As a follow-up to the Emerging Trends workshops at San Diego State University earlier this week, I was asked by participant who had just created new blog sites, “How do we find each others blogs” and by the planners, “Hey Alan, you’re an RSS guru, how can we syndicate them into one place?”. So I was charged up to find an answer.
One level worth doing, and quite easy to do, is to set up a Bloglines account so the new blogs can at least be browsed in one place. I tried to demonstrate that this is one approach teachers can do to either create a resource collection for students (gathering relevant RSS sources in their discipline or project area) OR to aggregate blogs created by students. The hitch with bloglines is having another email address available if you do not wish to mix it up with other sets of feeds (folder level sharing would be nice), but with my own server I can create one time use email addresses, and that is also what Yahoo mail works well for. I think I have created 10 or 12 Yahoo mail addresses in my life.
Okay, so I was able to create a Bloglines public collection for the Emerging Trends Initiative workshop, more or less what is needed to monitor them all in one place. It would be just swell, Wally, if Bloglines was capable of mixing these back as a single RSS feed. I could have used Blogdigger to do this, and likely will as I like that it can keep an archive of feed content.
But my approach is not always use what you know works, and to try and use something you don’t know and see if it works (ducking from the cries of grammar specialists on that last sentence).
I tried RSSMix but it keeps getting choked on something and produces errors and empty feeds.
Fortunately, D’Arcy Norman mentioned using Feed on Feeds, and one of my good signposts is if D’Arcy uses something, it must be worth checking it out.. though I think I had furled Feed on Feeds a while ago.
Basically, Feed on Feeds (hereafter shrunk to FoF) is a web server aggregator that runs on PHP and mySQL, and uses the Magpie RSS parser as well, all stuff I know. FoF is really meant as a personal aggregator, meaning for someone to add and delete feeds, and thus it is suggested putting it behind a password. That precludes using it to create a public viewable collection. But I managed to tweak things enough to remove the links and direct access to the editing tools.
First, this is my own FoF collection of the SDSU blogs. Like D’Arcy noted, pretty plain formatting, but CSS prettification can always come later. I was able to export the blog feeds from Blogline as an OPML file, load that on server, and then import as an OMPL URL into FoF (holy acronyms! That means I was able to create an XML file with all the address of the feeds I had saved into Bloglines, put that file on a web server, and then import them all in one click to the FoF database).
The mods I made include:
* Replace the default index.php (the control panel) with the view.php content, so the default URL produces the feed display, not the editor. In this file, I manual set the non editing flag to 1 to prevent the display of the editing form. I moved the include file for the view-menu file up top (see next step) to be used on all views. I also headed a little bit of a custom header, the h1 page title and bit of explanatory text.
* In the view-menu content file (the one that creates the list of links in the upprt right box), I removed all the links for marking/flagging feeds, and remove the links to the file that is for adding feeds. This reduces it to a navigation menu, and allows for the paged view mode.
* In the file for adding feeds, I add a check for a special key as a URL parameter as a cheap access gatekeeper:
// cheap gatekeeper
if ($_GET['key'] != '********') header('Location:index.php');
which means of the URL lacks a key=_________ in the query string, or the value is not correct, the viewer is bounced back to the main page.
* Finally, if there are any links to the old view.php file, I redirect them to the in-use index.php file with the same query string via having this as the contents of view.php:
<?php header("Location:index.php?" . $_SERVER['QUERY_STRING']); ?>
* I made parallel changes for the framed view version (that lists the sources on the left), but removed / commented out anything that is an editing function.
* As suggested in the FoF site, I created a cron script that calls the updating URL once every hour.
In summary, Feed on Feeds is very useful although it takes some PHP legwork to make a public viewable site
The post "Bloggregation" was originally pulled charred and crispy from a smoky charred oven at CogDogBlog (http://cogdogblog.com/2005/05/bloggregation/) on May 26, 2005.