- From: Addison Phillips [wM] <aphillips@webmethods.com>
- Date: Tue, 27 Jan 2004 14:12:44 -0800
- To: "Mike McKenna" <mgm@globalisation.org>, "I18n WSTF" <public-i18n-ws@w3.org>
See notes below. Some elision for clarity. Addison P. Phillips Director, Globalization Architecture webMethods | Delivering Global Business Visibility http://www.webMethods.com Chair, W3C Internationalization (I18N) Working Group Chair, W3C-I18N-WG, Web Services Task Force http://www.w3.org/International Internationalization is an architecture. It is not a feature. -----Original Message----- AP> 4.1.1 locale neutral was changed as noted "Time of day, although it is presented with different formats around the world, is measured the same way everywhere and a standardized single frame of reference is available to be used." maybe change to: Time of day, although it is presented with different formats and perhaps different calendars around the world, if using a standardized method is measured the same way everywhere and a single frame of reference is available to be used. AP> See new edits, which are pretty close to this. "... including shifting the time into the local time zone." maybe change to "... including shifting the time into the local time zone or even a local calendar, e.g., Arabic or a Japanese imperial calendar. AP> The example explicitly references only time-of-day. Are there any cases where calendar would play into this? AP> I can think of one thing additional thing (not added), which is that we might reference the use of 24hour vs. am/pm. 4.1.1.2 ... does not require extra locale information ... AP> Changed to "The Web service description does not need to include extra locale information..." 4.1.1.3 No additional locale attributes or information are required for this pattern a locale neutral service. AP> Changed to "No additional locale attributes or information are required for the locale neutral pattern." 4.1.2 ----- "This may be due to how the service's code is written, how the provider is configured ..." => awk, how about "This may be due to the way the service's code is written, the provider's configuration ...", or, "This may be due to the way the service's code is written, the way the provider is configured ..." 4.1.2 ... the service provides a specific localized set of behaviors ... => how about ... the service provides a set of specific localized behaviors ... AP> See new text... 4.1.2.1 ------- This is not really that huge of an example. But we could simplify it - order processing in local language only, one and only one service/wsdl provided per locale due to locale processing limitations. AP> How about something altogether simpler: list the names of the days of the week. Sort this list of strings.... We should discuss this block of stuff. Add "GetProductInventory" as a name for the service. See notes to 4.1.3.1 AP> Changed as follows: <head>Example: 'GetProductInventory' Service Returns List of Products in Stock</head> <p>A company is in 3 markets: the United States, Japan, and Germany. Each market is supported by a local warehouse. A Web service, 'GetProductInventory', is provided to indicate the availability of parts. For a given part number, the following information is returned: part number, quantity, language, description, size, and list price.</p> // removal of Andrea's remarks about use of booze in example I've got a good example I use: "wrench". Tools (and vegetables) are great because they really are different between almost every locale. Here are a few translations: AP> ... to be discussed... Are there still systems with 7-bit ASCII (answer: Cobol is still alive :-( ) AP> Not sure that we want to deal with this here. We have scenarios later?? // removed Andrea's comments not commented upon 4.1.2.2 It is unclear from the introduction to this section that you have one and only one service/wsdl per supported locale. Perhaps we should make a simple statement to that fact. AP> Might not be true. Can have several bindings in one WSDL, each with a different locale on its endpoint... Maybe should say both?? 4.1.3 "... locale preferences specified by the user ... user's ...." maybe change to "... locale preferences specified by the client. ... client's ..." The client my be human, or it may be another application or service. Ultimately, there is a human somewhere at the end-point, but I prefer "client" for these descriptions. "The service should therefore provide a way for the user's client's locale preferences to be communicated to the service and, if there is a response, the actual locale values used to perform the processing should be returned to the client." AP> How about (for symmetry with Service Determined): -- "In the Client Influenced pattern, the service provides a specific set of localized behaviors which which are tailored according to the locale preferences in the request. The service therefore requires a way for the requester to communicate the preferences and, if there is a response, the actual value used to perform the processing." -- // more removal "GetProductInventory" does not appear anywhere else in the doc. It looks like it is refering to the example in 4.1.2.1. Therefore, we should add "GetProductInventory" to 4.1.2.1 example. AP> Added ref as noted above. Same comment on "user" vs. "client" AP> ... vs. requester. Need to tighten this language. "Really this is something the provider should provide, ..." need to clarify why this is so, even if it's a reference to another section in the doc. (nit) I would remove "Really" AP> Done. -- Michael McKenna Globalisation Architect mgm@globalisation.org California Digital Library University of California Office of the President michael.mckenna@ucop.edu http://www.cdlib.org
Received on Tuesday, 27 January 2004 17:16:53 UTC