docs: add immediate division and prepared offspring mutation routes
Foundation / checks (checks) (push) Successful in 5s
Foundation / checks (client) (push) Successful in 4s
Foundation / checks (ecs) (push) Successful in 5s
Foundation / checks (network) (push) Successful in 4s
Foundation / checks (postgres) (push) Successful in 5s
Foundation / checks (tsan) (push) Successful in 4s
Foundation / checks (deploy) (push) Successful in 4m22s
Foundation / postgres-tools (push) Successful in 5s
Foundation / server-image (push) Successful in 5s
Foundation / image (push) Successful in 4s

This commit is contained in:
dingo
2026-10-21 11:32:22 -04:00
parent 0bbe650265
commit 9adaf6515b
3 changed files with 143 additions and 1 deletions
+125
View File
@@ -70,6 +70,125 @@ growth need separate expression classifications. Birth-time succession remains
one-way; nothing here grants possession of independent relatives or rescue after
death. New autonomous NPC mutations are outside this draft's authorization.
## Two mutation routes: evolve now or prepare offspring
Discussion addition, 2026-10-21: offer two ways to commit an earned adaptation.
The current-cell route gives immediate experience at the cost of dividing now;
the offspring route lets the player observe a genetic direction before committing
their controlled organism to it. This section records design intent and a proposed
resolution through offspring succession, not implemented reproduction rules.
| | Evolve now (current-cell experience) | Prepare offspring |
| --- | --- | --- |
| Trigger | Select the adaptation and perform a viable funded division now. | Select a heritable plan, then choose when to reproduce. |
| Player control | Continue immediately as the newly developed mutated offspring. | Keep playing the unchanged parent; a newborn can become the controlled organism at a later birth. |
| Main benefit | Immediate experience of the adaptation, with another organism added to the habitat. | Observe autonomous offspring and evaluate whether that genetic direction suits the habitat. |
| Main drawback | Commit to a smaller organism with fewer reserves and the adaptation's consequences immediately. | No immediate experience of the changed body; gather, reproduce, and evaluate before transferring control. |
| Body and reserves | Start from roughly half the dividing organism's mass and resources, subject to viable allocation and actual development costs. | Transfer the funded offspring investment when reproduction is initiated; retain the parent's remaining body and reserves. |
| Changing direction | Earn another suitable adaptation to shed, replace, or counter the trait; fund another viable development event. | Revise the future-offspring plan through subsequent earned choices; existing offspring retain their actual inherited forms. |
### Immediate division and identity
For the current-cell experience, the player continues as the **mutated half**;
the unchanged half becomes independent wildlife, or remains attached where supported.
"Unchanged" means retaining the established traits, not retaining full original
mass and reserves after those have been divided.
For structural changes, interpret this as **division plus immediate one-way
offspring succession**, rather than rebuilding the existing adult. The unchanged
organism retains the original permanent identity and expressed developmental basis;
the mutated offspring has a new identity and its actual inherited plan. The
relinquished organism keeps its established behavior and selected reproductive
plan, separately from its unchanged expressed traits, as required by ADR-0163.
No control can return to it after transfer or after the controlled offspring dies.
The roughly half-and-half allocation is a starting design target, not a rule that
every nutrient must be split blindly when that makes either body invalid. Specify
actual tissue partitioning, viable size, reserves, placement, and development costs
before implementation. This needs an explicit division contract; existing budding
semantics must not be silently reinterpreted as division.
The resources the player gives up remain in the other organism. They are not a
mutation fee or destroyed stock. Development work can consume E into heat, and
material changes must account for every channel. Two full-sized, fully stocked
cells cannot emerge from one without additional funded growth.
An adaptation that cannot currently fund a viable division may be prepared for
later; **Evolve now** must not create missing tissue or force an invalid split.
Show why it is unavailable, without hiding the adaptation from the planning route.
Immediate commitment is consequential: disliking a trait does not grant a free
undo, guaranteed shed offer, or free reconstruction. The player must survive,
gather what the replacement body needs, and earn a useful subsequent choice.
Reversal opportunities must nevertheless remain discoverable under the weighting
and preference-decay rules; a bad fit must not become a permanent hidden class lock.
### Testing a lineage through offspring
The prepared route is:
1. Earn and select an adaptation for the offspring plan without rebuilding the parent.
2. Gather the resources and choose a useful time/location to reproduce.
3. Keep control of the parent and watch the actual offspring feed, move, survive,
and interact under its inherited traits.
4. Keep or revise the retained plan for subsequent births.
5. At a later birth, optionally continue as that newborn and relinquish the parent.
The later control choice concerns a **newborn at that birth**, not possession of an
older offspring observed earlier. Observing a successful sibling reduces uncertainty;
it does not guarantee that the next offspring will survive changed conditions.
Later plan revisions never rewrite an already born organism.
The parent does not keep the offspring's allocated material. The advantage is
retaining a familiar body and controlling the timing of investment, rather than
avoiding reproduction costs. A plan can also be selected solely to produce useful
autonomous descendants, with no obligation to transfer player control.
### Cost of divergence from the parent
Offspring may cost more when expressing a different body requires more construction,
reorganization, or provisioning. Price those requirements directly rather than
charging an arbitrary amount for each different trait:
| Difference | Proposed cost consequence |
| --- | --- |
| Small pigment or regulatory adjustment | Modest cost only where the actual recipe or developmental work changes. |
| New propulsion, shell, or other machinery | Fund the required material, assembly E, and relevant newborn operating reserves. |
| Changed body layout | Additional developmental work where supported, even if final mass is similar; quantify it rather than inventing a complexity tax. |
| Removal or simplification | May make offspring cheaper by avoiding tissue construction and upkeep. |
| Several interdependent changes | Account for the resulting integrated body once; do not charge for shared tissue repeatedly or count plan-edit history as anatomy. |
Division transfers existing material into resulting organisms; it must not also
charge the full cost of building that same transferred material from nothing.
Any breakdown/reassembly work and losses need an explicit funded model. Do not
automatically reuse the old budding bill for the proposed division route.
Preview both resulting organisms, E/C/N/P/O allocations, actual changed traits,
construction requirements, and which organism the player will control. Distinguish
estimated future operating budgets from guaranteed survival. Experimenting with a
divergent offspring must remain affordable enough to support the early pacing target.
### Ordinary reproduction and colony growth
Division remains possible **without a mutation**; random offers and declines must
not gate the first three-minute birth or the ability to populate the habitat.
Mutation selection also need not force an immediate split: the prepared route
exists for players gathering resources, travelling, or avoiding danger.
Proposed early default: ordinary division separates similar offspring; a selected
adaptation changes the new offspring; retained attachment keeps the resulting cells
together. Further adaptations support transport and specialized colony roles.
Budding can remain an alternative reproductive strategy rather than being removed
from every lineage.
The catalog's existing "bud/budding" wording is retained while the maintainer
reviews its 150 entries. Read the pacing targets here as **first funded birth**
(division in this proposed early default). Catalog adaptations that affect offspring
size, provisioning, or birth rate need explicit compatibility with division versus
budding; this note does not silently apply historical budding rules to division.
Avoid a mandatory regrow-to-full waiting period after each split, while retaining
minimum viable bodies and real material/energy requirements for further births.
## Resources and producer rules
Abbreviations: **E** = usable energy, **C** = carbon, **N** = nitrogen,
@@ -549,6 +668,12 @@ consented playtest notes rather than silently expanded behavioral surveillance.
rather than inferred only from behavior?
- How are useful multi-change adaptations represented without hiding dependent
losses or turning the catalog into fixed classes?
- What exact tissue/reserve partition and viable-size rules support immediate
division, and which reproductive adaptations apply to division versus budding?
- How does the UI preview Evolve now versus Prepare offspring, including both
resulting organisms and the irreversible control choice?
- Which real developmental work makes a divergent offspring more expensive,
without double-charging transferred tissue or making experimentation too slow?
- Which traits can change physiologically during life, and which only develop in
offspring? Classify each operator explicitly before implementation.
- How do colony membership, resource ownership, control, death, and logout work
@@ -12,7 +12,7 @@ spec:
env:
- {name: REGION_SESSION, value: "dev"}
- {name: REGION_SESSION_LISTEN, value: "0.0.0.0:7400"}
- {name: REGION_SESSION_WINDOW, value: "dev-20261021T150708-c6255524:1792609628"}
- {name: REGION_SESSION_WINDOW, value: "dev-20261021T153222-6e90d67e:1792611142"}
- {name: REGION_M10_MAX_GRANT_LIFETIME, value: "14400"}
ports:
# QUIC (ADR-0090): UDP only; nothing listens on TCP 7400.
+17
View File
@@ -10110,3 +10110,20 @@ there is no new executable behavior to validate. No runtime or ecological
pacing result is claimed. The required ADR-0087 pre-push session-window refresh
changes desired Git configuration only; no deployment or live reachability is
claimed.
### 2026-10-21 - Two mutation routes discussion addition
Expanded the draft with Evolve now (funded division and immediate offspring
succession) and Prepare offspring (scheduled birth, observation, and optional
succession at a later birth). Documented roughly half-body allocation, permanent
identities, retained parental plans, no free undo, and divergence costs based on
actual development rather than trait count. Ordinary reproduction remains
independent of mutation offers. The 150 catalog entries are unchanged.
Validation for this documentation segment: catalog rows compared verbatim with
the preceding committed draft; sequential IDs, populated fields, local links,
and whitespace checked. No application code, public contract, or runtime behavior
changes; CMake/CTest, clang-format, clang-tidy, and sanitizers are not applicable
to this prose addition and were not run. Design viability and balance remain
unqualified. The AGENTS.md-required pre-push window update is desired Git
configuration only, not a deployment.