[ANN] Test cases for transparent metalinks

[Peter Poeml]

Hi,

I have set up a server with testcases for testing the transparent
negotionation of metalinks in the way we discussed:

http://www.poeml.de/transmetalink-test/

Basically, click on 4 files and see if what happens is the expected
result :-)

The README explains the test cases.

I hope I got them set up correctly; let me know if not, or if you feel
something's missing.

I hope this helps client authors to verify the implementation!

Peter
--
"WARNING: This bug is visible to non-employees. Please be respectful!"

SUSE LINUX Products GmbH
Research & Development

[Bram Neijt]

I just had enough time to post this on metamirrors.nl, so this message
is just to say: if you have see problems with
http://metamirrors.nl/node/189 let me know ;-)

Greets,
Bram

[Peter Poeml]

Hi,

On Fri, May 02, 2008 at 02:58:00PM +0200, Peter Poeml wrote:
> I have set up a server with testcases for testing the transparent
> negotionation of metalinks in the way we discussed:
>
> http://www.poeml.de/transmetalink-test/
>
> Basically, click on 4 files and see if what happens is the expected
> result :-)
>
> The README explains the test cases.
>
> I hope I got them set up correctly; let me know if not, or if you feel
> something's missing.
>
> I hope this helps client authors to verify the implementation!

I have added a hint to the http://www.poeml.de/transmetalink-test/README
which was brought to my attention by Hayden, who develops Retriever:

When the client sends "application/metalink+xml" in the Accept request
header, it also needs to add "*/*".
This is to ensure compliance with RFC 2616, Hayden found at least one
webserver (IIS) that sends an error message if the "*/*" is omitted.