SUBJECT: COBOL FCD Normative Reference to ISO/IEC 14652
AUTHOR: William M. Klein (firstname.lastname@example.org)
The COBOL International FCD review period is currently expected to begin February 1, 2001. I personally hope that both individuals who participate in L2 and the L2 (and/or L2-TAG) committees will give as much input as possible on the internationalization issues addressed in this document. However, there is one issue in particular that concerns me - especially as some information that I have seen so far seems to indicate that this MAY be a controversial issue both within the US and international communities.
It is my understanding that at the last SC22 plenary, explicit permission was "given" to allow normative references in ISO Standards to TRs (Technical Reports) regardless of whether they were Type 1, 2, or 3 TRs. However, in reviewing the October 2000 WG20 minutes, it seems as if the US - and possibly some other national bodies - have concerns about referencing TRs where "consensus" has not (and may never) be reached - particularly 14652. (This is not necessarily a question of whether or not they may – but whether or not they should make such references.)
The following normative text appears on page 64 of
Information technology — Programming languages, their environments and system software interfaces — Programming language COBOL
Reference number of working document: ISO/IEC/JTC 1/SC 22/WG 4 N 0147
Reference number of document: ISO/IEC FCD 1989:2001
"The set of extended letters consists of characters from the repertoire specified in ISO/IEC TR 10176, excluding any character that is defined as a basic letter, basic digit, or basic special character in the COBOL character repertoire, with the addition of character Middle Dot (code 00B7 in ISO/IEC 10646). Extended letters in identifiers are subject to the following rules:
a) A character in ISO/IEC TR 10176 is included in the set of extended letters if it exists in the implementor-defined compile-time coded character set.
b) Characters identified as special characters in ISO/IEC TR 10176 shall not be written as the first or last character of a user-defined word.
c) An uppercase extended letter is treated as though it were its corresponding lowercase extended letter when there is a character correspondence specified by the keyword 'tolower' in the category LC_CTYPE defined in ISO/IEC DTR 14652, with the following exceptions:
1. Latin small letter dotless I (0131) folds to Latin small letter I (0069)
2. Greek small letter final sigma (03C2) folds to Greek small letter sigma (03C3)
where the names and code values of letters are those defined in ISO/IEC 10646.
For purposes of matching, extended letters in identifiers are folded from uppercase to lowercase.
NOTE The specification of code values is a means of character identification and not a requirement for implementation of ISO/IEC 10646."
(This section may be found under the topic, "8.1.2 COBOL character repertoire")
My questions are:
1) Does L2 (and/or the L2-TAG) have any input on whether this is an advisable/desirable normative reference to 14652 (as well as the other ISO documents reference)? If not, what would you suggest that this might be changed to?
2) Is the technical/normative text correct and "reasonable" for the COBOL Standard? If not, will you be providing input to "us" (J4 and WG4) during the FCD Public Review period?
3) In general, will L2 be reviewing the "internationalization" issues included in ISO/IEC FCD 1989:2001 and providing input either to J4, the J4-TAG, or any other body/group in time for processing during this FCD Public Review period? Is there anything that I might personally do to facilitate this - or that I might pass on to others who might be able to give additional assistance?
J4 is currently scheduled to be meeting the week of February 5, 2001 - and I am certain that both J4 and the J4-TAG would be interested in any and all input and assistance that you (individually or as a group) might be able to provide.