<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<br>
<br>
<div class="moz-cite-prefix">On 4/7/16 5:01 PM, Cantor, Scott wrote:<br>
</div>
<blockquote
cite="mid:9846A6064BD102419D06814DD0D78DE1128B2612@CIO-TNC-D2MBX02.osuad.osu.edu"
type="cite">
<blockquote type="cite">
<pre wrap="">(Had to double-check, those particular ones aren't using the MDQ instance
on shibboleth.net, they're using the template variant against static paths on
svn.shibboleth.net).
</pre>
</blockquote>
<pre wrap="">
Oh, sorry. Just assumed. </pre>
</blockquote>
<br>
It's totally impossible to tell without looking at the individual
test methods since there are essentially 4 dynamic strategies: 1)
Template 2) Regex 3) MetadataQueryProtocol 4) well-known location.
For the tests that aren't testing any specific strategy per-se, like
the TLS trust engine tests, an arbitrary one is used. Generally
Template IIRC, since that's easier and less likely to fail for
reasons that don't have anything to do with what is being tested.<br>
<br>
<br>
<blockquote
cite="mid:9846A6064BD102419D06814DD0D78DE1128B2612@CIO-TNC-D2MBX02.osuad.osu.edu"
type="cite">
<pre wrap="">Are the MDQ tests using localhost? I guess I thought they were, so assumed that was the cause.
</pre>
</blockquote>
<br>
No, the ones that do use MDQ are hitting <a class="moz-txt-link-freetext" href="http://shibboleth.net:9000">http://shibboleth.net:9000</a>.
That's the MDQ instance running specifically to support the unit
tests. If it was localhost, then it wouldn't work from our
desktops, unless you fired one up locally...<br>
<br>
<br>
<br>
</body>
</html>