The Illusion of Internet Speed

The term "Internet time" became so hackneyed so quickly that it virtually died out. A quick search on Google News suggests that an entire month -- surely at least a couple of years worth of that Internet time -- can pass without seeing the phrase in the press. Perhaps that is because the drive for speed has become such an assumption that it barely needs saying. But the desire to move ever faster makes the mistake that the tools which are supposed to move things quickly actually do. And often they don't.

For example, Marshall Kirkpatrick on ReadWriteWeb mentions that WordPress is supporting RSSCloud so that bloggers can syndicate their work immediately rather than waiting the 15 to 60 minutes it takes for the words to disseminate throughout the Known Electronic Universe, otherwise known as the RSS aggregators:

RSSCloud is an element that's always been present in the RSS 2.0 spec but has drawn new attention with the rise of interest in the Real-Time Web. The element was just added to the WordPress code this afternoon. The implications of this big vote of support go beyond reading WordPress blogs; this is the kind of traction that new technologies can leverage to gain support in many different applications.

It assumes, for one, that something like Twitter is real-time. Anyone in the audience ever participate in a Twitter group chat? I have on a number of occasions and was annoyed by the dawdling experience. I'd type something that wouldn't show up for maybe five minutes and the Twitterverse was moving through some bend in its space-time. In fact, #editorchat is one discussion group that moved from Twitter to Friendfeed in an attempt to find something fast enough that the participants didn't doze off.

The problem is that real-time sounds good on the surface, but inevitably there is a dichotomy, because real-time is actually me-time. I want real-time because I want what I want now. However, the other party that has to do something based on my wants may not be so enthusiastic. The phone is left on voice mail. The email inbox contains dozens of unread messages. For one person to have me-time, another person has to provide them-time, and always at some cost. For online services, the cost is in infrastructure. The more me-time handled, the bigger the pipes, the faster the software, and the more responsive the support must be. It's expensive.

It's not as though real-time RSS is new. According to Dave Winer, who as an inventor of RSS should know if anyone does, a real-time capability went into RSS in 2001. But it wasn't until last month that Google Reader moved to a form of real-time information distribution. Why put it off? Because it's a pain in the rear for the service providers, who can't just do things in batch processing when convenient.

Why do it? Because people want to perceive that their every desire is met with alarming alacrity. However, the key word is "perceive." That's why Twitter is popular. People assume that the messages pop up as soon as they are available and shipped out. The same with FaceBook or MySpace. But it really is only as available as the servers can get through whatever temporary log jam they may be experiencing and push out the data. Nevertheless, it doesn't matter because most people only see what they receive and can't compare it to when other people receive messages -- or to when the messages are sent. So there's the perception of speed. And because the messages scroll off, they don't generally see the accumulation they get in email, so no blinding flash that their messages as likely are also part of someone else's stack.

Tech companies, especially web-based services, are going to face increased pressure to satisfy real-time demands. But in many cases, to save on infrastructure costs, companies will see to what degree the verisimilitude of speed can satisfy as a replacement for the actual thing.

Image via stock.xchng user konr4d, site standard license.