RSS feeds are marvellous. They represent a “notification” system for new content that has existed for quite some time now (born in 1999), enjoys solid support, and is used by millions of people, sometimes without even realising it, as in the case of podcasts. Yet here’s the rub: RSS makes advertising revenue and audience measurement far more complicated, particularly the sort of egregiously intrusive variety we must endure at present.

I remain rather cross with Radio France for severing their RSS feed, forcing podcast listeners to either visit their website or download their application. In either case, one can no longer “officially” download episodes to listen to later. Like a goblin hoarding treasure it refuses to share, they guard their content behind proprietary walls.

But that is neither here nor there.

RSS presents itself as a structured file made available by servers, containing information about the most recent “entries” (in the case of this blog, an “entry” would be an article, for instance).

An RSS client maintains a list of addresses for these files and consults them periodically to determine whether new articles have been published. Relatively straightforward, one might think.

For some years, I employed an RSS reader similar to Google Reader: a website that aggregates content from feeds to which I subscribe and presents them in an interface rather like email. Like a goblin servant fetching treasures from various lairs, it dutifully collected all my content in one place.

This solution presents both advantages and disadvantages, but I eventually abandoned it because I found the requirement to keep a browser window open with the aggregator tab perpetually active rather onerous.

I subsequently switched to the RSS client integrated into Thunderbird, my mail client. The interface proves extraordinarily cumbersome for managing subscriptions (additions, removals, and so forth), though local storage of feeds in an interface resembling email is rather pleasant. Nevertheless, subscription management was genuinely quite debilitating.

I have therefore spent considerable time seeking a replacement. I examined modern web based solutions (some are quite excellent and comprehensive), I looked at heavyweight clients, yet nothing really prompted me to declare “Ah yes, this is precisely what I require.”

My reflections surrounding RSS have simmered in the background of my mind for quite some time, until I finally established a list of my needs and desires.

None emerged as a clear and decisive victor.

And when I beheld all this, I thought to myself: “Might there not be a way to make this even simpler? Cannot I perhaps, at least for the time being, rid myself of everything that is merely rather nice to have?”

Then an idea struck me like a goblin’s cudgel: distil the very essence of what constitutes an RSS aggregator, namely something that alerts you when new content of interest arrives.

I harbour no great affection for mobile telephones and the manner in which they render us contactable at any moment, yet when I am at my computer, I am generally connected to IRC, an instant messaging protocol. Why not simply create a small robot commanded by message that sends you notifications when new RSS feed entries arrive?

One evening of programming later, I had created IRSSC (mirror). It does not accomplish much, one can merely subscribe, unsubscribe, and list one’s subscriptions, but it fulfils the minimal task I required of it. One does not always require thousands upon thousands of features. Like a goblin content with his small cache of shiny pebbles, simplicity brings joy.

Besides, writing IRC bots is rather enjoyable.