Developer notes for planned Stratax feature work.

Purpose

Tracks the expected build-out of the library from core infrastructure through containers, generic operations, numerical features, bindings, and documentation.

Main API

Phase 1: Core Infrastructure

Buffer

Shape

Strides

Slice

Types

Concepts

Exceptions

Config
Validation

Phase 2: Containers

Vector

Matrix

Tensor

Creation helpers

Conversion helpers

Phase 3: Generic Operators

Arithmetic

Comparison

Indexing offsets

Slicing
Logical operators

Bitwise operators

Broadcasting

Phase 4: Shape Operations

reshape

flatten
ravel

squeeze

expand_dims

transpose

swapaxes

permute_axes

Phase 5: Linear Algebra

Matrix multiplication

Dot product

Cross product

Determinant

Inverse

Solve

LU

QR

Cholesky

SVD

Eigenvalues

Eigenvectors

Norms

Phase 6: Input and Output

Stream printing
CSV I/O

Binary I/O

Serialization policy

Phase 7: Documentation and Examples

Developer docs for implemented core files

Developer docs for implemented containers

Developer docs for implemented ops

Developer docs for printing
User-facing guide

API reference

Example refresh

Later Phases

Calculus
Statistics
Random number support
Python bindings
Optimization

Validation Notes

The roadmap describes intended scope, not guaranteed current behavior.
Checked items should have tests before they are treated as complete.
Validation remains unchecked because the validation module is currently deferred.
Placeholder headers are intentionally not marked complete until they expose real behavior.
Empty and zero-dimension shape behavior is tested for implemented containers and ops.

Implementation Notes

Complete core and container correctness before expanding into larger numerical features.
Generic operators should work across all containers before specialized algorithms are added.
Shape operations should preserve storage invariants and make copy/view semantics explicit.
Optimization work should come after correctness tests and baseline behavior are stable.
Current implemented APIs are copy-based; view semantics are future work.

Future Work

Broadcasting
Views and slices
Type promotion rules
Sparse arrays
GPU backend
Automatic differentiation
FFT
Symbolic mathematics