Atari XL manual Development and Testing, WHY LW Came Into Being

The Last Word 3.0 Reference Manual

as critical a factor as it has always been with LW. After many years trying to get the Atari Macro Assembler to co-operate with SpartaDOS X, I finally wrote my own Macro Assembler. The result - MA65 – was used in emulation until quite recently to compile LW. It was only when the superb WUDSN IDE cross-development platform for Eclipse was published that I finally ported LW’s 20,000 lines of source code to the PC, where I continued to develop the software using the superb WUDSN plug-in for the Eclipse platform, along with the ATASM cross-assembler. It now only takes a second to compile the whole program. Nevertheless, MA65 remains my disk-to-disk assembler of choice on the Atari platform.

12.4 DEVELOPMENT AND TESTING

After nine years of thinking about writing a word processor, I finally began LW at the start of 1999. It took only 3 months to get a fully working version, and a further 2 months to produce one which was reliable. It was this test period, during which I used LW to write utility macros and the bulk of this documentation, that highlighted bugs and design faults for later correction. During this time probably hundreds of complete re- compilations were done. The program culminated in version 2.1 in 2000.

After version 2.1 was published (to an audience of none!) in 2000, there was a hiatus of eight years while my Atari lay in a storage box on top of a wardrobe. Then, in November 2008, a nostalgic conversation with a friend who was also interested in programming led me to retrieve the Atari. Amazed to find it still worked, I soon set about porting my files over to the PC using the fantastic new gadgets that had become available in the intervening time, and by the turn of the year LW 2.1 finally found publication. It drew such interest among Atarians that it became obvious the program would have to be updated to take advantage of all the new hardware that had been developed. 320K was now virtually the minimum amount of memory fitted to an Atari XE, while 1MB was common. Disk storage was almost unlimited thanks to SIO2IDE, SIO2SD and SIO2PC, and emulation meant the program development time could be dramatically lessened. The number one feature people wanted to see in a new word processor for the Atari8 was a fast 80 column display which didn’t use any hardware add-ons. The 80 column display only took a few weeks to write. Then I began to see other things I could improve...

12.5 WHY LW CAME INTO BEING

As well as being popular gaming machines, the Atari 8-bit line of computers have a huge catalogue of “serious” application titles, including many word processors. The most popular commercial offerings back in the Eighties included AtariWriter (and later, AtariWriter Plus), Paperclip, The First XLEnt Word Processor, and Superscript, although there were also dozens of other, less “heavyweight” word processors. Many public domain and open source text editors and word processors were also written for the Atari 8, including Speedscript and TextPro, the latter going through many incarnations and being one of the few word processors to offer tight integration with SpartaDOS, the advanced disk operating system by ICD. AtariWriter 80 and TurboWord used 80 column displays (most Atari 8 word processors were limited to the Atari’s 40 column screen), although these programs required special hardware in the form of the XEP-80 device.

I got my first Atari 8 – a 65XE – in the late Eighties while still at school and as soon as I got a disk drive I began using TextPro for word processing. At the same time, I started to learn to program: first using Atari BASIC, then Turbo BASIC XL, then C and finally Assembly Language. Having written a database, a drawing program, various

12-73