Repository navigation
docs: plan v4.3 and correct the stale roadmap status notes - #17
Merged
Merged
Conversation
The autumn roadmap had drifted from the code. LinearAlgebra Phase 1-2 was merged but still described as an open PR, a customer-driven physics track had taken over the autumn without appearing in the plan at all, and v4.1.0 and v4.2.0 were tagged with contents that do not match their release tables. Add Roadmap-v4.3.md, which plans the next release around the debt Phase 1-2 left behind: the decompositions were built but their call sites were never migrated, so eight files still solve dense systems with hand-rolled Gaussian elimination and CoupledOscillators keeps a private Jacobi eigensolver. Two of those migrations are accuracy fixes rather than cleanup - FittingSolver squares its condition number by going through the normal equations, and the Kalman filters form an explicit inverse of a symmetric positive-definite innovation covariance. The release also adds the root-finding module, whose absence leaves NewtonRaphson as the only root finder: it runs a fixed 100 iterations with no convergence check, no guard against a vanishing derivative, and a fresh Pascal matrix allocated per iteration. Mark the autumn plan as background rather than an active plan, and correct LinearAlgebraRoadMap where it no longer matched the code: Phase 1-2 is merged, and Phase 4 is partly done outside the plan - SparseMatrix has CSR storage and a Jacobi-preconditioned CG that drives Assembler2D, verified against the existing sparse and 2-D FEM tests. Documentation only; no code changes.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The autumn roadmap had drifted from the code. LinearAlgebra Phase 1-2 was merged but still described as an open PR, a customer-driven physics track had taken over the autumn without appearing in the plan at all, and v4.1.0 and v4.2.0 were tagged with contents that do not match their release tables.
Add Roadmap-v4.3.md, which plans the next release around the debt Phase 1-2 left behind: the decompositions were built but their call sites were never migrated, so eight files still solve dense systems with hand-rolled Gaussian elimination and CoupledOscillators keeps a private Jacobi eigensolver. Two of those migrations are accuracy fixes rather than cleanup - FittingSolver squares its condition number by going through the normal equations, and the Kalman filters form an explicit inverse of a symmetric positive-definite innovation covariance. The release also adds the root-finding module, whose absence leaves NewtonRaphson as the only root finder: it runs a fixed 100 iterations with no convergence check, no guard against a vanishing derivative, and a fresh Pascal matrix allocated per iteration.
Mark the autumn plan as background rather than an active plan, and correct LinearAlgebraRoadMap where it no longer matched the code: Phase 1-2 is merged, and Phase 4 is partly done outside the plan - SparseMatrix has CSR storage and a Jacobi-preconditioned CG that drives Assembler2D, verified against the existing sparse and 2-D FEM tests.
Documentation only; no code changes.