Downgrade: raise SciMLBase compat floor to 2.49 (2.0.0 floor unresolvable with solver stack)#87
Merged
ChrisRackauckas merged 1 commit intoJun 15, 2026
Conversation
…solvable at downgrade minimum) At the downgrade minimum, SciMLBase is pinned to its compat floor 2.0.0 (from `SciMLBase = "2, 3.1"`). The test target pulls in the full OrdinaryDiffEq / MethodOfLines / ModelingToolkit stack, which (unpinned) resolves to a recent DiffEqBase. Modern DiffEqBase declares `SciMLBase` lower bounds well above 2.0.0 (e.g. DiffEqBase 6.130.1+ needs SciMLBase 2.0.1+, and the resolved stack needs higher still), and OrdinaryDiffEqSDIRK at the resolved version has no SciMLBase 2.0.x in range. So SciMLBase 2.0.0 has an empty intersection with the co-installed solver stack -> Unsatisfiable for DiffEqBase / OrdinaryDiffEqSDIRK. 2.49.0 is the smallest SciMLBase that resolves with the current solver stack on the LTS (1.10) downgrade floor (2.48.x and below remain unsatisfiable). The 3.1 upper series is unchanged; no upper bound lowered, no test logic touched. Co-Authored-By: Chris Rackauckas <accounts@chrisrackauckas.com> Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 MINIMUM-VERSION (downgrade) job fails. The failure surfaces at the
Pkg.testsandbox re-resolve, but it is a pure resolution conflict (reproduced locally on Julia 1.10, the LTS downgrade floor):and, probing minimal floors directly,
Unsatisfiable ... OrdinaryDiffEqSDIRK [2d112036] ... restricted by SciMLBase to versions: uninstalled.Why the old floor was impossible
SciMLBase = "2, 3.1", sojulia-downgrade-compatpins SciMLBase to the floor of the first range,2.0.0.[compat], so they are not downgrade-pinned and resolve to recent versions, dragging in a recent DiffEqBase + OrdinaryDiffEq sublibraries.6.130.1-6.132-> SciMLBase2.0.1, climbing higher for newer DiffEqBase), and OrdinaryDiffEqSDIRK at the resolved version has no SciMLBase 2.0.x in range. So SciMLBase 2.0.0 has an empty intersection with the co-installed solver stack -> Unsatisfiable.(Master latest-versions CI is unaffected: the Core groups on
julia 1andjulia ltspass; this is downgrade-specific.)Fix
Raise the lower bound of the first range:
SciMLBase = "2, 3.1"->SciMLBase = "2.49, 3.1". The author intent to support both the 2.x and 3.1 series is preserved; only the unresolvable low end of 2.x is dropped. No upper bound lowered, no test logic touched.Resolve verification (local, Julia 1.10 LTS), resolution only
Mimicking the downgrade pin (SciMLBase pinned to floor, test stack left to resolve):
2.49.0 is the smallest co-installable floor with the current solver stack. The full test suite was not run locally; PR CI confirms.
Note: there is a separate, pre-existing QA-group failure on lts (Aqua
project_extras: Project.toml vs test/Project.toml mismatch) unrelated to this resolution fix.Ignore until reviewed by @ChrisRackauckas.