Issues for Mailing list of the Robotic Localization Service 1.0 (RLS) Finalization task Force
To comment on any of these issues, send email to rls-ftf@omg.org. (Please include the issue number in the Subject: header, thusly: [Issue ###].) To submit a new issue, send email to issues@omg.org.
List of issues (green=resolved, yellow=pending Board vote, red=unresolved)
Issue 12916: create sub packages
Issue 12917: Sequence diagrams should be added
Issue 12918: Separate XMI definition from RLS specification document.
Issue 12919: Figure 3 - Relative and Mobile c oordinate reference system
Issue 12920: The navigability notation(arrow head) should be added as follows
Issue 12921: The association end name should be added as follows
Issue 12922: add an explanatory note
Issue 12923: Table 5 RelativeDatum class
Issue 12924: Figure 4 - Mobile CRS operations
Issue 12925: Figure 5 - Identity information
Issue 12926: 7.3.2 Identity Information
Issue 12927: Figure 6 - Error Type
Issue 12928: Figure 7 - Error information
Issue 12929: Figure 9 - RoLo Architecture
Issue 12930: Figure 10 - RoLo Data Operation
Issue 12931: Figure 11 - RoLo Data format
Issue 12932: 7.4.1 Common data format TypeI.Cartesian coordinate system
Issue 12933: 7.4.1 Common data format TypeII.Polar coordinate system
Issue 12934: Figure 21 - RoLo Ability and RoLo Parameter Set
Issue 12935: Figure 22 - RoLo Service
Issue 12936: Table 132 - Stream Parameter Set class
Issue 12937: Table 133 - Stream Ability class
Issue 12938: Table 134 - RoLo Stream class
Issue 12939: Sample C++ Header
Issue 12940: Typos
Issue 13106: Relative/Mobile CRS name is confusing
Issue 13107: Relative/Mobile operations are not well defined
Issue 13108: SymbolicCS/CRS name is confusing
Issue 13109: Possible inconsistency between RoLo Data class attributes
Issue 13110: RoLo Attribute Set definition typo
Issue 13111: Possible inconsistency between RoLo Attribute Set attributes
Issue 13112: Typo in 6.3.4 Robotic Localization Architecture
Issue 13113: Typo in Position Element Operation class description
Issue 13114: RoLo Attribute Definition Set definition typo
Issue 13115: The union of a documents
Issue 13116: Change the class name of RoLo Architecture
Issue 13117: Angle unit representation
Issue 13118: Posistion description in Type II
Issue 13119: Name of Type II coordinate system
Issue 13120: Name of Orientation
Issue 13121: Definition of orientation in Type I, II, III
Issue 13122: Filter package for the Issue #12916
Issue 13129: posSpecRef in Error Element (Figure 9/Table 49)
Issue 13130: Mismatch in Error Type (Figure 6) and Error Information (Figure 7) definitions
Issue 13131: Missing 'public:' declaration in PSM RelativeDatum definition
Issue 13143: Typo in Position Element Concatenated Operation class definition
Issue 13144: Definition mismatch for Service Ability class in Figure 22 and Table 139
Issue 13145: Figure 8 may accompany an object diagram
Issue 13184: Filter Condition should be reduced and move to annex as informative
Issue 13185: Hierarchy of Error can be reduced
Issue 13186: XML mapping of RoLo Data and RoLo Element
Issue 13187: Hierarchical relationship for instances of CRS and RoLo Data Spec
Issue 13317: Simplication of Error Type (Figure 6):
Issue 13318: Typo in Table 22 (RoLo SymbolRef)
Issue 13319: Obligation of 'values' attribute of Matrix (Figure7/Table36)
Issue 13320: The type of 'usesOperation' attribute of RoLo Data Transformation (Table62)
Issue 13321: Association between RoLo Parameter Set and RoLo Parameter Definition (Figure 21)
Issue 13322: Inability of representing range values (Figure21/Table120)
Issue 13426: Conformance shall be specified in detail
Issue 13427: Some references are missing
Issue 13428: Terms and Definition shall be added
Issue 13429: Symbol should be described
Issue 13430: Some organization missing in "6.1 Supporting Organizations"
Issue 13431: Names missing in "6.2 Acknowledgements"
Issue 13432: Typos in 6.3 Background
Issue 13433: Overview in 6.3 Background needs to be modified
Issue 13434: Some description on Coordinate System usage shall be added
Issue 13435: Ambiguity in 'rla' attribute description in RelativeDatum class
Issue 13436: 'inStream' attribute in MobileDatum class shall be an object
Issue 13437: 'numDists' attribute in Lineare Mixure Model class may be unnecessary
Issue 13438: Typo in RoLo Attribute Definition class description
Issue 13439: Structural change for RoLo Service class
Issue 13440: Typo in Service Ability class description
Issue 13441: Inconsistency between PIM and PSM namings
Issue 13442: Inconsistency in class names
Issue 13443: Typo: argument type for methods in RoLo Service class
Issue 13451: Missing references
Issue 13998: RLS - typos
Issue 13999: (7.4.1) Timestamp for common data types
Issue 14000: (Annex A) Mapping between PIM and C++ PSM is not clear
Issue 14001: (7.5) In the filter condition section
Issue 14002: 7.5) Filter condition requires naming of RoLo data elements. However, the naming rules are not clear.
Issue 14003: Fig3,Tbl9,Fig22,Tbl134,Tbl140) Mobile Datum class, RoLo Stream class and RoLo Service
Issue 14004: Fig5,Fig9) RoLo Symbolic Position class and RoLo Position class
Issue 14005: (Fig7) Matrix class is specified as "DataType" but this is unnecessary
Issue 14006: Tbl119-Tbl129)
Issue 14007: Tbl124) Specifying RoLo Attribute Singlevalue class as an enumeration
Issue 14008: 7.6) RoLo service is described to have only one output stream, but this is sometimes too difficult to use
Issue 14009: (Tbl135/Tbl136)
Issue 14010: (Tbl135/Tbl136) The 'connect' methods
Issue 14011: Tbl134) The behavior of setParameter method in RoLo Stream class is unclear
Issue 14012: Tbl130) Two types of data passing mode specified in StreamType enumeration is not clear
Issue 14013: Tbl130,Tbl136) The 'setData' method in InStream class
Issue 14014: (C++-PSM) Methods shall better send return codes as exceptions
Issue 14015: Subset specifications (Fig22)
Issue 14016: Parent class wrong for RelativeCRS
Issue 12916: create sub packages (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary:
The package division
In RLS specification, there is only one package named RLS.
I think that it should be divided into some sub packages.
Ex.Coordinate system package, Data Format package, Filter package, Service package, etc.
Resolution: Modify Figure 2 as follows:
¡V Divide the RoLo package into four sub packages of Architecture
package, DataFormat package, FilterCondition package and Interface
package.
¡V Delete original dependency relationships between RLS and GIS
standards and add the following dependency relationships in the
diagram:
ƒÞ from Architecture package to ISO 19107
ƒÞ from Architecture package to ISO 19103
ƒÞ from Interface package to ISO 19103
ƒÞ from Interface package to ISO 19115
ƒÞ from DataFormat package to Architecture package
ƒÞ from Interface package to Architecture package
ƒÞ from Architecture package to Interface package
ƒÞ from RoLo package to ISO 19111
ƒÞ from FilterCondition package to Interface package FilterCondition package to ISO 19142
ƒÞ from FilterCondition package to ISO 19143
Add the following sentences(package descriptions) at the beginning of section
7 (Platform Independent Model):
The PIM consists of four parts:
1. Architecture package
The architecture package defines a new framework for representing location informatio
n required in the field of robotics. See section 7.3.
2. DataFormat package
The data format package defines how the defined data is represented for exchange amo
ngst RoLo modules. See section 7.4.
3. FilterCondition package
The filter condition package defines methods for filtering localization results. See secti
on 7.5.
4. Interface package
The interface package defines an API for data passing and configuration of RoLo modu
les. See section 7.6.
Change the following section titles as package names:
¡V Change the title of section 7.3 Representing Robotic Localization
Results into 'Architecture Package'.
¡V Change the title of section 7.4 Data Formats into 'DataFormat Package'.
¡V Change the title of section 7.5 Filter Condition into ¡¥FilterCondition
Package¡¦.
¡V Change the title of section 7.6 Service Interface into 'Interface Package'.
Add package name to following items:
- MobileDatum class (Table 9 and Figure 3): change type name of
¡¥inStream¡¦ attribute and ¡¥inStream¡¦ argument of ¡¥getInStream¡¦ method to
RoLo::Interface::InStream
- SpecificDataFormat class (Table66): change type name of ¡¥rla¡¦ attribute
to RoLo::Architecture::DataSpecification.
- StreamAbility class (Table 124 and Figure 22): change type names of
¡¥rlaList¡¦ and ¡¥dfList¡¦ attributes to RoLo::Architecture::Data and
RoLo::DataFormat::DataFormat respectively. OutStream class (Table 135 and Figure 22): change type name of
‘getResult’ method’s argument ‘value’ as RoLo::Architecture::Data.
- InStream class (Table 136 and Figure 22): change type name of
‘setData’ method’s argument ‘value’ as RoLo::Architecture::Data.
- RoLo Service class (Table 140 and Figure 22): change type name of
‘adjust’ argument ‘data’ as RoLo::Architecture::Data.
Modify C++ PSM as following:
- Add header files, Architecture.hpp and Interface.hpp, that represent
Architecture package and Interface package respectively as following:
// $Id: Architecture.hpp,v 1.3 2009/06/20 06:18:42 nishio Exp $
#pragma once
#include <RLS/RelativeCRS.hpp>
#include <RLS/MobileCRS.hpp>
#include <RLS/MobileOperation.hpp>
#include <RLS/Identity.hpp>
#include <RLS/ErrorType.hpp>
#include <RLS/Error.hpp>
#include <RLS/RoLoArchitecture.hpp>
#include <RLS/RoLoDataOperation.hpp>
// $Id: Interface.hpp,v 1.2 2009/06/20 06:18:43 nishio Exp $
#pragma once
#include <RLS/Ability.hpp>
#include <RLS/Service.hpp>
- Reorder header files in C++ PSM so that files are ordered based on the
packages they belong to. The files will be in the following order:
Returncode_t.hpp, Architecture.hpp, RelativeCRS.hpp, MobileCRS.hpp,
MobileOperation.hpp, Identity.hpp, ErrorType.hpp, Error.hpp,
RoLoArchitecture.hpp, RoLoDataOperation.hpp, DataFormat.hpp,
Interface.hpp, Ability.hpp, Service.hpp
- Move definitions in Error.hpp, ErrorBase.hpp, ErrorType.hpp,
Identity.hpp, MobileCRS.hpp, MobileOperation.hpp, RelativeCRS.hpp,
RoLoArchitecture.hpp and RoLoDataOperation.hpp into a namespace
“::RoLo::Architecture”.
Note: ErrorBase.hpp is added in issue 13185
- Move definitions in DataFormat.hpp into a namespace
“::RoLo::DataFormat”.
- Move definitions in Ability.hpp and Service.hpp into a namespace
“::RoLo::Interface”.
- Modify type of rla in SpecificDataFormat class (DataFormat.hpp)
as ::RoLo::Architecture::DataSpecification.
- Modify type of ‘inStream’ and argument ‘inStream’ of getInStream()
method in MobileDatum class (MobileCRS.hpp)
as ::RoLo::Interface::InStream. Modify type of ‘dfList’ as ::RoLo::DataFormat::DataFormat and type of
‘rlaList’ as ::RoLo::Architecture::DataSpecification (Service.hpp).
- Modify type of ‘value’ argument of setData() method in InStream
class (Service.hpp) as ::RoLo::Architecture::Data.
- Modify type of ‘value’ argument of getResult() method in OutStream
class (Service.hpp) as ::RoLo::Architecture::Data.
- Modify type of ‘data’ argument of adjust() method in RoLo Service class
(Service.hpp) as ::RoLo::Architecture::Data
Disposition: Resolved
Revised Text:
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Discussion: Reorganize the classes into three sub packages, Architecture package, Data
Format package, and Interface package. Architecture package includes classes
defined in section 7.3, Data Format package includes classes defined in section
7.4 and Interface package includes classes defined in 7.5.
Also, rename the titles of section 7.3 to 7.5 in order to make it clear which
package each class belongs to.
Issue 12917: Sequence diagrams should be added (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: In RLS specification, there is no sequence diagram.
I think it is difficult to understand behavior.
So I think it would be better if sequence diagrams were added.
Resolution: Add non-mandatory sequence diagram and accompanying description at the last
part of section 7.5, in order to show typical RoLo service usage.
Revised Text: Add the following description at the end of section 7.5. Using RoLo Service
Here we show several non-mandatory steps and sequence diagrams as examples. Typical steps of
using RoLo Services can be listed as following:
1. (optional) Obtain ability description by calling 'getAbility' method toward RoLo servic
e. An ability description obtained from RoLo service also includes descriptions on its s
treams. This step can be omitted if users already have sufficient information such by r
eading reference manuals.
2. (optional) Set up service and/or stream parameters through calling 'setParameterValues'
method. If the default settings are sufficient or if there exists no parameter to be config
ured, this step can be omitted. In complicated cases, users may need to repeatedly call
'setParameterValues' and 'getParameterValues' to set and to confirm parameter change
s.
3. Establish connection.
4. (optional) Set up initial position data by calling 'adjust' method with necessary data.
5. Perform data passing.
6. (optional) Occasionally, perform adjustment if necessary. Adjustment is an act to provid
e auxiliary information to the target module for improving the localization process..
7. Disconnect the connection.
Figure 22 to 26 show sequences of typical steps on using RoLo service. Note that in step 3, con
nection establishment can be initiated from two side; either from the service that outputs data
(OUT service) or from the service that accepts data inputs (IN service). Figure 23 and Figure 2
4 show typical connection sequences in both cases. Note that, disconnection of the established
connection (step 7) can be performed from both sides regardless of which side initiated the con
nection (Figure 26). As can be seen from Figure 23, Figure 24 and Table 93, 'connect' method of RoLo service has t
wo forms. The first is for establishing connection from OUT service to IN service (OUT servic
e initiates connection), and another is for the opposite where IN service initiates connection. As
RoLo services may have multiple input streams of different natures, when connecting from O
UT service to IN service the stream to be connected shall be specified. Thus, the first form of 'c
onnect' method has an additional 'target' argument.
Another factor that needs consideration is the type of data passing. In this specification, two da
ta passing types are provided as elements of StreamType enumeration: PUSH mode (OUT side
triggers data passing) and PULL mode (IN side triggers data passing). For example, most GPS
receivers output data in PUSH mode, that is, measurement results are outputted continuously in
some frequency. These two types of data passing can be performed regardless of which side in
itiates connection, as far as both modules have the ability to perform data passing in the specifi
ed type. Figure 25 shows typical steps for performing data passing for the two directions. As ca
n be seen from the sequence, in PULL mode, the IN service triggers data passing by calling 'ge
tData' method. And in PUSH mode, the OUT service triggers data passing by 'setData' method.
PUSH type data passing can also be understood as a callback from OUT side to IN side. Thus,
when using PUSH mode and when connection is established from IN side, the 'source' argume
nt cannot be omitted. Without this, the RoLo output stream on OUT side cannot know where to
make callbacks for data passing. However, when connection is established from OUT side, thi
s 'source' argument is not required for the sake of making callbacks, as the RoLo input stream i
s given back as an 'inStream' argument.
Disposition: Resolved
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12918: Separate XMI definition from RLS specification document. (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: The division of documents
The RLS Specification FTF Beta1(dtc/2008-07-01) includes XMI definition(Annex B).
I think that the XMI definition had better separate from RLS specification document.
Resolution: Separate XMI from the main RLS specification document
Revised Text: Remove Appendix B (Sample XMI) from the specification document and provide
the XMI as separate machine-readable file.
Actions taken:
September 24, 2008: received issue
October 15, 2009: received issue
Discussion:
Issue 12919: Figure 3 - Relative and Mobile c oordinate reference system (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: The association between Relativedatum and ReletiveCRS should be reverse.
-In RelativeDatum class, attributes(rla,base) should replace to derived attributes.
-In MobileDatum class, attributes(inStream) should replace to derived attributes
Resolution: Fix Figure 3 according to the suggestions. However, for the second and the third
items, follow the resolution in issue 12920.
Revised Text: In Figure 3 (Relative and Mobile coordinate reference system), modify UML
diagram as follows:
– Reverse the direction of association between RelativeDatum and
RelativeCRS.
Disposition: Resolved
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12920: The navigability notation(arrow head) should be added as follows (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -RoLo Architecture class side of association between RelativeDatum and RoLo Architecture.
-RoLo Data class side of association between RelativeDatum and RoLo Data.
-InStream class side of association between MobileDatum and inStream.
Resolution: Fix Figure 3 according to the suggestions. However, for the first item, it is better
to make this an aggregation. For the second and third items on the associations
RelativeDatum - RoLo Data and MobileDatum – InStream,it is better to make
these relations as compositions.
Revised Text: In Figure 3, make the following modifications:
- Make the association between RoLo Architecture and RelativeDatum as
aggregation on RelativeDatum side with association end name ‘rla’ and
with multiplicity [0..1]
- Make the association between RelativeDatum and RoLo Data as
composition on RelativeDatum side with association end name ‘base’ and
with multiplicity [0..1]
- Make the association between MobileDatum and InStream as composition
on MobileDatum side with association end name ‘inStream’
- Change the C++ PSM definition of ‘rla’ attribute in RelativeDatum class
(RelativeDatum.hpp) as a pointer
- Change the C++ PSM definition of ‘inStream’ attribute in MobileDatum
class (MobileCRS.hpp) as a non-pointer variable
- Change the C++ PSM definition of the argument ‘inStream’ of
getInStream() method in MobileDatum class (MobileCRS.hpp) as single
pointer
Disposition: Resolved
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12921: The association end name should be added as follows (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -rla : RoLo Architecture class side of association between RelativeDatum and RoLo Architecture.
-base : RoLo Data class side of association between RelativeDatum and RoLo Data.
-inStream : InStream class side of association between MobileDatum and inStream.
Resolution: See issue 12920 for disposition
Revised Text:
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12922: add an explanatory note (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: Because the following association is meanings to limit relations defined in superior,
we should add an explanatory note.
Relativedatum - ReletiveCRS
RelativeCartesianCRS - SC_CertesianCS
RelativePolarCRS - SC_PolarCS
MobileCRS - MobileDatum
MobileCartesianCRS - SC_CertesianCS
MobilePolarCRS - SC_PolarCS
Resolution: Fix Figure 3 and relevant tables to show that these attributes derived from parent
classes are restricted in these child classes.
Revised Text: Add a note to Table 2 as following:
Note: Values for the attribute ‘usesDatum’ which is derived from parent class shall
be limited to instances of StaticRelativeDatum or its inherited classes.
• Add a note to Table 3 as following:
Note: Values for the attribute ‘usesCS’ which is derived from parent class shall be
limited to instances of SC_CartesianCS [ISO19111] or its inherited classes.
• Add a note to Table 4 as following:
Note: Values for the attribute ‘usesCS’ which is derived from parent class shall be
limited to instances of SC_PolarCS [ISO19111] or its inherited classes.
• Add a note to Table 6 as following:
Note: Values for the attribute ‘usesDatum’ which is derived from parent class shall
be limited to instances of MobileDatum or its inherited classes.
• Add a note to Table 7 as following: limited to instances of SC_CartesianCS [ISO19111] or its inherited classes.
• Add a note to Table 8 as following:
Note: Values for the attribute ‘usesCS’ which is derived from parent class shall be
limited to instances of SC_PolarCS [ISO19111] or its inherited classes.
• In Figure 3, add subset description to the following relations:
- StaticRelativeCRS - StaticRelativeDatum (subset usesDatum)
- MobileCRS - MobileDatum (subst usesDatum)
- StaticRelativeCartesianCRS - ::ISO19111::SC_CartesianCRS (subset
usesCS)
- StaticRelativePolarCRS - ::ISO19111::SC_PolarCRS (subset usesCS)
- MobileRelativeCartesianCRS - ::ISO19111::SC_CartesianCRS (subset
usesCS)
- MobileRelativePolarCRS - ::ISO19111::SC_PolarCRS (subset usesCS)
Disposition: Resolved
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12923: Table 5 RelativeDatum class (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -The last of Description is over in "Holds a RoLo Data that".
Is there no sentence after that?
Resolution: Fix the description line by deleting the sentence in Table 5 beginning by "Holds a
RoLo Data that".
Revised Text: Modify the description of Table 5 (RelativeDatum) as follows:
Datum for static relative coordinate reference system.
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12924: Figure 4 - Mobile CRS operations (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: The relationship between Mobile2Static Operation and ::ISO 19111::SC_CRS should replace
to association(solod line).
-The relationship between Mobile2Static Operation and MobileCRS should replace to
association(solod line).
-The relationship between Static2Mobile Operation and ::ISO 19111::SC_CRS should replace to
association(solod line).
-The relationship between Static2Mobile Operation and MobileCRS should replace to
association(solod line).
-In the relationship between Mobile2Static Operation and ::ISO 19111::SC_CRS,
the association end name "+source" (::ISO 19111::SC_CR side) should replace "+target".
-In the relationship between Static2Mobile Operation and MobileCRS,
the association end name "+target" (MobileCRS) should replace "+source".
-The another directed association should added between MobileCRS and Mobile2Mobile Operation.
(The direction is from Mobile2Mobile Operation to MobileCRS.)
And the navigability notation(arrow head) and association end name(+source,+target)
should be added to MobileCRS class side.
-In Mobile2Static Operation class, derived attributes(+source,+target) should added.
-In Static2Mobile Operation class, derived attributes(+source,+target) should added.
-In Mobile2Mobile Operation class, attributes(+source,+target) should replace to
derived attributes.
Resolution: Fix Figure 4 according to the suggestions in the summary
Revised Text: Modify figure 4 as following:
„O Correct the relationship between Mobile2Static Operation and ::ISO
19111::SC_CRS to show a derived attribute of Mobile2Static Operation class
"target".
„O Correct the relationship between Mobile2Static Operation and MobileCRS to
show a derived attribute of Mobile2Static Operation class ¡§source¡¨.
„O Correct the relationship between Static2Mobile Operation and ::ISO
19111::SC_CRS to show a derived attribute of Mobile2Static Operation class
"source".
„O Correct the relationship between Static2Mobile Operation and MobileCRS to
show a derived attribute of Mobile2Static Operation class ¡§target¡¨.
„O Draw two associations between Mobile2Mobile Operation and MobileCRS to
indicate the ¡¥source¡¦ and ¡¥target¡¦ attributes of Mobile2Mobile Operation class.
Disposition: Resolved
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12925: Figure 5 - Identity information (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -In RoLo Symbolic Position class, attributes(direct,indirect) should replace to derived attributes.
-The navigability notation(arrow head) should be added as follows:
-SymbolicCRS class side of association between DirectSymbol and SymbolicCRS.
-DirectSymbol class side of association between DirectSymbol and RoLo Symbolic Position.
-RoLo Symbol Ref class side of association between DirectSymbol and RoLo Symbolic Ref.
-DirectSymbol class side of association between DirectSymbol and RoLo Symbolic Ref.
-DirectSymbol class side of association between DirectSymbol and RoLo Symbolic Position.
-RoLo Symbol Ref class side of association between RoLo Symbol Ref and RoLo Symbolic Position.
-The association end name should be added as follows:
-direct : DirectSymbol class side of association between DirectSymbol and RoLo Symbolic Position.
-indirect : RoLo Symbol Ref class side of association between RoLo Symbol Ref and
RoLo Symbolic Position.
-Because the following association is meanings to limit relations defined in superior,
we should add an explanatory note.
NumericIdentityCS - NumericIdentityCRS
SymbolicCS - SymbolicCRS
IdentityCRS - IdentityDatum
Resolution: Fix Figure 5 according to the suggestions. However, the associations RoLo
Symbolic Position – DirectSymbol and RoLo Symbolic Position – RoLo Symbol
Ref shall be compositions both on RoLo Symbolic Position side. For the
associations SymbolicCRS – DirectSymbol and DirectSymbol – RoLo Symbolic
Ref, it is better to make them aggregations that indicate relevant attributes.
Revised Text: Modify figure 5 as follows:
¡V Make the association between DirectSymbol and RoLo Symbolic Position
as an composition on RoLo Symbolic Position side and add association
end name ¡¥direct¡¦ with multiplicity [0..1]
¡V Make the association between RoLo Symbol Ref and RoLo Symbolic
Position as an composition on RoLo Symbolic Position side and add
association end name ¡¥indirect with multiplicity [0..1]
¡V Make the association between DirectSymbol and SymbolicCRS as an
aggregation on DirectSymbol side and add association end name
¡¥coordinateReferenceSystem¡¦ with multiplicity [0..1] (also by issue 13998)
¡V Make the association between DirectSymbol and RoLo Symbolic Ref as
an aggregation on RoLo Symbol Ref side and add association end name
¡¥point¡¦ (also by issue 13998)
¡V Add subset description for the following relations :
ƒÞ SymbolicCRS - SymbolicCS (subset usesCS)
ƒÞ NumericIdentityCRS - NumericIdentityCS (subset usesCS)
ƒÞ IdentityCRS - IdentityDatum (subset usesDatum)
Add the following explanatory note to Table 18:
Note: Values for the attribute ¡¥usesDatum¡¦ which is derived from parent class shall
be limited to instances of IdentityDatum or its inherited classes.
Add the following explanatory note to Table 19:
Note: Values for the attribute ¡¥usesCS¡¦ which is derived from parent class shall be
limited to instances of NumericIdentityCS or its inherited classes.
Add the following explanatory note to Table 20:
Note: Values for the attribute ¡¥usesCS¡¦ which is derived from parent class shall be
limited to instances of SymbolicIdentityCS or its inherited classes.
Disposition: Resolved
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12926: 7.3.2 Identity Information (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: I think that explanation may be added about attributes(direct,indirect)
in RoLo Symbol Position class.
Resolution: Add the following sentence in "Description" of Table 23:
This class is used as a data holder for accessing symbolic information transparently,
whether it is directly held or indirectly referenced.
Revised Text:
Actions taken:
September 24, 2008: received issue
October 15, 2009: resolved, close issue
Issue 12927: Figure 6 - Error Type (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -In RoLo Error Type Operation class, derived attributes(+source,+target) should added
Resolution: withdrawn by submitter
Revised Text:
Actions taken:
September 24, 2008: received issue
October 28, 2008: closed issue
Issue 12928: Figure 7 - Error information (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: The constraint {ordered} should be added to attribute(particles) in Particle Set class.
-The multiplicity [0..1] should be added to attribute(errprType) in RoLo Error class.
-The navigability notation(arrow head) should be added as follows:
-Particle class side of association between Particle Set and Particle.
-Probability class side of association between Particle and Probability.
-Weighted Distribution class side of association between Linear Mixture Model
and Weighted Distribution.
-Covariance Matrix class side of association between Gaussian and Covariance Matrix.
Resolution: Fix Figure 7 according to the suggestions. However, as Particle class and
‘particles’ attribute in Particle Set class is removed in issue 13185, items in
summary related to these will not be revised. Also, for the relations Weighted
Distribution – Linear Mixture Model and Covariance Matrix – Gaussian are better
to be shown as compositions
Revised Text: Modify figure 7 as follows:
– Add multiplicity ‘[0..1]’ for ‘errorType’ attribute in RoLo Error class
– Remove ‘dists’ attribute from Linear Mixture Model class and instead,
make the association between Linear Mixture Model and Weighted
Distribution to show composition on Linear Mixture Model side and show
association end name ‘dists’ with multiplicity [1..*] and indicate that this is
ordered.
– Remove ‘covariance’ attribute from Gaussian class and instead, make the
association between Gaussian and Covariance Matrix to show composition on Gaussian side and show association end name
‘covariance’.
Disposition: Resolved
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12929: Figure 9 - RoLo Architecture (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -In RoLo Element class, attributes(spec) should replace to attributes.
-In RoLo Architecture class, attributes(spec) should rename to elemSpec, and should replace
to derived attribute.
-In RoLo Position class, attributes(symbolic,numeric) should replace to derived attributes.
-In Error Element class, attributes(err) should replace to derived attributes.
-In RoLo Data class, attributes(elem) should replace to derived attributes.
-The association end name should be added as follows:
-symbolic : RoLo Symbolic Position class side of association between RoLo Position and
RoLo Symbolic Position.
-numeric : ::ISO_19107::GM_Position class side of association between RoLo Position and
::ISO_19107::GM_Position.
-err : RoLo Error class side of association between Error Element and RoLo Error.
-elem : RoLo Element class side of association between RoLo Data and RoLo Element.
-The navigability notation(arrow head) should be added as follows:
-GM_Position class side of association between RoLo Position and GM_Position.
-RoLo Symbolic Position class side of association between RoLo Position and
RoLo Symbolic Position.
-RoLo Position class side of association between Position Element and RoLo Position.
-RoLo Error class side of association between Position Element and RoLo Error.
-RoLo Error class side of association between Error Element and RoLo Error.
-RoLo Element class side of association between RoLo Data and RoLo Element.
-CS_CoordinateSystem class side of association between RoLo Element Specification
and CS_CoordinateSystem.
-SC_CRS class side of association between RoLo Element Specification and SC_CRS.
-RoLo Error Type class side of association between RoLo Element Specification and RoLo Error Type.
-RoLo Error Type class side of association between Error Element Specification
and RoLo Error Type.
-RoLo Element Specification class side of association between RoLo Architecture
and RoLo Element Specification.
-RoLo Architecture class side of association between RoLo Data and RoLo Architecture.
-Because the following association is meanings to limit relations defined in superior,
we should add an explanatory note.
Position Element Specification - Position Element
Error Element Specification - Error Element
Resolution: Apply the necessary modifications to Figure 9 as specified. However, the
following relations are compositions and do not show derived attributes:
RoLo Position – ::ISO19107::GM_Position
RoLo Position – RoLo Symbolic Position
Position Element – RoLo Position
Position Element – RoLo Error
Error Element – RoLo Error
RoLo Data – RoLo Element
RoLo Element – RoLo Element Specification
Thus, for these relations, replace the associations to indicate composition with
corresponding attribute names.
Revised Text: Revised Text:
Modify Figure 9 as follows:
¡V Make the following associations to show attributes as aggregations:
ƒÞ RoLo Data ¡V RoLo Architecture for ¡¥rla¡¦ in RoLo Data
ƒÞ Position Element Specification ¡V RoLo Error Type for ¡¥errType¡¦ in
Position Element Specification
ƒÞ Error Element Specification ¡V RoLo Error Type for ¡¥errType¡¦ in Error
Element Specification
ƒÞ RoLo Architecture ¡V RoLo Element Specification for ¡¥elem¡¦ in RoLo
Architecture. This attribute shall be renamed to ¡¥elemSpec¡¦.
ƒÞ Position Element Specification ¡V SC_CRS for ¡¥crs¡¦ in Position
Element Specification
¡V Make the following associations as aggregations and add subset
description for the following associations:
ƒÞ Position Element ¡V Position Element Specification on Position
Element Specification side. Subset description shall be ¡¥subset
spec¡¦.
ƒÞ Error Element ¡V Error Element Specification on Error Element
Specification side. Subset description shall be ¡¥subset spec¡¦.
¡V Make the following associations to show compositions for
corresponding attributes:
ƒÞ RoLo Element ¡V RoLo Element Specification for ¡¥spec¡¦ in RoLo
Element
ƒÞ RoLo Data ¡V RoLo Element for ¡¥elem¡¦ in RoLo Data
ƒÞ Error Element ¡V RoLo Error for ¡¥err¡¦ in Error Element
ƒÞ Position Element ¡V RoLo Error for ¡¥err¡¦ in Position Element
ƒÞ Position Element ¡V RoLo Position for ¡¥pos¡¦ in Position Element
ƒÞ RoLo Position - ::ISO19107::GM_Position for ¡¥numeric¡¦ in RoLo
Position
ƒÞ RoLo Position ¡V RoLo Symbolic Position for ¡¥symbolic¡¦ in RoLo
Position
Add the following explanatory note to Table 51 (Position Element):
Note: Values for the attribute ¡¥spec¡¦ which is derived from parent class shall be
limited to instances of PositionElementSpecification or its inherited classes.
Add the following explanatory note to Table 52 (Error Element):
Note: Values for the attribute ¡¥spec¡¦ which is derived from parent class shall be
limited to instances of ErrorElementSpecification or its inherited classes.
Modify C++ PSM as following: – Change ‘crs’ in Position Element Specification class to be a pointer
Disposition: Resolved
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12930: Figure 10 - RoLo Data Operation (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -In Position Element Operation class, derived attributes(+source,+target) should added.
-In RoLo Data Operation class, derived attributes(+source,+target) should added.
-The relationship between Position Element Operation and Position Element Specification
should replace to association(solod line).
-In Position Element Single Operation class, attributes(usesOperation,UsesErrorTypeOperation)
should replace to derived attribute.
-The navigability notation(arrow head) should be added as follows:
-CC_SingleOperation class side of association between Position Element Single Operation
and CC_SingleOperation.
-RoLo Error Type Operation class side of association between Position Element Single Operation
and RoLo Error Type Operation.
Resolution: Correct the errata of UML diagram in Figure 10 (RoLo Data Operation) according
to the suggestions in the summary except for the first two issues, as they are
already in the diagram
Revised Text: Modify Figure 10 (RoLo Data Operation) as follows:
– Replace the relationships between Position Element Operation and
Position Element Specification by association (solid line).
– Replace 'usesOperation' and 'usesErrorTypeOperation' attributes of
Position Element Single Operation class to be derived attributes and
change the associations between relevant classes as aggregations.
Add relevant attribute names as association end names. As for
‘usesErrorTypeOperation’, also show multiplicity [0..1].
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12931: Figure 11 - RoLo Data format (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -Because the following association is meanings to limit relations defined in superior,
we should add an explanatory note.
User Defined Data Format - RoLo Architecture
Resolution: Actually, this association is required between Specific Data Format and RoLo
Architecture, and not between User Defined Data Format and RoLo Architecture.
This association is for the ‘rla’ attribute of Specific Data Format class. Thus,
instead of adding note for limiting this association, we shall better correct the
association to be between Specific Data Format and RoLo Architecture that
shows the derived attribute ‘rla’ of Specific Data Format class. The new
association is also reported in issue 13998.
Revised Text: Modify Figure 11 as follows:
- Remove the association between User Defined Data Format and RoLo
Architecture
- Add an association between Specific Data Format and RoLo
Architecture as aggregation which indicates the derived attribute ‘rla’ in
Specific Data Format class
Disposition: Resolved
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12932: 7.4.1 Common data format TypeI.Cartesian coordinate system (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: In Table 69 - Common data format type I, the Unit of Orientation is "degree",
but the explanation of Orientation is "The unit of this parameter is radians.".
Which is correct?
Resolution: See issue 13120 for disposition
Revised Text:
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12933: 7.4.1 Common data format TypeII.Polar coordinate system (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -In Table 70 - Common data format type II, the Unit of Position is only "radian".
It should replace "meter,radian".
-The explanation of Position is "The unit of each value is radian.".
It should replace "The unit of r is meter and that of alpha and beta is radian.".
-Table 70 - Common data format type II, the Unit of Orientation is "degree",
but the explanation of Orientation is "The unit of this parameter is radians.".
Which is correct?
Resolution: See issue 13120 for disposition
Revised Text:
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12934: Figure 21 - RoLo Ability and RoLo Parameter Set (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -In RoLo Attribute class, attributes(def,value) should replace to derived attribute.
-In RoLo Attribute ValueList class, attributes(value) should replace to derived attribute.
-In RoLo Attribute Definition Set class, the multiplicity of attribute(attribute)
should replace to [1..*].
-The navigability notation(arrow head) should be added as follows:
-one side of self association at RoLo Attribute Set class.
-one side of self association at RoLo Attribute Definition Set class.
-RoLo Attribute class side of association between RoLo Attribute Set and RoLo Attribute.
-RoLo Attribute Definition Set class side of association between RoLo Attribute Set
and RoLo Attribute Definition Set.
-RoLo Attribute Definition class side of association between RoLo Attribute Definition Set
and RoLo Attribute Definition.
-RoLo Parameter Set class side of association between RoLo Ability and RoLo Parameter Set.
-RoLo Attribute Definition class side of association between RoLo Attribute
and RoLo Attribute Definition.
-RoLo Attribute Value class side of association between RoLo Attribute and RoLo Attribute Value.
-RoLo Attribute SingleValue class side of association between RoLo Attribute ValueList
and RoLo Attribute SingleValue.
-RoLo Attribute Type class side of association between RoLo Attribute Definition
and RoLo Attribute Type.
-There are associations as follows but there is no attribute in C++ PSM.
-between Stream Parameter Set and Stream Parameter Definition
-between Stream Ability and Stream Parameter Definition
-between Stream Ability and Service Ability
-between Stream Ability and Service Parameter Definition
-between Service Parameter Set and Service Parameter Definition
Resolution:
Revised Text:
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed no change
Discussion: As RoLo ability and its related classes are restructured in issue #13322, these
modifications are not necessary any more.
Disposition: Closed, no change
Issue 12935: Figure 22 - RoLo Service (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -In RoLo Streame class, attributes(ability,parameterSet) should replace to derived attribute.
-In RoLo Streame class, the type of augument(ability) at getAbility()
should replace to "Stream Ability".
-In RoLo Streame class, the type of augument(parameter) at getParameter()
should replace to "Stream Parameter Set".
-In RoLo Streame class, the type of augument(parameter) at setParameter()
should replace to "Stream Parameter Set".
-In Stream Parameter Set class, the attribute(dataFormat) should rename to "df".
-In Stream Ability class, the type of attribute(dfList) should replace to "RoLo Data Format".
-The navigability notation(arrow head) should be added as follows:
-Stream Parameter Set class side of association between RoLo Stream and Stream Parameter Set.
-StreamType class side of association between Stream Parameter Set and StreamType.
-StreamType class side of association between Stream Ability and StreamType.
-Stream Ability class side of association between RoLo Stream and Stream Ability.
-Service Ability class side of association between RoLo Service and Service Ability.
Resolution: As RoLo service and its related classes are restructured in issue #13439, these
modifications are not necessary any more. Disposition: Closed, no change
Revised Text:
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12936: Table 132 - Stream Parameter Set class (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: In Derived From section, "RoLo Parameter" should replace "RoLo Parameter Set".
-In obligation section of attribute df, "M" should replace "O".
-In obligation section of attribute streamType, "M" should replace "O".
-In Condition section, the explanation is "If the RoLo Architecture can be identified by
the specified data format, the rlaList attribute may not necessary be specified explicitly.",
what is "rlaList attribute"?
Resolution:
Revised Text:
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed no change
Discussion: As Stream Parameter and its related classes are restructured in issue #13439,
these modifications are not necessary any more.
Disposition: Closed, no change
Issue 12937: Table 133 - Stream Ability class (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: In obligation section of attribute dfList, "M" should replace "O".
-In obligation section of streamTypeList, "M" should replace "O".
Resolution:
Revised Text:
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed no change
Discussion: As Stream Ability and its related classes are restructured in issue #13439, these
modifications are not necessary any more.
Disposition: Closed, no change
Issue 12938: Table 134 - RoLo Stream class (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -In type section of argument parameter of setParameter(),"Stream Parameter"
should replace "Stream Parameter Set".
-In type section of argument parameter of getParameter(),"Stream Parameter"
should replace "Stream Parameter Set".
Resolution:
Revised Text:
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed no change
Discussion: As Stream Parameter Set and its related classes are restructured in issue
#13439, these modifications are not necessary any more.
Disposition: Closed, no change
Issue 12939: Sample C++ Header (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -In header definition(P.58), "#include <vector>" should added.
-RoLoAttributeValueSet class(P.58)
-The class name "RoLoAttributeValueSet" should replace "RoLoAttributeValueList".
-The type of value "std::set<RoLoAttributeSingleValue>"
should replace "std::vector<RoLoAttributeSingleValue>".
-RoLoAttributeType(P.59)
-ISO19111::IO_IdentifiedObject should added as super class.
-UniformGaussian(P.62)
-The super class "RoLoErrorType" should replace "ErrorDistribution".
-DirectSymbol(P.65)
-The attribute "dimension:Integer" should be added.
-ErrorElementSpecification(P.69)
-The attribute "errType:RoLo Error Type" should be added.
-StreamParameterSet(P.72)
-In super class "RLS::RoLoAbility", "RLS::" should be deleted.
-StreamAbility(P.72)
-In super class "RLS::RoLoAbility", "RLS::" should be deleted.
Resolution: All suggestions except the one for ErrorElementSpecification is now unnecessary
due to changes made in other issues.
Revised Text: Modify C++ PSM as follows:
- Add public class variable ‘errType’ whose type is ‘RoLoErrorType*’
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 12940: Typos (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: -In Table12, "Fixed2Mobile Operation class" should rename "Static2Mobile Operation class".
-In Table18, Derived From section, "CD_Datum [ISO19111]" should replace "CS_SingleCRS [ISO19111]".
-In Table 21, occurrence section of "coordinate", "1" should replace "N ord".
-In Table 36, occurrence section of "values", "1" should replace "N ord".
-In Table 54, occurrence section of "elemSpec", "N" should replace "N ord".
-In Table 55, occurrence section of "elem", "N" should replace "N ord".
-In Table 60, Derived From section, "RoLo Architecture Operation"
should replace "RoLo Data Operation".
-In Table 60, attribute type section of "childOperation", "RoLo Architecture Operation"
should replace "RoLo Data Operation".
-In Table 60, description of "childOperation", "Ordered list of RoLo Architecture Operation
to be appied." should replace "Ordered list of RoLo Data Operation to be applied."
-In Table 61, Derived From section, "RoLo Architecture Operation"
should replace "RoLo Data Operation".
-In Table 62, Derived From section, "RoLo Architecture Single Operation"
should replace "RoLo Data Single Operation".
-In Table 63, Derived From section, "RoLo Architecture Single Operation"
should replace "RoLo Data Single Operation".
-In Table 65, Derived From section, "IO_IdentifiedObject" should replace "RoLo Data Format".
-In Table 125, "ord" of value, "M ord, N" should replace "M , N ord".
-In Table 134, parameter type section of the argument "parameter" in getParameter(),
"RoLo Parameter" should replace "Stream Parameter Set".
-In Table 134, parameter type section of the argument "parameter" in setParameter(),
"RoLo Parameter" should replace "Stream Parameter Set".
-In Table 135, parameter type section of the argument "value" in getResult(),
"RoLo Data Set" should replace "RoLo Data".
-In Table 136, parameter type section of the argument "value" in setResult(),
"RoLo Data Set" should replace "RoLo Data".
Resolution: Correct the typos according to the suggestions in summary of this issue.
However, the typos in Table 125 and 134 does not exist any more due to
changes by issue 13322. Therefore, we simply discard these two items.
Revised Text: Modify the title of Table 12 to be "Static2MobileOperation class".
Modify the Derived From of Table 18 to be "SC_SingleCRS [ISO19111]".
Modify the occurrence of 'coordinate' attribute in Table 21 to be "N ord".
Modify the occurrence of 'values' attribute in Table 36 to be "N ord".
Modify the occurrence of 'elemSpec' attribute in Table 54 to be "N ord".
Modify the occurrence of 'elem' attribute in Table 55 to be "N ord".
Modify Table 60 (RoLo Data Concatenated Operation) as follows:
– Modify the Derived From to be "DataOperation".
– Modify the type of 'childOperation' attribute to be "DataOperation".
– Modify description for the 'childOperation' attribute to be the following:
An ordered list of RoLo data operation to be applied. Target RoLo data
specification and source RoLo data specification for succeeding operations shall
match.
Modify the Derived From of Table 61 to be "DataOperation".
Modify the Derived From of Table 62 to be "DataSingleOperation".
Modify the Derived From of Table 63 to be "DataSingleOperation".
Modify the Derived From of Table 65 to be "DataFormat".
Modify the type of argument 'value' of 'getResult' operation in Table 135 to be
"Data".
Modify the type of argument 'value' of 'setData' operation in Table 136 to be
"Data".
Disposition: Resolved
Actions taken:
September 24, 2008: received issue
October 15, 2009: closed issue
Issue 13106: Relative/Mobile CRS name is confusing (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: Although the relative/mobile CRS are defined as 'relative' and
'mobile' in Figure 3, operations are defined as 'static' and 'mobile'
in Figure 4. The names are confusing, and shall reflect the
static/dynamic nature of the coordinate systems.
Resolution: Introduce new classes (RelativeCRS, RelativeDatum, DynamicRelativeCRS and
DynamicRelativeDatum) to reflect the static/dynamic nature of the relative
coordinate systems separately. RelativeCRS is renamed as StaticRelativeCRS,
and names of related classes are changed in a similar manner. Make MobileCRS
to be derived from DynamicRelativeCRS.
Based on this change, Mobile CRS is a special case of Relative CRS. Therefore,
change the title of section 7.3.1 Relative and Mobile Coordinate Reference
Systems to be 'Relative Coordinate Reference Systems' to avoid confusion.
Revised Text: Change the title of section 7.3.1 Relative and Mobile Coordinate Reference
Systems to be 'Relative Coordinate Reference Systems'.
Change description at the first and second paragraph of section 7.3.1 as
following:
In this section, relative coordinate reference systems are defined which may lack
fixed relation with the earth or users have no interest in referencing them to other
coordinate reference systems. We categorize relative coordinate reference systems in
two types, static and dynamic. A coordinate reference system on mobile platforms,
mobile coordinate reference system, is defined as a dynamic relative coordinate
reference system. That is, the relation with other coordinate reference systems may
change by time.
The GIS standard on spatial reference system [ISO19111] allows the definition and
use of such relative and mobile coordinate reference systems. However, there is no
specific model or description on these systems. As these systems are quite commonly
used in the field of robotics, here we explicitly define structures and operations
specific to these coordinate reference systems. Although here we only define coordinate reference systems based on two coordinate systems of frequent usage,
SC_CartesianCS and SC_PolarCS, users may define derivatives of relative or mobile
coordinate reference system based on the coordinate system of their interest.
Add the following note to Table 2:
Note: Values for the attribute ‘usesDatum’ which is derived from parent class shall
be limited to instances of RelativeDatum or its inherited classes.
Add the following new tables after Table 2:
Table 3 - RelativeDatum class
Description: Datum for relative coordinate reference systems.
Derived From: CD_EngineeringDatum [ISO19111]
Table 4 - StaticRelativeCRS class
Description: Abstract class for representing relative coordinate reference systems that have static relation with other CRS(s).
Derived From: RelativeCRS
Note: Values for the attribute ‘usesDatum’ which is derived from parent class shall be limited to instances of StaticRelativeDatum or its i
nherited classes.
• Add the following new tables after Table 5:
Table 8 - DynamicRelativeCRS class
Description: Abstract base class for representing dynamic relative coordinate reference systems.
Derived From: RelativeCRS
Note: Values for the attribute ‘usesDatum’ which is derived from parent class shall be limited to instances of DynamicRelativeDatum or i
ts inherited classes.
Table 9 - DynamicRelativeDatum class
Description: Datum for dynamic relative coordinate reference system.
Derived From: RelativeDatum
• Modify Table 3 to 5 as following:
Table 5 - StaticRelativeCartesianCRS class
Description: Static relative coordinate reference systems based on Cartesian coordinate system.
Derived From: StaticRelativeCRS
Note: Values for the attribute ‘usesCS’ which is derived from parent class shall be limited to instances of SC_CartesianCS [ISO19111] o
r its inherited classes.
Table 6 - StaticRelativePolarCRS class Description: Static relative coordinate reference system based on polar coordinate system.
Derived From: StaticRelativeCRS
Note: Values for the attribute ‘usesCS’ which is derived from parent class shall be limited to instances of SC_PolarCS [ISO19111] or its i
nherited classes.
Table 7 - StaticRelativeDatum class
Description: Datum for static relative coordinate reference system.
Derived From: RelativeDatum
Attributes
dataSpec DataSpecification O 1 A RoLo data specification indicating allowed structure for the ‘base’
attribute. If the coordinate reference system in target holds no relat
ion with other coordinate reference systems, this may be omitted.
base Data O 1 A RoLo data for determining relation to other coordinate reference s
ystem. Typically, this data includes spatial position for origin and p
ose for axis direction. If no relation with other coordinate reference
systems is required, this may be omitted.
• Modify the ‘Derived From:’ of Table 6 to be ‘DynamicRelativeCRS’.
Modify the ‘Derived From’ of Table 9 to be ‘DymanicRelativeDatum’
Modify figure 3 as follows:
– Add new abstract classes 'StaticRelativeCRS' and ‘DynamicRelativeCRS’
that inherit from RelativeCRS class.
– Rename RelativeDatum class to StaticRelativeDatum.
– Add new abstract class 'RelativeDatum' that inherit
from ::ISO19111::CD_EngineeringDatum class.
– Add new class ‘DynamicRelativeDatum’ that inherit from RelativeDatum
class
– Make StaticRelativeDatum class to be derived from RelativeDatum class
– Make StaticRelativeDatum to be included in StaticRelativeCRS class, not
RelativeCRS class.
– Make RelativeDatum to be included in RelativeCRS class, and add subset
note ‘subet usesDatum’
– Rename RelativeCartesianCRS class and RelativePolarCRS class to
StaticRelativeCartesianCRS and StaticRelativePolarCRS respectively,
and make them to be derived from StaticRelativeCRS.
– Make MobileCRS class to be derived from DynamicRelativeCRS class.
– Make MobileDatum class to be derived from DymanicDatum class
Make the following changes to C++ PSM:
– Rename RelativeDatum class to StaticRelativeDatum.
– Add the following definitions to RelativeCRS.hpp class RelativeDatum
: public ::ISO19111::CD_EngineeringDatum
{
};
class StaticRelativeCRS
: public RelativeCRS
{
};
class DynamicRelativeCRS
: public RelativeCRS
{
};
class DynamicRelativeDatum
: public RelativeDatum
{
};
– Make StaticRelativeDatum class to be derived from RelativeDatum class
– Make StaticRelativeDatum to be included in StaticRelativeCRS class, not
RelativeCRS class.
– Rename RelativeCartesianCRS class and RelativePolarCRS class to
StaticRelativeCartesianCRS and StaticRelativePolarCRS respectively,
and make them to be derived from StaticRelativeCRS.
– Make MobileCRS class to be derived from DynamicRelativeCRS class.
– Make MobileDatum class to be derived from DymanicDatum class
Disposition: Resolved
Actions taken:
November 21, 2008: received issue
October 15, 2009: closed issue
Issue 13107: Relative/Mobile operations are not well defined (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: Static2Mobile/Mobile2Static operations defined in Figure 4 are defined
to have SC_CRS as source/target domain. However, as SC_CRS is a base
class for both RelativeCRS and MobileCRS, these operations may take
MobileCRS for both source and target. This is confusing and shall be
well defined.
Resolution: Add some additional description to relevant tables for showing limitation that
‘static’ CRS shall not be an instance of DynamicRelativeCRS or inherited classes
Revised Text: Add the following line to table 11:
Note: Values for the attribute ‘target’ shall not be an instance of
DynamicRelativeCRS or its inherited classes.
• Add the following line to table 12:
Note: Values for the attribute ‘source’ shall not be an instance of
DynamicRelativeCRS or its inherited classes.
• Modify the class description in table 12 as following:
Description: Transformation operation from other static, non-mobile coordinate
reference systems to mobile coordinate reference systems.
Actions taken:
November 21, 2008: received issue
October 15, 2009: closed issue
Issue 13108: SymbolicCS/CRS name is confusing (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: In 6.3.2 Identity Information, symbolic identity CS/CRS are defined as
'SymolicCS/CRS' where numberic ones are defined as
'NumericIdentityCS/CRS'. Symbolic ID CS/CRS classes shall also be
named in the same manner.
Resolution: In order to make the naming consistent, rename SymbolicCS to
SymbolicIdentityCS and SymbolicCRS to SymbolicIdentityCRS.
Revised Text: Modify the caption of table 16 as following:
SymbolicIdentityCS class
• Modify the caption of table 20 as following:
SymbolicIdentityCRS class
• Modify type of attribute ‘coordinateReferenceSystem’ in table 21 to
SymbolicIdentityCRS.
Modify figure 5 as follows:
– Rename SymbolicCS class as SymbolicIdentityCS
– Rename SymbolicCRS class as SymbolicIdentityCRS
– Modify the type of ‘coordinateReferenceSystem’ attribute in DirectSymbol
to be SymbolicIdentityCRS class
• Modify C++ PSM as following:
– Rename SymbolicCS class as SymbolicIdentityCS
– Rename SymbolicCRS class as SymbolicIdentityCRS
– Modify the type of ‘coordinateReferenceSystem’ member variable in
DirectSymbol to be SymbolicIdentityCRS class
Disposition: Resolved
Actions taken:
November 21, 2008: received issue
October 15, 2009: closed issue
Issue 13109: Possible inconsistency between RoLo Data class attributes (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: In the RoLo Data class definition (Table 55), there are two attributes
which defines the elements of the target RoLo Architecture: rla (class
RoLo Architecture) and elem (class RoLo Element). This allows
inconsistent specification in the two attributes. Description shall be
added to notice users not to lead any inconsistency
Resolution: Place additional sentences to ‘elem’ attribute description for showing that
consistency is required.
Revised Text: Modify the description for attribute ‘elem’ in Table 55 as following:
An ordered list of RoLo elements. Numbers, orders and types of the RoLo elements
shall match that of the corresponding RoLo data specification.
Actions taken:
November 21, 2008: received issue
October 15, 2009: closed issue
Issue 13110: RoLo Attribute Set definition typo (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The 'attribute' attribute in RoLo Attribute Set class definition
(Table 127) is inconsistent with the UML diagram (Figure 21).
This shall be defined as 'O' (optional), not 'M' (mandatory).
Resolution: As RoLo Attribute Set and its related classes are restructured in issue #13322,
these modifications are not necessary any more.
Disposition: Closed, no change
Revised Text:
Actions taken:
November 21, 2008: received issue
October 15, 2009: closed no change
Discussion: As RoLo Attribute Set and its related classes are restructured in issue #13322,
these modifications are not necessary any more.
Disposition: Closed, no change
Issue 13111: Possible inconsistency between RoLo Attribute Set attributes (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: In the RoLo Attribute Set definition (Table 127), there are two
attributes which leads to the elements of the RoLo Attribute
Definition Set: def (class RoLo Attribute Definition Set) and
attribute (class RoLo Attribute). This allows inconsistent
specification in the two attributes. Description shall be added to
notice users not to lead any inconsistency.
Resolution:
Revised Text:
Actions taken:
November 21, 2008: received issue
October 15, 2009: closed no change
Discussion: As RoLo Attribute Set and its related classes are restructured in issue #13322,
these modifications are not necessary any more.
Disposition: Closed, no change
Issue 13112: Typo in 6.3.4 Robotic Localization Architecture (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: tino
The following sentence in the thrid paragraph of 6.3.4 is wrong:
In such case, the Error Element Specification (derived from RoLo
Element class) ...
This shall be as following:
In such case, the Error Element Specification (derived from RoLo
Element Specification class) ...
Resolution: This issue is actually about 7.3.4, not 6.3.4.
Correct the erratum in Section 7.3.4 of RLS Specification Beta1 (robotics/08-07-
01). Note that the section number #6.3.4 in this issue is based on the previous
version of RLS Specification (robotics/08-05-02).
Revised Text: In a sentence beginning with ‘In such case, the Error Element Specification ‘ in
section 7.3.4, replace ‘RoLo Element class’ to ‘RoLo Element Specification
class’.
Actions taken:
November 21, 2008: received issue
October 15, 2009: closed issue
Issue 13113: Typo in Position Element Operation class description (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The description for Position Element Operation class definition (Table
56), 'RoLo Elements' (twice) shall be replaced to 'Position Elements'.
Resolution: Correct the errata
Revised Text: Replace every ‘RoLo Elements’ to be ‘RoLo position elements’ in Table 56.
Actions taken:
November 21, 2008: received issue
October 15, 2009: closed issue
Issue 13114: RoLo Attribute Definition Set definition typo (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The 'attribute attribute in RoLo Attribute Definitino Set class
definition (Table 121) is inconsistent with the UML diagram (Figure
21). This shall be defined as 'O' (optional), not 'M' (mandatory).
Resolution:
Revised Text:
Actions taken:
November 21, 2008: received issue
October 15, 2009: closed issue
Discussion: Discussion:
As RoLo Attribute Set and its related classes are restructured in issue #13322,
these modifications are not necessary any more.
Disposition: Closed, no change
Issue 13115: The union of a documents (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity:
Summary: The RLS Specification FTF Beta1(dtc/2008-07-01) includes Sample C++ Headers(Annex A).
I think that the C++ Header is PSM model, so we should include it in the specification
as 8.1 C++ PSM
Resolution: Include PSM in the specification as section 8.1.
Revised Text: Copy the content of Annex A at the end of section 8.1.
Delete Annex A (Sample C++ Headers) from the specification.
Actions taken:
November 25, 2008: received issue
October 15, 2009: closed issue
Issue 13116: Change the class name of RoLo Architecture (rls-ftf)
Click here for this issue's archive.
Source: Samsung Electronics (Dr. Yeonho Kim, nobody)
Nature: Uncategorized Issue
Severity:
Summary: In the revised proposal, especially in Fig. 8 and 9, the class name of RoLo Architecture is confused with the name of RoLo architecutre that consists of the RoLo data specification and RoLo data.
I suggest to substitute the class name of RoLo Architecture to RoLo Data Specification.
Resolution: Change the class name of RoLo Architecture to DataSpecification. However,
when the expression ‘RoLo Architecture’ is used in a sentence to describe an
object of this class, rewrite as ‘RoLo data specification’ instead of
‘DataSpecification’ to improve readability. Also change the names of attributes
that are typed to this class from ‘rla’ to ‘dataSpec’. However, for the ‘rla’ attribute
in RoLo Data class, change this to ‘spec’ so that the name will be consistent with
other elements of RoLo architecture.
Revised Text: Replace every ‘RoLo Architecture’ phrase as ‘DataSpecification’ when used as
a class name, both in the specification document and UML diagrams.
- the type of 'rla' attribute in Table 5 (RelativeDatum)
- the sentence beginning by "Classes for describing the structure or the
meaning ..." in the first paragraph of section 7.3.4
- the caption of Table 54 (RoLo Architecture)
- the type of 'rla' attribute in Table 55 (RoLo Data)
- the type of 'source' attribute in Table 59 (RoLo Data operation)
- the type of 'target' attribute in Table 59 (RoLo Data operation)
- the type of 'rla' attribute in Table 66 (Specific Data Format)
- the type and description of ‘rlaList’ attribute in Table 133 (Stream Ability)
- Class name of ‘RoLo Architecture’ in Figure 3
- The type of ‘rla’ attribute in RelativeDatum in Figure 3 The type of ‘rla’ attribute in RoLo Data in Figure 9
- Class name of ‘RoLo Architecture’ in Figure 9
- Class name of ‘RoLo Architecture’ in Figure 10
- Class name of ‘RoLo Architecture’ in Figure 11
- The type of ‘rlaList’ attribute in StreamAbility in Figure 22
Change the following occurrence of ‘RoLo Architecture’ as ‘RoLo data
specification’:
- the description for 'rla' attribute in Table 5 (RelativeDatum)
- the last paragraph in page 16
- the sentence beginning by "The RoLo Data and RoLo ..." in the second
paragraph of section 7.3.4 (Robotic Localization Architecture).
- the description for 'rla' attribute in Table 55 (RoLo Data)
- the description for 'elem' attribute in Table 55 (RoLo Data)
- the description of Table 59 (RoLo Data Operation) (twice)
- the description for 'source' attribute in Table 59 (RoLo Data operation)
- the description for 'target' attribute in Table 59 (RoLo Data operation)
- the description for 'childOperation' attribute in Table 60 (RoLo Data
Concatenated Operation) (twice)
- the description of Table 61 (RoLo Data Single Operation)
- the description of Table 62 (RoLo Data Transformation)
- the description for 'usesOperation' attribute in Table 62 (RoLo Data
Transformation) (twice)
- the description of Table 63 (RoLo Data Mapping Operation) (four times)
- the description for 'sourceIndex' attribute in Table 63 (RoLo Data
Mapping Operation) (twice)
- the description for 'targetIndex' attribute in Table 63 (RoLo Data
Mapping Operation)
- the description of Table 65 (Encoding Rule)
Replace every attribute name ‘rla’ as ‘dataSpec’ in the following:
- Table 5 and Figure 3 (RelativeDatum)
- Table 66 and Figure 11 (Specific Data Format)
Rewrite the description for ‘rla’ attribute in Table 66 (Specific Data Format) as:
“Specifies a RoLo data specification that this data format can handle.”
Modify C++ PSM as following:
- the type of 'rla' class variable in RelativeDatum class as
DataSpecification (RelativeCRS.hpp) The name of ‘rla’ class variable in RelativeDatum class as ‘dataSpec’
(RelativeCRS.hpp)
- Class name of RoLoArchitecture as DataSpecification
(RoLoArchitecture.hpp)
- the type of 'rla' class variable inRoLoData class as DataSpecification
(RoLoArchitecture.hpp)
- The name of ‘rla’ class variable in RoLoData class as ‘spec’
(RoLoArchitecture.hpp)
- the type of 'source' and ‘target’ class variable in RoLoDataOperation
class as DataSpecification (RoLoDataOperation.hpp)
- the type of 'rla' class variable in SpecificDataFormat class as
DataSpecification (DataFormat.hpp)
- The name of ‘rla’ class variable in SpecificDataFormat class as
‘dataSpec’ (DataFormat.hpp)
- the type of ‘rlaList’ class variable inStreamAbility class as
DataSpecification (Service.hpp)
- The name of ‘rlaList’ class variable in StreamAbility class as
‘dataSpecList’ (Service.hpp)
Disposition: Resolved
Actions taken:
November 25, 2008: received issue
October 15, 2009: closed issue
Issue 13117: Angle unit representation (rls-ftf)
Click here for this issue's archive.
Source: Samsung Electronics (Dr. Yeonho Kim, nobody)
Nature: Uncategorized Issue
Severity:
Summary: In Section 6.4.1, the units of angle in Type I, II and II are not represented consistently .
I suggest to represent the unit of angle for Type I,II with radian, for Type II with degree.
All the parts of the document should be changed appropriately.
Resolution: Disposition: See issue 13120 for disposition
Revised Text:
Actions taken:
November 25, 2008: received issue
October 15, 2009: closed issue
Issue 13118: Posistion description in Type II (rls-ftf)
Click here for this issue's archive.
Source: Samsung Electronics (Dr. Yeonho Kim, nobody)
Nature: Uncategorized Issue
Severity:
Summary: The desrciption of position in Table 70 is not commonly used.
Following the ISO 31-11, the description of the position of type II in Table 70 would be better to be [r, ?, f].
Resolution: Disposition: See issue 13119 for disposition
Revised Text:
Actions taken:
November 25, 2008: received issue
October 15, 2009: closed issue
Discussion:
Issue 13119: Name of Type II coordinate system (rls-ftf)
Click here for this issue's archive.
Source: Samsung Electronics (Dr. Yeonho Kim, nobody)
Nature: Uncategorized Issue
Severity:
Summary: The name of Type II, polar coordination is typically defined as a two dimensional coordinate system.
The coordinate system would be better to be called as spherical coordinated system that is the three dimensioinal extension of the polar coordinate system
Resolution: Disposition: See issue 13120 for disposition
Revised Text:
Actions taken:
November 25, 2008: received issue
October 15, 2009: closed issue
Issue 13120: Name of Orientation (rls-ftf)
Click here for this issue's archive.
Source: Samsung Electronics (Dr. Yeonho Kim, nobody)
Nature: Uncategorized Issue
Severity:
Summary: In Table 69, 70, 41, the name of orientation (roll, pitch, yaw) is not relevant to the definition of the orientation that is described below each table.
It would be better to be removed from the tables.
Resolution: Extend common data formats each have two different interpretations so that they
can cover both of xyz-Euler's angle representation and roll-pitch-yaw
representation of orientation (two orientation representations for each Cartesian,
Polar, and Geodetic coordinate system). We thus have three common data
formats (syntax) each with two different RoLo data specifications (semantics).
We also add some figures in the specification to avoid misunderstanding or
ambiguity regarding the definition of position and orientation for each common
data format.
Also, based on issue 13999, change the types for timestamps in the common
data formats from a real number to two integers, which correspond to standard
UNIX C structure, time_t. Normally, the first integer indicates the number of
seconds elapsed from the epoch, and the second one indicates sub-seconds.
The unit of sub-seconds is normally nano second that is compatible to most
recent POSIX C time_t struct.
Revised Text: see dtc/2009-06-03
Actions taken:
November 25, 2008: received issue
October 15, 2009: closed issue
Issue 13121: Definition of orientation in Type I, II, III (rls-ftf)
Click here for this issue's archive.
Source: Samsung Electronics (Dr. Yeonho Kim, nobody)
Nature: Uncategorized Issue
Severity:
Summary: For the Type I, II, II, he definition of the orientation is ambiguous.
More precise definition for the orientation is required (with figures if needed)
Resolution: See issue 13120 for disposition
Revised Text:
Actions taken:
November 25, 2008: received issue
October 15, 2009: closed issue
Issue 13122: Filter package for the Issue #12916 (rls-ftf)
Click here for this issue's archive.
Source: Samsung Electronics (Dr. Yeonho Kim, nobody)
Nature: Uncategorized Issue
Severity:
Summary: Before making a decision of including the filter condition as a sub package, we
have to decide whether the filter condition would be included in the main part or in the Annex part as an option.
At the last meeting in Ottawa, we concluded that the issue of filter condition should be decided after reviewing more real-world examples. (refer robotics/2008-06-14)
I suggest to discuss and decide this issue in December meeting.
Dr. Noda, please show us those examples in December meeting for us to decide this issue.
Resolution:
Revised Text:
Actions taken:
November 25, 2008: received issue
October 15, 2009: closed no change
Discussion: As a result of discussion in Robotics DTF, we had a conclusion that it is better to
make the filter condition to be informative and non-mandatory. However, based
on ‘Policies and Procedures of the OMG Technical Process’ (document number
pp/09-01-02, version 2.8, 21/Jan/2009), we cannot remove significant
functionality from the specification (section 4.4.1.1). Thus, this issue is closed
without change.
Disposition: Closed, no change
Issue 13129: posSpecRef in Error Element (Figure 9/Table 49) (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: * posSpecRef in Error Element (Figure 9/Table 49) shall be a list of
references to objects
The attribute posSpecRef in Error Element class is now defined as a
list of CharacterString, which is used as a reference to objects of
Position Element class. This shall better be defined as a list of
references.
Resolution: Currently, pointers that indicate which RoLo position element specifications are
related to a RoLo error element specification are modeled to use ‘specID’
attribute in RoLo element specifications. Instead, as described in the summary, it
is much better to directly refer to the RoLo position element specification
instances. Thus, modify 'posSpecRef' to be an ordered list of references to
Position Element Specification instances and delete 'specID' attribute from RoLo
Element Specification class.
As RoLo Data Mapping Operation class depends on this, also modify
DataMappingOperation class definition to use the new data structure.
Additionally, some changes in C++ PSM are also required. Remove ‘specID’ from ElementSpecification class
(RoLoArchitecture.hpp)
– Change the definition of ‘posSpecRef’ class variable in
ErrorElementSpecification class (RoLoArchitecture.hpp) as following:
::std::vector<PositionElementSpecification*> posSpecRef;
– Change the name of ‘sourceIndex’ class variable in
DataMappingOperation (RoLoDataOperation.hpp) as
‘sourceElemSpecs’
– Change the name of ‘targetIndex’ class variable in
DataMappingOperation (RoLoDataOperation.hpp) as ‘targetElemSpecs’
– Change the definition of ‘sourceIndex’ and ‘targetIndex’ class variable in
DataMappingOperation (RoLoDataOperation.hpp) as following:
::std::vector<ElementSpecification*> sourceElemSpecs,
targetElemSpecs
Revised Text: Modify Figure 9 as follows:
– Replace the 'posSpecRef' attribute of Error Element Specification class by
an association between Error Element Specification and Position Element
Specification which shows an aggregation at Error Element Specification
side. Add association end name ‘posSpecRef’ to this association and
show multiplicity [1..*] and indicate this association to be ordered.
– Delete the 'specID' attribute from RoLo Element Specification.
Modify Figure 10 as follows:
– Add RoLo Element Specification class
– Add an association between RoLo Data Mapping Operation class and
RoLo Element Specification class to show composition on RoLo Data Mapping Operation side, with an association end name
‘sourceElemSpecs’, multiplicity [1..*] and indicator to show that this is
ordered.
– Add an association between RoLo Data Mapping Operation class and
RoLo Element Specification class to show composition on RoLo Data
Mapping Operation side, with an association end name ‘targetElemSpecs’,
multiplicity [1..*] and indicator to show that this is ordered.
Modify Table 47 as follows:
– Delete 'specID' attribute from the table.
Modify Table 49 (Error Element Specification) as follows:
– Modify the type of 'posSpecRef' attribute to be
"PositionElementSpecification".
– Modify the description for 'posSpecRef' attribute as follows:
An ordered list of references to RoLo position element specifications showing which
positional data the RoLo error contained in the RoLo error element is related to. The
referred RoLo position element specifications shall be contained in the same RoLo
data specification as this class instance.
Modify Table 63 as following:
– Modify ‘Description’ as following:
Description: Definition of an operation for transforming data between different
RoLo data specifications that simply maps elements in the source RoLo data
specification to elements in the target RoLo data specification. Only the structures
of the RoLo elements are altered, and the data content itself are not changed. With
RoLo error elements, the reference to the RoLo position elements shall be modified
appropriately. The two attributes contained are lists of references to RoLo element
specifications in source and target RoLo data specifications that defines how the
mapping is to be performed.
– Change the name of attribute ‘sourceIndex’ to ‘sourceElemSpecs’
– Change the type of attribute ‘sourceIndex’ to ElementSpecification class
– Modify the description of ‘sourceIndex’ attribute as following:
Ordered list of RoLo element specification references within the source RoLo
data specification which is to be mapped to the RoLo element specification in
the target RoLo data specification represented by the ‘targetElemSpecs’
attribute value at the same position. The numbers of 'sourceElemSpecs’
attribute shall match that of ‘targetElemSpec’ attribute.
– Change the name of attribute ‘targetIndex’ to ‘targetElemSpecs’
– Change the type of attribute ‘targetIndex’ to ElementSpecification class
– Modify the description of ‘targetIndex’ attribute as following:Ordered list of RoLo element specification references within the target RoLo
data specification.
Modify C++ PSM as following:
– Move the definition of RoLo Error class to an separate file
ErrorBase.hpp as following:
// $Id: ErrorBase.hpp,v 1.1 2009/06/20 06:18:42 nishio Exp $
#pragma once
#include <RLS/ErrorType.hpp>
namespace RoLo
{
namespace Architecture
{
class ErrorType;
class Error
: public ::ISO19111::IO_IdentifiedObjectBase
{
public:
ErrorType *errType;
};
}
}
– Remove ‘specID’ from ElementSp
Actions taken:
December 1, 2008: received issue
October 15, 2009: closed issue
Issue 13130: Mismatch in Error Type (Figure 6) and Error Information (Figure 7) definitions (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: There are mismatches in the hierarchical structure of Error Type
classes and Error Information classes, which shall correspond to each
other. One solution will be to:
- add ET_ErrorDistribution, derived from 'RoLo Error Type'
- make ET_Gaussian, ET_UniformGaussian, ET_ParticleSet and
ET_MixtureModel to be derived from ET_ErrorDistribution
Resolution:
Revised Text:
Actions taken:
December 1, 2008: received issue
Issue 13131: Missing 'public:' declaration in PSM RelativeDatum definition (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: 'public:' is missing in RelativeDatum class definition in
RelativeCRS.hpp (PSM C++ header).
Resolution: Correct typo as described in the summary.
Revised Text: Insert the following line sentence immediately after '{' of RelativeDatum class
definition in PSM C++ header:
public:
Actions taken:
December 1, 2008: received issue
October 15, 2009: closed issue
Issue 13143: Typo in Position Element Concatenated Operation class definition (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: In The Position Element Concatenated Operation class diagram (Figure
10), the childOperation attribute is described to be a reference to a
list of RoLo Element Operation. This does not match its definition in
Table 57, and should be replace to be a reference to a list of
Position Element Operation.
Resolution: Correct the typo as suggested in the summary
Revised Text: In Figure 10, remove 'childOperation' attribute and add an association end
name to the aggregation between Position Element Concatenated Operation
and Position Element Operation.
Actions taken:
December 5, 2008: received issue
October 15, 2009: closed issue
Issue 13144: Definition mismatch for Service Ability class in Figure 22 and Table 139 (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The inStreamAbility and outStreamAbility attributes of Service Ability
class are listed in Table 139 but not in Figure 22.
Resolution: See issue 13322 for disposition
Revised Text:
Actions taken:
December 5, 2008: received issue
October 15, 2009: closed issue
Issue 13145: Figure 8 may accompany an object diagram (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: This is actually not an issue, but Figure 8 "Representation of error
information related to multiple localization data" shall better be
accompanied with an UML object diagram.
Resolution: For better understanding of the specification, add an non-mandatory UML object
diagram into Figure 8 so that the original figure is accompanied with the UML
object diagram.
Revised Text: see dtc/2009-06-03 pages 85
Actions taken:
December 5, 2008: received issue
October 15, 2009: closed issue
Issue 13184: Filter Condition should be reduced and move to annex as informative (rls-ftf)
Click here for this issue's archive.
Source: AIST (Dr. Itsuki Noda, i.noda(at)aist.go.jp)
Nature: Uncategorized Issue
Severity:
Summary: There is an activity in ISO TC211 to develop UML diagram of
filter condition encoding. So, we should refer it directly
instead to define our own UML diagram.
* Filter Condition will be useful specification for various
application of RLS. But it domain is slightly different from
RLS's main target. In addition, ISO TC211's definition is not
finalized yet. So, the specification should be included in the
annex as an informative.
Resolution: As described in issue 13122, Filter Condition cannot be moved to be an
informative part for now, and thus will remain in its original position as mandatory.
However, as an identical ISO standard which can be used for RoLo service is
under process, descriptions will be modified to use the ongoing ISO draft. Here,
we modify the interface for treating filter conditions in Robotic Localization
Service. In the beta1 specification, section for filter condition (7.5) defined an
additional method for RoLo service class to treat filter condition data. As pointed
out by issue 14001, this is not a good way from the viewpoint of interoperability.
Thus, we define a much simple method that treats filter condition as a parameter
defined in ability description for RoLo streams. This allows users to check
whether filter condition functionality is supported or not and provides an easy
mean to specify conditions.
In addition, as the new filtering functionality based on ISO standard requires XML
notation of RoLo data (also from issue 13186) and naming rules (from issue
14002), we define two informative annexes, XML-PSM and naming rules. With
this addition, some description about what is given as annex is also added in
Section 1 “Scope”.
Revised Text: see pages 86 - 105
Actions taken:
December 20, 2008: received issue
October 15, 2009: closed issue
Issue 13185: Hierarchy of Error can be reduced (rls-ftf)
Click here for this issue's archive.
Source: AIST (Dr. Itsuki Noda, i.noda(at)aist.go.jp)
Nature: Uncategorized Issue
Severity:
Summary: Particle Set in Error can be defined as a subclass of Linear
Mixture Model. In this case each particle is denoted as an
impulse distribution
Resolution: Delete Particle class and reorganize classes in Figure 7 to reduce the hierarchy
of Error classes. Also add some description for improving understandability.
Revised Text: Modify the first paragraph of ¡¥RoLo Error¡¦ section as following:
RoLo errors are objects for holding error information in different representations. Here
we define some frequently used forms. Users may extend these classes to implement th
eir own RoLo error containers, accompanied with appropriate RoLo error type definitio
ns.
Modify Figure 7 as follows:
¡V Modify Weighted Distribution class as following:.
ƒÞ change the class name to WeightedModel
ƒÞ delete 'dist' attribute.
ƒÞ rename 'pos' attribute to be 'posElem' and replace it to be derived
attribute so that this class has a composition relationship with
PositionElement class on the WeightedModel side. Add association
end name ¡¥posElem¡¦.
¡V Delete Particle class from the figure.
¡V Make ParticleSet class to be derived from LinearMixtureModel class
and delete its attributes 'numParticles' and 'particles'.
¡V Rename 'dists' attribute of LinearMixtureModel class to be 'models' and
replace it to be a composition on LinearMixtureModel side. Add association end name ‘models’, show multiplicity [1..*] and indicate that
this shall be ordered.
Delete Table 40.
In Table 41,
– Modify the class description to be "Error represented by a set of particles.
As for the 'models' attribute derived from LinearMixtureModel class, the
'posElem' attribute shall either have no 'err' attribute or have an RoLo error like
an impulse response (such as a Gaussian distribution with zero standard
deviation). Normally, this is used for representing distributions by Monte Carlo
approximation, where distributions are approximated by a finite number of
random samplings.".
– Modify Derived From to be "LinearMixtureModel".
– Delete 'numParticles' attribute and 'particles' attribute.
In Table 43,
– Modify the table caption to be "WeightedModel class"
– Modify “Derived From” to be "IO_IdentifiedObjectBase [ISO19111]".
– Delete 'dist' attribute.
– Rename 'pos' attribute to 'posElem' and modify its type to be
"PositionElement".
– Modify “Description” as following:
A distribution with a weight. Recall that a PositionElement object can be
interpreted to represent a probability distribution. Its 'pos' attribute is treated as
the expected coordinate value and its 'err' attribute as the shape of distribution.
Thus, in this class the combination of 'weight' and 'posElem' attributes denotes
a weighted distribution.
In Table 44,
– Rename 'dists' attribute to 'models' and modify its type to be
"WeightedModel", its occurrence to be "N ord", and its description to be
"List of weighted models to be combined.".
In Table 45, modify the class description to be as following:
A distribution represented by a linear mixture of Gaussian distributions. The
models attribute derived from LinearMixtureModel shall have a 'posElem' attribute
whose 'err' attribute is restricted to be an instance of Gaussian class.
Modify C++ PSM as following:
– Modify Weighted Distribution class (Error.hpp) as following:
class WeightedModel
: public ::ISO19111::IO_IdentifiedObjectBase
{
public:
PositionElement posElem; ::ISO19103::Probability weight;
};
– Delete Particle class (Error.hpp).
– Make ParticleSet class to be derived from LinearMixtureModel class
and delete its attributes 'numParticles' and 'particles'.
– Change definition of 'dists' class variable in LinearMixtureModel class
as following:
::std::vector<WeightedModel> models;
Disposition: Resolved
Actions taken:
December 20, 2008: received issue
October 15, 2009: closed issue
Issue 13186: XML mapping of RoLo Data and RoLo Element (rls-ftf)
Click here for this issue's archive.
Source: AIST (Dr. Itsuki Noda, i.noda(at)aist.go.jp)
Nature: Uncategorized Issue
Severity:
Summary: * We should have a rule to denote RoLo Data in XML format. In
order to do it, we need to have an template of XML schemas for
the RoLo Data XML format and the mapping rule from RoLo Data
Specifications.
* This rule will be a part of annex as informative
Resolution: See issue 13184 for disposition
Revised Text:
Actions taken:
December 20, 2008: received issue
October 15, 2009: closed issue
Issue 13187: Hierarchical relationship for instances of CRS and RoLo Data Spec (rls-ftf)
Click here for this issue's archive.
Source: AIST (Dr. Itsuki Noda, i.noda(at)aist.go.jp)
Nature: Uncategorized Issue
Severity:
Summary: * Some applications need to handle unlimited number of CRSs and
their corresponding structures of RoLo Data Specification.
For example, suppose that a human counting application that can
connect camera systems that are mounted various rooms. Because
the origin of CRS used in each camera system is different from
each other, corresponding RoLo Data Specification is different.
It means that the application need to recognize unbounded number
of RoLo Architectures.
In order to avoid such situations, we can utilize hierarchical
relations among CRSs. In the case of above example, we define a
generic CRS, by which a CRS used in each camera system is
defined a kind of a derived CRS of the generic CRS. In the same
time, the human counting application is designed to accept the
generic CRS and corresponding RoLo architecture. Then, the
application can handle each derived CRS and its corresponding
RoLo architecture.
* In order to realize such functionality, we should introduce the
following mechanisms to define relations among CRSs and among
RoLo architectures.
* To define a hierarchical relation between two CRSs, we introduce
a CRS Hierarchical Relation class. In the class, baseCRS and
subCRS are specified. The baseCRS is a generic CRS and the
subSRC is a derived CRS.
Whole relations of CRSs shall form a lattice structure.
* To define a hierarchical relations between two RoLo Data
Specifications, we introduce a class for RoLo Data
Specifications, in which baseSpec and subSpec is specified. The
baseSpec is a generic RoLo Data Specification and subSpec is a
derived RoLo Data Specification.
Whole relations of RoLo Data Specifications shall form a lattice
structure.
* To define a hierarchical relations between two RoLo Element
Specifications, we introduce a class for RoLo Element
Specifications, in which baseSpec and subSpec is specified. The
baseSpec is a generic RoLo Element Specification and subSpec is
a derived RoLo Element Specification.
Whole relations of RoLo Element Specifications shall form a lattice
structure.
Resolution: Instead of introducing a new class to represent hierarchical relationship among
instances, use the don’t-care functionality in the original specification. However,
as there were not enough description in the original specification and also as
there were some errors, do some reorganization to clarify this.
Revised Text: see dtc/2009-06-03 pages 111 - 114
Actions taken:
December 20, 2008: received issue
October 15, 2009: closed issue
Issue 13317: Simplication of Error Type (Figure 6): (rls-ftf)
Click here for this issue's archive.
Source: ETRI (Dr. Jaeyeong Lee, PhD, jylee(at)etri.re.kr)
Nature: Uncategorized Issue
Severity:
Summary: Error Type classes are defined for each Error Information class in Figure 6. However it seems to be unnatural and too complex to define such classes just for discriminating error types. Implementation also is hard.
More natural way would be defining one class that is an enumeration of various error types like the case of RoLo Attribute SingleValue in Figure 21.
Resolution: Instead of the way suggested in the summary, we decided to use only a single
class for describing RoLo error types and use hierarchy of RoLo error type
instances to represent the relationships instead of the class hierarchy in the
original specification. By this change, we can decrease the complexity of defining
classes that is never instantiated while preserving the original functionality and
extendibility.
Revised Text: see dtc/2009-06-03 pages 115 - 117
Actions taken:
January 22, 2009: received issue
October 15, 2009: closed issue
Issue 13318: Typo in Table 22 (RoLo SymbolRef) (rls-ftf)
Click here for this issue's archive.
Source: ETRI (Dr. Jaeyeong Lee, PhD, jylee(at)etri.re.kr)
Nature: Uncategorized Issue
Severity:
Summary: In Table 22, 'Point' attribute of RoLo SymbolRef should be renamed by 'point' for namning consistency.
Resolution: Correct the typo
Revised Text: Modify the attribute name in Table 22 from 'Point' to 'point'.
Actions taken:
January 22, 2009: received issue
October 15, 2009: closed issue
Discussion:
Issue 13319: Obligation of 'values' attribute of Matrix (Figure7/Table36) (rls-ftf)
Click here for this issue's archive.
Source: ETRI (Dr. Jaeyeong Lee, PhD, jylee(at)etri.re.kr)
Nature: Uncategorized Issue
Severity:
Summary: In Table 36 (Matrix) the obligation of 'values' attribute is mandatory but it is represented as optional in Figure 7. Which one is correct?
Resolution: The obligation of 'values' attribute is mandatory. Therefore, fix Figure 7 to make
'values' attribute to be mandatory
Revised Text: Modify Figure 7 as following:
– Modify multiplicity of 'values' attribute of Matrix class to be '1..*'.
Actions taken:
January 22, 2009: received issue
October 15, 2009: closed issue
Issue 13320: The type of 'usesOperation' attribute of RoLo Data Transformation (Table62) (rls-ftf)
Click here for this issue's archive.
Source: ETRI (Dr. Jaeyeong Lee, PhD, jylee(at)etri.re.kr)
Nature: Uncategorized Issue
Severity:
Summary: In Table 62, the type of 'usesOperation' attribute of RoLo Data Transformation should be "Position Element Operation".
Resolution: Modify Table 62 as described in the summary. Also modify the description for this
attribute to fit this change.
Revised Text: Modify attribute type of the 'usesOperation' attribute in Table 62 to be 'Position
Element Operation'.
Replace every appearance of ‘RoLo Element Specification’ as ‘RoLo position
element specification’ in the description for 'usesOperation' attribute in Table
62.
Actions taken:
January 22, 2009: received issue
October 15, 2009: closed issue
Issue 13321: Association between RoLo Parameter Set and RoLo Parameter Definition (Figure 21) (rls-ftf)
Click here for this issue's archive.
Source: ETRI (Dr. Jaeyeong Lee, PhD, jylee(at)etri.re.kr)
Nature: Uncategorized Issue
Severity:
Summary: In Figure 21 (RoLo Ability and RoLo Parameter Set), the aggregation association between RoLo Parameter Set and RoLo Parameter Definition is not fully specified. If it is for derived 'def' attribute, we should add association end name in Figure 21 and an explanatory note to limit the value of derived attribute in Table 128 (RoLo Parameter Set).
Resolution:
Revised Text:
Actions taken:
January 22, 2009: received issue
October 15, 2009: closed no change
Discussion: As RoLo ability and its related classes are restructured in issue #13322, these
modifications are not necessary any more.
Disposition: Closed no change
Issue 13322: Inability of representing range values (Figure21/Table120) (rls-ftf)
Click here for this issue's archive.
Source: ETRI (Dr. Jaeyeong Lee, PhD, jylee(at)etri.re.kr)
Nature: Uncategorized Issue
Severity:
Summary: The 'choice' attribute of RoLo Attribute Definition class (Table 120) can only represent a list of discrete values. In case of ranged attributes that have minimum and maximum value, it cannot be well expressed.
One possible resolution is to introduce a new class, RoLo Attribute RangeValue, which is derived from RoLo Attribute Value and has 'minValue' and 'maxValue' attributes of type of RoLo Attribute SingleValue.
Resolution: solution suggested in the summary is sufficient for this issue. But similar
issues may rise if we kept on using the current definition of RoLo Attribute
SingleValue class defined as an enumeration of types which is not extendable by
users. Thus, in order to make the RoLo Attribute structures more extendable and
flexible, we restructure these classes to use type templates. In addition, some
more sentences are added to clarify how RoLo attribute and related structures
are to be used.
Also from issue 14006, there are too many classes of similar functionality.
Especially, usage of RoLo Attribute Definition Set was unclear and distinction
between RoLo Attribute and RoLo Parameter (Set) was not clear. Thus, the
following reorganization of class structure is proposed:
- Remove the RoLo Attribute SingleValue, RoLo Attribute Value, RoLo
Attribute ValueList, RoLo Attribute Type classes that were used to represent
variations of type/value pair. InStread, Make RoLo Attribute class to be a
template class. By this change, RoLo Attribute can be used for any time that
is required for attributes.
- Separate ‘attribute’ and ‘parameter’ explicitly, and describe their differences.
Here, attributes are ‘constant’ parameters that users cannot change.
Parameters are user-changeable attributes. As requested by summary, we
will prepare several template types for the datatype of parameter. This is for
convenience, and users can always extend this to any types on demand. Simplify the attribute set/attribute definition set class structures for clarity.
Some attributes (such as min/maxOccurence in RoLo Attribute Definition) are
not necessary any more due to the introduction of templates (users can
always define types with these occurrence limits).
Revised Text: dtc/2009-06-03 pages 123 - 132
Actions taken:
January 22, 2009: received issue
October 15, 2009: closed issue
Issue 13426: Conformance shall be specified in detail (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: Conformance criteria described in chapter 2 is too simple and too
strict. Shall be described in more detail.
Resolution: Add further description showing the conformance conditions.
Revised Text: Modify Chapter 2 as following:
Any implementation or product claiming conformance to this specification shall supp
ort the following conditions:
„O Implementations shall provide the interfaces described in section 7.5 Interface Pack
age.
„O Implementations shall provide their ability descriptors and the necessary attribute d
efinitions described in section 7.5 Interface Package.
„O Data treated by implementations shall follow the data structure described in 7.3 Arc
hitecture Package and the data formats described in 7.4 Data Format Package. This
does not mean that modules shall be able to treat every structure or formats describ
ed herein. However, every module shall support at least one of the common data for
mats and the relevant data structure.
„O Implementations shall support the return codes described in section 7.2.
Disposition: Resolved
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13427: Some references are missing (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: Reference described in chapter 3 is not enough, and shall be
added. Non-normative references may also be helpful.
Resolution: Add necessary references. Also, split chapter 3 to include both normative and
non-normative references.
Revised Text: Modify Chapter 3 as following:
3.1 Normative References
The following normative documents contain provisions which, through reference in t
his text, constitute provisions of this specification. For dated references, subsequent a
mendments to, or revisions of, any of these publications do not apply.
[ISO19103] International Organization for Standardization, Geographic information
– Conceptual schema language, 2005
[ISO19107] International Organization for Standardization, Geographic information
– Spatial schema, 2003
[ISO19111] International Organization for Standardization, Geographic information
– Spatial referencing by coordinates, 2007
[ISO19115] International Organization for Standardization, Geographic information
– Metadata, 2003
[PER] International Telecommunication Union Telecommunication Standardization S
ector, Specification of Packed Encoding Rules (PER), ITU-T Rec. X.691 (2002) / IS
O/IEC 8825-2:2002
[UML] Object Management Group, OMG Unified Modeling Language (OMG UM
L), Superstructure, Version 2.2, OMG document number formal/2009-02-02, 2009 3.2 Non-Normative References
[ISO/DIS19142] International Organization for Standardization, Geographic informat
ion – Web Feature Service, DIS, 2009
[ISO/DIS19143] International Organization for Standardization, Geographic informat
ion – Filter Encoding, DIS, 2009
[Wikipedia] Wikipedia, the free encyclopedia, http://www.wikipedia.org/
Disposition: Resolved
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13428: Terms and Definition shall be added (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: There are certainly some terms and definitions required for this
specification.
Resolution: Add necessary definitions on terms used in this specification
Revised Text: see dtc/2009-06-03 pages 136 - 139
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13429: Symbol should be described (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: Chapter 5 "Symbol" shall be described, especially for the data format
section.
Resolution: Add necessary symbol descriptions
Revised Text: Modify Chapter 5 "Symbol" as following:
Symbol
x, y, z Cartesian coordinate
r, è, ö spherical coordinate
ö, ë, h geodetic coordinate (latitude, longitude, height)
á, â, ã orientation
x, y, z a fixed Cartesian coordinate system
X, Y, Z a rotating Cartesian coordinate system
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Discussion:
Issue 13430: Some organization missing in "6.1 Supporting Organizations" (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: Not sure whether this section is necessary, but if so, some more
organizations shall be added here.
Resolution: Add another section describing the organizations which submitted the original
specification. Also, from issue 13998, add submitter information
Revised Text: Modify the first sentence in “6.1 Supporting Organizations” as follows:
The following organizations supported parts of this specification:
• Add the following sections before “6.1”: Submitters
The initial submissions that this specification is based on were submitted by the followi
ng people:
Kyuseo Han, Electronics and Telecommunications Research Institute (ETRI)
Yeonho Kim, Samsung Electronics Co., Ltd.
Shuichi Nishio, Japan Robot Association (JARA) / Advanced Research Institute Int
ernational (ATR)
Submitting Organizations
The following organizations made the initial submission that this specification is based
on:
Electronics and Telecommunications Research Institute (ETRI)
Japan Robot Association (JARA)
Samsung Electronics Co., Ltd.
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13431: Names missing in "6.2 Acknowledgements" (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity: Minor
Summary: Some more people shall be added here.
Resolution: Add names to be acknowledged.
Revised Text: Add the following name to the list in section 6.2.
„O Jaeyoeng Lee, Electronics and Telecommunications Research Institute
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Discussion:
Issue 13432: Typos in 6.3 Background (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: There are some simple typos in "6.3 Background." For example, the
phrase 'wit this specification' at the end of page 6 shall be 'with
this specification.'
Also, for the purpose of this specification, it is better to describe
explicitly that this specification can be applied to fields other than
robotics, such as GIS or LBS.
Resolution: Correct the typo and add some description in section 6.3.
Revised Text: Modify the paragraph which begins by “A Geographic Information System
(GIS) is one of the …” in section 6.3 as following:
Geographic Information System (GIS) is one of the most popular and established
systems that treats location information. Many spatio-temporal location related
specifications have been standardized in the International Organization for
Standardization (ISO/TC211), and there already exist versatile production
services based on these standards such as driving navigation systems or resource
databases. However, current GIS specifications are not powerful enough to
represent or treat information required in the field of robotics.
• Modify the paragraph which begins by “As can be seen from these
examples, …” in section 6.3 as following:
As can be seen from these examples, operations in robotics require a much more
detailed representation of location information. Still, interoperability with the
current GIS systems shall be supported. In this proposal, we define a new
framework for representing and treating location information suitable for robotic
use, by extending existing GIS specifications. Using the GIS specification as a
basis of the proposal will make it easy for robots to interconnect with existing
GIS-based systems and utilize existing geographic datasets. This will also ease the
use of this specification in the emerging fields of next-generation GIS systems,
sensor network systems or location based systems where advanced positioning
methods and complex data processing similar to robotics usage is required. Figure
2 illustrates the existing GIS standards that are related with this specification.
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13433: Overview in 6.3 Background needs to be modified (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The paragraph beginning with 'In order to fulfill the requirements...'
in 6.3 needs to be updated
Resolution: See issue 12916 for disposition
Revised Text:
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13434: Some description on Coordinate System usage shall be added (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: In this specification, the CS_CoordinateSystem class from ISO19111 is
used in a rather special manner than its original intention, such as
the assumption for using it in n-dimensional spaces (where n >
3) or using axis definitions in a more flexible manner. Although this
does not violate the original ISO specification, it is better to have
some notes.
Resolution: Add some description which clarifies the difference between the usage in this
specification and the implicit assumption made in the ISO 19111 specification.
Revised Text: Add the followings as a last paragraph of “7.3 Representing Robotic Localization
Results”, before “7.3.1 Relative and Mobile Coordinate Reference Systems.”
Note that, although the ISO 19111 specification assumes every CS to be 2 or 3
dimensional [ISO19111], in this specification, we do not assume any limitation on the
number of dimensions on any coordinate systems. This is to enable representation of
complex data such as feature points defined over multi-dimensional space. Also note that
this does not violate the ISO 19111 standard where no formal limitation is specified on
the number of dimensions. One issue is how to treat the attribute bounded to specific
feature in the real space such as axisDirection (type CS_AxisDirection) in
CS_CoordinateSystemAxis which is a mandatory attribute and where the type is defined
as an finite enumeration of direction names such as ‘north’ or ‘south’. It is clear that
these values are not suitable for some robotics usage such as for relative or mobile
coordinate reference systems. We thus recommend that implementers and users of this
specification to simply ignore this attribute and to set this value as the first element in the
enumeration, ‘north’, if necessary. This is a safe solution as we cannot expect GIS
systems to treat data based on this specification correctly; we only expect data from GIS
systems to be treated on systems based on this robotic specification.
Disposition: Resolved
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13435: Ambiguity in 'rla' attribute description in RelativeDatum class (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The description for 'rla' attribute in RelativeDatum class (Table 6)
is somewhat ambiguious. The second sentence 'If the base attribute
contains structural information, ' shall be modified as following: 'If
the base attribute contains structural information, or if the
coordinate reference system in target holds no relation with other
coordinate reference systems,'.
Resolution: Modify the description for 'rla' attribute according to the suggestion in the
summary.
Revised Text: Modify the description for the 'rla' attribute of Relative Datum class in Table 5
as follows:
A RoLo data specification indicating allowed structure for the ‘base’ attribute. If the
coordinate reference system in target holds no relation with other coordinate
reference systems, this may be omitted.
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13436: 'inStream' attribute in MobileDatum class shall be an object (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: 'inStream' attribute in MobileDatum class shall be an object, not n
reference to an object.
There seems to be no reason to make the 'inStream' attribute to be a
reference.
Resolution: Although the ‘Description’ part of MobileDatum explicitly says that ‘inStream’
attribute is an instance, it is described as a reference in the description of the
attribute. Fix this inconsistency and modify ‘inStream’ attribute in MobileDatum as
non-reference.
Revised Text: In Table 9 (MobileDatum), replace the description for ‘inStream’ attribute as
following:
InStream instance used in this datum.
Modify the definition of ‘inStream’ attribute in MobileDatum for C++ PSM as
following:
::RoLo::Interface::InStream inStream;
This change also include revisions by issue 12916.
Disposition: Resolved
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13437: 'numDists' attribute in Lineare Mixure Model class may be unnecessary (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: 'numDists' attribute in Lineare Mixure Model class may be unnecessary
The 'numDists' may not be necessary, as this value can be obtained
from the 'dists' attribute (which is a list).
Resolution: Delete 'numDists' attribute from Lineare Mixure Model class.
Revised Text: Delete 'numDists' attribute from Table 44.
Delete 'numDists' attribute from the class diagram for Linear Mixture Model
class in Figure 7.
Delete ‘numDists’ attribute from Linear Mixture Model class definition in C++
PSM (Error.hpp)
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13438: Typo in RoLo Attribute Definition class description (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: 'Of minOccurence...' shall be modified to be 'If minOccurence...'.
Resolution:
Revised Text:
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Discussion: As RoLo Attribute Definition class has been revised in #13322, no revision is
made here.
Disposition: Closed, no change
Issue 13439: Structural change for RoLo Service class (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The RoLo Service class may be modified not to inherit OutStream
class, as to clarify its functionality. In order for this change, the
following modification are required: 1. Change description at the
beginning of 7.5 Service Interface, 2. UML diagram fix, and 3. add
some attributes that were inherited from OutStream class. (Originally
proposed by Mr. Han, ETRI)
Resolution: In the beta1 specification, the RoLo service class which denotes the service
interface for Robotic Localization Service was described to be inherited from
OutStream class which denotes data output stream. This issue points out that
although from the point of functionality this is sufficient, from the meaning of the
classes this is somewhat misleading. Thus, we decided to change the structure
of RoLo service class so that it will not be inherited from OutStream, but instead
will contain OutStream instances as well as InStream instances.
Based on the proposed solution, we made a reconstruction of RoLo Service
class and related classes. Based on this change, some additional description is
also added for clarity.
Revised Text: dtc/2009-06-03 pages 152 - 156
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13440: Typo in Service Ability class description (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The phrase 'Service Parameter class' in the Service Ability class
description shall be 'Service Parameter Definition class'. UML diagram
shall be fixed, and also a subset note shall be added to the relation
between Service Ability class and Service Parameter Definition class
in the diagram.
Resolution:
Revised Text:
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed no change
Discussion: As ServiceAbility class has been revised by issue #13322, this issue is closed
without change.
Disposition: Closed, no change
Issue 13441: Inconsistency between PIM and PSM namings (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: As class names with blanks are not allowed in most programming
languages (or XML defnitions), different class names are used in PIM
and PSM. This should be fixed. One way is to remove blanks from PIM
class names.
Resolution: In addition to the item pointed out in the summary, some class names have the
¡¥RoLo¡¦ prefix and some does not (issue 13442). In order to resolve these
inconsistencies, the following changes are performed:
„O Remove blanks from PIM class names
„O Remove 'RoLo' prefix from PIM and PSM class names
„O Instead, use global namespace 'RoLo' in place of 'RLS'. As most class
definitions belong to some packages (from issue 12916), two namespaces
will be used for class definitions.
For example, ¡§RoLo Element Specification¡¨ class will be renamed as
¡§ElementSpecification¡¨ or ¡§::RoLo::Architecture::ElementSpecification¡¨ in full
resolution.
However, this change makes descriptions in natural language harder to read. To
make sentences simple, we follow the resolution from #13116 such that
whenever class names are used to indicate its instances, the names are written
in small letters in divided forms with a ¡¥RoLo¡¦ prefix. For example, ¡§xxx is a list of
references to PositionElement class instances¡¨ are rewritten as ¡§xxx is a list of
reference to RoLo position elements.¡¦
There also exist inconsistencies in attribute names (from issue 13998). To
reduce this inconsistency, we will modify attribute names based on the following
rules:
¡E Shorten the attribute names whenever it is possible, but keep it long when
it is not easy to infer the original name. Use the same attribute names across the specification whenever they
indicates the same meanings.
• Make attribute name to be plural when multiplicity of the attribute is able to
be more than 1.
For clarification, we will enclose the names of attributes and methods with a pair
of single quotation marks whenever they appear in natural language sentences.
Revised Text: dtc/2009-06-03 pages 159 - 168
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed issue
Issue 13442: Inconsistency in class names (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: Some classes have 'RoLo' prefix in their names, and others do
not. These namings shall be modified to follow some rule.
Resolution:
Revised Text:
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed no change
Discussion: See issue 13441 for disposition
Issue 13443: Typo: argument type for methods in RoLo Service class (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The argument types for 'setParameter' and 'getParameter' methods in
RoLo Service class shall be 'Service Parameter Set' class, not
'Service Parameter' class.
Resolution:
Revised Text:
Actions taken:
February 4, 2009: received issue
October 15, 2009: closed no change
Discussion: This issue is closed without change due to the changes in #13439.
Disposition: Closed, no change
Issue 13451: Missing references (rls-ftf)
Click here for this issue's archive.
Source: Shibaura Institute of Technology (Mr. Takeshi Sakamoto, sakamoto(at)globalassist.co.jp)
Nature: Uncategorized Issue
Severity: Minor
Summary: The RLS specification refers to some other specifications.
These references should be added. For example,
*[UML] Unified Modeling Language, Superstructure Specification version 2.1.2
http://www.omg.org/technology/documents/formal/uml.htm
*[XML] .....
Resolution: See issue 13427 for disposition
Revised Text:
Actions taken:
January 27, 2009: received issue
October 15, 2009: closed issue
Issue 13998: RLS - typos (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: - (Fig7/Tbl36) Reference to ISO19103 missing for Matrix::values
- (Fig7) Matrix.values shall be drawn as an composition to ISO19103::Number
- (Fig3) RelativeDatum.rla is not described as optional
- (Fig3) Description of RelativeDatum.base is hard to understand. Please clarify.
- (Fig5) RoLo Symbol Ref.point and DirectSymbol.coordinateReferenceSystem shall be described as association name. Also these association shall be arrowed.
- (Fig10) Position Element Concatenated Operation.childOperation, RoLo Data Concatenated Operation.childOperation and RoLo Data Transformation.usesOperation shall be described as derived attributes.
- (Fig11) association between User Defined Data Format and RoLo Architecture shall be drawn as an arrowed association between Specific Data Format and RoLo Architecture
- (Tbl33) Description shall say this class is abstract
- (Tbl35) Description shall say this class is abstract
- (Tbl35) Delete unnecessary 'Attribute' row from table
- (Tbl37) Add note indicating that covariance matrix need to be a square matrix
- (Tbl39) Uniform Gaussian, from its meaning, shall be a child class of Gaussian class. Thus, its 'stddev' attribute is unnecessary and 'covariance' attribute in Gaussian class shall be used. However, restriction shall be noted that this 'covariance' attribute can only be a matrix where row/col dimensions are limited to 1
- (Tbl43) reference to ISO specification missing in 'Derived From'
- (Fig7/Tbl43) For Weighted Distribution class, reference to ISO specification missing for type name (Probability) of attribute weight
- (Fig7) Weighted Distribution class's 'weight' shall be described as composition
- (Fig7,Tbl45) 'dists' attribute for Mixture of Gaussian class shall be restricted to Gaussian class, from its meaning
- Section number "6.2" in the sentence starting as "Note that derived attributes" in 7.1.1 is wrong; should be "7.2"
- (Tbl8) The word "Polar" shall be "polar" in the description line
- (Tbl46) Reference to ISO specification missing for typename of 'numeric' attribute
- (Fig9) 'elem' attribute in RoLo Data class shall be described as composition
- (Tbl55) Typo in description for 'elem' attribute: Ordered list for the... -> Ordered list of the...
- (Tbl36) Describe that dimensions of matrix shall always be 1 or more than 1; that is, matrix will never be 0-dimensional
- Information on Submitters and Submitting Organizations shall be added
- (Fig5,Tbl21) 'dimension' attribute for DirectSymbol class is redundunt; it can be obtained from the list 'coordinate'.
- (Fig22,Tbl134) The names of setParameter/getParameter methods are strange. It is the parameter values that is to be set/get, not the parameter itself. It shall be renamed as setParameterValue/getParameterValue
- (Fig22,Tbl135) The name 'getResult' in OutStream class is not consistent with the 'setData' method in InStream class. This shall be changed to 'getData'
- Attribute names are inconsistent throughout the specification. Some are plural and some are not plural, even if it denote a list of values. Some are abbreviated and some are not. This inconsistency shall be fixed.
Resolution: Fix typos based on the suggestions
Revised Text: see dtc/2009-06-03 pages 173 - 178
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 13999: (7.4.1) Timestamp for common data types (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: (7.4.1) Timestamp for common data types are described as 'POSIX time' but it is not clear which type shall be actually used
Resolution: See issue 13120 for disposition
Revised Text:
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14000: (Annex A) Mapping between PIM and C++ PSM is not clear (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: (Annex A) Mapping between PIM and C++ PSM is not clear
Resolution: Describe how C++ PSM was mapped from PIM
Revised Text: Replace the first paragraph of section 8.1 (C++ PSM) as following:
In this section, we show a PSM in C++ language based on the PIM described in
section 7. This PIM-PSM mapping is based on the following rules:
• The return values of methods are assumed to be mapped as exceptions. Thus, in
this PSM, no explicit description is given.
• When methods had only a single 'out' argument, it was mapped as return value
of the corresponding function.
• The 'in' arguments to methods are mapped as method arguments with the 'const'
modifier.
• Arguments which were based on non-primitive types are passed by reference.
• An attribute or an argument that is marked to occur more than once and is
marked as unordered is mapped to '::std::list'. If marked as ordered, it is mapped
to '::std::vector'.
• When an attribute is shown as an aggregation or as a derived attribute, or when
an argument indicates a reference to other object, it is mapped as a pointer.
• CharacterString is mapped as '::std::string'.
The following shows the resulting C++ header files.
Note: the above content reflects the resolution from issue 14014.
Disposition: Resolved
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14001: (7.5) In the filter condition section (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: (7.5) In the filter condition section, an extention to RoLo Service class is described which enables RoLo Services to specify filter conditions. However, this is not a good way to implement, and will lead in lack of interoperabiility. Instead, it is much better to make it so that conditions can be specified through some parameters
Resolution: See issue 13184 for disposition
Revised Text:
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14002: 7.5) Filter condition requires naming of RoLo data elements. However, the naming rules are not clear. (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: 7.5) Filter condition requires naming of RoLo data elements. However, the naming rules are not clear.
Resolution: See issue 13184 for disposition
Revised Text:
Actions taken:
June 18, 2009: recerived issue
October 15, 2009: closed issue
Issue 14003: Fig3,Tbl9,Fig22,Tbl134,Tbl140) Mobile Datum class, RoLo Stream class and RoLo Service (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: Fig3,Tbl9,Fig22,Tbl134,Tbl140) Mobile Datum class, RoLo Stream class and RoLo Service class holds public attributs and at the same time provide methods for obtaining the value of the same public attributes. This is redundunt and is losing flexibility of dynamically creating the returned values
Resolution: Make the attributes as protected
Revised Text: be ¡¥protected¡¦
„O In Figure 22 and Table 134, modify ¡¥ability¡¦ attribute in RoLo Stream class to
be ¡¥protected¡¦
„O In Figure 22 and Table 140, modify ¡¥ability¡¦ attribute in RoLo Service class to
be ¡¥protected¡¦
„O in C++ PSM, modify the following:
„« Move ¡¥inStream¡¦ variable declaration in MobileDatum class definition to
¡¥protected:¡¦section
„« Move ¡¥ability¡¦ variable declaration in RoLoStream class definition to
¡¥protected:¡¦section
„« Move ¡¥ability¡¦ variable declaration in RoLoService class definition to
¡¥protected:¡¦section
Disposition: Resolved
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14004: Fig5,Fig9) RoLo Symbolic Position class and RoLo Position class (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: Fig5,Fig9) RoLo Symbolic Position class and RoLo Position class are specified to be 'union' but it is better to leave them unspecified. 'Union' is hard to implement for some programming language
Resolution: Stop using ‘union’ as suggested in the summary, For the C++ PSM, as we found
that some compilers are unable to treat ‘union’ with elements that owns
constructors, we will also stop using union and make the elements as pointers.
Implementers need to be aware that only one of the elements shall be effective.
Revised Text: In Figure 5, remove the ¡¥Union¡¦ stereotype from RoLo Symbolic Position
class and add multiplicity [0..1] for its attributes ¡¥direct¡¦ and ¡¥indirect¡¦.
„O In Figure 9, remove the ¡¥Union¡¦ stereotype from RoLo Position class and add
multiplicity [0..1] for its attributes ¡¥symbolic¡¦ and ¡¥numeric¡¦.
„O In C++ PSM, redefine RoLoSymbolicPosition and RoLoPosition as following:
class RoLoSymbolicPosition
: public ::ISO19111::IO_IdentifiedObjectBase
{
public:
DirectSymbol* direct;
RoLoSymbolRef* indirect;
};
class RoLoPosition
: public ::ISO19111::IO_IdentifiedObjectBase {
public:
RoLoSymbolicPosition* symbolic;
::ISO19107::GM_Position* numeric;
};
Disposition: Resolved
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14005: (Fig7) Matrix class is specified as "DataType" but this is unnecessary (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: (Fig7) Matrix class is specified as "DataType" but this is unnecessary
Resolution: Remove the stereotype as suggested
Revised Text: In Figure 7 (Error Information), remove the stereotype ‘DataType’ from Matrix
class.
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14006: Tbl119-Tbl129) (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: Tbl119-Tbl129) A set of class related to attribute/parameter/attribute definition is defined, but usage of these classes are not clear. For example, difference between 'parameter' ad 'attribute' is not clear, or how 'attribute definition' classes are related with 'attribute' classes is not clear. Please clarify. Also, there seems to be too many classes of simliar meanings.
Resolution: See issue 13322 for disposition
Revised Text:
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14007: Tbl124) Specifying RoLo Attribute Singlevalue class as an enumeration (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: Tbl124) Specifying RoLo Attribute Singlevalue class as an enumeration is hard to implement for some programming languages. It also lack extendability for new data types. It is better to specify it as template class.
Resolution: See issue 13322 for disposition
Revised Text:
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14008: 7.6) RoLo service is described to have only one output stream, but this is sometimes too difficult to use (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: 7.6) RoLo service is described to have only one output stream, but this is sometimes too difficult to use. It should allow multiple output streams, such as those of server daemons, of the same nature.
Resolution: See issue 13322 for disposition
Revised Text:
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Discussion: Increasing the number of output streams that a RoLo service can handle may
violate the FTF rule. Thus, we will not make changes for this issue. However, the
resolution for issue 13439 may allow RoLo services to own multiple output
streams.
Disposition: Closed, no change
Issue 14009: (Tbl135/Tbl136) (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: (Tbl135/Tbl136) The 'connect' method for initiating service usage is defined for 'stream' classes. However, in most existing systems or languages, connection initiation is performed toward services, and not toward streams. This should be moved as a method for RoLo Service
Resolution: See issue 13439 for disposition
Revised Text:
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14010: (Tbl135/Tbl136) The 'connect' methods (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: (Tbl135/Tbl136) The 'connect' methods lacks the facility to let the targetting module to know about which module is making the connection. This knowledge is required for some methods such as 'getConnectedStream' method in RoLo Stream class. Some means shall be provided to make this specification independent to underlying middleware
Resolution: See issue 13439 for disposition
Revised Text:
Actions taken:
June 18, 2009: closed issue
October 15, 2009: closed issue
Issue 14011: Tbl134) The behavior of setParameter method in RoLo Stream class is unclear (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: Tbl134) The behavior of setParameter method in RoLo Stream class is unclear. When this method is specified an unknown parameter, it shall return some error code. But which type of error code is returned is not described. Please clarify
Resolution: Clarify the behavior of ‘setParameter’ method by adding description
Revised Text: Add the following to the description of ‘setParameter’ method in Table 134.
If some nonexistent or inconfigurable parameters were specified,
UNSUPPORTED_PARAMETER or BAD_PARAMETER will be returned
respectively.
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14012: Tbl130) Two types of data passing mode specified in StreamType enumeration is not clear (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: Tbl130) Two types of data passing mode specified in StreamType enumeration is not clear. It is currently defined to be relative to the stream in concern, but shall be defined by which side of data flow is holding the trigger for data passing.
Resolution: As specified in the summary, PUSH/POP in StreamType enumeration is not clear.
Here we alter the definition to make them clear.
Revised Text: Modify Table 130 (StreamType Enumeration) as following:
PUSH Indicates that data passing is performed in PUSH mode, i.e. OUT side triggers dat
a passing.
PULL Indicates that data passing is performed in PULL mode, i.e. IN side triggers data
passing.
Disposition:
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14013: Tbl130,Tbl136) The 'setData' method in InStream class (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Clarification
Severity:
Summary: Tbl130,Tbl136) The 'setData' method in InStream class may be used for 'PUSH' type data passing. However, it is quite common that input side streams can not process all the data received. In such case, some flow control mechanism is required. Such mechanism is required for practical use. This should be defined in the specification, not depending on the middleware's ability of flow control.
Resolution: Flow controls for ‘getData’ and ‘setData’ methods in RoLo streams originally
assumed that flow control will be done by the underlying middleware. However,
as pointed out in the summary, it will lead to middleware dependency which is
not what we want. Thus, here we prepare some methods for controlling the flow.
We add three methods, ‘activate’, ‘deactivate’ and ‘isActivated’ to OutStream
class.
Revised Text: Add the following methods in Table 135 (OutStream class)
activate Activate stream output. Only meaningful on PUSH mode.
deactivate Deactivate stream output. Only meaningful on PUSH mode.
isActivated Query whether this stream is activated or not.
out status Boolean If activated true, otherwise false.
„O Add the following methods to OutStream class in Figure 22 (RoLo Service)
„« activate(): Returncode_t
„« deactivate(): Returncode_t
„« isActivated(out status: Boolean) : Returncode_t
„O Add the following method definitions to OutStream class in C++ PSM
(Service.hpp)
„« void activate(); „« void deactivate();
„« bool isActivated();
Disposition: Resolved
Actions taken:
June 18, 2009: received issue
October 15, 2009: closed issue
Issue 14014: (C++-PSM) Methods shall better send return codes as exceptions (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: As in most C++ implementations, it is better if methods will return
the return codes (Returncode_t) as exceptions.
Resolution: All the class methods (functions) in beta1 specification’s C++ PSM were
returning Returncode_t value. Thus, the attributes marked as ‘out’ were handled
as by-reference arguments that were normally implemented by pointer passing.
Besides the C++ style pointed out in the summary, passing pointers are not a
good manner for effective development. Thus, we will change C++ PSM
definitions so that methods will use exceptions for passing return codes. The ‘out’
attributes will be passed as return values whenever possible.
Revised Text: Modify method definitions in C++ PSM as following:
„O ¡¥getInstream¡¦ in MobileDatum class
const InStream& getInStream();
„O ¡¥getAbility¡¦ in RoLoStream class
const StreamAbility& getAbility();
„O ¡¥setParameter¡¦ in RoLoStream class
void setParameter(const StreamParameterSet& parameter);
„O ¡¥getParameter¡¦ in RoLoStream class
const StreamParameterSet& const getParameter();
„O ¡¥getService¡¦ in RoLoStream class
const RoLoService& getService(); ¡¥getConnectedStream¡¦ in RoLoStream class
const RoLoStream& getConnectedStream();
„O ¡¥isConnected¡¦ in RoLoStream class
bool isConnected();
„O ¡¥disconnect¡¦ in RoLoStream class
void disconnect();
„O ¡¥connect¡¦ in OutStream class
void connect(const InStream& target)
„O ¡¥getResult¡¦ in OutStream class
const RoLoData& getResult();
„O ¡¥connect¡¦ in InStream class
void connect(const OutStream& target);
„O ¡¥setData¡¦ in InStream class
void setData(const RoLoData& data);
„O ¡¥getNumInStream¡¦ in RoLoService class
int getNumInStream();
„O ¡¥getInStream¡¦ in RoLoService class
const InStream& getInStream(int index);
„O ¡¥adjust¡¦ in RoLoService class
void adjust(const RoLoData& data);
„O ¡¥getChild¡¦ in RoLoService class
const RoLoService& getChild(int index);
„O ¡¥getAbility¡¦ in RoLoService class
const ServiceAbility& getAbility();
„O ¡¥setParameter¡¦ in RoLoService class
void setParameter(const ServiceParameterSet&
parameter);
„O ¡¥getParameter¡¦ in RoLoService class const ServiceParameterSet& getParameter();
Disposition: Resolved
Actions taken:
June 19, 2009: received issue
October 15, 2009: closed issue
Issue 14015: Subset specifications (Fig22) (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The following associations shall be composition and shall have subset
specification.
RoLo Stream - Stream Ability
RoLo Service - Service Ability
Resolution: The RoLo abilities that RoLo streams and service instances hold shall be in an
aggregation relation, not in a composition relation. This is because several
instances of RoLo streams or services may share the same RoLo abilities. In
order to clarify this, we modify the UML diagram to show this explicitly.
Revised Text: Modify Figure 22 (RoLo Service) as following:
„O Change the association between RoLo Stream and Stream Ability to show an
aggregation on RoLo Stream side. Also, add association end name ¡¥ability¡¦
and remove the ¡¥ability¡¦ attribute from RoLo Stream class.
„O Change the association between RoLo Service and Service Ability to show
an aggregation on RoLo Service side. Also, add association end name
¡¥ability¡¦ and remove the ¡¥ability¡¦ attribute from RoLo Service class.
„O To both associations, add subset description ¡§subset ability¡¨
Modify C++ PSM as following:
„O Change the ¡¥ability¡¦ class variable definition in RoLoStream class
(Service.hpp) as following:
Ability *ability;
„O Remove the ¡¥ability¡¦ class variable definition in RoLoService class
Add the following note to Table 134 (RoLo Stream) Note: Values for the attribute ‘ability’ which is derived from parent class shall be
limited to instances of StreamAbility or its inherited classes.
Add the following note to Table 140 (RoLo Service)
Note: Values for the attribute ‘ability’ which is derived from parent class shall be
limited to instances of ServiceAbility or its inherited classes.
Disposition: Resolved
Actions taken:
June 19, 2009: received issue
October 15, 2009: closed issue
Issue 14016: Parent class wrong for RelativeCRS (rls-ftf)
Click here for this issue's archive.
Source: JARA (Mr. Shuichi Nishio, nishio(at)ieee.org)
Nature: Uncategorized Issue
Severity:
Summary: The parent class for RelativeCRS is described as SC_CRS but is
actually SC_EngineeringCRS.
Resolution: Correct the error.
Revised Text: Change the C++ PSM definition for RelativeCRS (RelativeCRS.hpp) to be
inherited from ::ISO19111::SC_EngineeringCRS
Actions taken:
June 19, 2009: received issue
October 15, 2009: closed issue