<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 12/9/15 11:54 AM, Cantor, Scott
wrote:<br>
</div>
<blockquote cite="mid:11B5A8C3-743C-45DA-8194-5634F5CCF8A1@osu.edu"
type="cite">
<pre wrap="">On 12/9/15, 11:49 AM, "dev on behalf of Rod Widdowson" <a class="moz-txt-link-rfc2396E" href="mailto:dev-bounces@shibboleth.netonbehalfofrdw@steadingsoftware.com"><dev-bounces@shibboleth.net on behalf of rdw@steadingsoftware.com></a> wrote:
</pre>
<blockquote type="cite">
<pre wrap="">
I'd start by looking at
org.opensaml.saml.metadata.resolver.impl.FilesystemMetadataResolver (but
remember that many of the objects it inherits from are in the impl tree)
</pre>
</blockquote>
<pre wrap="">
I might be misremembering, but I think Brent might have been intending to start moving some of the base classes up once the indexing stuff and the issues Marvin's working on get worked out. If not, I'd suggest it. Keeping things in -impl was a time-saver to avoid having to commit but I would like to see us commit at some point.</pre>
</blockquote>
<br>
I don't know that I've ever consciously intended to do this, but I
don't necessarily object. But gets into the whole OpenSAML api vs.
impl question.<br>
<br>
<blockquote cite="mid:11B5A8C3-743C-45DA-8194-5634F5CCF8A1@osu.edu"
type="cite"><br>
<pre wrap="">
That would work if the goal is to feed an octet stream, but I would expect a database approach is possibly more about avoiding the XML entirely.
</pre>
</blockquote>
<br>
<br>
I was wondering that as well. If the schema is just something like
a table with 2-ish columns like, entityID and XML document as blob,
that's one type of impl, probably pretty easy. But if the
individual entity data elements are normalized out to many tables,
that impl would be quite different.<br>
</body>
</html>