At a large institution like UCF, it is good to have a plan for emergencies. I set up a simple form that will update the main page at http://ucf.edu in an emergency so that important information can be realeased as fast as possible.

The main page is an html file that is copied every few minutes from our database driven application. This speeds up the website and cuts down on processor utilization considerably. A simple update to our cron job was added that checks if the site is in emergency mode and pulls from our other emergency page. This emergency page is created with a simple form and simple template file.

I created this page updater to be reliable and simple so that there is little turn arround time from emergency situation to information available. The form edits files in the filesystem instead of using a database that would require more complexity. There is a place to update the important information. That info is then put into the pre-built template when the user hits preview. Once the user is satisfied with the way it looks, there is a button to enable/disable the page. It updates the status that the cron job looks for and the main page will change in under a minute.

I say middle end (even though it’s not an end) since I didn’t work on the front-end skin or on the content-management back end. The new site was launched a few days ago (6/21/8) and uses an installment of InQuira Information Manager as the content management system. With the CMS comes a JSP tag library. I used the tag library to extract data from the CMS and format it in the front-end layout. I was also responsible for designing the structure of the channels, categories, and schema in InfoManager. Read the rest of this entry »

Workflow Management

To Be Done is a Web-based workflow tool that manages the collection, tracking, and processing of work requests. It is written in PHP and uses a MySQL database. It facilitates the collaboration between teams by enabling team members to create requests for other teams’ members to complete. Time-to-completion data is stored when a user completes a request and can be used to display totals, percentages, and averages of requests and hours in a report that can be generated automatically. The report that is generated can also display specific information per user and per course.

Read the rest of this entry »

Valhalla, which is implemented with PHP and a MySQL database, powers the Techranger Web site. It provides the Techranger team with the ability to manage all aspects of the site including: image galleries, user biographies, job applications, XHTML templates, multiple cascading style sheets and other site features.

For this project I created the data access classes as well as designing the database itself. I also wrote the job application, image gallery, project showcase, and user biography features.

To Be Done is a Web-based workflow tool that manages the collection, tracking, and processing of work requests. It is written in PHP and uses a MySQL database. It facilitates the collaboration between teams by enabling team members to create requests for other teams’ members to complete. Time-to-completion data is stored when a user completes a request and can be used to display totals, percentages, and averages of requests and hours in a report that can be generated automatically. The report that is generated can also display specific information per user and per course.

Similar to the request functionality, there is a tech support tracker which uses the same process with the addition of client information being stored.

I joined the project to develop the 2.0 version, which was a complete re-write since the original developer didn’t give us any comments or documentation before leaving.

My initial role in the project was to design the database and create the data access classes. After the interface developer left, I started fixing bugs as well as developing new features, such as the report generator. I also began broadening the scope of the project to allow for any team to assign tech support incidents or requests to any team.

I created an OpenID provider for CDWS. It connected to our LDAP server to authenticate users and provided them with an OpenID in the form of http://username.cdws.us

I used the Jan Rain PEAR library for the OpenID connections and followed the code they provided in an example. I modified the example code heavily to fit it into the CyTE framework for easier future development and maintenance.