During the late 1970's I was asked by Dr Eugene Ogden to help in the testing of his new Fern Identification book. What was different about Dr. Ogdens work is he had devised a synoptic key for doing his identifications.
Botanists usually use Dichotomous keys to do identifications where the user is asked about a particular trait after which you are asked about the next trait the order of the choices controlled by the author of the key. Dr Ogdens key allowed the user to select which character to identify first. The user kept track of his choices on sort of a score card eliminating the ferns that were not the fern you were examining.
Many years later when I became more experienced with computers I remembered this experience and thought with the computer I could allow the computer to keep track of the choices and do the hard work of searching for the choices in the data records.
My first experiment was on the AMIGA (IBM was only CGA No Pics) writing a graphic interface to identify gilled mushrooms. I was able to make it work but I never really completed the data and specimen pictures for this enormous project.
In 1993 Dr. Richard Mitchell and I discussed doing the Fern Identifier as a computer program. At that time IBM was EGA but we thought that by using the AMIGA to develop the graphics we could produce graphics good enough to write an identifier. We worked on the project off and on for several years without compensation while experimenting with various graphic methods and program logic. It wasn't until 1997 that we had collected the graphics and written the appropriate software for release.
Although this program contained less than 70 species to identify it took several years to collect the photos of all species. During that time the IBM platform adopted the SVGA format allowing for 256 color graphics. We went from DOS, to WIN3.1, to WIN95 in that time and I had to constantly update just to keep up with the changes in hardware.
Since that time I have written numerous identification programs. In the past year I have re-written the Fern identifier in the C# programming language.
Sometime in the fall of 2004 I was contacted in reference to doing a rewrite of the NYS Museum Fern Identifier. This project required a complete rewrite of the software written under Windows 98. So that it would be compatible with the on going developments in operating systems and more importantly hardware it required a complete rewrite to C#. Also all the images had to be converted to be compatable with C#. Although we had no contractual agreement I had been working on this project for about 10 years and the contracts in the past had been a real joke. On one occasion the first and second payment on the contract were long past due before the contract was even signed.
I agreed to work on the NYS Museum Fern Identifier project pending their new budget in March. I delivered a working program sometime before the first of March. Several months past with no contact or mention of payment. After several inquiries about payment I received a contract with very specific requirements for me but contained no specific date for payment. I had already invested almost $1500 in this rewrite for contracted labor, software, and hardware. I felt it was safer to “cut and run” from what had always been a bad experience working with these people.
I have recently rewritten the program again exclusively in C# eliminating the C console window. I have compiled under .NET Visual 2005 using the .NET framework 2.0. I have also loaded the data into a data array so that the program now runs much faster especially when running from a CD as the data is only read once.
I will not release this program without getting paid and the BRI seems to have little interest. So it remains in limbo.
In the month of Feb. so far I have had 72 hits on my Fern Identifier page. Notice that the Bri had two hits on Feb 6 with none since. It is probable that those two hits were mine.
