Defect Report Number: Z39.50MA-DR0001
Source of Report: Bob Waldstein
Submitted-to and Date: 12/95
Report Concerning: Close Service
Type of Defect: Editorial
Reference in Document: 3.2.11.1.1 and ASN.1
Nature of Defect: There is a reason of 'response to Close request'. But in the ASN, value 8 is 'peerAbort'.
Solution Proposed: Change 'peerAbort' to 'responseToPeer' in ASN.1.
Status/Secretariat Response: Endorsed by ZIG
Defect Report Number: Z39.50MA-DR0002
Source of Report: Mike Douglass
Submitted-to and Date: 12/96
Report Concerning: PeriodicQuerySchedule
Type of Defect: Technical
Reference in Document: EXT.2 ESFormat-PeriodicQuerySchedule
Nature of Defect: If the origin supplies a persistent query schedule that names a persistent query package rather than supplying a query, the origin need not supply any database names if they are in the persistent query. However, the only place to put them in the schedule is in the originPart. Shouldn't this field be in the notToKeep part with a corresponding field in the TargetPart? (In the same manner as the actualQuery field.)
More of a problem is additionalSearchInfo. Once the task package is built we lose the persistent query package name and there's nowhere to put the additionalSearchInfo.
Solution Proposed: The problem with the current definition is that the origin may omit the names, supplying a query package name (instead of an actual query) assuming that the same databases as are listed in the query package are implied. But then the target copies the actual query into the package (but not the database names), and the original query package name is not retained, and so the database names are lost. In the proposed revised definition below, if the origin supplies names, the target uses those; if the origin does not (in which case the origin has to supply a packageName and not an actual query) the target copies the databaseNames from the package.
Proposed Correction: The current definition looks like:
OriginPartToKeep ::=SEQUENCE{ activeFlag [1] IMPLICIT BOOLEAN, databaseNames [2] IMPLICIT SEQUENCE OF InternationalString OPTIONAL, resultSetDisposition [3] .... ...... OriginPartNotToKeep ::= SEQUENCE{ querySpec [1] .... originSuggestedPeriod [2] ..... expiration [3] ..... resultSetPackage [4] ...... TargetPart ::= SEQUENCE{ actualQuery [1] Query, targetStatedPeriod [2] .... expiration [3] .... resultSetPackage [4] .... lastQueryTime [5] .... lastResultNumber [6] .... numberSinceModify [7] ....Proposed Change:
OriginPartToKeep ::=SEQUENCE{ activeFlag [1] IMPLICIT BOOLEAN, resultSetDisposition [3] .... ...... OriginPartNotToKeep ::= SEQUENCE{ databaseNames [0] IMPLICIT SEQUENCE OF InternationalString OPTIONAL, querySpec [1] .... originSuggestedPeriod [2] ..... expiration [3] ..... resultSetPackage [4] ...... additionalSearchInfo [5] OtherInformation OPTIONAL} TargetPart ::= SEQUENCE{ databaseNames [0] IMPLICIT SEQUENCE OF InternationalString, actualQuery [1] Query, targetStatedPeriod [2] .... expiration [3] .... resultSetPackage [4] .... lastQueryTime [5] .... lastResultNumber [6] .... numberSinceModify [7] .... additionalSearchInfo [8] OtherInformation OPTIONAL}
Status/Secretariat Response: Pending ZIG endorsement.
Defect Report Number: Z39.50MA-DR0003
Source of Report: Marko Kusej
Submitted-to and Date: 12/96
Report Concerning: Explain
Type of Defect: Clerical
Reference in Document: REC.1
Nature of Defect:
The Explain record syntax defines in 'RpnCapabilities' the parameter 'operators' as SEQUENCE OF INTEGER. Which value representation are to be supplied for the different operators?
Solution Proposed:
There appears to be an INTEGER enumeration lost.
Change:
Defect Report Number: Z39.50MA-DR0004
Source of Report: Mike Douglass douglm@rpi.edu
Submitted-to and Date: 5/96
Report Concerning: ASN.1 for diagnostic format Diag-1
Type of Defect: Clerical
Reference in Document: ERR.2
Nature of Defect:
An identifier (including a defined value) should start with a lower case letter.
Therefore
should instead be
DiagnosticFormatDiag1 {z39-50-diagnosticFormat diag-1 (2)} DEFINITIONS ::=
Solution Proposed:
Defect Report Number: Z39.50MA-DR0005
Source of Report: Mark H Needleman mhn@stubbs.ucop.edu
Submitted-to and Date: 9/96
Report Concerning: Close
Type of Defect: Clerical
Reference in Document: 3.2.11.1 and 3.2.11.1.2
Nature of Defect:
In 3.2.11.1.2, Diagnostic Information (for Close) it says the target may include an optional text message, but in the table in 3.2.11.1 there is no note indicating target only.
Solution Proposed:
Status/Secretariat Response: Endorsed by ZIG
Source of Report: Makx Dekkers
Submitted-to and Date: 7/96
Report Concerning: Explain
Type of Defect: Technical
Reference in Document: Explain ASN.1, TargetInfo and DatabaseInfo
Nature of Defect:
Solution Proposed:
Status/Secretariat Response: Pending ZIG endorsement
Source of Report: Mike Douglass
Submitted-to and Date: 2/97
Report Concerning: Extended Services,
PeriodicQuerySchedule
Type of Defect: Technical
Reference in Document: TargetPart of ASN.1 for PeriodicQuerySchedule Extended Service
Nature of Defect:
Solution Proposed:
Status/Secretariat Response: Pending ZIG endorsement
Source of Report: Jack Smith jax@sanderson.net.au
Submitted-to and Date: 16 Sep 1997
Report Concerning: ASN.1 Capitalization
Type of Defect: Technical
Reference in Document: Appendices
Nature of Defect:
As an example, the definition of the SUTRS record syntax commences
with:
RecordSyntax-SUTRS
In the second line, SUTRS, as an identifier, must begin with a
lower-case letter.
The requirement for the initial lower-case letter is essential to
differentiate an identifier from a typereference or
modulereference, which in certain expressions would be otherwise
indistinguishable.
The following Module identifiers all exhibit the same problem.
RecordSyntax-SUTRS
RecordSyntax-generic -- For detailed semantics, see Appendix RET.
RecordSyntax-ESTaskPackage
ESFormat-PersistentResultSet
ESFormat-PersistentQuery
ESFormat-PeriodicQuerySchedule
ESFormat-ItemOrder
{Z39-50-extendedService ItemOrder (4)} DEFINITIONS ::=
ESFormat-Update
ESFormat-ExportSpecification
ESFormat-ExportInvocation
Solution Proposed:
RecordSyntax-SUTRS
RecordSyntax-generic -- For detailed semantics, see Appendix RET. RecordSyntax-ESTaskPackage
ESFormat-PersistentResultSet
ESFormat-PersistentQuery
ESFormat-PeriodicQuerySchedule
ESFormat-ItemOrder
ESFormat-Update
ESFormat-ExportSpecification ESFormat-ExportInvocation
Source of Report: Bob Waldstein wald@library.mt.lucent.com
Submitted-to and Date: Mon, 8 Dec 1997 11:47:08 EST
Report Concerning: Explain -- AccessInfo unitSystems
Type of Defect: Technical
Reference in Document: REC.1: Explain Record Syntax -- AccessInfo
Nature of Defect:
Solution Proposed:
Status/Secretariat Response: Pending ZIG endorsement
Submitted-to and Date: January 28, 2998
Report Concerning: Appendix RET
Type of Defect: Technical
Reference in Document: RET.3.4.1.2.1
Nature of Defect:
Solution Proposed:
April 15, 1998
..... SEQUENCE OF INTEGER
To:
......SEQUENCE OF INTEGER{
and (0),
or (1),
and-not (2),
prox (3)}
Status/Secretariat Response: Endorsed by ZIG
Z39.50-1995 Defect Report Z39.50MA-DR0004: ASN.1 Capatalization
DiagnosticFormatDiag1 {Z39-50-diagnosticFormat diag-1 (2)} DEFINITIONS ::=
As suggested.
Z39.50-1995 Defect Report Z39.50MA-DR0005: Close Diagnostic
Add "target only" in the "Note" column in the table in 3.2.11.1.
Z39.50-1995 Defect Report Z39.50MA-DR0006: Explain Icon
Defect Report Number: Z39.50MA-DR0006
In both TargetInfo and DatabaseInfo, an icon can be supplied. This field is a non-key brief element. Some Explain implementors do not want to retrieve the icon because of the effect on performance. So there should be a way to retrieve the (other) brief elements without the icon.
Define an element set name, 'brief-1', for TargetInfo and DatabaseInfo records, identical to 'brief' except to exclude the icon.
Z39.50-1995 Defect Report Z39.50MA-DR0007: "Last query" parameters in
PeriodicQuery
Defect Report Number: Z39.50MA-DR0007
In TargetPart of the PeriodicQuerySchedule Extended Service, the parameter
lastQueryTime and lastResultNumber should be optional, since they will not
have meaningful values between the time the package is created and the first
query is executed.
Change these two parameters to optional.
Z39.50-1995 Defect Report Z39.50MA-DR0008: ASN.1 Capitalization (2)
Defect Report Number: Z39.50MA-DR0008
Errors have been identified in certain module identifiers where an
object identifier is specified which has its initial letter in
upper case, contrary to the ASN.1 standard.
{Z39-50-recordSyntax SUTRS (101)} DEFINITIONS ::=
{Z39-50-recordSyntax SUTRS (101)} DEFINITIONS ::=
{Z39-50-recordSyntax GRS-1 (105)} DEFINITIONS ::=
{Z39-50-recordSyntax ESTaskPackage (106)} DEFINITIONS ::=
{Z39-50-extendedService PersistentResultSet (1)} DEFINITIONS ::=
{Z39-50-extendedService PersistentQuery (2)} DEFINITIONS ::=
{Z39-50-extendedService PeriodicQuerySchedule (3)} DEFINITIONS ::=
{Z39-50-extendedService Update (5)} DEFINITIONS ::=
{Z39-50-extendedService ExportSpecification (6)} DEFINITIONS ::=
{Z39-50-extendedService ExportInvocation (7)} DEFINITIONS ::=
Change these as follows:
{Z39-50-recordSyntax sUTRS (101)} DEFINITIONS ::=
{Z39-50-recordSyntax gRS-1 (105)} DEFINITIONS ::=
{Z39-50-recordSyntax eSTaskPackage (106)} DEFINITIONS ::=
{Z39-50-extendedService persistentResultSet (1)} DEFINITIONS ::=
{Z39-50-extendedService persistentQuery (2)} DEFINITIONS ::=
{Z39-50-extendedService periodicQuerySchedule (3)} DEFINITIONS ::=
{Z39-50-extendedService itemOrder (4)} DEFINITIONS ::=
{Z39-50-extendedService update (5)} DEFINITIONS ::=
{Z39-50-extendedService exportSpecification (6)} DEFINITIONS ::=
{Z39-50-extendedService exportInvocation (7)} DEFINITIONS ::=
Status/Secretariat Response: Pending ZIG endorsement
Z39.50-1995 Defect Report Z39.50MA-DR0009: Explain -- AccessInfo unitSystems
Defect Report Number: Z39.50MA-DR0009
In AccessInfo, unitSystems [11] is not OPTIONAL. It should be OPTIONAL.
Change it to OPTIONAL as suggested.
Z39.50-1995 Defect Report Z39.50MA-DR0010: Defect Report Number: Z39.50MA-DR0010: Schema Identifier
Source of Report: Mike Taylor mirk@ssl.co.uk
It appears that Appendix RET (specifically, in RET.3.4.1.2.1) recommends that the schemaIdentifier (TagSet-M element 1) if it occurs, should occur as the first element in a record (or subrecord). This recomendation is inconsistent with current schema development, in particular with the CIMI schema.
At the beginning of the last paragraph,
change:
This element, if provided, should normally occur as the first element ...
to:
This element, if provided, will often (but need not) occur as the first element ...
Status/Secretariat Response: Pending ZIG endorsement
Library
of Congress