For When You Can't Have The Real Thing
[ start | index | login ]
start > Why Twitter Has Failed

Why Twitter Has Failed

Created by dave. Last edited by dave, 15 years and 213 days ago. Viewed 3,578 times. #1
[edit] [rdf]
Twitter has failed because we are watching now.

Let me back up a second.

A couple of years ago, a door-to-door from Rogers came to the house. He said that he had come to thank me for being a loyal Rogers customer of their Internet and Cable TV offerings, and to save me some big money on my phone bill.

I said, no thank you, I'm not interested.

He took some guesses as to which services I was getting from Bell, and told me he could save me something like 30% off my phone bill.

I said, no thank you, I'm still not interested, and I'll tell you why.

Internet and cable TV are what you would call high-visibility services. When you are looking for them, it is very obvious if they are not there. It is even more obvious if you do like I do and run monitoring software from your home network to keep an eye on your hosting provider and your link to the office. When you do that, any little burp or drop out becomes noticeable, and is recorded for ever (or at least the next year or two).

When you turn on the TV, odds are you are doing so because you want to sit down and watch TV, either because you are looking for something specific, or you are terminally bored and want to fill some time. If the cable TV isn't working, you are either annoyed (because the program you wanted to watch is now unreachable), or… well, annoyed (because now you have to find something else to fill the time). These incidents are highly visible because in most cases I end up leaving the TV on just in case the cable starts working again. So there end up being long periods of time where I am in the same room with a non-functional cable TV, waiting for it to resume service.

What's that got to do with the phone?

Well, my perception of cable TV is that there are long periods of time where the cable TV doesn't work. And my monitoring software shows me every gap, pause, hiccup, and worse, that the internet has. So my perception is that from Rogers services, there are periods of time when their network is down.

The telephone is a special service. It is a connection to emergency response services. It is essential. My perception is that there is a small, but non-zero, chance that if I pick up my phone on a Rogers network, I won't get a functioning connection to the outside world.

(As an aside: now having a Rogers cell phone issued to me from work, my perception of Rogers' ability to run a communications network has not improved.)

Meanwhile, I can probably count on one hand the number of times that I have picked up my Bell phone and not had a functioning connection to the outside world. And that is over a lifetime of Bell phones.

The difference is I'm not constantly monitoring the state of the Bell network. I have no doubt that there have been many incidents where the Bell network (or at least, my local part of it) was down, and I never noticed because I had no need of the service at that time.

But my perception is that the Bell network will be there if I need it, while the Rogers network might not be.

This is exactly why Twitter has failed.

When Twitter was starting up, nobody really monitored the state of the Twitter network. But now there are monitors, blogs, notices and who knows what else. It seems like twice a week I get a RSS feed item saying "The database went down for some reason". Even if I was not trying to use Twitter at that time, I now know there's less reliability than might be immediately apparent.

I don't know if Twitter will recover from this perception of unreliability. Some services, no matter how long they go, never fully recover or supersede their histories -- witness the bad reputation that Sendmail has in some circles, despite its longevity and the amount of time it has gone without a serious vulnerability.

no comments | post comment
This is a collection of techical information, much of it learned the hard way. Consider it a lab book or a /info directory. I doubt much of it will be of use to anyone else.

Useful: | Copyright 2000-2002 Matthias L. Jugel and Stephan J. Schmidt