contentious.com

Yesterday I wrote about how feeds might play a role in emergency communication. I’ll admit, I thought I had a pretty good idea there. Well, I do have lots of good ideas, but that doesn’t mean I don’t occasionally miss something – but then, everyone does, so there’s no shame in that.

Fortunately, one beauty of weblogs is that this type of publishing makes it very easy to correct errors or omissions, continue a line of thought, or just plain change your mind. I know that admitting or correcting errors is scary to some folks, so here I’m giving an example of how easy and beneficial that process can be.

Here was my error: Feeds are not truly a “push channel” for communication. I sometimes forget this because I (and many other people I know) tend to use feeds as if they were “push” media…

WHAT IS “PUSH” MEDIA?

In a nutshell, “push” media or communication comes to you – you don’t have to go to (or retrieve) it. This offers the great advantage of immediacy – something crucial to emergency communication and certain other types of mass or niche communication.

E-mail is a “push” communication channel. So is the telephone. So are radio, TV, pagers, and text messaging. So is a neighbor knocking on my door, or a police car roaming my neighborhood blaring a warning from a loudspeaker. So is a newspaper getting delivered to my doorstep, although I probably wouldn’t read it immediately, or at all.

Feeds can provide the illusion of push because of how feeds and feed readers generally work:

• A feed is really just a file (technically, an XML file) that resides on a server connected to the internet. Whenever an online publisher posts fresh content, that file gets updated by the publisher’s blogging software, content management system, or even manually to show the fresh content at the top of the list.
• Your feed reader periodically checks that file (a process called “polling”) to see if there are updates. If so, your feed reader downloads the update and stores it so you can read it at your convenience. Most feed readers allow you to specify how often your list of feed subscriptions (or even individual feeds in that list) get check – often called the “polling interval.” For instance, my feed reader checks most feeds on my list twice a day. However, certain feeds get checked every two hours, and I also can manually tell my feed reader to check my feeds at any time.
• Since polling is automated, content seems to magically appear in my feed reader without me having to remember to go get it. That’s what can make feeds look like “push media.” But in fact, my feed reader went out on the net and retrieved every bit of content that it’s showing me. In other words, mone of that content was pushed to me – I (or my feed reader, acting as my agent) had to go get it.

Indeed, one pitfall of the colloquial term “feed” is that it implies “push.” To be fair, the geeky acronym “RSS” does not imply push media – but personally, I don’t think that’s sufficient justification for non-geeks to wrestly with yet another acronym.

For most publishing purposes, immediacy is not absolutely necessary – so you can treat feeds as if they were a push channel. This is true for most people who read online news, weblogs, data streams (like stock prices), and discussion forums, or who access multimedia content online such as podcasts or Flash presentations.

However, for some types of communication (especially emergency communication), seconds count.

TO FEED OR NOT TO FEED?

Think about it the need for immediacy from the intended recipients’ perspective. Do they think seconds count when it comes to your messages?

…If your message is “There’s a sale a Foley’s,” or “We have a new VP of marketing,” or “Here’s what happened in local news today,” probably not. For these types of messages, timeliness counts, but not immediacy. Feeds are very good for timeliness, but not good for immediacy.

…If your message is “IBM is up 13 cents/share” or “We just attacked Iran,” or “The Phila. transit strike has just been settled,” then maybe. Some people (such as stock traders, or commuters, or families of military personnel) might have an immediate need for certain types of information. Others might consider the same messages important, but not necessarily of immediate need.

…If your message is “There has been a chemical leak within two miles of your home. Stay indoors,” then yes, pretty much everyone in the affected area would want to get that message immediately.

MORE FREQUENT POLLING IS NOT THE ANSWER

It is possible to set your feed reader to poll all feeds, or selected feeds, more frequently. This does allow you to receive updates faster. However, it also imposes a larger burden on the servers where a feed is hosted.

For instance, say you set your feed reader to poll a certain feed every minute. Yes, you would get alerts from that feed virtually immediately. However, it probably wouldn’t take too long before the system administrator at that server where that feed resides looked at the server logs, saw you pounding frantically on her server, and then blocked your IP address due to abusive behavior. Then you wouldn’t get any content from that feed at all.

…As with any media, feeds are a tradeoff. I wouldn’t be surprised if at some point we had a revival of true online push media that behaves more like feeds than like e-mail. Back in the 1990s, Pointcast tried to do that. I haven’t looked into push media much lately, but I wouldn’t be surprised if some bright people are cooking up a better solution.

(Many thanks to Tom Vilot for reminding me that feeds are not push media.)