Minutes of the Face-to-Face of the WS I18N Task Force
November 2002
Links: Agenda — Meeting page — Friday, 22 November — Saturday, 23 November
Attendees: Addison Philips (chair), Martin D�rst, Takao Suzuki (notes)
Review of agenda
Getting organized, discussed about things to discuss:
- Setting up regular meeting
- Sending comments to SOAP. Should check the status. What they have done
based on comments previously sent to them.
- WSDL requirements. Should look at the comments.
Some other issues discussed:
- Interaction with the other WGs.
- Addison - has communicated with the Coordination group, and found
they would like us to coordinate directly with the other groups.
- What message we want to come up with for the other groups. Our
visions for our tasks.
- Privacy issues - i.e. transmitting personal data.
- Issues with enterprise/cooperate scenarios
- .NET vs. Web Services: What are the differences?
- WSI - external to W3C also works on WS.
- What do we do other than reviewing individual specs.
- Publishing Standards?
- Kentaroh's Draft.
- Modification to WS specs
- Locale language negotiation
- Best practices
- Shall we divide this into pieces? Not for now. Review the draft
and find out whether we should divide it into 2 docs.
- Can we recruit more participants? To be discussed further
Use Cases document
- Review Kentaroh's draft
- Ask for General Calls?
- First publication of Use Cases document - End of November or Early
December.
Locale
- The use of locale
- Generate use cases
- When do we need locale? Server side, client side.
- Not to use locale sensitive data in locale sensitive format
- Use locale only as needed. Most of the time you don't want
locale
- Locale registration - W3C not interested in registration business
Editing Kentaroh's draft
- Converting it from HTML to XML for easier maintenance
- Updating in this f2f meeting
- Started updating it (see http://www.w3.org/International/ws/ws-i18n-scenarios-edit
for actual notes/results.)
- Some sections need clarification from Kentaroh
- Some sections need further investigation
- Japanese Calendar Era transition scheme (Takao to investigate)
Conference call schedule - came up with proposal
- Frequency: Start with Bi-weekly, adjust later as needed
- 3PM PST (= 6PM EST, 8AM JST) Tuesday
Next Face to Face meeting - Opportunities/Possibilities
- W3C TP meeting in Boston in March 2003
- Unicode conf in Prague, Czech in March 2003
- Do we need another one earlier than that?
Any web services with UDDI (may find some from uddi.microsoft.com)?
Attendees: Addison Philips (chair), Martin D�rst, Takao Suzuki (notes),
Tex Texin
Recap of Friday's meeting
Review of agenda
- Locale dependency in locale independent data. Guessing the nature of
handling legacy data is not in the scope of our TF.
- 3 locales: Sender's locale, Receiver's locale, and locale to describe
where the event happens. For instance, Sender sends Airplane departure
time to receive, and in this case, the location of the airport would
likely be what's associated with the departure time, but not sender's or
receiver's locale.
- Do we need use case for sender's precision about data format on
receiver's side (i.e. decimal places)?
- Sender wants to influence what receiver does? Human interacted output
vs. data processing cases. Explicit requirement of service (i.e. service
contract). Implicit requirement (Service operates differently if the
language changes, aka an i18n bug).
- Formatting, Unit of data, what to do with the unit. Process is what to
do with its relationship
- Problem in the past: Server locale was allowed to influence the
data.
- Best practice vs. requirements. (i.e. Data structuring, price field
with no currency. don't think we can prohibit. This is an example of best
practice which is very nice to do, but we cannot force.)
Discussions about Activity proposals
- We could spend more time on use case
- Layout of upcoming activities
- Define how to develop i18n web services, and possibly make it
standards
- Come up with use cases and may publish "Best practices guide" document,
possibly make it Recommendation.
- Reviewing existing web services
- SOAP doc review
- WDSL requirements review
- Petitioning our work items maybe needed to work in parallel?
- Define what not to do, but still needs to be done in the future.
- Categorize use cases, coming up with models/generalized categorization.
When to implement? Not likely be working on it before the end of this
year, but in the following one.
- Extensibility, length measurement
- Two different targets: Creating new web services and Dealing with
existing web services
- Do we need XML code in the usage scenarios? They are being used to
demonstrate the locale dependencies. Some may be too suggestive, some
could be helped with graphics/pictures. Any one of them we want to
eliminate? To be reviewed later.
- IUC: Common XML locale format, an open source repository for locale
data. Does it make sense to change formats in Unix, Windows, etc.?
- Do we need more use cases with these scenarios? Data should be passed
with locale neutral schema.
- Non typically-supported case (i.e. date in Thai) may require special
format.
- Long term: What is supposed to happen to locale model, which is
different in various systems
- Help people express locale neutral general model, resulting predictable
format. Human readable messages: sender sends data, and receiver returns
error message in the sender's data locale (doc locale). Problem comes
when the receiver cannot recognize the data locale, or when there is
none. Error message may be returned in the receiver's language, and the
sender may not understand it.
- This TF needs to define what to cover, and what not to cover. To be
narrowed down. Our main forcus is on Web Services area. We are not going
to go looking for outside of Web Services area. For instance, we will
review SOAP, but we might or might not review XML Schema V2.
- The needs for header
- Use case requirements
- Existing systems that are exchanging proprietary data to use Web
Services.
- Some kind of generalized identifier. Need a standardized way to
exchange locale preferences. Most cases data are locale neutral, and
there should be not much needs for exchanging locale preferences.
- There is some justification to personalize data. If all services can be
expressed in neutral way, the receiver can decide the format or
personalize it.
- Shoe size is an example that to consider different data type. Locale
provides a hint to indicate the size (i.e. if you walk into a shoe store
in Japan, they would go with centimeter system. Online would be an
example that requires data type definition)
- Headers, Why needed, what problems are being solved. Declaring too much
in header may be waste as many of them may not be used at all.
- Continue to collect use cases
- Categorize the each of those.
- What qualifies use case?
- Requirement for submitting use cases.
- Detection, format, scenario definition, description, which often
includes examples.
- Type of use cases that we accept. Grouping scenarios.
- Boundary cases.
- Number of scenarios
- What to do to current doc before we can use it as use case examples.
- Needs to become in XML format
- Title may need to be changed.
- Kentatoh to be the primary editor. Martin to talk to him further to
define editing responsibilities between him and Martin.
- Still have chance to publish the first WD before the end of
November.
- Categorize currently available use case.
- Coming up with good name for each scenario
- Adding table of contents.
- Plan to manage submitted use cases
- How submitted use cases need to be triaged.
- Following up on the status (not reviewed, reviewed, accepted,
rejected, investigating) of each case
- Not over-organize
- Schedule
- Announce for call, use case submission
- Scenarios. Tex: Do we want to create matrix to make sure we are not
missing anything?
- Putting pointers to each other.
- Review XML protocol, which might give us additional requirements.
- Kentaroh's feedback.
- What docs to review.
- By the first conf-call, that's is being scheduled in 2 weeks, we will
clean up the use case doc, and publish for call.
- User case: Page format.
Review work
Reviewing http://www.w3.org/2000/xp/Group/xmlp-lc-issues.html,
and http://www.w3.org/2000/xp/Group/2/06/LC/soap12-part0.html.
Discussion about resolutions on the issues that Martin has previously
submitted:
The chair of the Internationalization Working Group is Misha Wolf
(Reuters). The chair of the Web Services Task Force is Addison Phillips
(webMethods, Inc.). The staff contact of the Internationalization Working
Group and the Web Services Task Force is Martin D�rst ([email protected]). The chair and staff contact
of the the GEO Task Force is Richard Ishida ([email protected]).
Addison Phillips ([email protected])
$Id: ftf200211-minutes.html,v 1.3 2002/11/23 14:50:54 duerst Exp
$