Hadn't looked at this thread for a while but got a reaction bump so checking back to see how things played out...
OCAT might hav value to new hacintoshers helping them gain a sense of the lay of the land of config structure, but no one seems to be treating it as such, maybe because typical OC users just want to get their specific kit to work, not become build experts.
The Sonoma release got me looking at OCLP again simply because the OCLP post-install for Broadcom quickly became the goto solution to wifi compatibility.
I was grateful to find that Dosdude1's great work years ago on macOS-on- Unspported patchers has taken on a new life under Dortania with OCLP.
I used OCLP to put Ventura on an old Macbook 9,1, and the experience went great: starting with excellent documentation at the right level of abstraction, it makes the process of building and installing easy and effective.
The OCLP approach has coalesced into a patcher app that does exactly what most users need:
- Gather and assemble a bootable macOS installer for a given build.
- Slipstream key support into the installation process (a traditional Windows admin approach).
- Patch the installation via a post-install step to deal with contingencies.
- Run an updater in the background to track improvements to patches.
This is an example of "recipes" that I was trying to explain in previous posts reviewing OCAT: OCLP exposes configuration options appropriate to the build task at hand while hiding as much underlying configuration detail as possible.
The olde dayz of the Beasts were a valiant stab at this, but various factors worked against it in the long run:
- Pre-Github. While I lament the business-model of Github, the site is super smart and it's taken the world by storm. And not only that but Dortania is an excellent use of Github.
- Uni/multi nomenclature never made sense to the uninitiated, and maybe never made sense to anyone.
- Hid too much detail.
- Config evolution got too complex for tool owners to keep up.
- Clover was not the future.
Well, nothing went wrong, it's just that history is always messy. But as progress grinds on, talented hard-working people continue to make progress. In case of OCLP, it so happened that just keeping actual macs working has substantial overlap with hackintosh. And the focused objective of unsupported Macs has helpfully guided the course of development.
Now, could OCLP be taken to the verge of hack config?
Imagine if OCLP was template driven: it might accept a characterization of a golden build and emit bootable drive for its parts list, reusing the essentially useful design of the OCLP patcher?
Then instead of lore being passed around on threads with Q/As being endlessly repeated in the forums and mods obsessing with categories for discussion (although paradoxically it's better here than at macrumors "unsupported" where all discussion is dumped into a single thread and everything is forgotten with each macOS release), the lore would get baked into the tool and the templates. This will free up forum energy to expand the catalog of golden builds and review experiences.
This is just rambling at the moment.