As follow-up, I guess I would have to count this as a core Java API bug, since the high-level functions (HttpURLConnection) allow you to routinely emit bad markup.
However, wouldn't it be prudent in Tomcat to recognize that something has gone wrong with the method name earlier? Should method names ever be allowed to contain numbers? How about carriage returns and other white space? So the root question is, should I write this up as a low-priority bug, or is the current behavior desired? -----Original Message----- From: Mark Thomas [mailto:ma...@apache.org] Sent: Tuesday, April 07, 2009 2:40 AM To: Tomcat Users List Subject: Re: Help with a Tomcat issue??? Jason Smith wrote: > Solved. You gave me a clue that helped solve it, though Tomcat could handle > this better (not put '0 POST' in for the method name in the first place). > > Sun's implementation of HttpURLConnector apparently creates a new > ChunkedOutputStream every time you call .getOutputStream(). In other words, > multiple calls to conn.getOutputStream().close(); cause corrupted output. > Calling .close() on the same output stream does not cause corrupted output. Glad you found the problem. Mark --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
