I have made changes to the Windows Registry to facilitate opening Word Documents Read/Write directly from a link within IE in a separate instance of Word. Information on the specific registry setting is here. http://support.microsoft.com/?kbid=867838 <http://support.microsoft.com/?kbid=867838>
I have been able to successfully test this behavior using the default File Content Store that comes with Slide 2.0. Clicking on a link like http://host/slide/files/Document.doc <http://host/slide/files/Document.doc> and watching the TCP traffic, I see IE make a GET request to Slide which returns the content of the file. The file is opening in Word because of the contentType header "application/word". Then I see Word make a LOCK request to Slide which returns a 200 OK response, then Word makes a GET request that returns a 304 NOT MODIFIED. Then Word opens the file Read/Write IE - GET Word - LOCK Word - GET However, when I try this with my own content store connected to my existing Content Management application, Word opens the document Read-Only. The first GET from IE and the first LOCK request from Word are exactly the same as described above - including the 200 OK response to the LOCK request. However, right after the LOCK, Word issues an UNLOCK request, before making the final GET. The file opens, but is Read-Only, because of the UNLOCK I'm sure. IE - GET Word - LOCK Word - UNLOCK Word - GET I cannot understand what makes Word issues the UNLOCK when using my content store versus the default Slide File store. Here is the XML response to the LOCK request, which is the same in both cases, save the value of the opaquelocktoken ==================== <?xml version="1.0" encoding="UTF-8"?> <D:prop xmlns:D="DAV:"> <D:lockdiscovery> <D:activelock> <D:locktype> <D:write /> </D:locktype> <D:lockscope> <D:exclusive /> </D:lockscope> <D:depth>0</D:depth> <D:owner><![CDATA[fallij]]></D:owner> <D:timeout>Second-179</D:timeout> <D:locktoken> <D:href>opaquelocktoken:5db646da4738896d5318dcb754ce331f</D:href> </D:locktoken> <D:principal-URL> <D:href>/slide/users/unauthenticated</D:href> </D:principal-URL> </D:activelock> </D:lockdiscovery> </D:prop> ====================== The are only two difference between the two sets of requests (as far as I can tell). 1. The LOCK request from WORD to my content store includes an additional request header "Authorization: Negotiate YIILgQYGKwYBBQUCoIILd......eQI5Rn8V3YJkS" 2. The URI path for the file in my content store does not include the ".doc" file extension. The URL looks like http://host/slide/12345 <http://host/slide/12345> where "12345" is a document id within my existing Content Store. The ID from the URI is simply used to return the correct information about the file when retrieveRevisionDescriptor and retrieveRevisionContent methods are called. Thanks Jonathan
