Magnus Lie Hetland wrote:
> I've named an XMLRPCServlet Main.py; when I refer to it as
> www.foo.bar/Main everything works just fine, but when I refer to it as
> simply www.foo.bar the client seems to block indefinitely...
> Is this common behaviour, or might it have something to do with my
> setup? (I could of course easily make a rewrite rule in Apache to fix
> this, but that seems unnecessary...)

What happens if you access www.foo.bar/ (note the trailing slash)?

WebKit tries to redirect requests for a directory without a trailing slash
to the same URL but including a trailing slash.  So it'll try to redirect
www.foo.bar to www.foo.bar/ .  But maybe xmlrpclib doesn't know how to
handle the redirect.  I don't know why it would hang, though.  I've never
tried to make an XML-RPC servlet be a Main.py before.

> And another thing; when I open an XMLRPCServlet in a browser, it gives
> me an empty web page back. Where does that originate (i.e. how do I
> override it to give some information like "this is not a web page...
> blah, blah, xmlrpc, etc.")?

When I wrote XMLRPCServlet, I never intended it to be accessible both from
XML-RPC and also through a browser.  If you look in the source for
XMLRPCServlet, it only responds to POST requests (which is the type of
request made by XML-RPC).  If you access it through a browser, you're doing
a GET request, which the servlet just isn't designed to respond to.

If you really want to have it respond to both XML-RPC and browser requests,
you'll have to do some more work in XMLRPCServlet to make that happen.
Doesn't seem worth the trouble to me -- I would just make 2 different
servlets, one for XML-RPC requests and one for browser requests.  (Both
servlets could inherit common functionality from a common base class to
avoid code duplication.)

- Geoff

_______________________________________________
Webware-discuss mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webware-discuss

Reply via email to