Making and using an OPML feed of student blogs

The ICTs for Learning Design course requires each student to create a blog. The blog is then a key part of the assessment, it’s actually how we’re meant to submit assignments. The public nature of the blogs has caused some concern, but we are now being encouraged to read and comment on our peer’s blogs.

With the number of students approaching 100, this is not simple. Especially given that the main method of sharing blogs is via a Moodle wiki page. Most other students have used Blogger’s/Blogspot’s “follow” button (Blogger is the recommended platform), but I use WordPress and like RSS/newsreaders. So the following documents the setting up of an OPML file with all the students’ blogs and a brief bit about using it.

The plan is to let the other students in the course know about this and see if any make use of it. Given time constraints, the apparently novelty of RSS/newsreaders, and the fact that we’re about to start our placements, I imagine uptake might be limited.

What is this RSS/OPML thing and why?

The folk at Commoncraft have a good video that gives the idea of RSS. The Youtube version is embedded below

OPML is a way of gathering a large collection of feeds into one place. It’s just a text file with a special format.

But when you import an OPML file into a news reader, you get a single place to observe and track what is going on. The following image shows the OPML file generated here imported into my news reader (a Mac application called NetNewsWire). Click on the image to see it larger. The smudged areas represent where I’ve applied a bit of privacy to the blog URLs for other students.


I find this a useful way to keep track of who is posting what. I’m guessing that most of the other students in the course will probably use Google reader as their news reader.

So, now to generate the OPML file.

Generate the OPML file

I could have generated the OPML file by hand, but that would have been painful. A quick Google and I found OPML Builder. You give it a list of URLs for RSS feeds and it generates the OPML file for you.

So, the process was

  • Copy and paste the student blog URLs from the Moodle Wiki.
  • Remove all the extra content.
  • Modify the blog URLs so they point to RSS/Atom feeds.
  • Feed it into OPML Builder.

One of the blog URLs was password protected, so I’ve removed it from the final feed. I’ll share a copy of the file on the course Moodle site.

Of course, if the course had been designed so that students used BIM to register their blogs, the academic staff could have just shared the OPML file that BIM generates.

Use the OPML file

To use the OPML file you have to import it into your news reader. This means your newsreader will examine the OPML file, extract all the URLs for blog feeds, retrieve them and analyse them to see what’s been posted. The results of that analysis shows up the news reader interface.

There are some instructions for importing an OPML file into Google reader. If I follow those instructions I get an image like the following.

Google reader and EDED20491 feeds

Further possibilities

Feeds provide a standardised way of sharing information. Using them simply to keep track of people posting to a bunch of different blogs is only a start to what can be done.

For example, Google reader allows people to share items they think are interesting in someway. The shared items are available, to some, as an RSS feed. One application of this would be for all the EDED20491 students to share all the posts (by other students) that they find interesting or good. A collection of all the students “shared” feeds could then be gathered. One way to see what is interesting.

The Google sharing also allows the adding of comments and tags, so explanations about why it is interesting or good etc could be added.

And that’s only scratching the surface.

Reflections on a 12 year-old course site – It would be harder now

For a variety of reasons, mostly due to some current study, I’ve retrieved from the rubbish tip of history a course website I helped design and teach back in 1999. What follows here are some reflections on what has and hasn’t changed since 1999. At the end there is a bit of speculation that the current context within universities would make it harder to generate this type of course site.

What’s changed or not

Bleeding edge is now standard?

This course site was designed/implemented with a couple of unusual undergraduate project students. Unusual in the sense that they were mature age students starting their second careers. One of them had been a multimedia designer in their previous career. Which is why the look of the site is so unusual, it’s based on the cover of the textbook.

The course design included:

There weren’t very many other course websites in 1999 within this institution, or elsewhere that approached this level of use of the web.

But since then universities have spent significantly greater funds on implementing “enterprise” level e-learning systems and process.. After all that money has been spent surely there must have been a significant increase in the number of course sites that are approaching what this course site did 12 years ago? Surely all that money has led to the development of systems and processes that actually automate, or at least significantly reduce the effort required to achieve, these tasks.

Well, let’s look at the four courses I’m studying this year. Even if I focus on the fairly simple task of having a well-integrated electronic version of the study guide, none of my current courses match this 12 year-old course. Two of them have a study guide page, but it remains separate from the study schedule. i.e. the study schedule doesn’t have links to the appropriate PDF. The other two had used ad hoc approaches with either Word documents or Moodle HTML resources.

And that’s without looking for good instructional design or attempts to modify standard practice to match the capabilities of the new technologies.

Barriers to sharing

This course was implemented using Webfuse, the system that formed the basis of my thesis. Webfuse course sites, including this one, were completely open by default. The only password protected area on this course site was the Staff section which we used to share assignment solutions and discuss some issues. In addition, Webfuse was a web publishing system. i.e. it generated static web pages. Mainly to minimise load on under-strength servers.

Even with it producing unprotected web pages, it isn’t a straight forward process to share/repurpose this course site. Webfuse had various assumptions that meant its web pages aren’t straight web pages. In order to share this course site, I had to manually update the pages to make them usable outside of Webfuse.

I didn’t complete this process on all of the pages in the course site. If you come across a page with broken images or a “This is a old course site” message, you’ve entered areas that haven’t been fixed up.

Using a system, any system, reduces in someway the ease with which the content can be re-used. Even using standards limit reuse to contexts which support those standards.

Some of these course sites were once available via the Wayback Machine, but at some stage in the last 5 years, the robots.txt file was changed on the server and this meant that the Wayback machine stopped making the sites available. An earlier iteration of this site is available on the wayback machine.

LMS enforced “quality through consistency”

This course site design would not be possible at the host institution, because it is now using Moodle. Moodle simply couldn’t support this design.

Moodle (like most LMS) are generally used by institutions to achieve quality by making everything look the same. At a minimum the look and feel for this course, which connects directly to the textbook and looks okay (especially in 1999), would have to be sacrificed to fit within the ugly constraints of the institutional Moodle template.

While “quality through consistency” helps bring those from below up to some minimal standard, it also constrains those who want to move beyond back to some minimal standard. Let alone the question of responding to the diversity inherent in teaching.

Beyond that, a structure and interface that moves beyond just a study schedule would all have to be lost (or at least significantly modified) to fit within the constraints of Moodle.

As would the open nature of the course and the content. As implemented by the institution, the Moodle course site would be available only to students enrolled in the course. As a consequence the value of the course and its resources doesn’t get known. A bit of a Google search can find a range of folk using the animations/lectures produced as part of this course: by William Stallings one of the authors of the *standard* OS texts; one of the lectures not sure of this context; good recommendation of the animations in the comments of this post etc.

External factors on teaching

If I were responsible for this course today, there is no chance in this era of ERA that I would be investing the time necessary to implement a course like this.

Space, bandwidth and dropbox

Back in 1999 we were concerned with bandwidth. So, only audio for online lectures and we also produced a CD-ROM mirror of the website. The site is no longer available via the institution that offered the course for various reasons, including saving space.

The site is now hosted on my free dropbox account with which I get 2Gb. The site is about 300Mb in size.

And this is just one simple indication of just how much better the technology has gotten in 12 years. Not to mention how online learning has gone from a novelty to be feared to an expectation.


If I had to teach this course today, in the current university context, I would be hamstrung by institutional e-learning policies and technologies as well as the focus on research. I would have to spend far more of my time trying to work-around limiting institutional factors, far more than I had to 12 years ago.

And that’s despite online technology being more broadly available, widely accepted and of a significantly better quality than it was 12 years ago.

Now, it is possible to describe this course as an example of a “lone ranger”/fred-in-the-shed doing his thing and an approach that doesn’t fit well with institutional systems and approaches. But it is also fair to say (I think) that for the majority of courses, the institutional systems and approaches are failing to provide something approach a minimal acceptable standard (which is what they claim to do).

Lone ranger doing his thing
Sledge to a computer

Two new laws for ERA/NAPLAN etc

For the last few years some colleagues and I have been watching the rise of managerialism within higher education, and more recently in K12 education.

Typically implemented as some vast activity – e.g. a nationalised literacy and numeracy test for all Australian school children, or the reduction of research quality to some numbers – that takes a complex task and reduces it to some sort of ranking.

In response, we’ve been banging on about Goodhart’s law

states that once a social or economic indicator or other surrogate measure is made a target for the purpose of conducting social or economic policy, then it will lose the information content that would qualify it to play such a role.

i.e. once you establish NAPLAN test results as indicating quality and linking it to funding, the value of NAPLAN test results as a measure of quality (if there ever was any value) is no gone as people will “play the game”.

It’s good (and somewhat frustrating) to know that we in Australia are not alone in this globalisation influences, accountability mad rush to task corruption.

Mark Guzdial identifes two additional but related “laws” (Goodhart’s and Campbell’s seem to be quite close) in commenting on this article from USA Today. An article that details some fairly widespread apparent examples of Goodhart’s law.

How long before one of the Australian newspapers or other media outlets does a similar analysis?

Amplify’d from
Dukenfield’s Law <>: “If a thing is worth winning, it’s worth cheating for.”
Campbell’s Law <>: “The more any quantitative social indicator is used for social decision making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes  it is intended to monitor.”