From fujimura@etl.go.jp Wed Oct 19 14:30:24 1994 Received: from etlpost (etlpost.etl.go.jp) by dkuug.dk with SMTP id AA16282 (5.65c8/IDA-1.4.4j for ); Wed, 19 Oct 1994 14:30:24 +0100 Received: from etlpom.etl.go.jp by etlpost (8.6.9+2.4W/2.7W) id WAA05136; Wed, 19 Oct 1994 22:29:24 +0900 Return-Path: Received: by etlpom.etl.go.jp (4.1/6.4J.6-ETLpom.MASTER) id AA11430; Wed, 19 Oct 94 22:29:23 JST Received: by etlcom.etl.go.jp (4.1/6.4J.6-ETL.SLAVE) id AA08881; Wed, 19 Oct 94 22:29:22 JST Date: Wed, 19 Oct 94 22:29:22 JST From: fujimura@etl.go.jp (Koreaki Fujimura) Message-Id: <9410191329.AA08881@etlcom.etl.go.jp> To: sc24@dkuug.dk Subject: Japan's comments on DIS 12089 X-Charset: ASCII X-Char-Esc: 29 ----------------- Title: Japan's comments on DIS 12089 -- Information Technology - Computer Graphics and Image Processing -- Encoding for the Image Interchange Facility (IIF) Source: National Body of Japan The National Body of Japan disapproves Draft Amendment ISO/IEC DIS 12089 with the following comments. -- TECHNICAL COMMENTS -- T1) Global: This draft defines only one encoding method which can handle a wide variety of packing schemes but puts a heavy burden on each parser. Because there is no concept of conformance level, an application which needs only a limited data structure has to implement the full capability now. The method shall be simplified (See T2) to lessen the load or the conformance level shall be introduced. T2) Page 8, EncodingRules: The way of structuring data allowing any combination of sequences of IterationAndAlignment and recursive use of EncodingRules is unnecessarily complex. We propose to separate the above mentioned two facilities as follows: EncodingRules ::= CHOICE { uniform-encoding [1] UniformEncoding, hierarchical-encoding [2] HierarchicalEncoding } UniformEncoding ::= SEQUENCE { iteration-and-alignments [0] SEQUENCE OF IterationAndAlignment, components-encoding [1] SEQUENCE OF ElementEncoding } HierarchicalEncoding ::= SEQUENCE { iteration-and-alignment [0] IterationAndAlignment, components-encoding [1] SEQUENCE OF ComplexEncoding } ComplexEncoding ::= CHOICE { elementary-component [0] ElementEncoding, complex-component [1] HierarchicalEncoding } T3) Page 13, ElementEncoding: It is not appropriate to use "IntegerEncoding", which may be unsigned or signed, for "non-negative-integer" and "signed-integer". The two lines in the production for ElementEncoding non-negative-integer [1] IntegerEncoding, signed-integer [2] IntegerEncoding, should be changed to non-negative-integer [1] UnsignedIntegerEncoding, signed-integer [2] SignedIntegerEncoding, The related changes are as follows: a) Change the definition of IntegerEncoding to IntegerEncoding ::= CHOICE { non-negative-integer [1] UnsignedIntegerEncoding, signed-integer [2] SignedIntegerEncoding } b) Add the definition of UnsignedIntegerEncoding and SignedIntegerEncoding UnsignedIntegerEncoding ::= number-of-bits INTEGER SignedIntegerEncoding ::= number-of-bits INTEGER T4) Page 16, FixedPointEncoding: This encoding is only exported and not referred from other components. It shall be a part of ElementEncoding by adding it to the production of ElementEncoding or RealEncoding. --------------- -- EDITORIAL -- E1) Title page: Page iii, Foreword: The third element of the title, "Encoding for the Image Processing and Interchange Standard (IPI)" shall be removed because the current title, consisting of four elements, is too lengthy. If the third element remains, it shall be changed into "Image Processing and Interchange (IPI)" because "encoding for" is duplicated with the fourth element and "standard" is not necessary. E2) Page 1, 1.1 Scope, the second paragraph: Page 3, 3 Definitions and abbreviations, NOTE: The phrase "Image Processing and Interchange Standard (IPI)" shall be changed into "Image Processing and Interchange (IPI)" because the word "Standard" is not included in the title of the original standard. E3) Page 1, 1.2 Field of application: The text ... whenever images are interchanged, according to the IIF-DF ... shall be changed into ... whenever images are interchanged using the entity ExternallyDefinedDataUnit, according to the IIF-DF ... because there are no need for referencing this standard if images are interchanged using only other types of data units. E4) Page 2: 2 Reference: The title of this clause shall be "Normative references". E5) Page 2: 2 Reference: The reference to [ISO/IEC 8632:1987] shall be removed because the 1987 version is superseded by [ISO/IEC 8632:1992] which is also referenced here. E6) Page 2: 2 Reference: The reference to [ISO/IEC CD 8825-2.2:1993] shall be removed because it is only referenced in a Editorial Remark (Page 4) to be removed at the time of publishing. E7) Page 7 -END: All the words "byte" shall be "octet" as are done in CGM Part 3, because the number of bits per byte is not limited to eight in ISO/IEC documents. E8) Page 8 - END: All the words "explicite" and "implicite" shall be "explicit" and "implicit". E9) Page 11-12, ComponentEncoding: The indentation in the line "components-encoding" in Example 1) and 2) shall be equal to that in the corresponding line "iterations-and-alignments". E10) Page 15, IntegerEncoding: In the mathematical definition for the signed case n-i 2 shall be n-1 2 E11) Page 16, FixedPointEncoding: The name does not match with its technical content. It should be changed into "LinearEncoding" or "ScaledEncoding". ------------------