The following article was written by Thierry Peycru, describing some of the background behind this new storage device for the Z88.
In 1994, I read an electronic newspaper about the new Flash Eprom technology. Immediately I thought of adapting it for the Z88. The first prototype was a card with an Intel 28F020, a 256K chip with the ETOX II technology. The PCB was single-sided. It worked well and I built two others and wrote some code inserted into BBC BASIC programs.
It was revolutionary for the Z88. Firstly, the file saving was very quick, the LCD was still on and its power drain was less than a RAM.
In 1996, I decided to make a 1M FlashCard using the Intel ETOX III technology of the 28F008SA chip. We designed the PCB layout with a friend, Xavier Thomas. It was double-sided with metalized holes and costs us much time and money. After resolving a lot of problems, it took 4 hours to built one card without any trouble... Four cards were done. It was difficult to consider mass production. We encountered a lot of problems with the slot connector which must be very precise.
So I gave my work to Vic Gerhardi who decide to produce the cards in a professional way. The layout was redesigned by Tony Firshman with professional software to a single-sided PCB with CMS capacitors and the latest Flash chip, the 28F008S5. It is much faster.
I developed the software as a machine code application run from BBC BASIC. A lot of bugs have now been fixed. The latest versions have been completely improved by Gunther Strube (Interlogic). This software is free. FlashStore (1.5.6) is now fast and very stable.
The next step will be an application directly run from the FlashCard itself. As we are developing it with Gunther it will probably be associated with his wonderful EazyLink. This version will be sold.
Some new commands are in development. A great interface including a window with filenames and a horizontal inverted bar for selecting files to be fetched will be supported (like in EP-Fetch).
A defragmentation command for packing files and erasing the deleted ones without re-formatting the whole card is planned.
We are also thinking about a LZW or LZSS compression algorithm. It could work in the background and would be fully transparent to the end user. The card could provide in that way more than 2Mb of capacity.
At the moment we are actually using the standard OZ EPROM file format. A new filing system will have to be defined in order to use the FlashCard as a hard-disk...
When disassembling the OZ kernel, I have found a secret jump (certified later by Paul Bond). It is possible to put a new OZ in slot 1 and to run it directly without replacing the internal ROM! Imagine an unbugged international OZ with 1M of applications in slot 1. All the improvments for OZ v5 are on paper...
All of these will depend on the interest of other people. Programmers are welcome...