ivailo ilionov <[email protected]> writes: > I have a few questions which will help me to write the application. > > I understand what webdav can provide ( i've read the rfc : ) ). I've > also read that > there are work arounds in the face of wiki's for example which also > deal with documents, authoring, versioning etc.. > I've also looked at a package ( project) davfs2. So, what do the > project developers > expect of the webdav support in the server to be used for ? To be > mounted as local > FS, or to easy development of web applications which require that > functionality ?
Mount a remote location as a local FS is just a use case. Another can be to have more control on files like FTP offers and plain HTTP doesn't. > I'll assume that locking on write is desired, so probably shared > locking will be used on resources ? > I'm not sure how the files should be described ( stored ). I know that > ( well, i've never heard actually ) for example the ext3 filesystem > doesn't support custom attributes. > Is there a native file system, which can help with this problem, or > should I describe > every file in an xml ? That is, instead of directly accessing the > file, on valid request, > the xml file which has the information about the resource and it's > attributes, and flags.. will be read? You must not depend from the underlying FS, but you must use your data structures. You can safely assume that there is only the MyServer process to access the webdav data and all the locking mechanism is handled inside the process, without access any structure on the file system. This information will not be persistent between server reboots, but it is not a problem. Persistence can always be implemented later if there is really need of it. Giuseppe
