Tell Me About RSSmeme
This isn’t going to be just any run-of-the-mill review from me. Why? Because the application I am going to review is different in several substantial ways:
- It isn’t a piece of software you install on your PC
- It isn’t a web application like Gmail or YouTube
- It really isn’t exactly meant to be a destination page like My Yahoo! or iGoogle
What is RSSmeme?
So, given all these things that RSSmeme is not, what the heck is it? It actually evolved, and continues to grow, from a very simple thought: Wouldn’t it be cool to tally up how many times a news post is shared in Google Reader and make a list of the most popular articles? And in fact, it wasn’t Benjamin Golub that originally had that thought.. it was probably the folks at ReadBurner or TechMeme.
RSSmeme sprung, virtually overnight, from a challenge Benjamin gave to himself to make a competing service, but written in Django, a development framework based on the programming language Python.
While all three services being similar in that they give you top news stories based on shared articles, they all have different directions that they are going.
- ReadBurner is a pure democratic post counter. Nothing gets special consideration.
- TechMeme uses a proprietary set of rules and contributions to determine the hot news of the day.
- RSSmeme not only counts shares, but also allows searching, parameterization, different views, and RSS everywhere. It also now has an API. I’ll dive down into these features in this article.
Default Home Page
When you first visit RSSmeme, you will notice that it gives you a default view — Most Popular Stories From The Last 24 Hours. A lot of people will find this to be a useful page, it simply shows you what has been shared the most in the last day. But, as you look more closely at the individual articles, you will notice that a lot of the article’s header and text are actually hyperlinks:
- The subject
- The author
- The site (or source)
- The tags
- The names of people who shared the article
- … and more
In fact, each of these hyperlinks (except the subject) sends a command back to the RSSmeme engine to refine your list of articles based on your selection. You can get a list of articles by author name, source, tag and sharer. There are really two reasons why this is awesome. First, it means you can easily refine how to view the list of shared articles.
Second, and this is the one most people don’t immediately notice (even though it’s in the site’s title), Every page you generate has its own special RSS feed.
Yes.. you get it now. Want to add an RSS feed for just shared articles written by Louis Gray? Nothing could be easier. How about the most recent articles from Gizmodo? Not a problem.
In fact, it’s almost addictively fun to link-surf through RSSmeme, pulling up different lists of articles by clicking on an author name here, a source there. And any page you can view in your browser, you can subscribe to as an RSS feed. Beautiful.
Searching
Searching has a couple of caveats in RSSmeme. You can search for article headline or text, or a user name. However, there isn’t a direct way to search specifically for an author or source name. And in fact, due to the nature by which RSSmeme compartmentalizes the data as it comes in, internally it refers to authors and sources with a specific ID number. If you know this number, it’s a cinch to pull up that info later.
What I recommend is, if you are looking for a specific author (say, yourself) or a specific source (web site), search for a recent article published by that author or source, click on the name, and then bookmark that search result for future reference.
For example, if you want a list of articles written by me, you can find it here.
Parameterization
Essentially, we have already been talking about this. Since RSSmeme analyzes the incoming RSS feeds for individual elements such as author, source and sharer, it naturally is able to generate a list based on any of these things. The user interface already holds a lot of this power. You can change the time period to search, and the minimum number of shares. If you watch the address bar closely, you can construct custom queries:
Want to find all stories in the last 48 hours by Gizmodo that have over 3 shares?
http://www.rssmeme.com/custom/48/3/?source=9Want to find all English articles in the last 72 hours that have over 15 shares?
http://www.rssmeme.com/custom/72/30/?language=en
If you are running up against a wall with the UI, you can explore the RSSmeme API as well. In fact, let’s talk about the API a bit.
RSSmeme API
Benjamin just revamped the RSSmeme engine, and in the process added support for an Application Programming Interface, or API. What this means to you is that you can write your own code to query RSSmeme and extract information out of it in any format and sort order you wish. Some other features (not inclusive):
- Different output formats, including the RSS standard ATOM
- Limit the number of hours to go back
- A specific feed ID
- Share minimum and maximum
- Language
- Sort order
- Source ID
The API is very flexible. In fact, Benjamin has already created two very useful RSSmeme API-based tools that are very handy. The first is over on the sidebar of my site.. it’s the 5 Top Shared Stories. A few others:
- FriendFeed-Personalized RSSmeme – This one is great, I’ve been using it for over a month now.
- Creating a Most Shared Stories Widget for your Blog – See above
- Mobile RSSmeme – You can find the result at m.rssmeme.com
Other people are also starting to take note of the API as well. I found a post by Evaristo on his blog, Soliloquio, about how to combine RSSmeme Filters for custom output.
With the advent of Google Reader’s support of sharing with a bookmarklet and notes, RSSMeme is now extracting the notes as well. Although not officially supported yet, Benjamin has already started generating some notes-based statistics.
Other Features
As I just mentioned above, one big up and coming feature is Google Notes. So far, RSSmeme is out front with being able to manipulate the notes on shared items. It’s really very cool to look at an article that has been shared 50 times and see the 20-odd notes that people have tacked on to it.
URL Canonicalization is a fancy term for removing duplicate articles. Without this essential ability RSSmeme would be cluttered with dozens of slightly-different copies of each article.
RSSmeme supports the Google REST language API, to filter articles by language.
There is a handy FeedBurner FeedFlare that you can drop in your blog and FeedBurner Feed to show how many times an article is shared, and the number of notes.
Finally, RSSmeme also uses Disqus as a comment system. All articles allow a viewer to create a comment about that article, and I can tell you that Benjamin tends to read all of them. At least, I’ve left a random comment here and there and he’s noticed them all so far. Probably not used much, but it is there if you want to start a discussion or if you notice a rendering bug that you want to bring to the developer’s attention.
The Future
I believe that Benjamin is really working out on the edge of useful RSS feed data manipulation and social media. It’s hard to say where the service will go in the future, but it is obvious that it has a lot of potential and can be used in a lot of different ways.
If I were to make a blue-sky sort of hypothesis, I could see RSSmeme being used in a variety of ways:
- Canonicalized clipping service: This is a no-brainer. Find a keyword, source or author, construct an API query and you have an always-ready de-duped stream of articles.
- Trend analysis: With the amount of atomization RSSmeme does with RSS feed data, it would not be hard to get trending on any number of variables.
- Shared items corruption detection: I just thought of this. There is a big problem right now with Google Reader’s sharing bookmarklet, in that it lets you edit the text you cut and paste directly. So someone could completely change the content of a shared article. However, if RSSmeme has received this article shared more than one time, it would be relatively easy for it to diff the versions. Or go to the source to pick up the original text.
- Digg replacement: Ok, this one is really pie-in-the-sky, but Digg is not democratic. It can be gamed. Because of the way that RSSmeme collects its data, it is much more difficult to skew the popularity of a particular article or source.
Final Notes
Obviously I am very excited by Benjamin Golub’s free offering. It has had its share of rocky times, but Benjamin has not only worked through this but consistently improved how RSSmeme works. The features he adds are well thought-out and generally enhance the overall usefulness of the site.
Also, Benjamin himself is very responsive to questions or feature requests. He is open to suggestions and seems completely committed to making RSSmeme a great site. I am very excited to see where he takes this technology!
Update: Sarah Perez has written an article that goes into more depth on the RSSmeme API over at ReadWriteWeb. Check it out!