I’ve grown hooked on Twitter. But of all the Beta, Web 2.0 applications I use – it is the most unreliable application. At first I thought it was my Twhirl client at home. I’m running it on the Air Linux alpha and I figured there were some issues and that’s why I couldn’t connect. Then I started just hitting the web – and discovered that no, it’s not my Twhirl client:
The server at twitter.com is taking too long to respond.
The entire site is DOA. Doh!
Not sure what’s going on but they need to restrict new account signups or something. I certainly can’t say it’s a useful application but it is fun to see what everyone is doing at any given moment. I just hope the developers at Twitter can fix their issues before I loose interest and find something else new and shiney to play with on the web…
Twitter has always had issues with uptime and scalability, so don’t expect that to improve anytime soon. I think most hard-core Twitter uses have come to reconcile themselves to the fact that it’ll be unavailable every once in awhile.
It seems twitter is down every other day. Sometimes for very short periods, but hours at a time others. They have already given their main developer the boot, and hired new teams. So far downtime is even worse. Rumors about about them dropping Ruby on Rails, but who really knows.
It sucks that it goes down so often, but it’s not like I rely on it for anything serious. It’s such a great service overall, I take the downtime in stride.
I think Brian said it perfectly and I believe most users have resolved to just learn to deal w/ Twitter being inconsistent. What I’ve found odd is how Pownce, which provides similar capabilities to Twitter, has never caught on, in spite of all of the Twitter issues.
I never really understood why twitter stayed with RoR, all up it’s relatively simple system with a hell of a lot of users.
My normal rule of thumb is to simply or eliminate heavy layers to increase system capacity / performance /stability..
RoR looks like one of those, i have oft read people saying that most RoR developers can’t / don’t understand the code in the framework and that only a few people in the that community do
Quite a good read over at techcrunch with regards to Twitter uptime:
http://www.techcrunch.com/2008/04/22/twitter-may-not-have-to-care-about-uptime-any-longer/
It is strange that Pownce hasn’t caught on. It may be simply that Twitter got there first and it’s hard to displace the incumbent technology, especially when you’ve got some many popular figures in the tech world using Twitter and talking about it (sometimes ad nauseum) on podcasts like “This Week in Tech”.
One of the panelists on that podcast made the point that changing the programming language behind Twitter may not matter much, that the bigger issue is handling all of the constant database reads and writes that Twitter must be doing to facilitate the service.
@dc – the Tech Crunch article was very interesting! It will be interesting to see if they do work the issues out, or if they get worse – will people continue to cling to the sinking ship?
I hope that Twitter gets sorted out because I really love it.
Here’s an interesting discussion of the real underlying problem with Twitter – and no it’s not ROR:
http://myles.eftos.id.au/blog/2008/05/27/a-stark-realisation/