<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 12:12 PM, Rod Widdowson
wrote:<br>
</div>
<blockquote
cite="mid:017801d132a4$dcb6ae00$96240a00$@steadingsoftware.com"
type="cite">
<blockquote type="cite">
<pre wrap="">That would work if the goal is to feed an octet stream, but I would expect
</pre>
</blockquote>
<pre wrap="">a
</pre>
<blockquote type="cite">
<pre wrap="">database approach is possibly more about avoiding the XML entirely.
</pre>
</blockquote>
<pre wrap="">
Good point. In that case one might want the thing to implement
BatchMetadataResolver (and hence extend AbstractReloadingMetadataResolver)
but rather DynamicMetadataResolver (and hence extend
AbstractDynamicMetadataResolver).
Now that would be very nice. Minimal in memory footprint...
</pre>
</blockquote>
<br>
<br>
Yes, making it be a dynamic resolver instead of batch would work out
nicely. I think that could work probably work with either the XML
or normalized schema approach. <br>
</body>
</html>