Why This Exists
Two related problems should not be confused: Wist offers embeddable, restricted application formulas, while UniversalToolchain lets language authors resolve an implementation from package-owned contributions before compiling source.
Application problem: controlled formulas (Wist)
Hardcoded formulas, rules, and domain logic start simple and become expensive fast.
At first, a condition in C# or a compact expression evaluator is enough. Then the product needs tenant-specific pricing, versioned validation rules, policy changes without redeploying the app, and a way to reject language features that should not exist in that domain. The problem is no longer only "evaluate this expression." The problem is owning a small runtime surface without turning every rule into application code.
Why a plain evaluator is sometimes not enough
A plain evaluator is useful while the shape of the language stays fixed.
It is usually enough when:
- the syntax is fixed,
- every accepted expression can use the same capabilities,
- there is no need to grow toward a domain-specific language,
- one execution surface is enough.
It starts to strain when the product needs custom syntax, restricted dialects, capability control, diagnostics, or a path from expressions into a more explicit runtime model.
Why a full language workbench is sometimes too much
Full language workbench approaches can be the right answer for large language programs.
They are often too heavy when the target is an embeddable .NET runtime layer. An application team may need parser, validation, translation, and execution hooks, but not a full platform story with heavyweight tooling, editor integration, and broad language governance before the first production rule can run.
Where UniversalToolchain fits
A handwritten parser, transformations and backend connected by typed .NET interfaces are usually sufficient for one fixed language. UT adds an independently inspectable configuration step when language implementations vary: dependency and capability resolution, single-owner slots, typed backend routes and exact runtime binding.
This additional planning machinery has a maintenance cost and does not guarantee semantic compatibility of arbitrary extensions. It is justified when packages and execution routes vary, rather than merely because a project contains a compiler. See the package model, architecture status and alternatives.
Good Wist entry scenarios
- Pricing formulas that start as arithmetic and grow into controlled product logic.
- Validation rules that need domain-specific constraints instead of arbitrary host-language behavior.
- Routing or policy logic where allowed capabilities differ by context, tenant, or environment.