Building an ODF Intranet Portal?

jeevesbond writes "I have been doing some feasibility work on creating a FLOSS Intranet Portal for ODF documents; the first task is to find existing projects that already provide some of the required functionality. The requirements are: version control — including diff and merge capabilities for ODF; integration with OpenOffice for check-in/out as a starting point; a Web-based CMS for group sharing of files (preferably one that can be extended to perform other tasks); and network authentication for the CMS (so users don't have to login twice). The eventual aim is to be able to bundle all this up in some way: 'apt-get odf-portal', for instance. Which FLOSS tools would you use for this job? How would you handle diffs and merges for ODF documents?"

Comments

knowledgeTree

[smoondog]

KnowledgeTree doesn't do everything you want, but it is an easy to use web application that supports plugins (addons).

Diffs

[JohnFluxx]

Store the odf's as unzipped folders. Zip it when the user downloads it, and unzip it when the user uploads it.

You'll probably want the download/upload down via webdav. User uploads say via webdav, you notice the upload, unzip it elsewhere, and commit. Should be easy :-)

As a slight twist, you could probably instead have the unzipping done as a pre-commit hook in svn server.

Re:Diffs

[jeevesbond]

Is there a way, when the user is browsing the WebDAV repo, that the unzipped document could still be shown as a single file? Is there some useful hook in SVN, or a way to keep both the zipped document and the directory in SVN at the same time (whilst hiding the directory)?

Thanks for the suggestions though, another avenue to explore!

Re:Diffs

[khanyisa]

I've tried doing this in svn - unfortunately OpenOffice.org changes random class strings in the XML on each save, so it doesn't work as well as expected.

You can do version comparisons graphically through OpenOffice.org though - TortoiseSVN contains some scripts to do this - using the change review mechanism, and I've found this very useful

Try this for a start

[narrowhouse]

http://dev.alfresco.com/

Their description:
"Alfresco is the leading open source alternative for enterprise content management."

I'm sure it doesn't do everything you want out of the box, but you wanted FLOSS for a reason, right?

Re:Try this for a start

[afd8856]

Maybe this: DocumentLibrary or CPS Project

New: FreePoint

[WED Fan]

Coined: FreePoint - The SharePoint Alternative

As a .NET and SharePoint developer, one of the biggest selling points for SharePoint is that integrates with Office, has a fairly easy to use web interface, and allows the user to design their experience and manage their own pages.

Here is the design challenge for a FreePoint tool:

• Option to use almost any SQL backend (MySQL,MS SQL,Oracle, etc.)
• Can be set up to use almost any web server
• Can be set up on any OS with web and SQL services
• Integrates with MS Office (you cannot alienate a huge install base)
• Integrates with OpenOffice
• Has a robust OSS workflow engine
• Can be extended with "web parts" (flakes, gadgets, whatever term you like)
• Can be programmed against using Mono, PHP, various languages an frameworks
• Biggy: Works with Active Directory (under Windows) or other such under other OS's.
• And others

Something that takes SharePoint, kills its short-comings, while expanding its usefulness, and opening it up, will be a SharePoint-Killer and get more OSS adopted. (We just had a client switch off a planned OpenOffice deployment after finding out about SharePoint.)

Re:New: FreePoint

[mpapet]

I am intrigued by your ideas and wish to subscribe to your newsletter.

Re:New: FreePoint

[jeevesbond]

Thanks for the excellent answer.

I'd actually like to steer clear of too many comparisons with Sharepoint, a clone will always be directly compared to Sharepoint, always living in it's shadow. Although most of your points are important though, a FLOSS product must provide an easy transition from any other system. Working with MS Office will be essential (although to get a feel for how the workflow will operate I thought OpenOffice would be best).

The current working title is: OpenDocumentPortal, how does that sound?

For the backend I've been looking at WebDAV with SVN and autocommit, for the frontend I'd like to create a Drupal Mod. The ideas here a great food for thought of course, I'm very interested to hear how others would do this project.

O3spaces

[Stefan73]

Have a look at http://www.o3spaces.org/. It seems that they pretty much match your requirements.

Re:O3spaces

[Kadin2048]

Except that it seems to be closed-source and proprietary, yeah. But if you can live with that, it's a very different question, because then you lose much of the motivation (besides upfront cost) for using OpenOffice/StarOffice instead of MS in the first place.

Re:O3spaces

[Stefan73]

Yes, you're absolutely right. It is closed source.

Convert Open Office docs to text?

[LWATCDR]

Does any one know of a program that will take an open office document and create a text file from it?
That would be a good step in the right direction. At least as far as searching the contents.

Re:Convert Open Office docs to text?

[John Hasler]

Package: odt2txt
Priority: optional
Section: text
Installed-Size: 100
Maintainer: Nelson A. de Oliveira
Architecture: i386
Version: 0.2+git20070106-2
Depends: libc6 (>= 2.3.6-6)
Filename: pool/main/o/odt2txt/odt2txt_0.2+git20070106-2_i386 .deb
Size: 16160
MD5sum: 499bfd9439fde00936d35cb2f0d58836
SHA1: 01d3cc5ad7cacdc6486a921bf5368d065b12f20e
SHA256: 35343b9a5bb4ebe48c97da61a5d5631bf48f3d161464339ef8 b1f1e2ab6bbf9f
Description: simple converter from OpenDocument Text to plain text
  odt2txt extracts the text out of OpenDocument Texts, as produced by
  OpenOffice.org, KOffice, StarOffice and others. It is small and
  fast, can output the document in many encodings and adopts to your
  locale.
  .
    Homepage: http://stosberg.net/odt2txt/

Re:Convert Open Office docs to text?

[Anonymous Coward]

You could probably write a 3 line bash script to do this (unzip, extract the ODF content.xml, regex to remove all tags) but if you're wanting a conversion engine try Docvert

Re:SharePoint vs OOO

[mpapet]

I know that it's desirable to have alternatives to the Microsoft crack pipe, but the reason your customer probably chose crackpoint was because they've got Microsoft desktops and domain controller. Whether you (or me) like it or not that's how they roll.

There is no "sharepoint killer."

Concentrating on being "better than product XYZ only GPL" is okay, but it won't ever overtake XYZ. Instead, concentrating on making the product insanely useful to many different niches, one niche at a time is generally a good plan. You aren't going to overtake crackpoint in this example, but you create something different where crackpoint stands as the clearly inferior choice in selected applications. Over time, the number of applications grows.

Your ideas about extensions and backend flexibility (unixodbc/win32odbc) are the way forward.

Finally, ActiveDirectory has an open counterpart that is excellent, it's called OpenLDAP. Many so-called identity management systems use LDAP as their authentication/identity management engine. They make money by putting a friendly GUI on top. If they don't use LDAP then they don't know what they are missing.

Emma

[oldbox]

The University of Georgia has developed a open source XML-rich framework for writing based on Open office and firefox called Emma that seems to fulfill most of your requirements, and a good deal more.

http://www.emma.uga.edu/

From the site:
  (Electronic Markup and Management Application) enables

        * writing, editing and posting compositions
        * collaborating on and evaluating texts
        * web-based collecting, modifying, distributing, rendering and archiving of student and professional writing
        * creating and maintaining portfolios

  puts people and texts together.

Plone

[jdstahl]

Plone is a powerful, extensible, easy to use CMS that will probably get you a good portion of the way there.