You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
To make c2uplc as good as it can be, we (meaning @gnumonik) want to determine precisely what will be good to have as part of Covenant. Things we can already expect to be present:
Requests to call a builtin
Data types and their encodings (or a statement that we don't care how it's encoded)
Specific catamorphisms for tearing down known structures (lists and Data at least) given a coalgebra
Accessors (or specialized catamorphisms) for tearing down ledger API types
Serial and parallel let bindings as distinct forms
Outstanding questions:
Anamorphisms (or indeed, any sort of non-static data generation) is hard in a Turner-total setting. Some questions need to be answered there.
Are there specific kinds of information we'd want to convey that would make code generation easier and more efficient? This can include both restrictions and permissions: for example, 'I don't care how this data type is represented as long as the interface is such').
Should we also provide ways to build up ledger API types? For some this could possibly make sense (for example, Value or POSIXTime), but not in general (I can't think why any script would want to build ScriptContexts).
How we should support case expressions in a way that'd aid compilation.
@gnumonik - for visibility, please put suggestions or discussions into this issue, which we will pin for posterity, even once decisions are made.
The text was updated successfully, but these errors were encountered:
To make
c2uplc
as good as it can be, we (meaning @gnumonik) want to determine precisely what will be good to have as part of Covenant. Things we can already expect to be present:Data
at least) given a coalgebralet
bindings as distinct formsOutstanding questions:
Value
orPOSIXTime
), but not in general (I can't think why any script would want to buildScriptContext
s).case
expressions in a way that'd aid compilation.@gnumonik - for visibility, please put suggestions or discussions into this issue, which we will pin for posterity, even once decisions are made.
The text was updated successfully, but these errors were encountered: