The informative section of SRFI 150 tells us how SRFI 99's ‘procedural’ layer might interact with SRFI 150’s refinement of the ‘syntactic’ layer. However, it doesn’t consider the ‘inspection’ layer of SRFI 150. It also leaves some other questions unanswered.

All three points allow the procedurally-defined or -accessed field names of an RTD to be:

> 	• an identifier naming the field in the style of this SRFI;
> 	• a string, character, boolean, number or keyword naming the field in the style of this SRFI;
> 	• a symbol naming the field in the style of SRFI 99 and SRFI 131;

Questions:

How does one pass an ‘identifier’, distinct from a symbol, to a procedural interface? Are syntax objects in the sense of syntax-case intended here? The normative section says that field names are compared using equal?, but if they’re syntax objects surely they have to be compared with bound-identifier=?.

What do the rtd-field-names and rtd-all-field-names procedures return, if not vectors of symbols? Vectors of ‘identifiers’ i.e. syntax objects?

Daphne