DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT <http://issues.apache.org/bugzilla/show_bug.cgi?id=29874>. ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND INSERTED IN THE BUG DATABASE.
http://issues.apache.org/bugzilla/show_bug.cgi?id=29874 Auto method retrial broken ------- Additional Comments From [EMAIL PROTECTED] 2004-07-03 21:57 ------- What irks me a little is that socket read timeouts on non-idempotent methods must be considered fatal, as an attempt to automatically retry an non-idempotent method may leave the application in an inconsistent state. As HttpClient can't know whether a method or a sequence of methods are idempotent, it should probably assume they are not. True. Technically we can probably retry all methods except POST and PUT, but I think in practice this would be dangerous. In this case rethrowing socket timeout as a recoverable exception sounds wrong, at least from the conceptual standpoint. I agree. I think the problem lies in the whole concept of recoverable exceptions. We are attempting to determine recoverability at the point the exception is thrown, but this decision should really be made by the code that catches the exception. Perhaps we should ditch the whole concept of recoverable exceptions and just base it on other factors. I do agree that it might make sense to have connect timeouts treated as recoverable exceptions. The trouble is that currently only method execution can be retried automatically, and connection cannot. Yes, this would be an enhancement. Mike --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
