Computing Assumptions That Collapsed: From an Appliance OS to Cursive Arabic Stroke Recognition

In the history of systems engineering, temporary constraints have a habit of hardening into sacred dogmas. When a computing layer is designed in an era where RAM is measured in kilobytes, writing system software requires a hundred-person team for three years, and recognizing handwriting or video demands a supercomputer, early engineers invent clever workarounds to survive those limits. Decades pass, the original hardware or cognitive bottleneck disappears completely, yet the entire industry keeps building on top of the old workaround—mistaking a historical compromise for a law of physics.

Inside our laboratory, we established a dedicated research track called Givens (المسلَّمات). Its sole mandate is to hunt down foundational computing assumptions whose historical premise has quietly expired, and rebuild those layers from scratch in days or weeks. Out of that laboratory came four unconventional infrastructure systems: the single-application appliance operating system (osapps), the adult cursive Arabic stroke-recognition engine (Raqeem / write-engine), the reactive geo-spatial datastore family (Locus and Motus), and browser-native edge vision platforms (Ishara and EyeChess).