[CL] Corrigendum Defect Reports

[Pat Hayes]

Submitter (include email): Pat Hayes (phayes at ihmc.us)
For review by: JTC1/SC 32/WG2
Defect report concerning: ISO 24707:2007

EDITORIAL DEFECTS

Qualifier: typographical error
References in Document: Page 3, clause 3.8
Nature of defect: '3.8' and 'dialect' should be on separate lines

Qualifier: Sense unclear
References in Document: Page 7, clause 5.1.1
Nature of defect: First sentence uses "shall", second sentence uses 
"will". It is not clear if this indicates a difference of sense. If 
it does, this difference should be clarified. If it does not, a 
single term should be used consistently.

Qualifier: wording misleading
References in document: Page 13, section 6.2, first sentence
Nature of defect: the phrase "mathematical structures" is potentially 
misleading as it can be construed as being contrasted with other 
'kinds' of structure. The qualifier "mathematical" is not necessary 
to the sense of the sentence, so should be deleted.

Qualifier: grammatical error
References in document: Page 14, fourth paragraph, second sentence
Nature of defect: "differ with"  should be "differ from"

Qualifier: grammatical error
Page 18, section 6.4, last sentence.
Nature of defect: Missing final stop.

Qualifier: typographical error
Page 20, section 6.6.1, last sentence
Nature of defect: Write "shall" in bold face

Qualifier: wording misleading
References in document: Page 21, section 7.1.2, paragraph 8, last sentence.
Nature of defect: "less complete" is like 'slightly pregnant'.
Change "... may be less complete..."  to  "... may be incomplete..."


TECHNICAL DEFECTS

Qualifier: needless restriction on irregular sentences
References in document: Page 16, second paragraph, fourth sentence.
Nature of defect: The qualification "... but shall not use irregular 
sentence forms to represent content that is expressible in Common 
Logic text." is unnecessary and potentially harmful to 
interoperability under certain conditions. It is not mentioned in 
section 7 defining conformity, so may be left over from an earlier 
version by mistake.
** It should be deleted.

Qualifier: misleading wording
References in document: Page 18, first paragraph, third sentence
Nature of defect: Reference to 'guarded' is inappropriate, as 
guarding is not mentioned in the CL main specification, and the CLIF 
de-sugaring of modules in Annex A uses only quantifier restrictions.
** Delete the words "or guarded"

Qualifier: erroneous/inconsistent statement
References in document: Page 18, first paragraph, last sentence
Nature of defect: this sentence seems to contradict the formal semantics.
** It should read:
"The module name indicates a local universe of discourse special to 
the text in the module, which can be asserted to be equivalent to a 
common universe of discourse associated with the dialect by sentences 
outside the module."

Qualifier: typographical error producing inconsistency
Reference in document: page 29, A.2.3.11
** The second production in the section should read
text = { phrase }
rather than
cltext = { phrase }

Qualifier: Underdefined mapping leading to internal inconsistency
References in document: page 31, table A.2, lines 4 and 5.
Nature of defect: The ordering of the names X1 ... Xn is not 
specified, and there is no natural way to specify it. Hence this 
'mapping' is undefined, and hence the CLIF semantics is undefined, 
for the guarded quantifier construction.
Solution proposed by submitter: Remove the 'guarded quantifier' 
construction from the CLIF specification in Appendix A. (Note: this 
solution was arrived at by a unanimous consensus among the technical 
committee which designed the specification, with no dissention in an 
open forum. The guarded quantifier construction is currently the 
subject of on-going research, is not required by the main CL spec and 
has not been implemented or used by any known CLIF project.)
** This requires the following editorial changes to the document, in 
addition to those noted elsewhere:
(a)  Delete lines 4 and 5 from table A.2 on page 31
(b)  In A 2.3.8, page 29: remove the words "and may be guarded;"
(c)  In A 2.3.8, page 29: remove the option [ interpretablename ] 
from the first production
(d)  In A.1, page 24: remove item 11.
(Note, this construct is also described in Annex C, but that is an 
error in any case as the guarded construction is not mentioned in the 
CL abnstract syntax.)

Qualifier: Consistency problem with external specification
References in document: Annex A
Nature of defect: The CLIF syntax uses several reserved names with an 
enclosed colon, such as "cl:imports". It has been pointed out by 
commentators that this is likely to give rise to clashes with a 
widely used XML convention called Qnaming, whereby a 'root' URI is 
abbreviated by a short prefix defined in the XML header, which is 
prefixed to a local name separated by a colon, such as 'rdf:type' 
being an abbreviation for the full URIreference 
'http://www.w3.org/1999/02/22-rdf-syntax-ns#type'.
Solution proposed by submitter: As this 'internal colon' convention 
serves no real purpose, simply eliminate it. This does not change the 
language essentially and would require only minimal changes to any 
extant CL parsers. Since these 'words' are reserved, it is better to 
not use up commonly used words such as 'text' or 'comment', so rather 
than eliminate the 'cl:' prefix entirely, we propose abbreviating it 
without the comma, producing entirely artificial neologisms.
** This requires the following editorial changes to the document:
throughout Annexes A and B, make the following textual substitutions:
'clmodule' for 'cl:module'
'clexcludes' for 'cl:excludes'
'climports' for 'cl:imports'
'cltext' for 'cl:text'
'clcomment' for 'cl:comment'
'clroleset' for 'roleset:'
These occur on pages 27 - 31, page 52 (table B.1, line E5, second 
part), page 53, lines E12, E14, E17, E19, E20.

------

Defects in Annex C will come later, they need more work.

Pat

-- 
---------------------------------------------------------------------
IHMC		(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32502			(850)291 0667    cell
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes