On Fri, Nov 21, 2008 at 9:22 AM, Julian Reschke <[EMAIL PROTECTED]>wrote:
> Marc Speck wrote:
>
>> Oh, ok. So you are saying that the following should work?
>>
> >
>
>> Index:
>>
>> jackrabbit-webdav/src/main/java/org/apache/jackrabbit/webdav/WebdavRequestImpl.java
>> ===================================================================
>> ---
>>
>> jackrabbit-webdav/src/main/java/org/apache/jackrabbit/webdav/WebdavRequestImpl.java
>> (revision 712647)
>> +++
>>
>> jackrabbit-webdav/src/main/java/org/apache/jackrabbit/webdav/WebdavRequestImpl.java
>> (working copy)
>> @@ -108,9 +108,7 @@
>> this.factory = factory;
>> this.ifHeader = new IfHeader(httpRequest);
>>
>> - String host = getHeader("Host");
>> - String scheme = getScheme();
>> - hrefPrefix = scheme + "://" + host + getContextPath();
>> + hrefPrefix = getContextPath();
>> }
>>
>> /**
>>
>
> Depends on where exactly hrefPrefix is used, of course. So in theory: yes.
Sure. I opened an issue https://issues.apache.org/jira/browse/JCR-1873 but
I do not know the implementation well enough to suggest a patch. There are
obviously more changes necessary than the one above.
>
>
> Though why is there the scheme and host in there? I'm somewhat puzzled...
>>
>
> Because somebody just looked at examples, and didn't read the spec
> carefully (as it frequently happens).
>
> Several WebDAV implementations already do return just the path (IMHO Apache
> moddav for instance), so clients have had to accept that for a long time.
>
Ok, thanks for the insights.
Marc
