[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: ebXML Requirements - November Minutes
Here are the minutes from our session at the November meeting. -- Michael C. Rawlins, Rawlins EDI Consulting http://www.metronet.com/~rawlins/
UN/CEFACT -OASIS ebXML Meeting @ San Jose CA - 11/17-19/1999
ebXML Requirements Group - Breakout Sessions
Minutes Summary
Chair: Mike Rawlins
Secty: Tom Warner
Editor: Mark Crawford
Member Address List:
Mike Rawlins, Team Leader rawlins@metronet.com
Tom Warner, Secretary thomas.warner@boeing.com
Mark Crawford, Recorder (Editor?) mcrawford@lmi.org
Kathleen Tyson-Quah jtq@granularity.co.uk
Tim Cochran tcochran@disa.org
Don Rudie rudied@dnb.com
Kenji Itoh kenji41@attglobal.net
Ravi Kacker rkacker@kraft.com
Kim Lambert klambert@air-transport.org
Jean Kubler jean.kubler@unece.org
Jon Bosak bosak@eng.sun.com
Marcia McLure marcia.mclure@mmiec.com
Doug Hopeman hopeman@xmls.com
Norbert Mikula norbert@datachannel.com
Garret Minakawa gminakaw@us.oracle.com
11/18 - 1st Breakout Session - Initial comments
Mike R. -
- Our work is to develop basic ebXML Requirements Documents
- Identify scope of ebXML
- Recognize OO-edi Stds differ from ebXML Stds
- Identify deliverables
Ravi K. -
- Why am I here? ...because this activity requires a business perspective?
- Perceived frustration of all XML diverse activities (observation)
- We (business people) want to break this XML Method of Operation (MO)
- For this 2 key things are needed...Promote Internet Commerce and develop B2B, B2C and C2B standard solutions across vertical and horizontal supply chains.
- Observations:
- Vertical XML presentations this week look well positioned (vertically), but how do we
bring these all under a single horizontal XML objective (umbrella)
- What horizontal (cross Industry) development is needed and how can we accomplish it.
- Suggestions - Lets look at one cross industry process such as exchange of a Purchase Order a Purchase Order Acknowledgment and an Invoice and develop the XML solution...This limits the scope and can be accomplished.
- Want to save the research already done in EDI.
- Maybe take advantage of AIA research for a short term solution
- We need to spell out the road map and direction of the XML work
- For Modeling
- For XML using EDI
Tom W.
- ebXML appears to have 2 given components...UML is for the business process analysis and design while XML is for the technical (functional) delivery mechanism .
- Sees three Possible "Requirements" coming out of ebXML:
- One might be to develop UML "marrying" to XML for CEFACT and OASIS population
- But the Business users need a workable solution now and can't wait for the UML work.
- Observation: Transaction specific DTD's and Schemas will be 80% boiler- plate, template stuff that will get sent, over and over, for every transaction.
Mark C.
- Every senior US government manager is being bombarded by software vendors selling proprietary solutions wrapped in XML and touting it as the language of the Internet
- Government EC/EB Trading Partner based cuts across every industry.
- US government cannot afford to Participate in every XML standards activity or implement all the different XML solutions being defined.
- US government needs something similar to ANSI ASC X12 and EDIFACT which cuts across all vertical and horizontal sectors and which is a one-stop-shop where government can participate and contribute business requirements.
Some Questions:
- What is the impact of W3C work on this ebXML Activity?
- Who will manage the Repository?
- Should Tags be....?
- Can we create a semantic meaning without a document (frame of reference)?
- DTD's need to be defined in a glossary
- Want Interoperability with a minimal amount of effort
Garret M.
- ebXML needs to build applications for its customer base
- These need to be compatible with dozens of "XML Standards"
- We want one XM L standard that all can use.
- Need a mechanism to insure cooperation across all groups identified
- We need to identify all the known XML initiatives
Kenji H.
- XML s becoming a hot topic around the world, but there is much
misunderstanding.
- We want one Standard under UNJ/CEFACT/Oasis in order to avoid differing
industry XML standards.
Jean K.
- (This work) is either too early or too late.
- If we had XML syntax when EDIFACT was started we whould not be here now.
- Semantics comes from the message design and not the syntax.
- The UN started to use SGML in 1989
- Likes the Level -0 approach to XML EDI (offered by AIA / Tom W.).
- Want to learn from X12 and EDIFACT
- We have a simpler mechanism to deliver the stuff for Trade Facilitation.
- Standardization on the intersection. Can standardize a Industry within the (global) standard framework.
- Must differ the Requirements from the design effort
- Business Needs equal Requirements Connectivity
Kim L.
- Is building XML now
- Want this group to build a framework that everyone can use.
- Suggest developing a Dictionary /Glossary of Terms to eliminate the
confusion in discussions.
Some words needing a glossary:
- DTD
- Semantics
- Interoperability
- Repository
- Directory
Doug H.
- How is industry going forward with XML
- 80+ different XML vendor / players
- Hope that ebXML can consolidate all the XML variations
- Cant afford to do 1500 variations of XML document/ messages
John B.
- Joined ebXML to learn about the EDI business.
- Had started the ebXML "idea" within OASIS
- Much disparity of XML EDI activity
- Want to do "it" in a uniform way...several ways are possible
Observations:
- If we all go away after this week then the proliferation of XML activity
will continue in OAISIS like organizations.
- If we go away after this week then we lose the chance to make use of all the
good expertise we have gathered here.
- I may agree with Tom W. that a long term and short term solution are needed.
- SME's are going to start doing the XML exchanges
- We need to choose some basic business domain that we can XMLize.
- Can we do it in a short period of time?
- A short term requirement for a purchasing deliverable.
Terms of Reference:
- Interoperability
- Open
- Simple
- Minimum Cost
- Show how we move forward
Don R.
- Working with a group in Europe who is replacing proprietary EDI with EDIFACT
in order to obtain universal understanding of content and meaning.
- Want an open forum to build a standard
- Now along comes XML and I see that ebXML is the best chance to save the
millions of dollars already invested in EDIFACT
- Want an XML standard that is sustainable. Don't want it to go away in
4-5 years.
- Not interested in the real short term
- Want everything in EDIFACT to be preserved in XML
- See CEFACT as long term because it is UN based
Marcia M.
- Represents the AIA SME's (small/ medium enterprises)
- Chaired the X12 Health Care modeling effort
- Thinks that she can make a difference here
- This looks like the best place for business process modeling.
- Want the cheapest, fastest, best solution for the SME's.
- Also want to lay down a good foundation (for ongoing work).
Tim C.
- X12 DISA has taken on 3 new business clients:
- The Open Travel Alliance is moving X12 to XML
- Mortgage Industry is interested
- All have an urgency to get something in place now
- All industries are anxiously awaiting the W3C Schema Std.
- Expect this ebXML effort to have the best bet to succeed
- Expectations: Hope that 20+ years of X12 and EDIFACT semantics get saved
and used.
Norbert M.
- Provides the tech support for OASIS and is here to make sure he hears the requirements correctly.
Mike R.
- Is an independent consultant
- Is and X12 member contributor
- Involved in XML
- Editor / co-chair of the X12C XML Tech Review Document
- Working on a Masters Degree - applying non functional requirements analysis
to EDI
- Can bring the results of his initial analysis work to this team
- Some Work Ideas for Requirements:
- Like to take a shot at a High -level Requirements outline.
- Need to draw a line between Business Operation View and the Functional
Service
- View (of the requirements)
- Believes that the "Functional Requirements" are outside this group's work.
- We need to deliver "ebXML Compliant Applications". (a number of criteria)
- Keep the SME in mind
- Open standards based process
- (interoperability is seen as a quality requirement)
- Education is needed
Next ebXML Meetings - (Facilities needed - 1 large room and 8 breakout rooms)
31 Jan - 4 Feb -- USA - Orlando FL
29 May - 2Jun -- Europe
7 - 11 Aug -- USA
6-10 Nov -- Japan - Tokyo
End of 1st Day Meetings
--------------------------------------------------------------------------------
11/19/99 - 2nd Day - ebXML Requirements Group
MIKE R.
Requirements are what you need, not what you need to do. In addition to non-functional requirements, after giving it some thought we will need some functional requirements as well. The ebXML needs to ensure that implementations do some things. (Discussion of functional vs. non-functional definition).
DOUG H.
"Interoperable" is the chief non-functional requirement.
MIKE R.
There is also another class of thing in the TOR. So far our discussions have focused on app to app. There's another phrase referring to app to human and human to app. It may something like flashing up a web page, but I've also heard of using XML to put tags in web pages to make it easier for search
engines to find text, so we may have requirements in that area that we need to consider too.
MARK C.
Concerned crossing boundaries between ebXML concerns and what should come out of W3C. Tags as part of database structure doesn't address what is being done in other teams.
JEAN K
For UN we want recommendations so that we use all benefits of EDI and EDIFACT to have Level O. . .
MIKE R.
Asked Jean to take responsibility for drafting his proposal as a requirement.
JEAN K.
Agreed.
MIKE R
The other important thing is that I have an idea for a TOC for a Requirements Document. I'll distribute it on the list. I think it imperative that this group produce a work plan as soon as possible. Having something before the next meeting isn't soon enough as other groups will be working from listserv. I propose target for initial draft of 30 days.
KENJI H.
Our group's result feeds into the work of the other groups, so we need to produce the requirements quickly.
DOUG H.
30 days is not out of the question.
MARCIA M
Proposed (offered) to work with Jean in drafting (AGREED).
MIKE R
Leave open the possibility for teleconference calls (for our interim work effort).
JEAN K
Do we have a list server?
MIKE R
We will. We need to actively solicit requirements from other groups. We expect this to be a collaborative process.
TOM W
List of Requirements (preliminary):
1. Precisely identify the business problem we are solving.
2. Provide the migration path from EDI to UML using a short term "core solution" (in 3 months) which
should also be the basis of the long term (18 month) solution.
3. Provide migration path from EDI to XML/EDI to UML/EDI.
4. Provide a Central Clearinghouse for consolidating all XML/EDI standards developments by the
various organizations involved.
5. Must be able to provide simple, continuing support to manage work, deliverables so users can find the
ebXML products.
6. Complete short term work in 6 months. Complete long term work in 18 months.
MIKE R:
What do you mean?
TOM W:
I want someone to link the two (EDI and XML...and UML if applicable).
DOUG H
Ray Walker said we must enable Mapping Implementation Guide.
MARCIA M
That is a long-term objective (Agreed).
KENJI H
Follow the scope/key issues in TMWG N104.
MARCIA M
After read N104, I conclude this is the deliverable, the guidelines and the short and long term. I'm not sure cost minimization was mentioned, but needs to be there.
DOUG H
That would be a non-functional requirement?
MARCIA M
The speed with which the short terms are accomplished: minimize cost and take EDI from larger clients to small and mid size companies.
TOM W
Both Kenji and Marcia agree recommendations in N104 be accepted as they are.
JEAN K
Yes, but there are things missing which must be made clear.
MIKE R
I see no disagreement with this being an excellent foundation.
DOUG H
I've only got one requirement. If I'm ABC corporation, why would I come to ebXML. We must provide a starting point for a company to come to, learn, research, design, implemenant any XML e-commerce process...that may be ordering, invoicing or who knows. We have to provide some boilerplate to start, otherwise they will go somewhere else. The only other comment, is that the term "framework" is too vague. We have to find something more concrete. Give me the beef.
TOM W
We need a glossary of terms (to explain words like "framework".)
MARCIA W
Build on the one in N104.
KATHLEEN T-Q;
1. Glossary.
2. Measurable objectives for ebXML to establish us as authoritative among other groups.
3. Some sort of process to "certify" conformance of XML schemas with our principles and results, to promote common objectives of interoperability.
MARCIA M
Measurable objectives and deliverables for long and short term requirements.
JEAN K
Comment by Jon Bosak that the idea is that when you put something on the web you do not know what will become of it.
MARK C
We will give high level requirements from client perspective and some questions. Reality is that client base will have mixed bag of IT. We are going to have legacy systems, client server systems, individual web systems. I need to make sure that they are interoperable. That any XML solution furthers
interoperability. More importantly, we are looking at this group to be the single body to take the lead in coalescing business requirements to further develop reusable, extensible, document structures. These products have to support interoperability and be backward compatible with EDI solutions. We don't have to retain tag names, segment names, etc. as long as they are cross-referenced somewhere. I don't want 200 tags for purchase orders. Having said that, if this is not going to be the ebXML group then who? The Federal Govt. will find someone else to do it. If its OASIS, Rossetta, etc. that's fine.
On a detailed level the other groups need to address:
Given that ebXML will take the lead as a standards body, should XML business standards follow existing X12 and EDIFACT transaction set (semantics)? What is the impact of ebXML business standards on W3C? What is impact of immaturity of various W3C specifications on short term and mid term work? What should be relationship be with W3C, ANSI and ISO? How should work be funded? Who maintains and controls access to repositories in the short and long term? Should tags reflect, embody, ignore X12 names, etc? Can you create semantic meaning without document structure? If you can't then what is impact on groups 2 and 3 operating independently?
TOM W
What do you mean by document structure?
MARK C
DTDs. You can't do anything without a DTD or schema somewhere.
MIKE R
There has been discussion of building blocks. Some groups are looking at a lower level of core components.
MARK C
Combination philosophical/technical/ business/modeling question.
Also, need to ensure that we clarify on interoperability that there are two levels: easy level which a technical solution can solve. More importantly, I need interoperability between my applications, so that what is exchanged is understood by both (parties) with minimal level of effort on my part. I don't want to create 500 maps, with a cadre of mappers and an XML server which is doing more than I'm doing today.
KIM L
How do you see your questions being used?
MARK C
These are questions to be put to the other groups. They are the ones that must provide the answers.
GARRET M
It is very important that we have some mechanisms to ensure cooperation among the various (ebXML) groups. We have lots of people saying they want the same thing, but also some saying that they will continue along the path of what they're doing. To get everyone to come together and make strong commitment, requires sharing and cooperation.
TOM W
We need radar map of all the major players doing XML stuff.
DOUG H
OASIS doing something like that now.
MARCIA M
That is already in N104. (see matrix in N104)
MIKE R
In time left we need to discuss how to move forward.
TOM W
If we normalize what we've done so far we could have a set of bullets to present as our draft.
MIKE R
I will find out arrangements for list server and let you know. First thing would be outline, headers of content and maybe a straw man of what we've talked about here. I have some other research to contribute for your review.
We'll put everything in it that everyone wants to be said and then sift it down. We can take out the trivial or obvious later.
MARK C
Volunteered to receive and edit all contributions.
DOUG H
For us to achieve something in 30 days, I think it reasonable for you (Mike) to task out Requirements to the group.
MIKE R
Straw man will condense what people have talked about. If you see a point that you would like to expand, please let everyone know that you will work on that point.
MARCIA M
We need a list of contact details as soon as possible.
MIKE R: Thanks
End 2nd Day
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC