3 ms·

- 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

- 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.

miguelrochefort 8y ago
That's exactly the kind of answer I was looking for! I'm surprised by how well your ideas align with mine. I don't think I've seen anyone else talk about how the UI should use task management mechanics. I've been repeating for years that GTD is the future of UI. I'm also a strong advocate of dumb peripherals. Input and output devices should not be processing devices. I also like the idea of things like external GPUs through Thunderbolt. Regarding connectedness, modules/devices shouldn't have to know that they're inside the computer, outside the computer, or even in a remote cloud-like location. Edge computing should be seamless and transparent. It's very important that data be portable. I'm not sure privacy is as important as we think it is (we would gain a lot by letting the world know.more about ourselves), but the option should probably exist nonetheless. The big problem is to have my data duplicated and out of sync in a million different places. I'd rather update a single source of truth, than run around and try to update all different databases (e.g., if my address or phone number changes). I'm mainly a UI/UX developer, and I was wondering if there would be a way to build and bring some of these improvements (especially the task management system workflow) to users, without building a custom OS. Perhaps as an app? I'll now look at Microsoft Sets. Seems interesting.

miguelrochefort 8y ago

That's exactly the kind of answer I was looking for! I'm surprised by how well your ideas align with mine. I don't think I've seen anyone else talk about how the UI should use task management mechanics. I've been repeating for years that GTD is the future of UI. I'm also a strong advocate of dumb peripherals. Input and output devices should not be processing devices. I also like the idea of things like external GPUs through Thunderbolt. Regarding connectedness, modules/devices shouldn't have to know that they're inside the computer, outside the computer, or even in a remote cloud-like location. Edge computing should be seamless and transparent. It's very important that data be portable. I'm not sure privacy is as important as we think it is (we would gain a lot by letting the world know.more about ourselves), but the option should probably exist nonetheless. The big problem is to have my data duplicated and out of sync in a million different places. I'd rather update a single source of truth, than run around and try to update all different databases (e.g., if my address or phone number changes). I'm mainly a UI/UX developer, and I was wondering if there would be a way to build and bring some of these improvements (especially the task management system workflow) to users, without building a custom OS. Perhaps as an app? I'll now look at Microsoft Sets. Seems interesting.