Gno Realm and Ownership Spec

[moul]

The issue is to follow up on the recently introduced document available here: https://docs.google.com/document/d/1eCgGPCJ8fMhNc_vc_pbgCxP10X7jGKLC89nBUGatsD4/edit

The ideal order to complete the GnoVM MVP (not only fixing bugs, etc.) is to at least finish and implement those breaking changes that should be shipped for the launch. These steps are ordered due to dependencies, and all of them are blocking the launch:

• Review and lock this document (the idea is to target a very high chance of not having to change anything, or just minor details)
• Once locked, open one or several "spec PRs" with all changes to structs and interfaces
• Review by Jae (at least) + anyone else who wants to
• Implementation

Additionally, this blocks things like pointer-based registries, see #2535, which is a pattern that deserves several rounds of experimentation before the launch, if possible. In other words, it's a high-priority blocking issue.

Assigning @thehowl and @ltzmaxwell to this, as discussed with Jae, so they can follow up on the topic and represent the EU and US teams. However, feel free to rely on the rest of the team, Jae, or me when needed.

The grantees and community are welcome to help, both on the Google Doc page or by proposing their spec PRs too; but for the sake of prioritization, the core team will directly work on it soon.

[moul]

Just added a new topic proposal.

[moul]

We recently discussed the possibility of not supporting local storage for remote realms. I understand this decision, as managing data would become more challenging and restrictive due to garbage collection, realm deprovisioning, or upgradeability. However, I appreciate this feature for its composability and the opportunities it offers. I'm currently brainstorming ways to adapt if we enable this additional inter-realm independence. The first experiment is underway in #3135.

Edit: #3135 is also adding several new gnovm/tests/files/... tests related to cross-realm functionality.