In 2005 Google decided they were going to attempt to unsettle the online spam ecosystem by recommending that web sites add an additional attribute to a link from site A to site B if they were not verified to be trustworthy. For instance, if someone left a comment on my site who I didn’t know or couldn’t vouch for – I would add a rel=”nofollow” to the link to their site.
While it sounds like a small change, the longer term plans were for search engines to not include those links when ranking a given web sites content. It was common place for spammers to comment spam thousand or millions of blogs, leaving a link to their preferred sites. Before the rel=”nofollow” attribute was introduced, the search engines had no way of knowing if a web site owner trusted that site and as such had to use other methods of measuring trustworthiness.
The intention is clear, by removing the incentive for a spammer by using rel=”nofollow” in the links – it was hoped it would have an impact on the insidious and incredibly aggressive spam ecosystem online. Unfortunately, spammers aren’t the kind of people that just roll over at the first sign of a battle and the war on spam raged on. As the rel=”nofollow’ attribute gained momentum over the years, spammers have subsequently sought out web sites which were considered dofollow – in that they have clean links that the search engines will count when ranking a web site.
As I’ve upgraded WordPress over the years, I’ve been carrying the same set of themes or templates forward through the upgrades and hadn’t bothered to upgrade to a newer base theme. Of course this meant that links from my site were dofollow links and were valuable to the spammers. So valuable in fact that my site has been listed countless times on different forums as being a dofollow blog – essentially proclaiming to the spammers of the world that they should target my site for high quality backlinks.
I’m happy to say that after a recent upgrade to WordPress 2.9.x, I ported my current theme over to a new base theme provided in the 2.9.x code base which uses rel=”nofollow” links by default. In fact, if I want to allow people to have clean links – I’ll need to specifically allow them by use of a plugin – which is fine by me.
This will no doubt upset the spammers out there that thought they’d be getting easy, free dofollow links from my site. While I understand why they want my clean links, I also hate having to deal with comment spam – so I’m going to leave my comments rel=”nofollow” from now on.
The WordPress development team released the next major update for the popular blogging platform, version 2.9 on the 19th December 2009 – code named Carmen. With it came over 500 bug fixes and enhacements from version 2.8.6 which was the previous latest build available.
When the WordPress team release a minor revision, moving from 2.8 into the 2.9 version space I like to wait until the next point release has been made before upgrading. Normally a raft of very subtle issues will arise when it hits the community and millions of people are using it instead of only tens of thousands.
On the 4th January 2010, WordPress devlopment team released version 2.91. after going through a beta and a release candidate which included about 25 bug fixes and enhancements over the initial 2.9 release.
Instead of doing a standard upgrade, which involves me taking a databae backup, unpacking the WordPress source and uploading it – I instead opted for a completely clean installation. This time around, I still took the backup (good practice) but instead of just uploading the source code – I deleted everything first and then uploaded a fresh copy of WordPress – which mean no extraneous files laying around.
At the same time I’ve gone back to the standard WordPress theme for a short period of time and installed a raft of very popular WordPress plugins to make things run a little more smoothly. In the next few days, I’ll be moving back to my old theme and will update my own WordPress plugins to make sure they work with the latest version of WordPress.
Following on from my 2008 web statistics, below is a summary of what traffic the site took in 2009.
In 2006 the site took about 95,000 visitors, increasing to 145,000 and declining to 135,000 respectively in 2007 and 2008. In 2009 the site took 106,930 visitors over the entire year which resulted in 136,525 pageviews. It’ll come as no surprise that from a traffic driving potential, a lot fewer people are interested in reading about my personal ramblings compared to technical style posts that I used to post.
While last year saw a couple posts catch a moderate amount of attention and punch through the metronomic rise and fall in traffic each day, in 2009 none of my posts really got any traction within the greater internet. Not surprisingly traffic did start to decline towards the end of the year, however I’m happy that it wasn’t obliterated like it was last year when I moved web servers within the same host.
The traffic breakdown, just like in 2008 shows the complete dominance that Google has within the web search market. Yahoo! are still the first non-Google search engine and is still delivering approximately 2.5% of the traffic the they were in 2008. The latest addition to the web search ecosystem is Microsoft’s Bing, which sits at position five. Of course, that isn’t a fair comparison since they haven’t been around for the entire year. If you count Bing, Live and MSN together they drove about 1750 visits for the year putting them in at fourth however by the end of May 2010 I expect Bing to have delivered 2000 visits – narrowing the gap against Yahoo!.
The most popular posts for the year were similar to 2006, 2007 and 2008 but with a few newcomers:
- Disable Options In A Select Dropdown Element
- Oracle RETURNING Clause
- HP Laserjet & Windows Vista Driver Support
- ORA-04030: out of process memory when trying to allocate <x> bytes
Removing those posts from the top of the list since they clearly dominate, changes things a little:
- Making HP Laserjet Printers Work In Windows Vista
- Oracle Dynamic SQL Using The DECODE Function
- ASP Error ‘ASP 0104: 80004005?
- ORA-06552: PL/SQL: Compilation Unit Analysis Terminated
- Australian Idol 2006 Contestants: The Real Contenders
However still none from 2009 were showing up in the list. Isolating the posts written in 2009 and the landscape is vastly different:
- Apple iTunes Store Account Signup Process Needs Work
- Windows Vista Business Double Clicking On Single Click
- Best Home Phone Plan & Telstra
- Apple iTunes Account Verification Has Poor Usability & User Experience
- Gold Coast Beach Weddings Are Spectacular
I find it telling that my two gripes about the quality of the Applie iTunes account sign up process are within the list. You’d assume a company with a market capitalisation of nearly USD$200 billion would have such a visible component of their business highly polished but it just goes to show everyone has their problems. Having a home phone plan comparison post residing at position three is just more evidence that the consumer is becoming more savvy by researching online, even when purchasing offline.
Onward and upward for 2010!
At the end of last year, I posted the 2008 blogging statistics for my site including the previous years. In keeping with that trend, below is the amended table which now includes 2009 as well.
||AVG. POST LENGTH
||TOTAL POST LENGTH
While I haven’t managed to get back to my peak of 2007 in the total number of posts, I’m happy that I’ve managed to increase it slightly over last year. In 2010, I’m going to try and increase it again to between seven and eight posts per month. While that doesn’t seem like that many, it can be difficult to write at least one post every four days – I challenge you to try it yourself.
For those that are interested in doing the same, I’ve made the blogging statistics SQL script available for you to use.
I’ve have a reasonable number of older posts, which are half written that I never got around to completing.
Instead of deleting them, I’m going to finish them off so they make sense and publish them.
I hope we can hit 88 miles per hour or it might get messy.