   They require the site to be secured which wouldn’t work for a blog.

I emailed a couple of the other community hosting options and got a very nice reply from FPWeb.   One of the services they offer gives you a single site collection on SharePoint Foundation which you can optionally give anonymous access to.   The site is limited in terms of which SharePoint features are available but they did offer the blog template.  If you’re not familiar with the differences check out this great guide which details the plans that are available.   The site provided uses SharePoint hostname site collections which allows me to have a fully qualified domain name (http://blog.vallery.net/) point to a single site collection on their web application.  

For my use case there were a few limitations that bit me:

• Deployed custom solutions must fit into a “Sandboxed Solution”.    This is true for any of the non-dedicated online SharePoint option though so not an issue with FPWeb specifically.    This means that many of the things I’m used to doing for my customer’s on-premise installations are not possible in the cloud.   Microsoft has a good list of the limitations of SandBoxed solutions worth reading.  The things I missed the most are:
  • Ability to create custom delegate controls.    This made doing a custom searchbox a bit more difficult and less flexible than I’d like.
  • Ability to integrate with external systems.   My current WordPress based blog will post to twitter automatically when I write a new article but given that no external HTTP requests are allowed I can’t write an event receiver to do that for me.
  • I have to deploy all style assets to the SharePoint “Style Library” because I can’t write to the file system.    This could potentially have a negative impact on performance because now they live in the database.
  • I wanted to create a few custom controls to handle some of the style elements.   For example the banner on the site isn’t as flexible as I’d like.   It should really pull the title, url, and description from SPContext.current.web.   I had to hardcode them into the masterpage which means my solution isn’t portable.

• Since FPWeb only provides SharePoint Foundation it makes branding more difficult.   There are a number of capabilities found in the enterprise publishing features that I typically use with my customers which were unavailable to me. 
  • In a custom masterpage you can use the SPUrl object to find a the URL of the site collection or site in order to reference your style library (to include CSS, JS, and images on the page).   This is not available if you don’t have enterprise so I had to hard code the path to the Style Library into the masterpage.   On my development box I created a blog site collection in a managed path off the root which made the path different than my site here at FPWeb (which is the root site collection).   This caused me to have to make a quick change every time I switched from my dev box to pushing out the solution to FPWeb.

• Search is limited given that it is only SharePoint Foundation.  This is fine because I only need a site scoped search anyways.   The trouble of course is that it turns out the control that is used is very different.   I styled the Enterprise version since that is what I had on my dev box.   Once I deployed to FPWeb I discovered that none of my styles worked.   I had to go back and refactor for the foundation search box.   It wasn’t a huge issue but it took me a while to figure out why my styles weren’t working!

Once I got my solution done with my custom style I needed to get all my old content off my WordPress blog and onto SharePoint.   I had some good conversations with Metalogix at their booth during the SharePoint conference in Anaheim so I was already familiar their "Migration Manager for Blogs and Wikis" product.   They were kind enough to give me a trial key which was enough to migrate my posts and comments from wordpress across.   The process was painless and just required confirming the source and destination.   It was able to capture metadata as well like category which was handy.  

I did discover a few issues with the way the data was migrated though.  I realized that the blog author for all of the posts and comments was listed as me.   I guess this make sense since the author field is tied to a people picker and needs to be a person.   None of the comment authors in wordpress are users in my SharePoint environment obviously.    Metalogix handles this by creating a couple of extra columns on the comments list which store the original comment author, name, and email address.   I plan on writing a custom XSLT file and view to parse this at some point to place this information onto the page.  I’ve also included these fields on the new comment dialogue so I can capture it for future comments.  For now though I’m listed as the comment author for all of the comments.

The only issue I had with the Metalogix tool is the moving of images.   This could easily have been my fault as well so I don’t want to blame the tool.   My posts had a number of embedded images which WordPress stored in the wp-content folder.   My expectation would have been that Metalogix would have grabbed those files and put them into SharePoint and updated the references.   It even looked like it attempted to do that however I received a bunch of error messages which indicated failure.   My images are still linking to their original source on my WordPress site (which is still up) so they are not broken.   One of my todo items is to migrate those across by hand and then update my old posts that reference them.     This will give me an opportunity to clean up my old posts anyways which I’ve been meaning to do for a while.

With everything up and running I put 301 redirects into the .htaccess file on my WordPress server for each post.   I pulled out a list of all the posts from the SharePoint list and matched them up to their original URL (which thankfully Metalogix stores in an additional column for me).   With all of these URLs known I’m able to redirect the deeplinks to their specific posts.   WordPress has the handy SEO optimized permalinks which SharePoint does not.

Overall I’m very satisfied with both FPWeb and Metalogix.  Without either of them this project would have been significantly more difficult.   If you can accept the limitations of hosting your SharePoint site in a shared environment then I highly recommend FPWeb.

In case anyone is interested in the actual development work that was involved with the branding in all this I’ve posted my Visual Studio project.  In total I spent about 15 hours on the entire project over my Thanksgiving weekend. You can download the project at http://vallery.net/wp-content/uploads/2012/01/ValleryBlog.zip.

Just when you started to get used to the idea of Web 2.0, the trend setters on the web start tossing around the term Web 3.0. Can’t they just let us get used to one version before pushing another one on us? Fortunately Web 3.0 is in its infancy, and nobody is certain of what it really means at this point. The direction of the internet and its users is a fickle thing. It’s hard to predict what they are going to want and use. The next big thing is all about the right idea, in the right place, at the right time.

I’ve been attending the International Conference on Weblogs and Social Media conference the last couple days and listening to some of these visionaries talk. While there certainly is no consensus about what the future holds, there are definitely some key trends to take note of. I decided to go this year because it just so happened to be local for me. Normally I wouldn’t have taken the time off of work, incurred the travel expenses, etc just to go to a conference like this. For some bizarre twist of fate they decided to hold the conference in my home town. Given my personal work in blog research, it seemed like a no brainer to go to this conference.

The topics presented at the conference were fairly diverse, ranging from sentiment analysis to gender bias in blogs. There were three invited speakers and a number of academics presenting published papers. I have to say that by far I got more out of the invited speakers than the papers, although there were a few the sparked my interest. The three invited speakers were Danah Boyd (a prominent blogger), Andrew Tomkins (a researcher at Yahoo!), and Evan Williams (founder of Twitter). Here are some of the highlights from my notes.

Danah Boyd – “MySpace is *my* space”

In the talk she gave us a brief history of social networking starting with Friendster and went on to discuss a number of specifics about MySpace.

• 93% of American teenagers have access to the Internet.
• 55% of online American teens 12-18 have a profile on a social network site (That are willing to admit it in front of their parents).
• 91% use it to talk to friends while very few use it to talk to strangers.
• The number one spot in the “top 8″ is dependent on culture. Some cultures dictate that it should be your significant other, some your best same sex friend, and others say family members like your cousins.
• 66% of teenagers use MySpace in a private fashion in order to avoid marketeers and adults (parents).
• Teenagers think something is fishy on a page without ads. They suspect at some point they will be asked to pay for something if they aren’t being force fed advertisement.
• Gadgetry has broken the gender barrier. More teen girls have game consoles, iPods, and trendy cell phones now.
• If parents are flipping out, they know it’s going to be fun!

In addition to the key points about MySpace, Danah also discussed the direction that she sees social networking going. The future to her is about mobile devices. We already see the prevalence of SMS in Asian cultures, and it is slowly becoming a phenomenon here in the US with the introduction of sites like Twitter. The problem is how this move will take place given the current atmosphere in the mobile market. The major cell phone carriers control every piece of software on their phones and every byte of data that goes over their network. There is no self-interest for them to open these platforms and networks up for development. They want to be in control so they can utilize the technology to make as much money as possible. Because of the competitive nature, there is not likely to be any major cooperation between the carriers, which will stunt social networking’s growth.

Andrew Tomkins – “Social Media, Storage, and Data Analysis”

Andrew is a PhD member of the Yahoo! Research team. His talk focused on several different topics, including the evolution of search, and Flickr.

Interesting things being done with search:

• Moving away from the “10 results” model and devoting more real estate to targeted results.
• Search is adding in shortcuts to specific verticals. For example, integrating weather and movie times right into the search results.
• Google Co-Op allows content providers to integrate their results in your Google search.
• The current approach that is used is naive, based mostly on regular expressions and filtering for target words.

Is search solved?

• Search is really good at finding content on relatively open and static pages. Search does not integrate results where content might be buried in forums or other social network sites.
• The data sits behind walled gardens and is generally unavailable. There is lots of money still to be made in search, but it will require a large capital investment and structured deals with the owners of the content to let search inside their walls.
• For example, the Yahoo! Answers data is fully crawlable but the several billion posts in Yahoo! Groups are locked up from the outside.

Andrew also showed off a demo off a timeline based view of Flickr tags. This can be seen over at Taglines. He also mentioned that they are developing a “game” that users can “play” that will actually generate metadata and tags for Flickr images.

Evan Williams – “The Evolution of a Social Media Platform: Twitter “

Evan is the founder of Blogger, Odeo, and Twitter. He is one of the leading entrepreneurs in social media and brings a unique perspective to the conference. Most of his presentation talked about Twitter and the things they are doing. Twitter started just 9 months ago, and went through a complete revamp in November of 2006. Over the last few months the number of “tweets” being sent into Twitter has doubled month over month.

• 2/3rds of the inbound tweets come from the web and IM, the remaining 1/3rd come from SMS
• They are currently working on an API that will soon be released
• While they don’t have hard data, their informal research indicates that the core demographic of Twitter is 30 year old web geeks, not teenagers as you might expect.

Evan also presented a quote from Lisa Reichelt about ambient intimacy that describes Twitter.

Ambient intimacy is about being able to keep in touch with people with a level of regularity and intimacy that you wouldn’t usually have access to, because time and space conspire to make it impossible. Flickr lets me see what friends are eating for lunch, how they’ve redecorated their bedroom, their latest haircut. Twitter tells me when they’re hungry, what technology is currently frustrating them, who they’re having drinks with tonight.

There was also a quick demo of Twittervision. This is a site developed by a 3rd party that integrates Google maps with Twitter. The result is that users who send a location code with their tweet (l:Boulder, CO for example) will be shown on the map display.

Summary

While the future is far from certain, one thing is for sure. The future of social media and “web 3.0″ will be focused around two key areas. Mobility and search. Developers needs to come up with better ways to get at the information you need and make it simple to do from mobile devices. While some predict the death of sites like Twitter, I think they are ground-breakers in their field. When blogging can become a commodity that is approachable to anyone, and all of that data is well organized and searchable, that is when we can say that web 3.0 has arrived.