<?xml version="1.0"?>
<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en">
	<id>https://docs.opendap.org/index.php?action=history&amp;feed=atom&amp;title=Use_of_MIME</id>
	<title>Use of MIME - Revision history</title>
	<link rel="self" type="application/atom+xml" href="https://docs.opendap.org/index.php?action=history&amp;feed=atom&amp;title=Use_of_MIME"/>
	<link rel="alternate" type="text/html" href="https://docs.opendap.org/index.php?title=Use_of_MIME&amp;action=history"/>
	<updated>2026-06-18T05:04:42Z</updated>
	<subtitle>Revision history for this page on the wiki</subtitle>
	<generator>MediaWiki 1.38.4</generator>
	<entry>
		<id>https://docs.opendap.org/index.php?title=Use_of_MIME&amp;diff=2228&amp;oldid=prev</id>
		<title>Jimg: New page:  I think that we should specify that MIME documents are used to hold the objects/responses.   James Gallagher - 26 Sep 2003  Ok, do you envision doing away with the separate blob response?...</title>
		<link rel="alternate" type="text/html" href="https://docs.opendap.org/index.php?title=Use_of_MIME&amp;diff=2228&amp;oldid=prev"/>
		<updated>2008-06-24T21:11:17Z</updated>

		<summary type="html">&lt;p&gt;New page:  I think that we should specify that MIME documents are used to hold the objects/responses.   James Gallagher - 26 Sep 2003  Ok, do you envision doing away with the separate blob response?...&lt;/p&gt;
&lt;p&gt;&lt;b&gt;New page&lt;/b&gt;&lt;/p&gt;&lt;div&gt;&lt;br /&gt;
I think that we should specify that MIME documents are used to hold the objects/responses. &lt;br /&gt;
&lt;br /&gt;
James Gallagher - 26 Sep 2003&lt;br /&gt;
&lt;br /&gt;
Ok, do you envision doing away with the separate blob response? &lt;br /&gt;
&lt;br /&gt;
Nathan Potter - 26 Sep 2003&lt;br /&gt;
&lt;br /&gt;
Nope. We need the BLOB to be separate because quasi-transports like SOAP require the entire response be complete before transmission starts. We want to support streaming, so we need to make the data available w/o involving SOAP, for example. Since MIME does not _have_ to include =Content-Length= we can stream a MIME document.&lt;br /&gt;
&lt;br /&gt;
Along these same lines, using the MIME/multipart for the BLOB (see BlobFormatting) we can support stuff like HTTP keep-alive (because it&amp;#039;s possible to tell when the stream ends since multipart specifies a trailing =boundary= marker). This is a big win for efficiency - we want to be able to support the more advanced features of HTTP even though we don&amp;#039;t want to tie the DAP4 to HTTP.&lt;br /&gt;
&lt;br /&gt;
James Gallagher - 26 Sep 2003&lt;br /&gt;
&lt;br /&gt;
Well, this is sort of related (perhaps change this topic header to &amp;quot;on-the-wire specification&amp;quot;?)&lt;br /&gt;
&lt;br /&gt;
I think we need a rigorous specification of what the &amp;quot;on-the-wire&amp;quot; protocol is. As part of that we need to know what is a legal request and response. Right now we have the situation that the HDF server can put out a DDS that breaks the Java parser. Which side has the bug?&lt;br /&gt;
&lt;br /&gt;
John Caron - 22 Oct 2003&lt;/div&gt;</summary>
		<author><name>Jimg</name></author>
	</entry>
</feed>