- From: olivier Thereaux <ot@w3.org>
- Date: Fri, 6 Aug 2004 11:51:22 +0900
- To: public-ws-chor-comments@w3.org
- Message-Id: <7FEA59C7-E753-11D8-80A3-000A95E54002@w3.org>
Dear WS-Chor WG, Here are a few notes and comments on the first public working draft of the WS Choreography Model Overview document. Analysis of the draft document was made with QA in mind, and comes from someone without advanced knowledge of Web Services. Generally speaking, the document is very easy to read, even for someone without particular knowledge of WS choreography, and especially for a first draft. With regards to the goal of introducing the topic to its audience, the document is a success. [QA: Status] Speaking of goal and scope, however, it is not altogether clear what the WG is aiming at with this document. The status only states that it [[ is not a chartered deliverable of the Web Services Choreography Working Group; however, it is the corner stone of the Choreography Definition Language, which is a chartered deliverable.]] Which in itself is fine, we know what the document is not, but not really what it is. At times it feels like a primer, at times like a requirements document, and at times, like a technical specification. [QA: Conformance] The document would thus benefit from a clear definition of its status and goal. Specifically, is it made to be implementable? Or is it just an informative overview? Unless its final status is a NOTE, this documentation will have to be implemented. It would be a good move to make its structure evolve towards a better definition of what is normative and what is not, what it means to conform to the document (the addition of a conformance section would be a first good step. Another step, when the conformance model is defined, will be to clarify how testing will be done... [QA : informative vs Normative] One particular aspect of making the document implementable will also be to clarify the usage of the diagrams. These diagrams are relatively clear and useful (in spite of not having a decoding guide, which should certainly be added), and they offer a graphical representation of the associated text. Or does, in fact, the text offer an explanation for the diagrams? How do you guarantee that both are equivalent (still keeping implementation in mind here)? Hardly feasible, perhaps. But then which holds the normative status, which the informative? Generally speaking, a sound practice would be to clearly mark normative and informative parts of the document. The document holds a lot of examples (which is great) but that might dilute the implementable/testable content. [QA: dependencies] Another point probably worth clarifying in future versions is the dependencies with other documents. I assume this will be addressed with an eventual clarification of the status. The current draft already has hints of dependencies definitions (e.g Compatibility with other Specifications. in 1.2) but it seems to lack formal definitions of the dependencies. The document states this overview will [[work alongside and complement]] other specifications. But How? Granted, the answer is partly available within the bulk of the document (e.g [[An Interaction always involves the exchange of information between two Roles in a Relationship that conform to a Message Exchange Pattern as defined by WSDL 1.2.]]). It might be a good idea to explain your dependencies and relationship model in more details. In a possibly related note, I found myself asking a few time "yes, but how?" when the documents states "this can do that". e.g [[An Import statement can contain references to a complete Choreography or part of a Choreography.]] in 3.3.2. This might, however, come from a lack of understanding of the technology, but if the "how" will be explained elsewhere, the specification should refer to that. [Editorial] As I wrote earlier, the document is particularly clear and readable. At times, however, the phrasing tends to be long and awkward. e.g [[It's purpose is to provide an information model *that* describes the *data* and the relationships between them *that* is needed *to* define a choreography *that* describes the sequence and conditions *in which* the data exchanged between two or more participants *in order to* meet some useful purpose.]] As a native french speaker I am often reminded that the French tend to use and abuse subordinate clauses when writing in english... Not calling anyone french, obviously, but the text of the specification tends to suffer from this trend, too. Using shorter sentences and strict structures might be a good idea. Also, in the sentence above: purpose... purpose is a bit awkward. Also: s/It's/Its/ in 1.2: s/participants take part in/participants in/ There are a couple typos (an ampersand missing in 3.4, and "choreogaphy") http://www.w3.org/2002/01/spellchecker?uri=http://www.w3.org/TR/ws- chor-model/ This tool actually seem to stumble on quite a few acronyms, you may want to use proper markup to expand them, and include a glossary of concepts introduced by the specification (especially important for an overview, I reckon). I hope these comments will help you make this document a very good specification. Feel free to contact me if any comment I made was not clear enough. Regards, -- olivier Thereaux - W3C / QA http://www.w3.org/People/olivier/
Received on Thursday, 5 August 2004 22:51:43 UTC