Use of the information contained in this unapproved document is at your own risk.

Last update: 20 April,2001

								1003.2-92  #48

 _____________________________________________________________________________


	Interpretation Number:	XXXX
	Topic:			shell parameter expansion
	Relevant Sections:	Table E-1
	Classification:	defect


Interpretation Request: (Defect Report)
-----------------------

	Please can you indicate whether an error (typographical or otherwise)
	has been made in Table E-1 (Volume II, lines 3273-3291 of Annex E)
	of IEEE Standard 1003.2-1992.

	Table E-1 seeks to clarify all possible combinations that arise
	from shell parameter expansions, as described in section 3.6.2
	of Volume I, lines 399-420. However, if lines 399-417 are read
	while ignoring the phrase "or null" (as lines 418-420 imply), the
	resultant table obtained differs from Table E-1 in the following
	places:

	- lines 3282,3283; column "parameter set but null":
		currently: "substitute parameter"
		should be: "substitute null"
	Rationale: the ultimate effect of the operation is the same,
	since the set parameter has the null value already; however,
	the proposed change makes this entry more consistent with
	similar entries of the same column at lines 3278-3279 and 3286-3287.

	- lines 3283,3283; column "parameter unset":
		currently: "assign null"
		should be: "assign word"
	Rationale: the indicated correction is a direct result of 
	reading section 3.6.2, lines 402-404 while ignoring the phrase
	"or null".

	It is agreed (lines 3268-3269 of Annex E) that historically
	this section of documentation has been "hard to understand".
	The above corrections (if adopted) will help to overcome this
	difficulty, but the present text only serves to compound the
	dilemna.

(Mr.D.R.Killick)


IEEE Interpretation for 1003.2-1992 
-----------------------------------
The rationale is in error.
Concerns about the wording of this part of the rationale for the
standard have been forwarded to the sponsor.
   

Rationale for Interpretation:
-----------------------------
Where there is a conflict between the standard and the rationale, the
standard takes precedence.

It is expected that a future revison of the standard will address
this issue.


 _____________________________________________________________________________