People are reporting errors when they update. Updates appear to vanish on the web (requires a page reload) and clients like Twitterrific & Twirl aren’t able to update at all. We’re working on a fix and intend to deploy it today.
Update (9:42a): One of our slave databases went down 10 minutes ago. We have to put the site into maintenance mode while we bring it back up. Thanks for your patience.
This is the same issue we see and hear again and again. While it’s understandable that a website can have errors, a “slave database” should not be the one thing that takes the whole system down. They say “We are working on problems”, but do we see results?
It’s probably time for Twitter to take a hard turn. They know they have problems. They know they have to deal with the mainstream clientelle. However if they continue down this path, they may loose more people.
It’s really simple: Take Twitter back into Beta. With that moniker on the site, things can be taken down and brought up when needed. Schedule a day where you plan to be down, and bring the site down, clean up code so when things come back up, you can attack.
Another thing that you can do while in ReBeta: Add an area where general questions are answered. That way people won’t ask the same questions over and over. Post the bugs and requests so we know whats being worked on.
For instance, a while back I wrote in and asked if Twitter could change the functionality of the “Followers” area. I didn’t like the fact I had to search through all my contacts to find those that I am not following. This is a very quick and easy fix by putting a filter on the database query.
I got an email back saying “Yeah, that’s a great idea. We’ll get on that”…….. That was about 2 months ago.
If you show accountability, people will respect the issues and give you a little more forgiveness. Why? Well simply because they will be in the know. And if you don’t hide anything from the users, they will accept a lot.
I really think that Twitter has to take a step back here. Schedule some major outages so items can get fixed. Put in redundancy so secondary databases don’t take away the full functionality of the program. Plan the Attack, then attack the plan I always say.