Running a reliable blog tool and ping service

Most people do not realize how ping notification services (“ping beacons”) can negatively affect user experience for a blog tool vendor. Blog authors usually welcome the publicity but are not sure who to blame when something goes wrong, if the user even knows what went wrong and where. Let’s start with some background on the process of a ping notification (“ping”).

  1. The blog author hits the “Post” button, publishing his or her blog entry to a publicly available website.
  2. The blog tool retrieves a listing of specified services that are either platform-defined defaults or user-specified preferences.
  3. The blog tool attempts to deliver a notification message — usually XMLRPC in format. — to all of the destinations retrieved in the previous step.
  4. The ping results in a timeout, success, or failure and that status is communicated back to the user.

What if the ping results in a timeout or failure? The blog author waits and waits and eventually might see a failure message from the blog tool letting him or her know the ping failed to successfully complete. Who does the author blame for the failure? Does he or she blame the owner and administrator of one of the ping beacons he or she copied from a list on a blog marketing site or a checkbox in the blog tool preferences? Or does he or she blame the blog tool for not effectively publicizing the author’s latest entry? A failed ping results in many user complaints and general dissatisfaction for blog tools.

Different blog tools have taken different approaches to the problem of errant pings.

  1. Hand-select ping beacons known to have a fast response time and infrequent failures.

    Technorati has a ping beacon with a fast response time and high level of uptime. We work with individual blog publishers to make sure if they add Technorati as a ping option that we are able to receive the volume of updates they generate and provide a reliable user experience for the authors using that blog tool.

  2. Utilize a ping distribution tool to receive one ping and distribute the information to a list of ping beacons.

    Ping-o-Matic was created by WordPress developers Dougal Campbell and Matt Mullenweg as a default ping option for all WordPress installations. WordPress and other tools send one request to publicize the updated status of the author’s blog to whatever destinations Ping-o-Matic chooses.

  3. Create a change file hosted on the servers of the blog tool company for parsing by interested parties.

    Blogger publishes updates to its own site as a chnges.xml file.

  4. Create a stream of updates and allow interested ping beacons to connect and listen to the stream.

    Six Apart currently publishes a stream of updates for LiveJournal journals and TypePad blogs. This stream includes the full content of a blog post.

A company such as Technorati would like to index as much new blog information as possible and will read and interpret many of the different methods of notification. The ideal situation of course is when the data comes directly to us either through a direct ping or through a relationship with the blog tool. Relationships with the authors of blog tools creates a direct point of contact for expediency and reliability and helps the company better understand the unique needs of a tool and its users.

If a blog tool chooses an alternate and indirect method of publishing updates Technorati is still active in ensuring the reliable delivery of legitimate update notifications. We’ve helped Ping-o-Matic by supplying hardware and tweaking some software code to make the service faster and more reliable and continue to help open and closed source developers help their users publicize their content.

Tags: , ,