Wagn 1.0 Blazes a Bold New Trail in Wiki Software

Written by Steven Walling / June 10, 2009 6:00 PM / 9 Comments

Today marks the 1.0 release of Wagn, a pioneering yet little-known software that wiki inventor Ward Cunningham has called "one of the freshest contributions to wiki since I coined the term."

Created by the non-profit Grass Commons and jump-started by a grant from the Meyer Memorial Trust, Wagn has been quietly honed into a tool that breaks new ground in collaborative software. What makes Wagn special is that it takes the wiki that you know and adds database structure and functionality. Also a simple CMS, Wagn can handle data like no wiki you've ever seen.

A Little Background

Now, wiki software is incredibly easy to write, but not so easy to write well. Tiny wiki programs, some as small as three or four lines of code, are commonplace. More comprehensive wikis like MediaWiki, Socialtext, Confluence and MindTouch have grown to become towering collaborative platforms through the addition of countless features.

But not one of these has changed the fundamental building blocks of a wiki: easily editable and linkable web pages. Wagn is easily editable and linkable by anyone, but what sets it apart is that it doesn't hinge on a fabric of web pages.

Wiki + Database = Wagn

What makes Wagn different - and the part that gives it its database characteristics - is that its basic unit is not web pages. It's cards. What is a card, you ask?

Just like a page, a card is a metaphor. Like playing cards in real life, Wagn cards can be nested together to form stacks (a single page might contain dozens of Wagn cards). Not only can cards be placed inside each other, but they can exist in many places at once, meaning nothing ever need be repeated.

Cards can be of any type, but all of them retain the character of a wiki in that they're fully editable and have a revision system. There some basic cards built-in, like Image cards or User cards, but anyone using Wagn can create cards types to suit their purpose.

The ability to create whatever cards you like, and then shuffle and reorganize them basically at will, means that people organizing knowledge and developing the patterns to fit an organization's needs can have access to a tool that is truly flexible. Other database features that Wagn employs include its own query language, created to call up dynamic lists of cards.

So What is Wagn Good For?

Wagn developer Ethan McCutchen said in a phone interview that originally Wagn was thought of with collaborative public knowledge works in mind, specifically a consumer education site. But as outsiders - especially in the non-profit community - began to use Wagn for their projects, it has adapted to a wide variety of use cases. Probably the most ambitious endeavor to use Wagn thus far is Connectipedia, a community-built site about non-profits in Oregon.

Programmers might appreciate that Wagn is one of the only mature, free and open source wikis built with Ruby On Rails. Wikis built on the framework are so rare, in fact, that even the official Rails wiki uses the PHP-based DokuWiki.

As yet, Wagn has not seen much enterprise use, but with the power and flexibility of the software, along with the unveiling of some inexpensive services that include both consulting to deploy behind the firewall and public or private hosting, the door has been opened to making Wagn work for anyone.

Comments

Subscribe to comments for this post OR Subscribe to comments for all ReadWriteEnterprise posts

What makes Wagn different - and the part that gives it its database characteristics - is that its basic unit is not web pages. It's cards. What is a card, you ask?

Yes I remember Apple Hypercard back in late 1980's but nice to see Ward Cunningham improve functionality of same ole same ole wikis.

Posted by: Steve Ardire | June 10, 2009 6:25 PM

Looks really interesting - nice to see someone giving what is becoming a well known paradigm for collaborative editing a bit of a rethink. I'll definitely be running this up and having a closer look.

Posted by: Damon Oehlman | June 11, 2009 2:46 AM

Well, actually the XWiki Open-Source enterprise wiki project has been soing something similar to this for a while now. The project was started back in 2003 and it has had the concept of "XWiki Objects" ever since.

A XWiki Object is a set of properties defined by a XWiki Class that can be put into a wiki page. It holds a set of data (for instance an object describing a product sheet would have a name string property, a price number property, a description textarea property...). Several objects can be put in the same page. The page can then be used to display object's content in a nice way based on a template. You can also write queries right into the wiki using the HQL query language.

More information about XWiki can be found on http://www.xwiki.org . This page is especially interesting as it shows what can be done with XWiki using these objects (you can basically create applications right into your wiki): http://platform.xwiki.org/xwiki/bin/view/Main/SecondGenerationWiki

XWiki is also a real comprehensive wiki (it's sometimes dubbed the "Open-Source equivalent to Confluence" as it's also a java-based wiki with *a lot* of features). The company backing the project (XWiki SAS - xwiki.com) has EMC & EADS among its customers.

Wagn looks really interesting too and it might be a source of inspiration for the XWiki project in the future.

Posted by: Guillaume Lerouge | June 11, 2009 8:32 AM

Thanks for rundown on XWiki Guillaume, I've definitely taken a look at it before, and it's a great product. It's cool that you've included in data-handling like XWiki Objects.

However, the fact that XWiki Objects are something "that can be put into a wiki page" would seem to make them fundamentally different than Wagn cards. Cards are essentially their own individual wiki pages put inside a web page (or eachother), so from my perspective it's a different paradigm.

Posted by: Steven Walling | June 11, 2009 10:28 AM

Well, XWiki also allows you to include pages within other pages which means the duplication issue can be taken care of too ;-)

However you're right, there seems to be only one "level" in Wagn (a card) versus XWiki's 2 "levels" (pages + objects). I haven't tested Wagn yet (too bad they don't have an easier-to-install demo package) but I'd probably come up with a better comparison if I were to test it thoroughly first :-)

The concept sounds appealing and they're obviously aiming at the same space XWiki is so I'm definitely gonna try it and see what choices they've made (the concept of documents embedded in other documents and "remixed" is often hard to explain to new users). Their UI / UX choices are probably interesting.

The main advantage of such systems is to make it easier for developers to create systems that provide more value to end-users. I think systems such as Wagn & XWiki make it easier for developers to create situated applications (see http://www.shirky.com/writings/situated_software.html). Such collaborative applications are likely to become more and more popular in the enterprise space (we regularly see customers asking for small apps to manage day-to-day processes) and platforms such as Wagn & XWiki make those apps really easy to write & deploy internally.

Thanks for the kind words about XWiki too, I sure hope we'll see it featured on RWW one day (don't know whether you saw the releases notes for XWiki Enterprise 1.9 yet by the way but we've added quite a lot of new stuff lately, cf http://bit.ly/23NO6)

Posted by: Guillaume Lerouge | June 11, 2009 12:17 PM

You're both definitely right that the "everything is a card" strategy has a lot of special consequences.

If you have a zip code card on an address card on a company card, and all of those cards will have the same key wiki traits: a unique name and readable url, a revision history, distinct permissions, etc.

Some really neat things come out of that approach, like the fact that if you build a new "view" of a card (which is sort of like a perspective on it: open, closed, name, content, etc), then you can apply that view to anything on the whole system.

Lots of systems have something somewhat similar, but few take it quite so far. For example, most things on Drupal are nodes, but many important things, like users, are not. So there are lots of neat things that you can do arranging and displaying "nodes" that won't work with users without a bunch of jiggering.

Great things seem to come out of staying faithful to everything-is-a-card even when it's tempting to do something more quickly by adding some new structure. My hunch is that more systems will go this way, and that Wagn is just starting to tap the potential.

Posted by: Ethan McCutchen | June 11, 2009 5:46 PM

So what does this do that mediawiki doesn't do through it's support of transcluding and all the functions supported by it's template markup (used in things like infoboxes)?

Posted by: qwertyu | June 13, 2009 7:41 AM

Hey, great to see Wagn getting some well deserved press. To echo above, yes, it's a really interesting approach.

To answer the above re: similar functions in mediawiki (or whatever), the functionality is all present on other platforms, but the guts are pretty innovative. To make an analogy, all the functions of a mediawiki page could be done with hardcoded HTML, but the ease of use really can't compare, can it?

Posted by: Jonathan | June 14, 2009 9:03 AM