RE: Teletex Terminal Identifierindraft-ietf-ldapbis-syntaxes-01

["Kurt D. Zeilenga" <Kurt@OpenLDAP.org>]

At 02:39 PM 2002-03-05, Steven Legg wrote:
>Kurt D. Zeilenga wrote:
>> I note that "The H-R column suggests whether a value in
>> that syntax would likely be a human readable string."
>> Even DirectoryString may not be human readable.
>
>Perhaps instead of indicating whether the native encoding for
>a syntax is human-readable we should be indicating whether it
>is a UTF8 character string.

I have no problem with this.

>If a client knows that the encoding
>is a UTF8 string then the unprintable characters have a particular
>meaning, e.g. (hex)0D is a carriage return. If the encoding isn't
>a UTF8 character string then (hex)0D shouldn't be assumed to be
>anything in particular, and treated accordingly. It probably isn't
>a carriage return.

I would suggest that we:
        - clarify that UTF-8 values can contain non-printables,
        - state that consistent handling of many non-printable
         characters cannot be expected,
        - non-printable characters should be avoided in
          absence of an agreement to their handling.

>> 
>> I think all that is needed with regard to this syntax
>> is an explicit statement that dollar sign
>> ("the following separator symbol") and backslash
>> should be escaped using the mechanism defined in RFC
>> 2252, Section 4.3, pp3.
>
>Dumping the raw octets into the encoding means it isn't (in general)
>a UTF8 character string. Using hexadecimal means it is.

I think it better to correct the H-R column then to change
the syntax's specification.

Kurt