Amos Programming

[Yury Morris]

AMOSBASIC is a dialect of the BASIC programming language for the Amiga computer. Following on from the successful STOS BASIC for the Atari ST, AMOS BASIC was written for the Amiga by Franois Lionet with Constantin Sotiropoulos and published by Europress Software in 1990.

The language was notable for its focus on media and game development capabilities, allowing users to easily create demanding multimedia software and games. It featured full structured code and numerous high-level functions for loading and manipulating images, animations, and sounds. These capabilities made it a popular choice among Amiga enthusiasts, particularly beginners, for creating video games (especially platformers and graphical adventures), multimedia applications, and educational software.

AMOS competed on the Amiga platform with Acid Software's Blitz BASIC. Both BASICs differed from other dialects on different platforms, in that they allowed the easy creation of fairly demanding multimedia software, with full structured code and many high-level functions to load images, animations, sounds and display them in various ways.

The original AMOS was a BASIC interpreter which, whilst working fine, suffered the same disadvantages of any language being run interpretively. By all accounts, AMOS was extremely fast among interpreted languages, being speedy enough that an extension called AMOS 3D could produce playable 3D games even on plain 7 MHz 68000 Amigas. Later, an AMOS compiler was developed that further increased speed. AMOS could also run MC68000 machine code, loaded into a program's memory banks.[1]

To simplify animation of sprites, AMOS included the AMOS Animation Language (AMAL), a compiled sprite scripting language which runs independently of the main AMOS BASIC program.[2] It was also possible to control screen and "rainbow" effects using AMAL scripts. AMAL scripts in effect created CopperLists, small routines executed by the Amiga's Agnus chip.

After the original version of AMOS, Europress released a compiler (AMOS Compiler), and two other versions of the language: Easy AMOS, a simpler version for beginners, and AMOS Professional, a more advanced version with added features, such as a better integrated development environment, ARexx support, a new user interface API and new flow control constructs. Neither of these new versions was significantly more popular than the original AMOS.[citation needed]

One of AMOS BASIC's disadvantages, stemming from its Atari ST lineage, was its incompatibility with the Amiga's operating system functions and interfaces. Instead, AMOS BASIC controlled the computer directly, which caused programs written in it to have a non-standard user interface, and also caused compatibility problems with newer versions of hardware.

Today, the language has declined in popularity along with the Amiga computer for which it was written. Despite this, a small community of enthusiasts are still using it. The source code to AMOS was released around 2001 under a BSD style license by Clickteam, a company that includes the original programmer.[3]

This article by Dan Polansky briefly looks at AMOS programming language for the Amiga line of computers (Motorola 68k-based computers strong on graphics and sound for games at their heyday), which is a structured and procedural dialect of BASIC with certain batteries included, especially those suitable for programming of simple games. The article intends to go beyond the Wikipedia article in providing a better idea about the language and supporting anyone curious enough to want to run AMOS in an Amiga emulator for Windows and play around with it.

Given the kind of facilities AMOS has, there seems to be no obvious reason why compiled AMOS could not be as fast as the C language. An empirical performance comparison is missing. It seems unlikely that compiled AMOS was as fast as C, but it is unclear what the case actually is and why that is. AMOS compiler itself was not written in C but in assembly[4].

One area where AMOS is expected to get slow is strings and their operations such as concatenation. I (Dan Polansky) remember writing a text format conversion tool in AMOS; I started using the built-in string facilities, but it was very slow and I could only make it perform well by switching to low-level facilities directly working with bytes and their sequences.

What is it? Amos (Analysis of Moment Structures) analyzes Structural Equation Models (SEM) by either drawing a path diagram or programming a command language. These models are also known as LISREL models, confirmatory factor analysis, analysis of covariance structures, path analysis, or causal modeling.

What can I use it for? Enrolled students, faculty, and staff can use Amos for university-related teaching and academic research purposes. Amos may not be used for commercial research or administrative purposes. Use that benefits outside organizations requires the purchase of a commercial license.

Where can I get support? OIT offers Full Support to University of Tennessee, Knoxville-area students, faculty, and staff for this software. For assistance, contact the OIT HelpDesk at 865-974-9900.

Jason has been programming a wide range of computers for considerably longer than he has been a journalist, during which time he has worked with just about every programming language - including assembler. C, Cobol and Pascal - athough AMOS remains his favourite language.

Like the toolbox, a programming language does nothing more than provide you with a set of tools in the form of general purpose commands - how you apply these tools is up to you. Because the commands offered by the language have been designed to be as general purpose as possible, a programming language places absolutely no restriction on the type of programs that you can write - all that is required of you is that are able to shape your ideas (essentially the wood that we used to build our chair!) into working programs using the commands provided.

The release of AMOS, way back in 1990, spawned a whole range of support products, not just from Europress, but from third party developers too. Compilers, alternative sprite editors, sound manipulation utilities and even a complete 3D graphics extension are all now available to the AMOS programmer, making AMOS the most flexible language available for the Amiga. Even AMOS itself is no longer a single product - in an attempt to make AMOS of interest to all Amiga users, AMOS is now available in three different flavours, each aimed at a particular user.

What makes Easy AMOS such an attractive proposition to beginners is its excellent manual which was written by computer industry veteran Mel Croucher, he of AutoMata PieMan fame (oops, I showed my age a bit there). Written more as a step-by-step walk-through of the Easy AMOS editor and language than a traditional reference manual, the Easy AMOS manual is ideally suited to beginners as it not only documents Easy AMOS, but also teaches the fundamentals of computer programming.

The enhancements offered by AMOS Professional go so much further than just the editor, though - it also offers a plethora of new commands (over 200 in total) that extend AMOS to over 700 commands! These include commands that allow AMOS programs to communicate with ARexx ports (and therefore other programs), load and play animations in standard IFF ANIM format (AMOS can actually run animations faster than Deluxe Paint!), perform double precision floating point maths (handy if you own an Amiga equipped with a maths co-processor chip), play Sound Tracker and MED modules and so much more besides.

In order to distribute AMOS programs to other users who may not already own the AMOS Interpreter, Europress have produced the AMOS Compiler, which can take any AMOS source program and convert it into stand-alone machine code. Like AMOS 1.35, this too has found its way onto magazine coverdisks, so keep your eyes peeled for back issues.

Our mission trips, service trips, and individual volunteer opportunities allow you to work alongside AMOS staff and community members to advance long-term goals identified by the communities we serve. All individuals who have a desire to serve and a willingness to learn are encouraged to come!

AMOS offers a variety of mission trip opportunities in Nicaragua for groups of all ages. Your service and financial support will enable us to implement life-changing projects for people in need and will advance our long-term work in the communities we serve.

AMOS offers a variety of service opportunities in Nicaragua for students of all disciplines. You will collaborate with us to implement projects that advance both our organizational goals and your professional goals. Come learn about community-based health care and gain experience in it firsthand. Contact us at deleg...@amoshealth.org to get started.

While recently Large Language Models (LLMs) have achieved remarkable successes, they are vulnerable to certain jailbreaking attacks that lead to generation of inappropriate or harmful content. Manual red-teaming requires finding adversarial prompts that cause such jailbreaking, e.g. by appending a suffix to a given instruction, which is inefficient and time-consuming.On the other hand, automatic adversarial prompt generation often leads to semantically meaningless attacks that can easily be detected by perplexity-based filters, may require gradient information from the target LLM, or do not scale well due to time-consuming discrete optimization processes over the token space. In this paper, we present a novel method that uses another LLM, called the AdvPrompter, to generate human-readable adversarial prompts in seconds, approximately 800 times faster than existing optimization-based approaches.We train the AdvPrompter using a novel algorithm that does not require access to the gradients of the target LLM. This process alternates between two steps: (1) generating high-quality target adversarial suffixes by optimizing the AdvPrompter predictions, and (2) low-rank fine-tuning of the AdvPrompter with the generated adversarial suffixes. The trained AdvPrompter generates suffixes that veil the input instruction without changing its meaning, such that the target LLM is lured to give a harmful response. Experimental results on popular open source target LLMs show state-of-the-art results on the AdvBench dataset, that also transfer to closed-source black-box LLM APIs. Further, we demonstrate that by fine-tuning on a synthetic dataset generated by Advprompter, LLMs can be made more robust against jailbreaking attacks while maintaining performance, i.e. high MMLU scores.

3a8082e126