Hi Gabo

> Either way, I'll get it fixed for 2.2 (the issue highlighted by your
original example)

Not in a rush. :) I doubt the client is going to use I.E. anyway. :D

Fingers crossed, oyt should be fixed now. I have a test matching your scenario. I'll still be tweaking few things around the way accept types are handled, but the test will protect again the regression.

If you could try and confirm it with a trunk (mvn install -Pfastinstall, cd api, mvn install, cd distribution, mvn install, get the cxf-distribution out of distribution/target) in the next couple of weeks then it would be good

Cheers, Sergey


Thanks.

Gabo

Sergey Beryozkin wrote:
Hi Gabo

I believe in your original post getObjects() had @Path("/"), right ?
In this last example the IE requests should be fine. What I meant was something 
like this :

  @GET
  @Path("/get")
  @WebMethod
  public MyObjects getObjects(){
  }

  @GET
  @Path("/get/")
  @WebMethod
  public Object getObject( @QueryParam("id")
          @WebParam(name="id")
          long id) {
  }

in which case /get?id=1 will be delivered to getObject(), while /get/foo?id=1 /get/bar?id=1 or will be delivered to getObjectProperty...

but it might not be acceptable for your specific case. Either way, I'll get it fixed for 2.2 (the issue highlighted by your original example)

Cheers, Sergey


Reply via email to