Popular links

Topic maps


Claudio Monteverdi

Alexander Johannesen

Technologies used
topic map logo
xSiteable logo
Mon, 16 Feb 2004 13:00:00 GMT

Notice! This blog is no longer updated as such, and the new spot to point your feedreaders and blurry eyes are https://shelter.nu/blog/

This also means no more comments here, and especially not you spammers, you filthy floatsam of the internet!

New project : topic maps Wiki

I've been planning a new project for some time, and this is the initial Release of Ideas. The project has been brewing for some time, but needs have arisen for me to actually get it up an running in a somewhat fast pace. I need a system for updating project areas with documentation, people involved, business cases, history and so forth (the usual suspects), more as a collobarotive system more than specialised editors of content. Very much like a Wiki, but fixing the (for me) appearrant flaws of current Wiki's.

The project is called Topiki, and is a Wiki- and topic maps based application for intranet/internet collaborative documentation / CMS. It works almost like other Eiki's, but has subtle Topic maps features that would make it ideal, and really is one of the big bugbears I have with Wiki's today. Here is the feature list;

  • Topic maps based Wiki (the definition)
  • PHP as the underlying technology for easy distribution (flat files, no database)
  • xSiteable notation, because it is powerful and works
  • Taxonomical classification, so that Wiki elements can be hiercally viewed
  • XTM, XFML, RSS and ATOM feeds of changes, announcements, pr. page and pr. site
  • Subversion and WebDAV (WebFolders on Microsoft platforms) support for versioning

Why is this project different from other Wiki's? Normal Wiki's have binary links to pages, so the relationships between them is non-exsistant / singular. This is a major flaw to fix. Also, normal Wiki's focus on having backend databases for storage, which hinders easy distribution and setup. Also, Topiki introduces scoped topics, various names for the same thing and faceted classification.

Anyways, I would really appreciate feedback on this before I get too gritty with it. All the problems from Wiki's, CMS and KM systems are wanted.

Update, 26th of February : A mail I sent to the Topic Maps mailing-list to explain a few queries;

The main idea of doing this project was to "plonk" two files in a file structure somewhere (with a property in it telling it what Project Manager [the magical software called Topiki Mama] it was attached to) using only the capabilities that comes with a PHP default installation. This basically means "install PHP if you haven't got it, download and extract Topiki somewhere, and plonk your two files (index.php and config.xml, or edit existing ones). That's it; run and use."

All links in the Topiki network will be binary when they are created, but all links (associations) can be edited to be of any association type. Same with links to resources (documents, images, etc).

At the moment I have a few classes that do all the magic stuff, so after installing Topiki to get a page to have Topiki capabilities it is as easy as ;

// Where is Mama?

define ( "TOPIKI", "/some/path/here/to/topiki/install/" ) ;

// Configure, and let Mama know where we are

require ( TOPIKI."topiki.php" ) ;

// Show thyself!

$project->display() ;


There is a template engine in there as well, so you can have existing code wrapped around this as well, and in your config.xml file you flag templating on or off.

Anyways, there will be at all times an XTM file in the Topiki Mama home directory that reflects the whole node-tree and its content. (Since Topiki uses notation for content, this is easily handled in the XTM) To extract information from your tree-nodes is a piece of simple XSLT or whatever you'd like to use. (I havn't thought much about XTM import, but I can rip out some stuff from xSiteable to do just that)

As for libraries and stuff, nothing but the default install (this is so that it is easier than easy to dump this in a intranet to get Wiki capabilities in there) I use flat files for storage and content (so no heavy-duty traffic, but it will do for most intranets). (At the moment I'm possibly, maybe, probably, definitly thinking of using PHP5, for both better OO system and a few underlying XML/XSLT goodies that would make a lot of code redundant and improve performance. Strong objections will be considered. :)

I have tons of ideas popping up right, left and center, so I'm trying to jot down a small spec, and somewhere in there lies a small PoC that can be tested and commented upon before the juicier bits gets done, but all in all I think it can prove itself interesting.

Comments still very much wanted.

Permalink (Mon, 16 Feb 2004 13:00:00 GMT)| Comments (0) | Topic maps