WonderOS | Hacker News Reader

WonderOS

wonderos.org

20 pointsdloss13 comments

I didn't understand much from the homepage, but looking at linked Lab Notes[0] I see 2 major takeaways:

1. "Atomic apps", which I understand to be apps that separate presentation from the data. 100x this. We used to have it, with files and programs that could open them, but with move to cloud apps, this feature of computing is disappearing. How do I make a directory that contains stuff from my google drive, onedrive, figma and youtube? Interestingly, there are now SaaSes that let you "index all your company resources", which try to tackle this problem.

2. "Graph OS". The author has really impressive Obsidian graph[1]. However, I don't believe this is a good approach for all apps, or for all people. I tried many times to make my notes reference each other, with backlinks and stuff, and I now believe that my brain is not wired for this stuff. Still, kudos for the author for making it work for them.

[0] https://alexanderobenauer.com/labnotes/000/

[1] https://alexanderobenauer.com/labnotes/014/

Windows and Mac did a huge disservice by making people think that a GUI (or a desktop) is part of an OS.

This looks like a nice way to think at common office tasks, but it doesn't need to be an OS - it's just a set of UI conventions participating programs would adopt along with the APIs to help those communicate.

I called this idea "data-centric computing" in the 90's, though that has a different meaning, about memory organization in game development, these days.

Basically data records live free from applications, so for example you can make a pasteboard that contains an image, an email, a couple of contact addresses, a map, or whatever, without having to go into an image editor, an email app, etc to get the data, or to have to link/embed those apps as in microsoft's 90's OLE/COM object embedding stuff.

It differs from 'everything is a file', in that files contain a whole bundle of different data, and are 'opened' by a particular application or class of application, and are effectively a 'save state' of the application, whereas 'items' or data records or whatever are independent of any particular app; they have a type, and any app/tool that can deal with that type can operate on them.

Sounds interesting, but after reading the entire page I’m still not sure I understand it.

I just read several articles and I still have no idea what, exactly, is he trying to achieve here.

It's supposed to be an OS but I think it's mostly about UI.

I read the whole thing, but I know just as much as when I started. This is the least amount of substance I've seen in a while.

"everything is an item" I looked into the further lab notes and I'm confused about how this is different from "everything is a file" in Linux.

After skimming https://alexanderobenauer.com/articles/os/1/ - I think the items are more like objects than files. Files have a uniform I/O interface, while items seem like they can have unique properties instead of having a uniform interface.

Based upon the description in the operator handbook, this reminds me of Microsoft’s Courier notebook project.

It's not actual software it's just a design right?

The author has a “lab” Patreon where he works out more of the specifics and also builds proofs of concept for the different functions he describes.

The prototypes he built so far are very impressive and he’s been using them as part of his workflow.

Why marketing before product?

I think his product is exploration/research/experimentation and it's funded by his Patreon. It's not the case that he's obviously working towards a saleable OS.

Reply on news.ycombinator.com