{\rtf1\mac\ansicpg10000\cocoartf824\cocoasubrtf100
{\fonttbl\f0\fswiss\fcharset77 Helvetica;}
{\colortbl;\red255\green255\blue255;}
\margl1440\margr1440\vieww11960\viewh6540\viewkind0
\pard\tx720\tx1440\tx2160\tx2880\tx3600\tx4320\tx5040\tx5760\tx6480\tx7200\tx7920\tx8640\ql\qnatural\pardirnatural

\f0\fs24 \cf0 Responses to issues against section 2.3 (Semantics)\
\
30. Removed general text on context from revised draft. The concept is 'replaced' by 'real world effect'. The suggestion re. transaction context is noted. However, transactions are positioned within the process model in revised draft.\
\
32. Removed paragraph. However, what a service *is* is a subtle issue. A service is an exposed functionality ... Personally, I find that kind of language to be obscuring more than enlightening.\
\
33. I find the comment incomprehensible. How can a service differ depending on the context -- surely, we are talking about different *services* if they behave differently.\
\
34. There is a new diagram\
\
35. I cannot conceive of a higher-level view of a service than that related to its purpose. (This service is for ...)\
\
36. The suggestion is architectural in nature. Not appropriate for a section on semantics. \
\
37. The semantics has been reworked into a different kind of layering -- data, process/behavior and real world effect\
\
38. New section on description and metadata in revised section\
\
39. The layers analysis was too architectural. Replaced with analysis as suggested above\
\
40. There is a new section on the limits of description. Also, in reality not everything will be written in a formal notation -- there is always going to be some informal documentation/description.\
\
41. Vocabulary removed. Some discussion on ontologies in data model section\
\
42. Offending section removed.\
\
43. Accepted.\
\
44. Autonomy section removed.\
\
45. Misunderstanding. However, like distributed deployed, deployment across ownership boundaries is a key reason for wanting to use SOAs.\
\
46. Autonomy section removed. It was intended as a placeholder -- never for the semantics section\
\
47. Removed\
\
48. Data model -- got it\
\
93  Autonomy section removed, pending suitable new location\
\
99  To say that semantics is meaning is tautological. Surely purpose is part of the semantics of a service -- hence the new analysis.\
\
102. In general, yes, metadata carries the machine processable implication\
\
103. Not appropriate to be too specific about purposes of metadata. However, new section does try to highlight the role of metadata\
\
104 Ontology is a deep and broad topic. I do not think that a reference model is the right place to call out specific services as being required.\
\
105. Have re-analysed semantics in terms of data/process/real world effect. (Actually there is much more to RWE than I can write) Context is too over-used a word\
\
106. Contracts do not necessarily imply an immediate transaction.\
\
107. Context has been re-analysed.\
\
108. Section re-drafted. Original comment due to mis-formatting of original section (i.e. it *was* a bulleted list)\
\
109 Ditto\
\
160. Ownership domain is a good candidate for a glossary entry (and PhD thesis)\
\
161. No, for a successful interaction they must have a congruent understanding -- not the same understanding. (The latter is impossible, the former sufficient)\
\
183. Cannot locate reference\
\
184. Have reworked section, no longer layered.\
\
185. Original was mist-formatted -- they were bullets. Now redone.\
\
186. Noted. Will rework in revised section (**Still to be done **)\
\
37? Good suggestion. -- single section on metadata\
\
\
\
\
\
\
}