EEP – Early Experience Program
An early-access program from Alaska Software – a kind of beta – in which selected enterprise customers can try out and use the new Xbase++ V3 code base in their own, real projects before the official release.
AAI – Agentic AI
A new framework within Xbase++ for embedding AI-supported assistants (“agents”) that act independently into business applications. The agents take on sub-tasks such as reading, classifying or drafting content, while a fixed control flow handles the actual execution.
AI / KI – Artificial Intelligence (German: Künstliche Intelligenz)
Artificial Intelligence (AI) is the branch of computer science that automates intelligent behaviour — building systems that reproduce human abilities such as learning, reasoning, problem-solving, language comprehension, and pattern recognition. Large Language Models (LLMs) are only narrow AI: because their parameters are fixed after training, they do not learn autonomously beyond a single conversation — a step toward general AI, but not yet its realisation.
LLM – Large Language Model
A language model trained on vast amounts of text (e.g. Claude or GPT) that can read and classify text, extract key details and draft new text. It does not always produce the same result for the same input and is therefore not predictable. At its core, an LLM is a probabilistic next-token predictor: it generates text by modelling statistical structure learned from data, without persistent memory or grounded understanding. Whether the internal representations this produces amount to more than pattern-matching remains an open research question.
GPT – Generative Pre-trained Transformer
A design and model series of language models from the provider OpenAI. Mentioned as one example of a language model that can be connected via Agentic AI.
FSM – Finite State Machine
Control logic that precisely defines which state a process is currently in and under which conditions it may move to the next state. Within Agentic AI, the FSM handles the control flow of the agent steps so that the behaviour remains traceable, verifiable and repeatable despite the unpredictable language model.
LOB – Line of Business
A term for business applications that map an organisation’s central, domain-specific processes, for example in finance, approval workflows or order processing. Such applications are the main area of use for Agentic AI.
MSA – Microservices Architecture
An architectural approach in which a large application is split into several small building blocks (services) that are developed and operated independently of one another. Xbase++ V3 supports this approach with its own feature set.
REST / RESTful – Representational State Transfer
A widely used architectural style for designing programming interfaces used for communication between programs over the web.
DSGVO – General Data Protection Regulation (German: Datenschutz-Grundverordnung, GDPR)
The European law for the protection of personal data. Mentioned in the newsletter as a possible framework to take into account when designing Agentic AI use cases.
V3 – Version 3
The third main version of Xbase++, currently in development, which introduces support for microservices and Agentic AI, among other things.
Part 2 – Additional architecture and development terms
API Gateway – Application Programming Interface Gateway
A central, upstream entry point that receives requests from outside and forwards them to the responsible internal service (microservice). An API gateway bundles cross-cutting tasks in one place that each service would otherwise have to solve individually, such as routing, authentication, rate limiting, connection encryption and logging. This decouples the calling programs from the internal division of the services. In the context of a REST/microservice architecture, the API gateway forms the outer interface in front of the individual services.
Microservice Mesh – Service Mesh
An infrastructure layer that governs communication between the microservices themselves. It handles tasks such as locating services (service discovery), load balancing, automatic retries, encryption of internal connections, and the collection of operational and measurement data (observability). The common technical term is “service mesh”; “microservice mesh” refers to the same thing. Distinction: while the API gateway controls traffic from outside to inside, the mesh concerns traffic between the services. In Xbase++ V3 your monolith can become a Mesh.
Code Scaffolding
The automatic generation of a basic source-code structure (a “skeleton”) from templates or a description. From a definition – such as an interface or data description – recurring building blocks such as service stubs, endpoints or data classes are generated in advance and then filled with the actual business logic. Scaffolding speeds up the project start and ensures a consistent structure; it produces starting code, but not a finished application. In Xbase++ V3 Microservices or unit-test scaffolded for example.
Asset Management / Asset Manager – management of reusable building blocks (“assets”)
A method by which functional building blocks in Xbase++ are no longer wired up by hand via individually linked static libraries (.lib), dynamic libraries (.dll) and include files, but are instead managed as self-contained, versioned “assets”. The asset manager integrated into the Workbench and available on the command line as xppcli adds an asset to a project target; in doing so, a copy of the respective asset version is placed into the project, the dependency is recorded in the project file (.xpj), and the building block is shown under the target’s dependencies. The project copy is not overwritten during updates. An asset can contain source code, DLLs, sample code and a readme description (examples: XbpHtmlViewer2, QR-code or CommandButton control); in the file system, assets are stored in dedicated .assets directories.
An early-access program from Alaska Software – a kind of beta – in which selected enterprise customers can try out and use the new Xbase++ V3 code base in their own, real projects before the official release.
AAI – Agentic AI
A new framework within Xbase++ for embedding AI-supported assistants (“agents”) that act independently into business applications. The agents take on sub-tasks such as reading, classifying or drafting content, while a fixed control flow handles the actual execution.
AI / KI – Artificial Intelligence (German: Künstliche Intelligenz)
Artificial Intelligence (AI) is the branch of computer science that automates intelligent behaviour — building systems that reproduce human abilities such as learning, reasoning, problem-solving, language comprehension, and pattern recognition. Large Language Models (LLMs) are only narrow AI: because their parameters are fixed after training, they do not learn autonomously beyond a single conversation — a step toward general AI, but not yet its realisation.
LLM – Large Language Model
A language model trained on vast amounts of text (e.g. Claude or GPT) that can read and classify text, extract key details and draft new text. It does not always produce the same result for the same input and is therefore not predictable. At its core, an LLM is a probabilistic next-token predictor: it generates text by modelling statistical structure learned from data, without persistent memory or grounded understanding. Whether the internal representations this produces amount to more than pattern-matching remains an open research question.
GPT – Generative Pre-trained Transformer
A design and model series of language models from the provider OpenAI. Mentioned as one example of a language model that can be connected via Agentic AI.
FSM – Finite State Machine
Control logic that precisely defines which state a process is currently in and under which conditions it may move to the next state. Within Agentic AI, the FSM handles the control flow of the agent steps so that the behaviour remains traceable, verifiable and repeatable despite the unpredictable language model.
LOB – Line of Business
A term for business applications that map an organisation’s central, domain-specific processes, for example in finance, approval workflows or order processing. Such applications are the main area of use for Agentic AI.
MSA – Microservices Architecture
An architectural approach in which a large application is split into several small building blocks (services) that are developed and operated independently of one another. Xbase++ V3 supports this approach with its own feature set.
REST / RESTful – Representational State Transfer
A widely used architectural style for designing programming interfaces used for communication between programs over the web.
DSGVO – General Data Protection Regulation (German: Datenschutz-Grundverordnung, GDPR)
The European law for the protection of personal data. Mentioned in the newsletter as a possible framework to take into account when designing Agentic AI use cases.
V3 – Version 3
The third main version of Xbase++, currently in development, which introduces support for microservices and Agentic AI, among other things.
Part 2 – Additional architecture and development terms
API Gateway – Application Programming Interface Gateway
A central, upstream entry point that receives requests from outside and forwards them to the responsible internal service (microservice). An API gateway bundles cross-cutting tasks in one place that each service would otherwise have to solve individually, such as routing, authentication, rate limiting, connection encryption and logging. This decouples the calling programs from the internal division of the services. In the context of a REST/microservice architecture, the API gateway forms the outer interface in front of the individual services.
Microservice Mesh – Service Mesh
An infrastructure layer that governs communication between the microservices themselves. It handles tasks such as locating services (service discovery), load balancing, automatic retries, encryption of internal connections, and the collection of operational and measurement data (observability). The common technical term is “service mesh”; “microservice mesh” refers to the same thing. Distinction: while the API gateway controls traffic from outside to inside, the mesh concerns traffic between the services. In Xbase++ V3 your monolith can become a Mesh.
Code Scaffolding
The automatic generation of a basic source-code structure (a “skeleton”) from templates or a description. From a definition – such as an interface or data description – recurring building blocks such as service stubs, endpoints or data classes are generated in advance and then filled with the actual business logic. Scaffolding speeds up the project start and ensures a consistent structure; it produces starting code, but not a finished application. In Xbase++ V3 Microservices or unit-test scaffolded for example.
Asset Management / Asset Manager – management of reusable building blocks (“assets”)
A method by which functional building blocks in Xbase++ are no longer wired up by hand via individually linked static libraries (.lib), dynamic libraries (.dll) and include files, but are instead managed as self-contained, versioned “assets”. The asset manager integrated into the Workbench and available on the command line as xppcli adds an asset to a project target; in doing so, a copy of the respective asset version is placed into the project, the dependency is recorded in the project file (.xpj), and the building block is shown under the target’s dependencies. The project copy is not overwritten during updates. An asset can contain source code, DLLs, sample code and a readme description (examples: XbpHtmlViewer2, QR-code or CommandButton control); in the file system, assets are stored in dedicated .assets directories.
