And so another month has rolled by and we are approaching the winter months. This time of year, at least in civilised parts of Australia, is a good time for staying indoors and getting to work on that pet PC project of yours. I have developed a short wish list which really need developing and, for those having difficulty in deciding on a project here they are:
I have a copy of an expensive ($800) debtors ledger from one of the larger Australian accounting package writers, and, frankly, the standard of the user interface on this package is embarrassingly bad. The sad thing is that the badness can only be assessed by comparing it with other types of application software, because it is good for an accounting package. By far the best database system available is Oracle, which is the nearest thing to a relational database as measured by its conformity to the Codd rules. This really is the future as far as database systems are concerned and its features leave any of the competition dead. In fact, Oracle (the company) doesn't even appreciate the fact that it has competition in the PC environment. However, because of this attitude, we are missing out on truly significant
software because of a price which relates to the mainframe and minicomputer environment in which Oracle
evolved, and not to the hurly-burly of the PC world. At $2400 ex tax, you need to be a very serious user
indeed before bypassing such offerings such as that we currently have for Foxbase+. And so to my final wish. (I always like to leave the best to last). How many times have you thrown a manual through the video screen because the stupid, idiotic compiler you were using failed to see the wider picture of your program and dumbly flagged an error which was only of interest on a small scale. (ie. even if it stuffed everything in the meantime). Regardless of which language an inferential compiler is written in, and almost without regard to cost, I will buy it. My idea is that you would tell the compiler in general terms what you want it to do and then any code you write would be examined to determine its conformity with that goal and would be modified by the compiler before compilation. Even 12 months ago such a project was laughed at as an in-joke, but with the advances in expert systems and AI perhaps such a development will not be that far off. In my experience it is the ideas which are the hard part, and coding is just the laborious and frequently boring task of making the idea happen. For those of you who are not turned on by software considerations, and were born with a soldering iron in their hands, please note that we have a volunteer, Wayne Coutts, who is willing to start a Hardware SIG. See elsewhere for details of how and when if you are interested. Also, Ray Watson is starting the Database SIG this month. Please note that this SIG is primarily for power users of dBASE, rBASE, Dataflex, and higher level (ie 4th generation) programming tools. Judging from the number of corporate members, this SIG should be a valuable resource for all. Geelong members are urged to contact Alan Edwards at Alcoa's Port Henry works if you would like to belong to a Geelong Regional SIG. I have given details of those members with Geelong mailing addresses to Alan, but I am sure that there are others with Melbourne work addresses who live in Geelong. Reprinted from the May 1987 issue of PC Update, the magazine of Melbourne PC User Group, Australia |
