Imported ADRs that Damian shared via gdrive

Started using https://github.com/npryce/adr-tools

Author: Nicolas Luck

.adr-dir

@@ -0,0 +1 @@
+doc/architecture/decisions

doc/architecture/decisions/0001-record-architecture-decisions.md

@@ -0,0 +1,19 @@
+# 1. Record architecture decisions
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+We need to record the architectural decisions made on this project.
+
+## Decision
+
+We will use Architecture Decision Records, as described by Michael Nygard in this article: http://thinkrelevance.com/blog/2011/11/15/documenting-architecture-decisions
+
+## Consequences
+
+See Michael Nygard's article, linked above. For a lightweight ADR toolset, see Nat Pryce's _adr-tools_ at https://github.com/npryce/adr-tools.

doc/architecture/decisions/0002-rewrite-second-iteration-in-rust.md

@@ -0,0 +1,27 @@
+# 2. Rewrite second iteration in Rust
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+We want to have holochain run in the browser (for Holo).
+
+Rust can compile to WASM. Rust is language for experts. Rust is trending.
+
+There is a WASM interpretor in Rust (WASMI).
+
+## Decision
+
+Refactor Holochain in the Rust programming language
+
+## Consequences
+
+We have to recode all holochain in Rust which will delay the next version release.
+
+It is an opportunity to refactor the architecture.
+
+We must have clear coding practice on how we manage memory ownership in Rust.

doc/architecture/decisions/0003-redux-architecture-pattern.md

@@ -0,0 +1,19 @@
+# 3. Redux architecture pattern
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+Context here...
+
+## Decision
+
+The new holochain architecture will follow a redux architecture in order for an agent to have one global state.  This reduce the complexity of handling state management which is already complex for a distributed app.
+
+## Consequences
+
+Holochain refactor must fit this new model of having State objects and Actions objects.

doc/architecture/decisions/0004-have-only-one-wasm-ribosome.md

@@ -0,0 +1,20 @@
+# 4. Have only one WASM Ribosome
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+There is a WASM intepretor in Rust which can enable hApps in many different languages. WASM is more performant than any other interpretor.
+
+
+## Decision
+
+Implement only one WASM ribosome for holochain and have it working for 2 different toolchains (Javascript and Rust or C++)
+
+## Consequences
+
+Might do another ribosome non-wasm compatible languages like Closurescript, Python or Ruby

doc/architecture/decisions/0005-alpha1-is-last-major-go-release.md

@@ -0,0 +1,19 @@
+# 5. Alpha1 is last major Go release
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+A complete Rust rewrite is planned (see ADR #0)
+
+## Decision
+
+Alpha1 go release is last major go release of holochain because energy will be focused on the new Rust version.
+
+## Consequences
+
+They will still be minor releases for bug fixing.

doc/architecture/decisions/0006-splitting-agent-into-front-house-back-house-or-not.md

@@ -0,0 +1,19 @@
+# 6. Splitting agent into front-house / back-house - OR NOT
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+Holo wants to have front-end part of holochain run in the browser
+
+## Decision
+
+For Holo, seperating the back end networking part of the agent from the front-end UI & logic part of the agent.
+
+## Consequences
+
+Consequences here...

doc/architecture/decisions/0007-abstraction-for-network-layer.md

@@ -0,0 +1,22 @@
+# 7. Abstraction for network layer
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+libp2p has issues. DHT is too tied to network.
+What we want from a network library: Public Key cryptography: Hash public key that corresponds to identity on the network. Seemless port use.
+Handles routing tables?
+
+## Decision
+
+Have an abstraction layer for the networking library so we are not dependent on a specific one.
+Investigate Promether and see if it fits all our needs, including beeing network protocol agnostic (not just IP).
+
+## Consequences
+
+Consequences here...

doc/architecture/decisions/0008-mobile-first-development.md

@@ -0,0 +1,21 @@
+# 8. Mobile first development
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+Mobile is the main platform today. We want to reach as much users as possible.
+
+## Decision
+
+Support only 64bits platforms and not any 32bits platforms because we may have issues with cryptography on 32bits.
+
+## Consequences
+
+Have mobile testing in place and always test on mobile before releasing.
+Development has mobile platform performance considerations.
+Holo binding helps achieve this in a short-term way.

doc/architecture/decisions/0009-support-64bit-architectures-only.md

@@ -0,0 +1,19 @@
+# 9. Support 64bit architectures only
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+Context here...
+
+## Decision
+
+Decision here...
+
+## Consequences
+
+Consequences here...

doc/architecture/decisions/0010-write-our-own-kademlia-dht-implementation.md

@@ -0,0 +1,20 @@
+# 10. Write our own Kademlia DHT implementation
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+No mature one in Rust?
+DHT is crucial to holochain and holochain has specific DHT needs.
+
+## Decision
+
+Decision here...
+
+## Consequences
+
+Support and maintain forever our Kademlia code.

doc/architecture/decisions/0011-build-hcadmin-app-with-qt-c.md

@@ -0,0 +1,22 @@
+# 11. Build HCadmin app with Qt/C++
+
+Date: 2018-05-16
+
+## Status
+
+Accepted
+
+## Context
+
+HCadmin app needs:
+Crossplatform, including mobile. Systems tray. Handle processes
+HCD a 'hidden' thing.
+Keep HCDev and app more features.
+
+## Decision
+
+Use QT framework for building the crossplatform HcAdmin - HC app manager app for end users
+
+## Consequences
+
+Dependence on QT