Some minor comments on Draft #7 (published: 2017/9/26).

  - It may be useful to allow fewer than two arguments for the string
    comparison procedurs (string=?, string<?, etc.), with the
    semantics that they return true when invoked with fewer than two
    arguments.  Such an interpretation may help avoid some
    special-casing when using 'apply' with these procedures (e.g., to
    check whether all strings in a given list are string=?, when the
    list could be empty or a singleton).

  - string-drop-while[-right]: In the description, I believe "initial
    prefix" should read "initial prefix/suffix" as in the description
    for string-take-while[-right].  Also, the next paragraph notes
    "different order of arguments" (compared with string-trim[-right])
    but it seems like the order is the same and only the number of
    required arguments is different.

  - It may be useful to include the "Unicode note" from SRFI 13 about
    the effects of reversal, as a reminder, in the description of
    string-concatenate-reverse.

  - string-fill!: The 'fill' argument is marked as optional (differs
    from R7RS).  If the difference is intentional, then a default
    needs to be specified.

  - very minor:

    - There is an unfortunate mismatch in the order of arguments of
      list-tabulate (SRFI 1) and string-tabulate (SRFI 13), and of
      some other similar pairs of analogous procedures.  Probably we
      are stuck with that mismatch, but a brief note (perhaps in the
      4th para of the Specification section, where SRFI 1 is briefly
      mentioned) of the difference may help (esp. those who may be
      familiar with SRFI 1 and not with SRFI 13).

    - There seems to be some excess document indentation
      from the "Selection" heading up to string-take-while.

    - (perhaps related to above) validator.w3.org gives errors on the
      SRFI doc.

Regards,

-chaw