SnowBlog: The Snowball Factory Blog

We are disappointed to announce that we have closed TweetPo.st to new users until further notice. This was a hard decision for us, but our top priority is delivering a high quality user experience for our current users and this is the only way we can do that at this point.

If you have already signed up for TweetPo.st, you will continue to be able to use it and we hope most of the issues that have cropped up in the last few weeks will soon be fixed. If you haven’t yet signed up for TweetPo.st, we really appreciate your interest and hope to open back up soon (please follow @tweet_post to be kept up-to-date on our progress).

The brief history of TweetPo.st has been a series of lessons in building products on other people’s platforms. The first version (called TweetFace until Facebook made us change it a couple weeks after release) was basically a public prototype built in a week about a year ago. It grew out of our desire to have ‘a smarter way to update Facebook from Twitter’ for our personal use and we only did it because it seemed so easy at the time. We didn’t even tell anyone about it other than our friends, but somehow people started to notice. We were happy that folks liked it and even happier that we didn’t have to worry about it while we worked on other things.

But then last fall, we began to get reports of strange behavior when new people tried to signup. When we finally had time to do some serious investigation, it turned out that Twitter had changed things in a way that made it impossible for TweetPo.st to add any more new users without us having to basically rewrite most of it, which we didn’t have the time to do then. It was wasn’t until this February that we were able to go back in and basically build a whole new TweetPo.st from the ground up using the latest APIs from Twitter and Facebook.

Based on our research and planning, we blocked out a couple weeks to implement and test the new version. But very soon into it, we realized that things weren’t going to be that easy — specifically, some of the Facebook APIs we had planned to use were broken and all of them were really slow and/or unreliable. In practical terms, this meant we had to do a whole bunch of extra work and testing to make things work correctly for our users. Unfortunately, the very issue with the Facebook APIs is their unpredictability and so at a certain point we had to make a best guess at how to deal with it and then see what happened.

The good news is that after seeing how the application has performed (or not) in the real world over the last couple months, we have a better sense of the pattern of the underlying issues. The bad news is that the long-term fixes for these issues are again going to require significantly more time than we can give right now. So, we have made some smaller fixes that should restore performance for existing users in the coming days and we have decided to close TweetPo.st to new users until we can make the long-term fixes to properly support them.

We can’t tell you how much we appreciate the patience and understanding of our current users. We recently added Tilly to our team to help with customer care across all our products starting with TweetPo.st, and she has been totally impressed with how cool you all have been. And to anyone finding this because they want to use TweetPo.st, we’re even more bummed than you that we can’t let you in yet (and we recommend Selective Tweets as a lightweight alternative in the meantime). We only want to build products people love to use and sometimes that requires us to go for more love from fewer people. We’re hard at work on a new version of awe.sm, which is going to be our main focus for the next few months. And hopefully, we’ll be able to do some more work on TweetPo.st soon after that. Thanks again for your support!