- Ecosystem would be based around a task management 'backbone' - the OS would be Activity based, instead of function (app) centric. I.e. apps would be required to 'plug in' to the backbone, and UI would be centered around the current event being undertaken rather than an unlinked view-per-function (app UIs would be secondary interfaces, not primary). Microsoft Sets would have started to get us thinking along these lines.
- Browser tabs and apps need to be synonymous. Google pulled 'merged tabs and apps' from Chrome on Android a couple of years ago. This was arguably a significant step backward in UX.
- Crowd-sourced Peer review / code review system for apps. This community would have to be carefully curated in order to prevent a 'trigger happy / aggressive' team culture (such as Stack Exchange suffers from)
- Privacy-first: Privacy and data handled to allow for HIPAA/GDPR/ISO:27001 considerations right from the get-go (apps to use a shared annex system for data & data exchange that must be complied with). Blackberry corporate user based absorbed.
- iOS strictness of 'screen real estate zoning' system to reserve critical areas, homogenize menu bar UI etc.
- Docking / shared screens / OS merging has been poorly attempted in the last 10 years. Desktops should just be dumb monitors with docking pads by now. Screen real estate zoning would be catered for at different levels from 0.Xcm - Sports stadium boards.
- Modular OS: More compartmentalisation and fragmentation right from the offset. The setup of the OS should allow access to toggling of all modules and their variables. A user could deploy a full featured OS, or a dumb 'issue reporting screen' where they limit functionality & modules to just a single source of traffic through a port, basic OS functionality and graphical text representation. This would allow for more use cases being appropriately covered by the OS. Essentially, the modularity of the 80s/90s PC market and the user-friendliness of the iPad could have come together by now. PC components shouldn't have exposed pins etc. by now.
- Ideally, there would be a hardware team working closely, with similar culture. Project Ara would have been appropriate for a corporate culture... and if scope had been wider (swallowing the thin client market, amongst other areas). MSPs / IT departments would have thrived on modular computing - Efficiency of plug & play, cost savings of reusability and smaller module scope, speed of deployment to end users, security of not having unnecessary components active.
- Canary-ridden & transparent documentation, code and licenses. Use OS market share power to drive inter-OS adoption of encrypted, secure, open source protocols. Government or third party intervention has loud triggers attached to it.