With phoebe's addition of accessor_name for indexes, the name field in various RawModuleDefs become redundant. Currently, the module bindings are allowed to specify names for RawIndexDef / RawConstraintDef / RawScheduleDef. These names are (almost?) completely auto-generated. The user-specified name for indexes is already its own field, accessor_name, distinct from these.
The problem right now is that we have to re-implement name generation in both the C# and Rust module bindings. It's a pain to make these consistent. It's much easier to just not bother, and then have the host generate all names uniformly during RawModuleDef validation. Names are almost entirely an internal implementation detail anyway.
With phoebe's addition of
accessor_namefor indexes, thenamefield in various RawModuleDefs become redundant. Currently, the module bindings are allowed to specify names forRawIndexDef/RawConstraintDef/RawScheduleDef. These names are (almost?) completely auto-generated. The user-specified name for indexes is already its own field,accessor_name, distinct from these.The problem right now is that we have to re-implement name generation in both the C# and Rust module bindings. It's a pain to make these consistent. It's much easier to just not bother, and then have the host generate all names uniformly during
RawModuleDefvalidation. Names are almost entirely an internal implementation detail anyway.