[Using Sakai] WebDAV & Sakai

[Marshall Feldman]

Laura is exactly right, although her comments point to a long-term
solution but I'm hoping to resolve this in the next week or two.

Universities are fond of saying their faculties do teaching, research,
and service and that the three functions all contribute to each other.
But although not entirely, Sakai and other LMS's tend to silo not only
the three functions but within the teaching function courses from each
other and within individual courses separate activities from each other.
For example, if for my research I am reading an article on my iPad and
think, "Say, this would be a great supplemental reading for both the
research methods course and the theory course I'm teaching this
semester," how many steps must one go through to share the article this
way? It should be as easy as one step: "Share with xxxx: Research
Methods and yyyy: Theory for Practitioners." But with Sakai and other
LMS's today, we're talking a dozen steps at least. And this is about as
simple an example as one can imagine.


On 2/15/15 4:23 PM, Laura Gekeler wrote:
> Well now you've gone and done it Marshall... all this WebDAV stuff
> aside, what we're really pointing out here is that we believe file
> system access to files in a Sakai course site repository should be
> accessible outside of Sakai, isn't that it?  We'd like to use end
> points and sharing from many different places, including:
>
>   * mounted and behaving like a native part of the OS file system.
>   * using 3rd party tools like Cyberduck
>   * and inside Sakai too, we should be able to copy or link to a Sakai
>     file in another course site, if we want.
>   * How about from inside an ePortfolio system, so students can choose
>     the same file they submitted for an assignment, as part of a
>     portfolio they're building?
>   * What about learning analytics? At some point we're going to need
>     to search files from their linked endpoints (or otherwise
>     traceable copies) in our analytics data warehouse in order to do
>     text mining of assignments in the spring of 2015, of sophomores,
>     which received A's ....
>
>
> WebDAV isn't the solution (interim maybe), but something like WebDAV
> expresses the requirements for the beginning of a solution. Such a
> solution will be at the core of a learning ecosystem.
>
> IMO
> Laura
>
>
> Laura Gekeler
> LMS Administrator //Concurrent Instructor
> Teaching and Learning Technologies
> University of Notre Dame
> P:(1) 574-631-2402
>
>
> On 15 February 2015 at 16:08, Marshall Feldman <marsh at uri.edu
> <mailto:marsh at uri.edu>> wrote:
>
>     Chuck,
>
>     I understand your point. For years Microsoft claimed to work with
>     WebDAV, but Microsoft's implementation was notoriously buggy.
>
>     OTOH, Cyberduck is primarily a file transfer program that can use
>     any of several protocols, WebDAV being one of them. WebDAV, OTOH,
>     is a way of using files on a web server as part of a distributed
>     file system. Since in any file system one can move files,
>     Cyberduck may do a very good job with this aspect of WebDAV on
>     Sakai. But just because it can use this facet of WebDAV
>     successfully does not mean Sakai's implementation is complete or
>     without bugs.
>
>     So as an end user, I'm still at this point caught between several
>     parties pointing fingers. My local support person claims that
>     Sakai doesn't really support WebDAV's distributed file system
>     feature because Sakai keeps its files on varying machines in a
>     cluster. Apple claims its OS works with WebDAV, and I've used it
>     successfully with other WebDAV servers. So this also points to
>     Sakai as the culprit. Some of the responses to my original query
>     have implied that our local configuration of Sakai is at fault.
>     Your reply implies that Sakai's implementation of WebDAV works,
>     but Cyberduck functionality is not proof, as I said.
>
>     So, I'll try a third rephrase of my question. I am concerned about
>     Sakai providing "fully functional" WebDAV, to any operating system
>     or app anywhere. By "fully functional" I am including the
>     distributed file system aspect. By "app," I'm allowing for a more
>     comprehensive app than Cyberduck, one that would actually make the
>     local operating system see the WebDAV server as a mounted volume,
>     thereby working around any bugs in the native WebDAV
>     implementation. (For example, a company called OpenText
>     <http://connectivity.opentext.com/products/network-file-system.aspx>
>     sells a product that lets PC's share distributed files using NFS.
>     It effectively adds this feature to Windows' native file system
>     access.) So now the question is, "Can Sakai work as a fully
>     functional WebDAV server with any operating system anywhere,
>     whether through the operating system's native WebDAV
>     implementation or through a third-party add-on?" Or more
>     specifically, "Can Sakai provide distributed file system access
>     through WebDAV to any operating system whatsoever, whether or not
>     a special app provides this capability, so that files on Sakai are
>     readable and writable through the native file system's built-in
>     file access capabilities?" In other words, is the distributed file
>     system aspect of WebDAV successfully implemented in Sakai at all?
>
>         Marsh
>
>     On 2/15/15 3:01 PM, Charles Severance wrote:
>>
>>     On Feb 15, 2015, at 1:21 PM, Marshall Feldman <marsh at uri.edu
>>     <mailto:marsh at uri.edu>> wrote:
>>
>>>     So, to rephrase my original question, does Sakai support WebDAV
>>>     client-server mounting at the level of OS X or Windows? In other
>>>     words, does Sakai support genuine WebDAV as implemented on the
>>>     two most common operating systems currently used in academia?
>>
>>     Marshall - Sakai has a WebDav that we are confident works with
>>     Cyberduck and *may* work with other operating systems.  There is
>>     no "formal commitment" that we will do "whatever" the OS/X or
>>     Windows operating system happens to do.
>>
>>     Having worked in standards and interoperability for a long time -
>>     just saying "we are genuine WebDav" is not at all a guarantee of
>>     interoperability.  
>>
>>     Our WebDav code is based on an WebDav from 2004 and we have made
>>     a few improvements  since then.  The operating systems feel no
>>     compunction to be careful in using WebDav to maintain
>>     interoperability with anything other than their own
>>     implementations.  CyberDuck is more interoperable because since
>>     they are neither Microsoft nor Apple - they need to be a little
>>     more careful how they implement WebDav so as to remain interoperable.
>>
>>     Since our webdav is so old, it would take a lot of resources to
>>     try to find and fix all the issues to guarantee it works with
>>     Mac/Windows.
>>
>>     If we had a lot of resources, we might start over with a more
>>     modern WebDav protocol implementation like this:
>>
>>     http://milton.io/
>>
>>     So for now, we can use WebDav if it works but we don't consider
>>     "WebDav not working on Windows" a "bug".  We need to work on
>>     things like getting rid of iframes :)  Of course if someone felt
>>     strongly about fixing it - they could take the initiative and
>>     build us a new WebDav :)
>>
>>     /Chuck
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://collab.sakaiproject.org/pipermail/sakai-user/attachments/20150215/2f756a2e/attachment.html