snip..
>I am not sure that I have ever done a formal two week last call
>for changes to the Printer MIB document. Let this e-mail serve
>as the last call. The Host Resources MIB has been published as
>an Internet Draft and I want to be ready at a moments notice
>to start pushing the Printer MIB forward as soon as something
>happens to the HR MIB.
Hi folks, Tuesday (31 March 1998)
Our response to the working group 'last call' on the current Printer MIB
draft, announced by Lloyd Young (co-chair) on Thursday (19 March).
This note proposes a SIMPLE update for text and name string localization
in the latest Printer MIB draft, WITHOUT adding any new objects or new
textual conventions.
Comments are of course very welcome.
Cheers,
- Tom Hastings (Xerox)
- Ira McDonald (High North)
>----------------------------------------------------------------------<
Background
==========
1) Many vendors' implementations of RFC 1759 localize strings besides
'xxxDescription' via 'prtGeneralCurrentLocalization' (although the
list of localized strings differs between vendors).
2) From April through June of 1997, Ira McDonald (High North) and Tom
Hastings (Xerox) proposed complex solutions to the localization of
the text and name strings in the Printer MIB (8 string objects are
now localized, while 24 string objects have indeterminate charsets
and languages), eventually abandoned by the working group co-chairs.
3) Later in 1997, the IETF/PWG Internet Printing Protocol (IPP) Working
Group added complete charset and language support to their working
documents for ALL text and name string attributes, and clarified the
role of 'keywords' within string attributes (invariant US English
tokens, encoded in US-ASCII).
4) January 1998, the IESG adopted Harald Alvestrand's excellent work
"IETF Policy on Character Sets and Languages", BCP 18 / RFC 2277,
which applies as a MANDATORY policy to all protocols on the IETF
'standards track', which transfer text or name strings.
5) In the seven SNMPv3 MIBs (RFC 2271-2275), the System Application MIB
(RFC 2287), the Lightweight Directory Access Protocol (LDAP, RFC
2251), the Service Location Protocol (SLP, RFC 2165), and in almost
ALL other new IETF RFCs with support for accessing text or names,
strings are ALWAYS specified in the UTF-8 transform of ISO 10646, in
compliance with the IESG policy (RFC 2277).
6) The original Printer MIB (RFC 1759) and the current Printer MIB
draft are NOT compliant with the IESG policy (RFC 2277).
>----------------------------------------------------------------------<
Our Proposal
============
1) In section 2.2.1.1 'International Considerations', on pages 13-14
[REPLACE entire section with...]
This Printer MIB complies with all requirements and recommendations of
the IETF Policy on Character Sets and Languages [CHAR-POL]. The IANA
Charset Registration Procedures [CHAR-REG] introduces the term 'charset'
for a coded character set (CCS, eg, ISO 10646) bound to an associated
character encoding scheme (CES, eg, UTF-8).
The localization portion of the general printer sub-unit is
responsible for identifying the natural language, country, and
charset in which all text and name strings are expressed.
The available localizations are represented by the Localization Table.
All implementations of this Printer MIB SHALL support at least one
localization per printer with the UTF-8 charset.
All implementations of this Printer MIB SHALL support the [UTF-8]
charset and MAY support other charsets, with the following restrictions:
1) All supported charsets SHALL be strict supersets of [US-ASCII], ie,
charsets in which code positions decimal (0-127) are identical to
[US-ASCII].
2) Control characters (code positions decimal 0-31 and 127) SHALL NOT
be used unless explicitly allowed by the DESCRIPTION clause of a given
TEXTUAL-CONVENTION or OBJECT-TYPE defined in this Printer MIB.
Note: Charsets which are strict supersets of [US-ASCII] include
ISO 8859-x, Shift-JIS, JIS X0208-1983, and UTF-8.
Charsets which are NOT strict supersets of [US-ASCII] include
IBM EBCDIC, ISO 10646 in UCS-2 (two-octet) form,
ISO 10646 in UCS-4 (four-octet) form, and all of the 7-bit
national charsets which conform to ISO 646 EXCEPT [US-ASCII] and the
IRV.
Certain string objects MAY contain 'keywords'. These 'keywords' SHALL
be restricted to be defined in visible [US-ASCII]. These 'keywords'
SHALL be written (by SNMP managers and SNMP agents) and SHALL be
returned (by SNMP agents) in visible [US-ASCII]. These 'keywords' SHALL
NOT be localized by SNMP agents.
NOTE: Since all charsets in the MIB SHALL be a superset of [US-ASCII],
keywords values SHALL be the same representation for any charset
specified by prtGeneralCurrentLocalization, thereby simplying parsing
by management applications.
Certain string objects MAY contain 'names'. These 'names' SHALL NOT be
be restricted to be defined in visible [US-ASCII]. These 'names'
SHALL be written (by SNMP managers and SNMP agents) and SHALL be
returned (by SNMP agents) in the charset specified by
prtGeneralCurrentLocalization. These 'names' MAY be
localized by SNMP agents to the language (and country) specified by
prtGeneralCurrentLocalization.
Some objects may contain keywords and/or names, depending on
implementation and usage as shown below. When retrieving such an object
from the MIB, management applications SHOULD assume that the value is a
keyword and localize the language for its user. However, if the
management application does not recognize the keyword, it SHOULD treat
the value as a name and display the name as it is to its user:
Object Requirements
------ ------------
PrtLocalizationLanguage SHALL be keywords defined in ISO 639
PrtLocalizationCountry SHALL be keywords defined in ISO 3166
prtGeneralCurrentOperator MAY contain keywords defined in this MIB
prtGeneralServicePerson MAY contain keywords defined in this MIB
prtChannelInformation MAY contain keywords defined in this MIB
localized by prtGeneralCurrentLocalization
charset language (and country)
------- --------
prtGeneralPrinterName SHALL MAY
prtGeneralSerialNumber SHALL MAY
prtCoverDescription SHALL SHALL
prtInputMediaName SHALL MAY (keywords and names)
prtInputName SHALL MAY (keywords and names)
prtInputVendorName SHALL MAY
prtInputModel SHALL MAY
prtInputVersion SHALL MAY
prtInputSerialNumber SHALL MAY
prtInputDescription SHALL SHALL
prtInputMediaType SHALL MAY (keywords and names)
prtInputMediaColor SHALL MAY (keywords and names)
prtOutputName SHALL MAY (keywords and names)
prtOutputVendorName SHALL MAY
prtOutputModel SHALL MAY
prtOutputVersion SHALL MAY
prtOutputSerialNumber SHALL MAY
prtOutputDescription SHALL MAY
prtMarkerSuppliesDescription SHALL MAY
prtMarkerColorantValue SHALL MAY (keywords and names)
prtMediaPathDescription SHALL SHALL
prtChannelProtocolVersion SHALL MAY
prtInterpreterLangLevel SHALL MAY
prtInterpreterLangVersion SHALL MAY
prtInterpreterDescription SHALL SHALL
prtInterpreterVersion SHALL MAY
prtAlertDescription SHALL SHALL
localized by prtConsoleLocalization
charset language (and country)
------- --------
prtConsoleDisplayBufferText SHALL SHALL
prtConsoleDescription SHALL SHALL
2) In all other sections
[REMOVE all (9) references to 'NVT-ASCII', which are now unnecessary
given the above revised section 2.2.1.1]
3) In all other sections
[REMOVE all (12) references to ASCII charset restrictions, except
for those occurring in 'keyword' descriptions]
4) In 'prtGeneralCurrentLocalization' object
[REVISE description to align with this proposal]
5) In 'prtLocalizationTable' and subordinate objects
[REVISE description to align with this proposal]
6) In section 'References', on page 176
[ADD the following references...]
[CHAR-POL] IETF Character Set and Languages Policy, BCP 18,
RFC 2277, January 1998.
[CHAR-REG] IANA Charset Registration Procedures, BCP 19,
RFC 2278, January 1998
[UTF-8] UTF-8, a transformation format of ISO 10646, RFC 2279,
January 1998.
[ISO-639] ISO/IEC 639, Code for the representation of names of
languages, 1988.
[ISO-3166] ISO/IEC 3166, Code for the representation of names of
countries, 1993.
[ISO-10175] ISO/IEC 10175, Document Printing Application
(DPA), 1996.
[ISO-10180] ISO/IEC 10180, Standard Page Description Language
(SPDL), 1995.
[REMOVE the following reference...]
[NVT-ASCII] ...
>----------------------------------------------------------------------