Comments on Progress (Milestone 3)

I thought that I would drop a few notes about how things are going with respect to the Geoserver project.

First of all, regarding Milestone 3. User support has been put into the latest CVS versions. At the moment, this means that if you have a user account in the Geoserver sphere, you can use that account to validate yourself in order to request documents. (Nevermind that there are no tools for creating users at the moment... Coming soon...) The way it basically works is this. Each feature in the database has a 'Owner' field that corresponds to a user ID in the 'users' table (0 if there is no Owner). If you want to view the features that belong to another user, you have to log on as that user. (Examples are in the 'Sample Queries.txt' file.) Features that have an owner of '0' are accessible to everyone. This setup is by no means complete, as I intend to add more Unix-like ACL's, but it is a start. I'm also considering making the authentication use some sort of challenge-response scheme, or have the server generate a random key that is used to encode a password to improve security. Nothing has been decided yet, and these may be version 2.0 features. At the moment, the username and password are sent in clear-text. And at the moment, there are no concepts of groups. So, in summary, the user / group areas of this project require quite a bit more work before they can be considered close to stable.

Diversion: one of the reasons for the lack of completion on the users and groups is that I've been playing with SVG the past week. I'm amazed by the types of things that can be done using a "mere" vector graphics format. So, this weekend, rather than code users and groups as a good boy would, I created a rudimentary GDTP -> SVG converter. It is nothing really special at the moment (it only converts gml:LineString objects at the moment), but it is a start to what I hope to develop into a robust SVG-based display system for Geoserver applications.

Finally, regarding the schedule of things, I had slated for a bugfix / feature-frozen release on the 25th of November. I think that this will not be reached for a couple of reasons. The first reason is that I have a virtual memory implementation due for my operating systems class due next week, and I expect it to consume my time this upcoming week. Also, I'll be away from my development boxes this weekend as I visit my better half's family for Thanksgiving. However, I intend to do what I can to hit that crucial 1.0 release as soon as possible. Good progress is being made, so be patient with the overworked, starving, student.

Posted by Chris J. Karr 2001-11-18