Scripting News for 4/21/2008

A new strategy for Twitter outages 

Okay, I’ve bit the bullet, I’m going to change what I do when Twitter is down.

1. When Twitter is down I will post updates to an RSS feed.

http://twitter.scripting.com/daveRss.xml

2. You may follow this feed in any tool that can follow an RSS feed. These include FriendFeed, Jaiku and of course many others.

I want to accumulate a list of services that can follow RSS feeds in a Twitter-like fashion (i.e. river of updates). If you know of others, please post a comment here, with detailed user-level instructions for following a feed. I will try to figure out how to add a feed to my profile on FriendFeed. (Any help would be appreciated.)

I feel pretty good about this. I have a nice tool that I might use even when Twitter is up. 🙂

Update #1: I’ve fixed FriendFeed so it’s now following, in addition to the feed for scripting.com, and my Twitter and Flickr accounts, also my “rainy day” feed, above. So, if Twitter goes down, you can just start using FriendFeed instead if you want to follow my tweets. Pretty cool. Let’s hope they know how to scale! 🙂

Update #2: Here’s a screen shot that illustrates items from the rainy day feed showing up in my FriendFeed stream. If you follow me over there, you’ll get the updates automatically, you don’t have to do a thing.

Update #3: Here’s a screen shot of the outliner-based tool that I use to author the rainy day feed. I’ve turned off the connection betw Twitter and FriendFeed, so I’m now committed to using this tool to author my Twitter stream. I think it’ll be okay. Already my posting volume is back to normal, even though the Twitter outage persists.

If this were a normal day on Twitter… 

It wouldn’t be a normal day, because tomorrow is the Pennsylvania primary and a lot of polls are coming out right now, and they’re presenting an interesting story. But that story can’t be told, because our main communiation platform, Twitter, is down.

Yes, it serves me right, and I, of all people, know better, than to build a network on a single point of failure, depending on one company, that is known for producing unreliable systems in an industry with incredibly thin skin (how can they get better if they won’t listen). In other words, this is hardly Murphy’s Law, it was easily predictable. It was likely.

In my own defense, we were lulled by months of relatively reliable service from Twitter into believing they were on the path to even more reliability. I stopped encouraging potential competitors to enter the market, now look where we are. No second source (Pownce, as much as I like it, is not a replacement for Twitter, neither is FriendFeed, though I thought they might be, but they said they weren’t interested when I visited with them).

We need some big infrastructure companies to get into this game. One is not good enough. We also need standards so that tools that are built to work with one work with the others. There’s no time for a standards body, so if you’re getting into this business, please, just use the Twitter API, as imperfect as it may seem.

No matter how reliable Twitter seems in the future I won’t change my mind about this. We need them to have serious competition, so we have a Plan B when something like this happens, which it will, not a matter of probability, imho.

Update: A new strategy for dealing with Twitter outages. A “rainy day” RSS feed of downtime tweets. Important if you depend on my twitterings.

The Twitter outage persists 

Monday morning, the outage that started Friday night, is still going on.

As a test, I posted a picture last night before signing off, only 62 reads, which is very low for a link that should have gone to over 8000 people. (Admittedly it was Sunday night in the Calif, but Twitter is a world wide thing, and it’s already 7:30AM on the east coast of the United States.)

Also, still no notice of the outage and what they’re doing to clear it on the Twitter blog.

http://blog.twitter.com/

Not good. 😦

PS: Amazon CTO Werner Vogels checks in on the Twitter outage.

Leave a comment