It appears that the problem is related to how certain libraries are doing the URLEncoding of the text of the tweet. Specifically, how spaces are being encoded.
For example, take the tweet "this is a test". This can be URLEncoded in either of two ways, depending on how your library works: 1. this+is+a+test 2. this%20is%20a%20test If your library is encoding spaces in the first style, your requests will fail. If your library is encoding spaces in the second style, they appear to work correctly. For those of you using libraries written in .NET, your library is probably using HttpUtility.UrlEncode() to encode the text of the tweet. This encodes spaces as "+" symbols. If you replace this with HttpUtility.UrlPathEncode, the text is encoded properly with the "%20" sequences and the tweet posts as expected.
