<?xml version="1.0" encoding="iso-8859-1"?>

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"

"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<html xmlns="http://www.w3.org/1999/xhtml">

<head>

<title>cwm - a general purpose data processor for the semantic web</title>

<link href="/StyleSheets/base.css" rel="stylesheet" type="text/css" />

</head>

<body xml:lang="en" lang="en">

<p><a href="/">W3C</a> | <a href="/2000/01/sw/Overview.html">Semantic

Web</a></p>

<h1>Cwm TO DO</h1>

<p></p>

<p>Here are things we know need doing. Help is always appreciated. Maybe we

should put this in bugzilla sometime, and certainly in RDF.</p>

<p>Cwm hackers with access (Tim Berners-Lee, Dan Connolly, Sandro Hawke, Eric

Prud'hommeaux...) please feel free to edit this page. (There is a system team

work item to allow more flexible acecss to pages like this by collborators

using CVS.)</p>

<table border="1">

<tbody>

<tr>

<th>What</th>

<th>Description</th>

<th>State

<p>(blank =</p>

<p>no work)</p>

</th>

<th>Plans</th>

</tr>

<tr>

<td>Installation</td>

<td>Write setup.py in the traditional instal style</td>

<td></td>

<td></td>

</tr>

<tr>

<td>Datatypes</td>

<td>Parsers, seruializers and builtin functions, and lots of test

cases</td>

<td>timbl started

<p>2003-03</p>

</td>

<td></td>

</tr>

<tr>

<td>Web demo</td>

<td>A few tools on the web for doing different forms of processing,

such as data forma conversion, query processing, rules.

<p>A ,n3 suffix for the web site for looking at any RDF file in

N3.</p>

<p>Needs Safe Mode.</p>

</td>

<td>(old version by DanC now out of date.)</td>

<td></td>

</tr>

<tr>

<td>Safe mode</td>

<td>A mode which puts cwm in a sandbox, not doing web fetches etc, or

running rescource-limited. (Use python sandbox?)</td>

<td></td>

<td></td>

</tr>

<tr>

<td>Diff and synch</td>

<td>Find the diffrences between two graphs, express it in terms of

identifyable nodes, send it around as a news item.</td>

<td>(some work, on the shelf, -timbl)</td>

<td></td>

</tr>

<tr>

<td>Remote query</td>

<td>Remote query by looking up N3 query URL encoded and appended to a

service URI</td>

<td>Work started timbl, EricP, DanC</td>

<td></td>

</tr>

<tr>

<td>Test files</td>

<td>Extend test/retest.py to handle the RDF core tests. [done].

<p> Migrate CwmTests where appropriatete from cwm command line args

to decalrative descriptions of functions involved, for use by other

software.</p>

</td>

<td>(started timbl)</td>

<td></td>

</tr>

<tr>

<td>Performance</td>

<td>Reduce number of hash tables, and profile and speed this thing up!

Maybe clone lyyn.RDFStore with different stored optimized for

different cases. Example: Don't index the store if no query/inference

needed in command line.</td>

<td></td>

<td></td>

</tr>

<tr>

<td>API convergence</td>

<td>APIs for RDF exist in many products - workon convergence. (rdflib,

Mozilla, etc etc)</td>

<td></td>

<td></td>

</tr>

<tr>

<td>Guessing parser</td>

<td>The guess code in llyn refactored as a parser which makes an

educated guess at the format and dispatches to the appropriate one.

Use of that as the default mode for cwm.</td>

<td></td>

<td></td>

</tr>

</tbody>

</table>

<hr />

<address>

Tim BL, with his director hat off<br />

$Id: cwm_TODO.html,v 1.14 2003-03-25 20:01:04 timbl Exp $

</address>

</body>

</html>