#187: Working Out-of-the-box WebDAV

Contents
  1. Motivation
  2. Proposal
by Sidnei da Silva last modified Nov 06, 2008 12:25 AM

This proposal should track requirements/improvements/documentation for the GSoC 2007 WebDAV Proposal

Proposed by
Sidnei da Silva
Seconded by
Alexander Limi
Proposal type
Architecture
State
in-progress

Motivation

WebDAV should work out-of-the-box when Plone is installed with the major WebDAV Clients

  • OS X Finder
  • Windows WebFolders
  • Enfold Desktop
  • Nautilus
  • Konqueror

Proposal

  • Default marshallers for built-in content types should be sane.  IE, should be 'file-oriented', not 'metadata-oriented'.
  • No OS X resource forks should be visible. If possible, avoid creating them.
  • Allow WebFolders to work, document hacks if needed.
  • If time permits:
    • UI integration for switching default marshallers
    • UI integration for multiple export formats
    • "Download folder as ZIP"
    • "Upload as ZIP"
    • "Export search results as ZIP"


Support for adding metadata

Posted by Alexander Limi at May 23, 2007 08:30 PM
It would be great if there was a standardized way of adding metadata on upload (FTP or WebDAV) — something like:

http://server/folder/anotherfolder/+subject:screenshot,examples/+description:Screenshots+from+our+experiment/

Upload images into that (virtual) directory, and they will automatically be tagged as screenshot, examples and have a description attached to them.

OS X resource forks

Posted by Ricardo Newbery at Jun 24, 2007 12:11 AM
A discussion about the OS X resource fork issue...
http://www.nabble.com/OS-X-[…]-uploads-t3957484s6741.html

WebDAV and Versioning

Posted by Matt Howell at Jun 26, 2007 12:35 AM
I seem to remember there being some issues with using WebDAV clients in conjunction with the CMFEditions / Versioning. Do you know if this was ever cleared up? In a quick Google search, I found this old post from Alan Runyan that seemed to point to issues in CMFEditions:

http://lists.enfoldsystems.com/[…]/000097.html

Based on recent correspondence (approx. 6 months ago) it sounded like CMFEditions & Enfold Desktop (as an example) still did not work well together?

Also, this is somewhat off topic, but would it be difficult to implement a CIFS (Common Internet File System) interface, similar to Alfresco (which supports both CIFS & WebDAV), allowing users to mount resources as Windows shares? Or are users typically comfortable with the WebDAV experience?

Preferring WebDAV to CIFS

Posted by Graham Perrin at Dec 29, 2007 11:53 AM
On one hand: as Microsoft Office 2004 is terribly incompatible with WebDAV volumes, so CIFS may be more appealing.

On the other hand: end users of CIFS are more likely to be blocked by firewalls that are beyond the control of Plone administrators.

Considering these factors and others, I think WebDAV is the smartest approach.

Framework vote

Posted by Martijn Pieters at Dec 21, 2007 10:47 PM
+1 (low hanging fruit as the work is already done)

Framework team vote

Posted by Andreas Zeidler at Dec 22, 2007 10:01 AM

Framework team vote

Posted by Tom Lazar at Dec 22, 2007 03:07 PM
+1 (if the implementation should prove too invasive or too slow we can always still bump this to 3.2 but i don't see any reason to reject this plip per se for 3.1)

Framework team vote

Posted by Danny Bloemendaal at Dec 22, 2007 04:13 PM
+1

Framework team vote

Posted by Raphael Ritz at Dec 23, 2007 08:34 PM
+1 - any improvement on the WebDAV side is much appreciated IMHO. And I know that Sidnei knows what he is doing.

(sorry for the late voting folks. I didn't realize this PLIP was submitted for 3.1 - me bad)

3.3 Framework team vote

Posted by Martijn Pieters at Oct 26, 2008 05:01 PM
+1

Framework team vote

Posted by Danny Bloemendaal at Oct 27, 2008 03:09 PM
+1

Framework team vote (for Plone 3.3)

Posted by Andreas Zeidler at Oct 27, 2008 05:01 PM
+1. the PLIP was accepted as such for Plone 3.1, so imho there's no question about that.

however, i'd like to reinstate my comments about unanswered questions and completion of the PLIP itself (see http://lists.plone.org/[…]/001623.html). i'd very much like to see the PLIP page updated regarding that.

also, and needless to say, the implementation needs to be thoroughly reviewed, as in this was the reason for not including it into Plone 3.1 in the end — see the discussion in http://dev.plone.org/plone/ticket/7732 for more info.