After discovering podcasts I went back and had a second go at trying to find an RSS aggregator setup that actually worked for me. Initially what I wanted was something that would send me an email whenever an RSS feed was updated. I had a look at rss2email but it didn’t work for me (I forget why). After a while I discovered Rawdog, which generates a HTML file with a newest-first list of all the new articles in the RSS feeds I’m subscribed to. I quickly realised I could use an existing shell script I wrote to email me when there are updates. The shell script (webmonitor.sh) is normally called from a cron job to monitor web pages on external sites. The script stores a local copy of each web page. Whenever a page changes, the script converts the old and new web pages to plain text with elinks, then runs ‘diff -c’ over the plain text files. The output from diff is then emailed to me. This made sense when I only had a few RSS feeds in my subscription list, but now my mailbox tends to pile up with messages with diffs of my Rawdog file. I use the Mutt mail reader though, so deleting them all is just a few keystrokes. Ordinarily I’d turn off the email part of it, but it does sometimes help to tell me the timestamp in my Rawdog file to resume where I left off. I still lose track of where I’m up to though.

I don’t know if ordinary aggregation software does this (none I’ve looked at do), but I tend to think they should mark news articles as perpetually unread until you tick a box next to an article’s title to tell the aggregation software that you’ve read an article and don’t want to see it again. There could also be another tick box that’s used for bookmarking articles so you can go back to them, similar to circling an article in a newspaper with a highlighter. In my view, RSS aggregators really need to look and feel like newspapers, although sometimes this just isn’t practical.

As I type this, I’m wondering if it would be useful to collect RSS feeds on a regular basis then import them into a CVS repository, so you could keep track of changes using CVS? You could then keep an entire history of an RSS feed in a CVS tree, and more easily keep an eye on changes, etc. The biggest problem would be having an easy way of doing CVS checkouts of different revisions of a feed, then displaying it in a sensible way, like a proper aggregator. The checkout part isn’t difficult, but I think I might need to write something that converts an RSS feed to rendered HTML in a similar (better?) way that Rawdog does.

I’d forgotten how useful CVS is. Maybe I should be using it to keep track of web sites too…

 

Leave a Reply

Your email address will not be published. Required fields are marked *

Anti-Spam Quiz: