Books

Current writing projects and published books in computer science.

Current book projects

  • Training the Delta

    Training the Delta is a practical guide to adapting existing AI models instead of training new ones from scratch. Its central idea is to teach only the smallest useful “delta,” whether that is a classifier, a few model layers, or a compact language-model adaptation. Examples progress through images, audio, text, time series, tables, graphs, and mobile or browser applications. The book distinguishes feature extraction, fine-tuning, LoRA and QLoRA, and retrieval-based approaches. It also emphasizes data quality, evaluation, and choosing the simplest method that solves the problem. All projects are intended to remain feasible on ordinary hardware, without large-scale pretraining or multi-GPU systems.

  • Building & Computing

    Building & Computing explores the boundary between physical construction and computation through simple mechanical models. It asks which elementary operations are needed to move, sort, connect, copy, and recycle material. Simulated machines provide links to Turing machines, tag systems, cellular automata, and von Neumann’s work on self-reproducing automata. The book studies how instructions, copying, construction, and self-replication can be expressed in mechanical form. It also considers broader questions of self-assembly, folding, timing, robustness, and scale. The result is intended as a conceptual bridge between computer science, biology, and engineering.

  • CPU Design

    CPU Design takes a bottom-up, hands-on approach to computer architecture using small, transparent machines built from standard logic components. It begins with input and output, registers, counters, memory, buses, and control logic before assembling complete 4-bit processors. Successive designs explore different memory architectures, addressing methods, stacks, subroutines, and lookup tables. The same design language is then extended to graphics and neural-processing units. A further section considers analog and hybrid forms of computation. Breadboard experiments and simulations keep the internal workings visible, with understanding rather than performance as the primary goal.

  • All is Computation

    All is Computation grows from a proposed interdisciplinary lecture series with Banaras Hindu University. It follows the foundations of computation from Gödel and Turing to von Neumann, cellular automata, and modern computer architecture. Formal languages and information theory broaden the account of what computation means. The discussion then turns to biological information processing, DNA, protein synthesis, and self-replication. Its final perspective connects computation, life, mind, and the debate around artificial general intelligence. Practical tutorials, models, and simulations are intended to accompany the conceptual and historical narrative.

Published books

  1. Variations on a Theme: JavaScript

    2023 · 220 pages · ISBN 978-1-7277-3314-3

    Web edition · PDF · Amazon

  2. Variations on a Theme: Physical Computing

    2020 · 123 pages · ISBN 979-8-6965-7777-7

    Web edition · PDF · Amazon

  3. Variations on a Theme: Java

    2018 · 162 pages · ISBN 1-7277-5750-5

    Amazon

  4. Variationen zum Thema: Android

    2018 · 320 pages · ISBN 1-7245-2613-8

    Web edition · PDF · Amazon

  5. Variationen zum Thema: Algorithmen

    2017 · 204 pages · ISBN 1-5454-6746-3

    Web edition · PDF · Amazon

  6. Variationen zum Thema: Internet

    2016 · 156 pages · ISBN 1-5377-6546-9

    Web edition · PDF · Amazon

  7. Variationen zum Thema: Java

    2016 · 164 pages · ISBN 1-5306-5184-0

    Web edition · PDF · Amazon

  8. Social Web

    2012 · Collaborative Wikibook

    Wikibook · PDF

  9. Game Creation with XNA

    2011 · Collaborative Wikibook

    Wikibook · Print version

  10. Introduction to Software Engineering

    2011 · Collaborative Wikibook

    Wikibook · PDF