On Thu, Feb 11, 2010 at 2:57 AM, <[email protected]> wrote: > > http://codereview.appspot.com/207059/diff/1/14 > File > > java/gadgets/src/main/java/org/apache/shindig/gadgets/AbstractSpecFactory.java > (right): > > http://codereview.appspot.com/207059/diff/1/14#newcode134 > > java/gadgets/src/main/java/org/apache/shindig/gadgets/AbstractSpecFactory.java:134: > response.getHttpStatusCode()); > On 2010/02/10 22:17:50, zhoresh wrote: > >> If the external system return 500, should we as well? >> maybe 504? >> > There is a small subset of 500s---e.g., 50{2,4}---that we should not > demur from using. I think a gateway- or proxy-related one is fine.
Agreed. In my latest CL (http://codereview.appspot.com/207060/show) I've taken this advice to use SC_BAD_GATEWAY in situations where the request itself was legitimately formed but Shindig nevertheless couldn't correctly process the request, in an attempt to honor the spirit of HTTP status code 400 (http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html): "The client SHOULD NOT repeat the request without modifications". Having done this, I wonder if there are circumstances in which SC_BAD_REQUEST is being used when BAD_GATEWAY should be. If you've got a moment to peek around I'd be interested in your thoughts. Cheers, John > > http://codereview.appspot.com/207059/show >
