The Archive Agents communicate via HTTP URLs and return results as XML documents. The previews are returned in binary JPEG format.
URLs are expected to be in the UTF-8 character format. Reserved characters should be encoded using the Quoted Printable standard. (More information can be found in the HTTP specification.)
The XML documents returned from the Archive Agent should be encoded using UTF-8. Agents may return data in UTF-16 format, in which case that should be specified in the Content-Type header:
Content-Type: text/xml; charset=UTF-16
Using UTF-16 encoding may not be ideal due to the large overhead of transmitting the data.
It is recommended that Archive Agent implementations support GZIP transfer encoding of the data. This wil dramatically reduce the download time of the XML documents. To enable GZIP transfer compression, use Accept-Encoding header in the request to the server:
The server signals that the data is in fact compressed using the Content-Encoding response header: