Hi,

I have couple of minor comments for the SRFI.

- Shouldn't text->utf16 and utf16->text accept/require endian? Or at least
  specify which endian it returns?
- Most of the procedures accept textual object but names start text. I found
  this a bit confusing since if procedures start type- prefix, usually they
  take that type of objects.
- Shouldn't text-index-right, text-skip, and text-skip-right return idx
  instead of k? Or should text-index return k since Rationale says it may
  return invalid index of the textual object?
- Why text-index, text-index-right, text-skip, and text-skip-right return
  only idx or k? Shouldn't they return idx-or-false like corresponding SRFI-13
  procedures?
- It might be even more convenient if text-concatenate-reverse also takes
  start argument.
- On references, both R6RS libraries and rationale don't have space. (I first
  saw Srationale.) It might be more friendly if there's space in between.

Cheers,

--
_/_/
Takashi Kato
E-mail: ktakashi19@gmail.com