On 03/18/2012 05:49 PM, Gary Gregory wrote: >> One additional thing: the fix for the QuotedPrintableCodec changed >> (slightly) the behavior of the encoding/decoding. As I have seen in some >> other classes, similar changes have been documented in the javadoc of >> the affected methods as well (e.g. Base64#encodeBase64String). >> >> Do you think we should do it also in this case? > > How would state this in the javadocs. I am now worried about > compatibility. Would it better to back out the change?
After some thinking, I believe this change has to wait till 2.0, as it would break compatibility. A string encoded with the changed QuotedPrintableCodec can not be read by an application using codec-1.6 or earlier (if it contains a soft line break). Also the decode behaves slightly different (more correct), as it will ignore any non-encoded CR/LF character. In prior version these special chars were treated like normal chars. Thomas --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
