About TurboBBS - by Bob Maxwell

     Those active on the local BBS scene are probably aware of 
the Turbo BBS at 738-7811. Those who use it may even know that 
the source code for the program was written by me in Turbo 
Pascal, and the program is offered as "shareware". It has 
received remarkably wide distribution: I have heard from users 
all over the U.S., as well as Australia, Britain, Finland and 
Saudi Arabia.

     Why has it been that popular? It isn't all that great as 
bulletin boards programs go... it lacks a number of features, and 
really doesn't do anything unique. The key to its success is most 
likely in its nature, rather than its content: it is available 
(cheap) in a high-level source code that is easily adapted for 
different uses. Furthermore, the compiler it requires is quite 
popular (also cheap), and produces tight, fast object code.

     The original program was not written to be distributed. It 
was simply my first (and so far, only) major project using the 
Turbo Pascal compiler. I wrote the program because (1) I wanted 
to start a BBS, and (2) the compiler seemed to be well adapted to 
the task. The program was written in two weeks of evenings and 
weekends, with file transfers finally being debugged about a 
month later. It was written for a Kaypro 2 (Z80 based CP/M 
machine), using two 191K floppy drives.

     In publicizing the new BBS, I announced it on the Borland 
International (publishers of Turbo Pascal) SIG on the massive 
CompuServe system. I was requested by Borland representatives to 
release the BBS source code: they had implied to a magazine 
article writer that they had a Pascal BBS program available on 
the SIG - and they didn't. After some consideration, I released 
my code - Version 1.00 - on CompuServe and my own BBS.

     Version 1.00 had a serious flaw: it was specific to the 
hardware of the Kaypro, and the code accessing said hardware was 
well scattered within the program. This made modification 
difficult. The hardware calls only dealt with the modem port and 
real-time clock (if there was one), and I endeavoured to separate 
these routines into a special file. Once that was done, not only 
was the program portable to other CP/M systems, but also to IBM! 
It was MS-DOS capability that helped its popularity immensely.

     The second version released of the BBS was 1.05... a vast 
improvement over its predecessor, but still well-stocked in 
shortcomings. It is still the current release, and has been in 
circulation for well over a year. It seems that further 
additions of features detract from the modularity of the program, 
and there are many problems deeply locked in the heart of the 
code - attempting to remove them would create a bug farm of 
massive proportions.

     I have learned a great deal from TurboBBS version 1, and am 
dreaming up version 2... This is a slow process, since I have all 
sorts of ideas for it. What makes it slower is the fact that I 
have already written a BBS, so the sense of adventure that drove 
the original's development is missing. In addition, there is 
little economic drive, since the contributions received have not 
been overwhelming - BBS programs are not quite as popular as word 
processors or spreadsheets. Fortunately, I had no dreams of 
getting rich from this endeavour!

     What I value most from writing and distributing TurboBBS is 
the people I have met because of the program and the BBS I run 
using it. People look at the expense of time and money in writing 
and running a bulletin board, and say I'm crazy. They're probably 
right, but you gain only by giving of yourself.