Earlier this year, I was informed (by Lotharek) that the U1MB/Rapidus compatibility problems (which have blighted many installations for over four years) have been confirmed to be caused by the Rapidus accelerator itself. The specific cause of said compatibility issue – though apparently observed via tech analytics to be Rapidus itself – remains unknown or undisclosed in any detail, but since it exists, and since the attitude of Rapidus developers still appears to be ‘Works here, so everything is fine’, ‘Your U1MB firmware must be to blame’, ‘the U1MB hardware must be to blame’, ‘Your opinion has no crediblity’, or any one of several other repetitive rebuttals, I have decided to withdraw all Rapidus support from the U1MB firmware and driver collection. This includes the ability to activate and deactivate the Rapidus core in software, and the ability (via the RAPIDISK.SYS driver) to double the HDD bandwidth on SIDE2/XEL-CF systems in the presence of the 20MHz 65C816 CPU.

When I first received Rapidus in 2016, I immediately found it puzzling that the issues I could observe on both stock (with SIDE2) and any randomly selected U1MB equipped machine were regarded as difficult or impossible not only to reproduce, but even to entertain. As a competent hardware installer, it was obvious to me that one statistical certainty is that any other European PAL Atari user with a Rapidus, U1MB and half a dozen Atari machines would be able to replicate the issue if an attention span of more than twenty minutes could be maintained. Nevertheless, my inability or unwillingness to ship a single randomly selected machine from my collection to Poland for ‘diagnosis’ was promoted as the single immovable obstruction in the way of actually finding a solution to the problem. Moreover, I was accused of undertaking a smear campaign against the hardware (for which I nevertheless developed software), constantly criticising the device (the installation of which I repeatedly and in great detail documented on video, presumably advertising the product in the process), and of not wanting the problem to be fixed (regardless of the hours of troubleshooting I undertook and stability fixes I tested).

In December 2020, the U1MB hardware or fimware apparently remains a prime suspect (in the eyes of Rapidus developers) as a cause of problems, regardless of the fact issues were observed with both generations of U1MB firmware, and even with or without U1MB present at all. Moreover, while going about my general business of answering operation queries from users of U1MB/Rapidus combos (since I provide software explicitly to support the interaction of both devices), I cannot do so without being harried, contradicted and fact checked by the Rapidus firmware developer (about whose software contributions to Rapidus I do not even have any complaints; indeed, they deserve nothing but praise). I have therefore chosen to entirely wash my hands of Rapidus, and this means withdrawing all U1MB plugins and drivers related to Rapidus operation. Only by doing this can I wholly avoid involvement in any future discussions of the Rapidus hardware, and thus avoid being drawn into tiresome quarrels or being obligated to provide technical support for problems which no-one can clearly identify or fix.

I realise the Rapidus plugin and HDD accelerator were probably of niche interest, so hopefully the deprecation of both will not cause anyone great inconvenience.

Meanwhile, I will no longer be installing Ultimate 1MB and Rapidus in the same machine, since I consider the combination to be problematic and therefore resistant to any level of quality assurance.