A major update to the “Where are they?” area as well as a short history of it

At long last I have completed my update to the “Where are they?” section. The update covers both the 1992 and 1993 versions. Over time, I have I kept adding more and more automation to the process of ebay auction tracking. This has made it less time consuming to maintain and, at the same time, has helped to keep it usable. Actually, this section has undergone four iterations over the three plus years that it has existed. Not all of them were visually identifiable to the users of the site so I thought I would start with a brief, if somewhat technical, history.

When I first started this area I was typing the auction results directly into an HTML document. This was both tedious and limiting in that only one page existed which did not allow sorting and provided very few possibilities for analysis. That approach was fine when the list had only a few entries. As it grew I realized the need to allow a more flexible interface into the data. This led to the second iteration which included the use of a database. For that approach, I entered the data into an offline Lotus Notes database and wrote Java code to extract data into XML format and then used XSLT on that data to produce a variety of HTML pages each with a different sort order. Each sort was still a static page that was uploaded to my web server. So, while it got slightly easier to use, the additional steps made it was more tedious to maintain.

My need to reduce some of the tediousness led to the third iteration. I wrote a Java class that would take an auction ID as input, go out to ebay to parse the data directly from their web server and automatically populate it into that same offline Lotus database. This greatly reduced the data entry burden (and errors) as I only had to gather the auction IDs, visually verify the data and add my comments. But, it did nothing to reduce the drawbacks of the static nature of the pages. After three years of data collection, the page size for each one of the ten available sorts was approaching 1/2 Mbyte.

This led to the fourth and current iteration which is being released at this time. The database was migrated from Lotus Notes to MySQL so it could reside on my Web Server. I then wrote a Java application to provide the data in a dynamic way that is user-driven and also allows me to maintain the data directly through the web site. I still enter the auction ID, but, the server goes to ebay, parses the data from the page and presents it back to me where I can verify it and add comments. When I simply click a button it is then submitted into the database for all to see. Viewing is more refined with the list broken into multiple page with the ability to sort as well as to select a subsection of the results based on various data points such as color, reserve and transmission. Everything the web users sees is dynamically generated just for them.

Of course, I have always collected more information than I have made available. Previously, I had archived each auction to a PDF file. With this latest go ‘round, I save most of the auction listing into the MySQL database and also store the images on my Web server. Right now that expanded detail is only available to me as the site administrator and only for auctions beginning in June of this year. I am toying with the idea of making this additional information available to others but, as an enhancement to my not-so-successful donation model, I might grant access to this expanded detail to only those that help support this site. If you have comments on this updated area or any interest in the expanded version, please let me know.