Project
The project began in 1987 when the large amount of data available was starting to make my written ledgers unworkable. Ledgers and simple computer 'logging' such as textfiles, database and spreadsheet systems become unworkable. Mainly due to the fact that they copy a ledger system that is a chronological record rather than a collection base for thousands of small items all assembled out of order and subject to editing as new information becomes available. To distribute information from such a system becomes so labour intensive any requests are deterred and the information jealously guarded. I required a system that would allow simple data entry without having to understand the complexity of rolling stock history. The system itself would join together all the histories. This would eliminate the 'static' nature of hand written work, the errors inherint in cross referencing and the time consuming ordering of data items. The system as envisaged was developed to provide:
Program development and data entry proceeded simultaneously for eight years. In that time some 230,000 vehicle history items were typed in that were being referenced by some 82,000 vehicle identities. In fact one system in current use is handling in excess of 200,000 vehicles for 500,000 data items. That database has built an Australia wide rolling stock history system on top of the data entries I started.
Upon request, the program will start tracking forward and backwards across all the code, number and ownership changes to present a single continuous history paragraph. This is assembled as each request is made to ensure that any recent data entry will be automatically included. Such is the scale that it is impossible to ascertain a history until it has been output.
Once it was realised that the formatted data could be used to tackle other questions, then the development included such features as an 'On Register' check which could accurately return a list of vehicles in service at a nominated date. Obviously this requires all the data to be input. Histories for a class could be output numerically by date and by event providing a very quick 'snapshot' of a class history over 80+ years.
To design the system it was necessary to understand what all the terms meant that were in rail use. Several terms needed specific meanings as applied to the Victorian rail scene
Development
The program was partially developed in DOS with the dBase III+ instruction set and development progressed to through the various 'Clipper' versions Summer '87 to Clipper 5.2. The program was called Register and was developed in DOS because 'Clipper' contained a powerful set of file utilities, could expand database capability and was very fast. Under DOS, the program included a self written file manager and text editor. This was to ensure that the user did not have to rely on any third party programs. Progression to a Windows environment was planned but work never progressed after 1997 when FTS (File Tracking System) became a more pressing issue.
With FTS development finished in March 2003, attention returned to Register and the future development path. Concepts from 1992 had the Register program track images applicable to rolling stock via a 'token' system accessing a separate database. With FTS established, attention could be turned to integrating the two systems. This became stimulated by a pressing need to distribute a Windows environment program that could track histories and photos. Since 1988, no cross referencing history program had been successfully developed. The primary focus of all this development was to reduce data entry and the double handling of all data applicable to histories and photographs. This is only significant once thousands and thousands of entries need to be types, basically by a single user.
Windows development started in March 2003 with a name change from Register to HTS (History Tracking Program). The name change was a reflection in use and to provide a similar name to FTS.
Anticipated development path will include a vehicle history output which can include a cross check on FTS indexed images if required. FTS data and images will be placed into the HTS output in chronological order and as links to image files.
Development has been halted due to the conversion of FTS to 32bit using xBase++. Once this project is done, attention will turn to HTS.
The following output is the history paragraph generated for a request for car 'M 107'. The search has dragged in the other four 'identities' the car was known as prior to the 'M 107' identity. Histories, references and photographs can be added to the individual car histories as required. Note: the information is not stored as the output shows. Data is encoded to allow the program to choose decision paths. The asterisk (*) on the left of 'M 107' informs the user the starting point for the request. This is useful when lists of class numbers are generated. The paragraph below is known as a 'Step History'.
Vehicle.............SV # Date...... Loc History............SV Notes
AA 85.VA - 12/10/1888 PIK Built new WH 88/9169
circa 1910 - To A 63.VC -
A 63.VC - 13/ 9/1913 - To AC 130.VA -
AC 130.VA - circa 1913 - Traffic Van fitted for passengers
31/ 1/1920 - To ACM 107.VA -
ACM 107.VA - 31/ 5/1924 - To M 107.VB (3/5/24)
*M 107.VB - 27/ 8/1969 - Out of Workshops Serviced
7/ 2/1974 - Off Register -
circa 1974 - On Register -
The output below shows the 'Local' history, so named as it is relevant only to the identity in question.
Vehicle.......... # Date...... Loc History.............. Notes................................................................
M 107.VB - 31/ 5/1924 - Ex ACM 107.VA -
- 27/ 8/1969 - Out of Workshops Serviced
- 7/ 2/1974 - Off Register -
- circa 1974 - On Register -
The program allows the user to add 'best available' data to the system for correction as more significant information becomes available. This compares with typed records where the mammoth task of editing does not bear thinking about and the scale of operations is reduced accordingly. In most cases presentation of material is 'withheld', ie not assembled, for fear that future editing work will prove excessive.
Email contact
pjv101@netspace.net.au