GSA
Headquarters
18th
& F Streets, N.W, Room 5141
Washington
DC 20405
Please send all
comments or corrections to these minutes to Glenn Little at glittle@lmi.org.
Mr. Owen Ambur: OK, we might as well get started. In view of the fact that we’re expecting many late arrivals and we have a lot on the agenda, I think we’ll dispense with introductions and go right into the presentations. First, we have several announcements. Brand has an update on the Web Services Working Group. Brand?
Mr. Brand Niemann: On December 10, the Web Services Working Group is having a one-hour meeting in connection with XML 2002. Then we’re going to recombine with Susan’s workshop [Susan Turnbull and Collaboration Expedition] on January 14 next year. We hope the agenda will cover presentations in four areas. This time, the vendor was ZOPE, which is an open source application server, specializing in content management, portals, and custom applications. This time we had ZapThink, and we had Kevin Williams talk about XML Design Collaboration and Registry Software, which is a key tool for collaboration resources. We also had Joe Molitoris from Mitre talk about Mitre’s PIXIT Web Services efforts.
We’re working on a white paper…when we had the brainstorming that led to the formation of the Web Services Working Group in July, John Dodd and Davis Roberts approached Barton Smith and me and said “We’re your ‘representatives from industry’ working council.”
We’re working on a white paper for the Federal Enterprise Architecture (FEA). We’re presenting it tomorrow. John’s presenting it at the Solution Architect’s Work Group (SAWG), Bob’s group [Bob Haycock’s group from the Office of Management and Budget FEA Program Management Office], which I’ve joined. At the last SAWG meeting, they said they wanted to document mapping the enterprise architecture to Web Services. It has four authors—me, John, Kevin, and Bob. We’ll present the draft at the January 14 combined meeting of the XML Work Group and the Web Services group.
At the last group meeting, Marion [Royal] proposed that Eliot [Christian] and I support the Geospatial OneStop, and Marion and I support the Business Compliance OneStop.
We’re proposing to help architect the Small Business OneStop with Web Services, and demonstrate that it’s interoperable with Web Services with a component of Geospatial. One of the tools they might need would be stuff like locational services. We’re thinking of it as a test for Web Services and two of the OneStops.
Mr. Ambur: Thanks Brand. OK, let’s go right into Bob Haycock’s address for the XML 2002 conference.
Mr.
Bob Haycock
Mr. Haycock: I appreciate the time and the invitation once again to come and meet with group. I’ve been asked to keynote at the XML 2002 conference in Baltimore next month. Marion volunteered me for that. In conjunction with Marion and Brand and Jim Benson, we’ve put together a draft of the talk for the conference.
When we met two weeks ago, I was thinking it might be useful for me to dry run this with you and get your feedback. Is it useful, interesting? Are these things important to talk about? Are there other things I can add that are relevant, that you want to include in the talk? That’s the purpose of this—to give me some practice. This is going to be a big group. This will help give me feedback from folks involved with this topic in the federal government, and make my talk more relevant. Feel free to stop me and ask questions, comment, ask for clarification, say, “I think this would fit,” etc. I’ll take that information and modify it. I haven’t sent it to the conference folks yet. I’ve resisted until this meeting. Once I’ve talked with you and gotten feedback, I’ll send it off. Tomorrow it should be on their websites.
I have about 45 minutes on the agenda. This takes about 1/2 hour. The XML and Web services part is new. I’m going to make clear that I’m not an XML expert. Some of the stuff, I’ve already done.
Slide 2 [Table of Contents]: I’m going to say I’m the Manager of the FEA component of the federal government, and talk about what we’re trying to do, not as an XML expert. That’s ‘s how it’s structured—“Here’s XML, here’s what we’re trying to do in government,” etc. The last slide isn’t comprehensive—just some examples. If you want others, I’ll put them in there as well.
If I’ve said this before, please bear with me. I’ve spoken a lot. I modify a little as I go. Initially I talk about the FEA—what it is.
Slide 3 [The Federal Enterprise Architecture is a business-focused…]: As it’s being developed by OMB, it should provide OMB and federal agencies with information they can use in the budget process. We are using a Business Reference Model (BRM) to identify major initiatives by function and sub function to see if there’s redundancy. You’ll see it more. It’s the first time this has occurred in the federal government. As we get more into it, you may see this happening more. From that standpoint, it allows OMB to identify cross-agency initiatives—where we can leverage technologies that other agencies have bought at the local, state, and federal level, rather than reinventing the wheel. The bottom line is to improve services to citizens, or do things on behalf of citizens—environmental protection and defense are citizen centric. The Objective of the FEA is focused on the way the citizen sees services. It’ll give us the opportunity to see if we can leverage services the way they’re applied to the citizen. It should help to eliminate organizational and program obstacles without forcing reorganization. Reorganization cause more trouble than they solve. It may give us an opportunity to do virtual reorganizations, and take advantage of technologies that let that occur.
It’s a business framework, not just IT. IT is an asset that’s applied to improve performance. You can apply other assets, like “people” things, not just IT. The two top layers—the BRM and the Performance Reference Model (PRM)—have nothing to do with technology. After that, you look at technical issues that support it.
The FEA should lay the foundation for improvements in many areas. How the budget is looked at across government—we’re starting to look at it by business line and function across agencies. “Grants” is a good example. Within the FEA, grants are similar. We’re thinking the grant process is similar across agencies. The outcome you’re hoping for is different, but the processes of doing it should be similar. If so, can you use a similar technology, rather than agency stove piping? We think that’s where Web Services comes in.
We think it’ll help with Homeland Security. 22 pieces and parts of agencies are brought together to form Homeland Security. There will be many parts coming together, all of which have infrastructure, applications, and business processes that need to be integrated seamlessly, to see who comes into the country—whether it’s INS [Immigration and Naturalization Service], Customs, or the Coast Guard. The way to do it is to look at what’s involved in the business processes. It should help with aligning those technologies to get the best service. It should help with performance measurement. It’s directed in terms of business line approach. It’s different from the past OMB view, which was on a program-by-program basis. The PRM is consistent, but looks from a business line viewpoint, looking for common performance metrics to judge across programmatic and organizational lines to support the outcomes you want to achieve.
It supports cross-agency collaboration. That’s the specific purpose of FEA. It was created primarily for the 24 EGov initiatives, the beyond that the next 24, then 48, and ultimately the entire federal government.
Then there’s improved service to the citizen, and finally there’s Component Based Architecture. In other words, what services do you need to facilitate the business process? What architecture do you put in place to do that? That’s where XML comes in.
Slide 4 [The FEA is being constructed through a collection …]: For the FEA itself, right now the only reference model is the BRM. Many people think that’s it. We’re working on the meat. The others—Performance, Component, Data, Technical—are in various stages of completion. The Service Reference Model is done. It’s been reviewed by the SAWG twice. It’s ready to go out. We’re briefing Mark Forman today at noon. That’s all that’s held it up, is waiting for his concurrence. We’re looking at all agencies on the Service and Technical Reference Models (TRMs) in early December. We’ll probably ask agencies to begin to map their service components (pieces of applications) and their TRM, which essentially is an Interoperability Reference Model. Since every agency has a technical model, we didn’t want to create a giant one. We said, “What’s it take to interoperate using a component and service architecture approach?
We may change the TRM name to TRM Interoperability View. I’m open to suggestions. So this is the FEA and where we’re going. The objective is to get a first version of each reference model out in agency hands, ready for use for the FY 2005 budget process, in January of the next calendar year. Some thing delaying it.
Mr. Ambur: All models, including the Data Reference Model?
Mr. Haycock: Probably not the Data Reference Model—principally because it’s a struggle to see what’s useful. That’s where XML comes in. What’s useful at the government-wide level? We’re thinking an XML schema repository organized by business lines, where you can see where business functions and sub functions interoperate. That would be contained in the Data Reference Model, probably in an XML format. We’re encouraging XML as the vehicle to exchange information.
We’re writing a white paper laying out that thinking now. It’ll be done in the next week or so. Then we’ll put it out to the SAWG. I’d like to have you all join in on that. I’m thinking of an all-day off-site meeting on the DRM to brainstorm through that white paper, to come up with the DRM. It’s doubtful we’ll have it by mid-January.
Slide 5 [The draft FEA Performance Reference Model …]: The PRM again attempts to begin to identify and align performance metrics by business line, function, and sub function. It’s organized in five categories of metrics, all of which point to a strategic outcome for that business line. In the education line, for example, “Leave no child behind.” What does that mean in terms of the process, technology, people metrics, and other assets? We’ll attempt to build it in five categories that begin at the process level. So you can begin to see the process.
This is the hardest to do, because it’s not the way the government does business. There is a series of discussions in OMB on how to align the way OMB does budget and performance metrics with the PRM, and have the PRM do it. We’ll probably take it slowly; do proofs-of-concept over the next two months. For the [FY] 2005 process, we’ll pick off several business lines and test them. Then we’ll build it further for the [FY] 2006 process.
Mr. Ambur: With respect to short-term versus long-term objectives, there are things that could be accomplished quickly, like having agencies make their GPRA plans available in a standardized XML format on their websites. Legislation on the Hill calls for website standards, and regardless of whether it is enacted or not, OMB has indicated it may issue such standards. Assessing how agencies are complying with such standards may be some of the “low hanging fruit” that can be gleaned quickly and easily, via automated means.
Mr. Haycock: We’ll have to do that and engage OMB, and grow and evolve, because it’s new and complicated. It’s a combination of Balanced Scorecard and other performance approaches.
Slide 6 [The FEA Business Reference Model (BRM) is a…]: The BRM is three business areas—services to citizens; support delivery of services; and internal operations/ infrastructure. Services touch citizens some way. Support and infrastructure are internal, facing those externals. We think you’ll see resources moving from the bottom up. The bottom layer is inter- and intra-agency, to denote cross-agency services, which are now minimal and stove piped. The thinking is, they’re standard across government, and you don’t need to do them 18 times. You move down the line and consolidate, then use that money further up in the pyramid, pushing resources into the citizen-facing resources. When I’ve talked to business management groups about this, I’ve generally received a hostile reaction.
Ms. Theresa Yee: Why?
Mr. Haycock: Because they’re on the bottom of the cake. Not that they’re not important, but they’re important in supporting the functions that should be delivered. These are important, but there are different ways to do them. We’re beginning to see where we have duplication and can leverage across other agencies.
Slide 7 [The draft FEA Service Component Reference Model (SRM)…]: The SRM gets into component and service-based architectures. It attempts to organize by service layers (customer service, process business management, digital, business analytical, business intelligence, etc.) so each component of a service is available for essentially plug and playing an application. It’ll give us way to organize components so agencies and developers can identify where a service or component is available as they build applications. I think we’ll eventually see a component service directory. It’s already being done internationally. Denmark and Sweden and Canada already doing some of that. Commercially, there are a number. We’ll leverage off of all that within the federal government.
Mr. Ambur: It seems that this is where Web Services map directly to FEA. Will Brand’s white paper address that relationship?
Mr. Brand Niemann: We hope to.
Mr. Ambur: It would also be good to delineate which services are inherently governmental, versus commercial/industrial in nature, and determine how best for government to leverage commercial/industrial service components.
Mr. Haycock: Right. I agree, it’s a good idea to proof-of-concept with the EGov initiative, and build all way through the architecture from the business model through the technical level. We can learn, “Does this actually work?”
Slide 8 [The SRM is supported by multiple Access and Delivery Channels…]: We think the BRM and PRM are more effective for OMB than for agencies, but the SRM and TRM are more useful for agencies, so we look at things in a componentized way.
Mr. Ambur: If OMB were “eating its own dog food” in terms of customer focus, which end of the FEA model would we start with, the high-level Business Reference Model or the lower-level Data and Information Reference Model, which represents the actual, day-to-day business processes?
[Mr.
Haycock declined to make any comment that might be construed as a policy
statement.]
Slide 9 [The SRM Supports Investment Planning…]: It should serve multiple access channels. We hope to identify access and delivery channels—how they tie to service layers and component groupings of the architecture. So if you needed customer relationship management, a better example is Pay.gov, we view it as a component. It’s government-fabricated, but it’s component-based design, runs on an interoperable infrastructure, and can interoperate with any agency, so if an agency needs that, they could make that available. There are issues of service agreements, support, and trust, but it’s the same issues as with Web Services in general. “Will the services be there when I need them,” so we need to identify what access channels are available for Pay.gov—what infrastructure channels it needs, from component to infrastructure, so when agencies want to get something, they know what they’re getting.
This takes the SRM through implementation. I don’t know whether I’ll keep it [the slide]. “Is the business function going across the government? We can go here and see. Maybe the components or complete applications are out there, and we just need to go get them. “Will the component support the business need?” “What access channels can be used?” “How do I buy and plug it into my application?”
Slide 10 [The draft FEA-TRM will provide an effective means…]: This is an overview of the TRM. It’s broken into three parts—Service Access and Delivery (What’s out there being used anywhere in world that facilitates the business?; the Service framework; and Service Platforms.
Slide 11 [The draft FEA Data Reference Model (DRM), while still…] For the DRM, we use Homeland Security to illustrate it. There are specific Border Control, Intelligence Gathering, anti-Terrorism, and Law Enforcement sections. There are common information and data requirements across these functions. We developed a homeland security reference model based on the President’s Homeland Security strategy. We can start to identify common information within the functions. Within Border Control, for example, there’s a Watch List between the FBI, the Border Guard, INS, etc. “What information do they need to share?” That same information can be provided by—or consumed in—other functions.
Mr. Marion Royal: Your XML and data-management audience will look at this in terms of vertical and horizontal. You talk about specific functions: they’ll look at the vertical domain of, say, Border Control—and within that are platforms. It might be confusing. Something to think about.
Mr. Ambur: Did you just develop a BRM, or did you compile a DRM too?
Mr. Haycock: BRM.
Mr. Ambur: If there’s a driving force to build a DRM, homeland security should be it, and many of the data elements required for homeland security are enterprise elements that will apply across other domains as well.
Mr. Haycock: There are groups working across each of these areas. “Border” since August. I’ve been working with them. They’re developing four products: concept-of-operations, activity trees, note connectivity, and information exchange matrices (C4ISR). Each of the other groups is in various stages of the same products. The idea is, by the end of the calendar year, to have consistent information across these four areas to push through the technical levels.
Mr. Ambur: With reference to the second to the last bullet on this slide, indicating that security is a priority, you may want to consider including another bullet explicitly referencing records management as well. Many people may think that security encompasses records management but it is good to keep in mind exactly what we mean when we talk about information security.
Mr. Haycock: So that’s it for the DRM.
Slide 12 [Collectively, the FEA reference models can be leveraged…]: This gives a high level example of the leveraging. Going back to the previous slide, the new EGov Border Control initiative—one of its functions is acceptance of cargo. “What are activities that need to occur in the cargo acceptance business process?” One might be looking at a license from a state government. If it’s at the Mexican border, with a California license, can you go to California and check the plate? Once you identify those activities—“What services are required to make it work?” “What components can be put together,” “What interoperability architectures are required to make it work?” you go back to the Border Control Officer to let the truck across border.
We may have 2-3 states, and 4-5 federal agencies involved in that chain. It’s difficult to do now, but not impossible. But the structure we’re in doesn’t facilitate it. That’s essentially what we’re trying to do with FEA.
Mr. Ambur: It goes back to a virtual, rather than structural, reorganization.
Mr. Haycock: I would argue strongly for Web Services.
Slide 13 [Table of Contents, continued]: The focus here is on XML and Web Services, and how we see them in the federal government.
Slide 14 [With over 135 XML and Web Service standards…] This is a chart from Brand. The point is, lots of XML Web Service stnadards are out there, changing as we sit here, so we’re starting and figuring out where to go, what we should sign on to (Susan is talking about J2EE vs. .NET, vs. ZOPE) The federal government should help facilitate, and sign on to commercially viable standards, but not create our own, One of the significant changes of the last 5-10 years—that’s to facilitate the ability to use open- source interoperability components that are already out there.
I’m from a small bureau, a small agency out west. This has been a challenge to get my mind around all that’s going on…there are a lot of the same faces in the meetings. There are a lot of groups, both the CIO Council, subgroups, sub- subgroups, and other entities. Susan has helped me. There’s lots of activity in IT in terms of where we should go, and which standards we should follow. The message is to begin to get that focused; “What are our objectives as a government?” “How do we select and use standard approaches?”
Slide 15 [The Federal Government has established…] Here’s a short list of some of the activities going on. The first one is the Architecture and Infrastructure Committee (AIC)—it’s going to be three new subcommittees—one on Architecture Governance, one on Component Architectures, and one on Emerging Technologies. Our thinking is, they’re end-to-end.
The Architecture Governance committee will look at governance and structure, and how we look at architectures that facilitate working together. FEA gives you a way to look across those architectures.
The Governance committee thinks that through. The Emerging Technology committee will collect many emerging technologies, like XML, Web Services, those kinds of things, and address how we begin to think about using them.
The Component Architecture group focuses on mature technologies that fit, and how we use them. A component registry/repository would be one, to process linkages to state and international, business repositories. “What are the mechanisms to put in place?” There are procurement issues around that. Those are needs to be dealt with.
Slide 16 [Specifically, the Federal CIO Council has reorganized…] Of the three subcommittees of the AIC, I’ve highlighted Emerging Technologies as the focal point for XML and Web Services.
Slide 17 [Led by the CTO of OMB, the Solution Architects…] This lays out the structure of the SAWG. One area is dedicated to XML and Web Services. We don’t have an architecture. We’re trying to goad Brand [Niemann] to become our expert. We have Marion [Royal], Roopangi [Kadakia], and Lou Sanford. The idea is to have one or two experts for each area, and leverage as we engage with initiatives.
Slides 18 & 19 [So What are Web Services in the Federal…]: This slide I don’t know whether I’ll include. It talks about stuff you already know—“What’s the definition of XML, Web Services,” etc. I’m interested in your feedback.
Slide 20 [Analysis of FY01 IT Investments against…]: This gets at what we believe the BRM shows. We took it from Mark Forman’s “What are the Opportunities for Web Services?” We’ve laid it out by business line; how many agencies participate; and how much money is spent. There are a lot of possibilities—some opportunities in some of these to leverage to reduce costs or get a bigger bang.
Mr. Royal: Are you saying you’re spending this much money, but how would Web Services or XML reduce costs, and by what percentage?
Mr. Haycock: I can’t say right now, and I’ve gotten this question before. This points out possible opportunities at a high level for those technologies to be applied to leverage the money. How much can be reduced? Mark Forman believes quite a bit can, by reducing redundancy.
Mr. Royal: You’re saying, “We’re spending this amount of money, and if it’s used more effectively, we get savings and efficiency and other resources?
Mr. Niemann: Mark [Forman] said in Geospatial it’s about $10 billion, and he thinks we could cut it in half.
Mr. Haycock: “Grants” is another area. They have multiple systems and databases. All e-grants could be a portal, then send email to an agency and everything occurs the way it has been.
Mr. Royal: The other thing you could say is, under the President’s management initiatives, “These are the areas we’re supposed to look at.”
Slide 22 [Web Services can provide the basis for…]: Federal Asset Sales is EGov. Assuming it needs a payment component, we can go to financial management services and tap into Pay.gov without developing it all over again.
Slide 23 [There are several fundamentals that will contribute…]: Here we identify common functions, interdependencies, interrelationships, and evaluate barriers to information sharing. Then we implement in a way that addresses both the opportunities and risks of a “networked” environment. Everyone knows the opportunities and risks. One of the main risks is security. We’re hearing it everywhere. If it has to work in business, that will be solved. On the third bullet, Owen said, “focus on small accomplishments rather than big bangs.” I think you’ll see it happen that way. Right now, with BRM we’re seeing a big bang, but as we move through the layers of the architecture, we’ll have to take it in small bites.
Mr. Royal: At the same time, we’re building the capability to put the pieces together.
Mr. Haycock: The framework—We’re viewing it as, getting the framework will be version 2 of the BRM and version 1 of everything else, except DRM.
Mr. Royal: Even small chunks—we see where they go in the big picture.
Slide 24 [The Federal Government has (and is) piloting…]: The government is piloting four projects. Brand is involved with two, piloting XML and Web Services. I’d like to highlight these and others if there’s time. That would help.
Ms. Jane Smith: I’ve done a lot of work for 10 years with the Census. In less than a month, we’re launching “Economic census.” We’d be happy to serve as a resource. We’re presenting at the XML 2002 conference. I can provide more information if you’d like.
Mr. Haycock: My email is rhaycock @omb.eop.gov. Please give me any feedback you can.
Unidentified participant: The business
approach looks like it would gloss by business managers and, say, go to IT
people. I don’t know how you address the fact that it needs to go to managers.
Mr. Ambur: The context is for Bob’s presentation is the XML 2002 conference, where the attendees will be more technically inclined, but OMB will have to try to get the point across to less technical program managers.
Mr. Haycock: My boss says OMB has to transform along with everyone. The point is, why do we say, “You have to start with the BRM?” IT is just a business enabler. I agree it’s hard, because when something like this comes across, my commissioner just sends it to me.
Ms. Meng Lin: In regard to the comments on the organization—this structure is implemented by people. If it’s going to be coordinated by agencies, people in different agencies have different requirements. How do you get people coordinated, if the chain of command doesn’t understand the cross-agency effort and the manpower needs in implementing?
Mr. Haycock: It’s a major hurdle. It’s why this whole thing is in OMB, to be honest. They have the power of the budget, so it gets peoples’ attention. They’ve required agencies to map their functions to the BRM. It’s already happening. We recently gave extra points in the budgetary review process to agencies that came up with cross-agency initiatives. That’s why we’re putting the FEA out there and making available, so agencies can see it and use it for the 2005 budget process, before OMB demands it.
End Presentation
Mr. Ambur: Next Bruce Bargmeyer will update us via teleconference concerning the metadata registries forum he’s organizing. Bruce, are you on the line?
Mr. Bruce Bargmeyer: Yes. We’re having an Open Forum for standards harmonization for registry standards in January. The standards effort relates to ISO/IEC 11179 metadata registries, UDDI registries, XML registries/repositories, database catalogs, CASE tool repositories, software component registries, ontological registries, and Dublin Core registries.
UDDI registries and XML registries are both in OASIS now. I recently spent a while with the committee for business object promotion. 11179 part 3 standard should be integrated by the end of the year. We’re working with DoD and EPA on putting 11179 registries together with XML registries, and there should be another meeting next week. At the last XML Working Group meeting, I got an assignment to draft a letter to OASIS to discuss the issue of UDDI with OASIS UBL registries. They’re receptive, but suggest we talk with the chair and co-chair of UDDI and the OASIS UBL committee, so the conversation has been started with Katherine Breininger, who is the ebXML co-chair.
They’re saying we should write the letter, but beforehand we should agree on what it says. There’s some dancing around, but they can see they need to do it. I hope to have something agreeable to all. It may extend to the time of the Open Forum, because the involved people will be at the Open Forum.
Mr. Ambur: That’s very encouraging. Of course, the outcome is more important than the letter itself.
Mr. Bargmeyer: Yes. The notion of a letter on its way is more important than the letter being there. Next week the W3C advisory committee meets. They’ll be talking about how XML works with these registries.
The Open Forum on Metadata Registries is January 20-24 of 2003, follow link from XML.gov, or www.metadata-standards.org/openforum2003. It has an excellent agenda. The chairs of UDDI and XML and 11179 registries, and editors of other standards are giving presentations, then it tracks showing how we’re practically using the registries and the connections between them. It covers Defense, Commerce, EGov, and others. If you know of groups that you think might be interested, I’d appreciate your sending a notice to them to look at the website.
Mr. Royal: Were you able to get Ron Schultz to speak about UDEF?
Mr. Bargmeyer: No.
Mr. Royal: I suggest you put it on the XML.gov mailing list, because it is an important forum. I wanted to mention to this group that we’ll present the business case for a government-wide ebXML registry/repository.
Mr. Ambur: Next on the agenda Steve Vineski update us on his effort to scope out enhancement of the Developer’s Guide.
Mr. Steve Vineski: In the August meeting, I committed to a review about shepherding the CIO Council’s XML Developer’s Guide, and combining it with the Department of the Navy Guide and EPA XML guidance to reconcile the three, review it by this Working Group, and produce a document for XML policy by the federal government. I put together a work plan on how to do this. LMI [Logistics Management Institute] was the author of those three documents. They would do a mapping of the guidelines, the commonalities, and where they disagree. Also, the documents are a little dated, so there’s an additional need to examine them. I thought the first task was to ask LMI for that product. I had a brief discussion with Mark Crawford, but it’s not yet confirmed.
Based on the mappings, we’d then have conference calls to allow people in this Working Group to ask questions about guidance, and based on those, interested people would need to submit written commits to me or LMI and we’d post them on the listserve. That would be the basis for developing final guidelines for a revised Working Group Developer’s Guide. Then we’d ask LMI to come in and present it, and also make it available to voluntary standards bodies for comments, and ask for written comments on the guide as a whole, then do a final revision for chairs and send it to OMB. This is a draft. We’re looking for feedback. We need to talk to LMI. We had an initial discussion at EPA for funding. I’d like to agree on an approach, then look at the cost. That’s about it.
Mr. Royal: The final one wouldn’t be completed until May?
Mr. Vineski: May.
Mr. Royal: That coincides with UBL 1.0. It’s around the same time frame; we should consider that. I notice that in the deliverables you specify MS Word and Adobe Acrobat. Can we have an XML version?
Mr. Vineski: I’ll put that down.
Mr. Royal: I’d be happy to work with you on that.
Break
Mr. David Colton:
“XML In The Federal Marketplace: What Are The Next Steps?” is an ITAA [Information Technology Association of America] event being held on how to take XML in the federal space and widen it to management and appropriations funding concepts. Representative Tom Davis is the keynote speaker. He’ll be setting out his vision for what he wants XML to accomplish. Per a conversation with Owen, we’ll avoid marketing—no ads about solutions. We’re looking at 200-250 senior people at the Reagan Center. It’s a half-day event. The date is Friday, January 24. The agenda is not finalized yet. We’d like to share, and to the extent we can come up with something useful, we’d be delighted to consider suggestions.
Mr.
Jean Paoli
XDoc
Architect
Mr. Paoli: I’m Jean Paoli, XML architect at Microsoft. I’ve worked in the field of semi-structured documents for 15-20 years. In the 80’s and the 90’s I worked in Europe on SGML , SGML authoring, and government and aerospace industry projects using SGML. I moved to Microsoft in ’96. I helped create the XML 1.0 standard, and my team (the MSXML team) and myself helped create a lot of the core stack of XML standards (XSLT, Namespace, etc). I was the manager of the MSXML team and the representative on W3C for Microsoft for multiple XML standards. My team built XML support (MSXML) in Windows and Internet Explorer (IE). At first, I concentrated inside Microsoft on explaining how to move the XML vision a step ahead on the infrastructure: we wanted to focus on the server—on the data and transport side. After IE 5.0, once we’ve established the goal of moving data freely between systems as the core of our business (.Net or WS), I moved to Office, inside Microsoft, because my heart is on the client. Now the next level is to bring XML to the desktop—to bring the creation/analysis of XML to the masses. We failed on that in the SGML time and I want to make this finally happen.
This is the first presentation to more than two or three people from our alpha program. I’m going to give you an overview for Office 11 (which is our internal code name), then focus on XDocs. I encourage you to see the XML support of all the Office tools that we are providing in Office 11. We’ll have an important presence in the XML 2002 conference in December. We’ll have more information and technical details at that conference. I hope to do 10 or 15 minutes here on slides, then 45 minutes on a demonstration of XDocs.
Slide 2 [Today’s Topics]: The first two bullets are the first 15 minutes.
Slide 3 [XML Overview]: Many forget that the goal of XML is to have a universal data format—platform independent—with which users create and use their own tags to describe data. With HTML, one couldn’t create new tags. In ‘96, a bizarre thing happened: XML was started in order to let users create and define their own tags. Now we all know about it. I repeat it, because the entire presentation about Office is about that: The power of XML is to let users define their own tags. In XML we use a formalization to help describe new tags: this formalization is called a schema and there is now an accepted W3C standard: the W3C XML schema standard that helps describe the vocabulary and grammar of new sets of tags.
For the clarity of the presentation and of use-cases, I like to differentiate between customers that define their own schemas because they want system interoperability within an organization (between their IBM machine and other machines), and customers that define their own schemas because they need cross-organizational communication within an industry—where it’s more for interoperability between constituencies. Those are two user needs, but it is fundamentally the same technical problem to solve.
The goal is easy data exchange between disparate systems; simple repurposing and reuse of data. It repeats the original XML vision, but is the basis of what we’re going to introduce on the desk top.
Slide 4 [History]: I started exchanging ideas with the Office 2000 team to put in some XML support in order to jumpstart the interest in the industry for XML support in the desktop. The Office team wanted to “round trip” their documents in HTML. You couldn’t do that using pure HTML, because you needed to save data and HTML handles only presentation, so they agreed to do it in XML. If you look in the HTML of any office document, those files embed some XML metadata information. It’s a beginning step but it was very early on in the history in XML so it was at the very early innovations stages.
The publicity from this quickly had people on Wall Street saying, “I’m doing a spreadsheet, and I want it on the server, because I’m using Excel.” So in Office XP, we created a consistent format for spreadsheets, so any spreadsheet can save the data as XML, and people can dissect it on the server. That’s the next logical step for the desktop, but the XML that’s created follows a Microsoft Excel design. It’s an advance, but it’s not the core of XML, which is to let customers—not Microsoft—define their new tags. That’s why I started working with lots of teams to arrive at a new model that is embraced in the majority of Office 11 tools: support XSD schemas. Support for generic XSD was a huge investment in Office 11. You can look at the surface of each box picture [on the slide]; the surface is proportional to our investment of XML in each version.
To shift the office model to embrace a new document model (support XSD), supporting customer schemas (where you define the schema as our customer, not Microsoft—I don’t know what use you’ll require for the document you need), there’s a great example: Web Services. Web Services have entities defining information they can deal with. Microsoft doesn’t know them. The person that creates the Web Service knows them. The big shift is, we put in wide support of customer-defined schemas.
This mean that you, the customer, define tags in XSD, and we support them in Office 11. The core is that the new document model supports something in the original XML vision—separation of data from presentation, separation of data from the tool that creates that data. Today that’s not the model you find in the mass market. It’s still theory, not practice. I visited a friend in France. He had IE because it’s mass marketed, and so he can use XML with MSXML. But there is not an equivalent wide presence of XML in the desktop. This is still not a reality. That’s the goal that I want to achieve for XML on the desktop.
Huge effort occurs on data validation, which ensures that you respect the XML rules for the schema. The result is that when you interact with a Web Service, it’s important when the document you’re using has a data structure. Now the software gives you more capability than ever. If the document has structure, the entire structure could go to the Web Service. So if you author half a page using XML structures, then do “tab” or a user function, this big chunk of data can go to a Web Service—say you enter the number of the truck, etc. It can go to a Web Service, a database, it can do some lookups (it doesn’t have to be a Microsoft database), and it can come back with another XML document and insert it as a second page of the document you’re creating.
Slide 5 [New
Document Model]: The customer schema is at the center. It
supports the XML native data model. It’s also a new way of thinking about
documents, because it’s not just about presentation, it’s about data. It will
help customers interact with the information they have. It solves the “islands
of data” problem.
Mr. Royal: Could you go back a slide? What is “sparse?”
Mr. Paoli: The fundamental XML data model is very wide: it spans from relational, to semi-structured, to very flat. People usually think of rows, columns, etc. Documents do not fit necessary in rows and columns.
Mr. Royal: I’m envisioning an Office document containing XML fragments from various locations—so you can support multiple schemas in a single document?
Mr. Paoli: In XML, it’s called a namespace. You can support multiple namespaces in single document. Word would be thought of as an integrator of that information.
Mr. Royal: Like a binder?
Mr. Paoli: Yes, but I don’t know whether we gave it all the support to do that.
Unidentified participant: So the XML is there, and XDocs is like a parser? Using XSLT to display it, it can consume a Web Service inside the document, with a WSDL and a utility?
Mr. Paoli: Yes. I’ll show you that. So the information worker (we used to call him the “knowledge worker”—he’s not paid to develop software, he is not technical and does not know XML, he’s paid for another type of job) can participate in the vision. Formerly it was limited just to IT people and developers. The information worker can participate because he’s connected to data centers now. The new vision is to think of users taking data from the back end, massaging it, sending it elsewhere using SOAP and your own XSD schema. The result is that Office is now a great client for XML Web Services.
Slide 6 [XML in Next Version Office]: Now this new model (customer-defined XML schemas) is supported mainstream. It’s in almost all of our Office 11 tools, so finally we have the possibility and the tools to move the world of the desktop toward XML, so the data can be reused. People ask, “Do you have an XML authoring tool?” I have six or seven of them. We think now in terms of scenarios, not technologies. All those tools understand XSD now, so the job is to pick the tool, not the technology. If the goal is to report, pick Word; if the goal is to analyze data, use Excel. For Access, if you have a database in Access, we added tools to let you create a hierarchical XSD schema from your Access database.
Unidentified participant: Does it have front end support?
Mr. Paoli: No. It uses a lot of XSLT and XSD. Front Page enables dynamic data-driven websites. These days websites use back-end databases. Front Page lets you, say, open an XML file and design how it’ll be rendered. In this new version, Front Page becomes a WYSIWYG XSLT tool. It lets you define the XSLT that will be applied to information on the server and then sent to browsers that don’t support XML.
I don’t yet know how the tools will be packaged. I don’t have any pricing or marketing information.
Every tool now can persist internal data in XML. For example in Excel, let’s say you have your own XSD to define what your government agency captures every day. So Excel, without any coding, has you go to “Open XSD.” You see the normal grid, then on the right is a new pane, which loads the XSD, then you can drag and drop each tag to the cell where you want it. Then you can open or import the file. The data is loaded in the spreadsheet.
Then you can read the data and do charts. Between this and our support of SOAP, you can think of Excel as a great Web Services front end. On one hand, you can bring data from a database that supports SOAP. Then they go inside Excel. It takes maybe only three hours to develop. You can analyze it, do it in, say, a bar chart, and update it every so often. It’s seamless. Excel is the best to analyze data, and Word is the best for documents. You can take any .doc file, without any customer schema, and save it as XML. This word file saved as XML could be parsed by any other XML tool. This is interesting, but for me, the most important thing is to allow one to generate one’s own schema. You can do it in the version of Word in Office 11.
Mr. Ambur: In opening and saving previous Word documents in XML format, how far back can you go in terms of versions of Word?
Mr. Paoli: Once you open it in Office 11, it works on any document. It’s independent of XML.
Mr. Ambur: With respect to records management, a long-term concern is outdated file formats. NARA has focused on wrapping proprietary formats in XML to assure their accessibility over time, as software becomes obsolete and is no longer available or supported. Maybe instead of thinking in terms of wrapping proprietary file formats in XML they should be thinking about making XML the native file format instead.
Mr. Paoli: It’s independent.
Mr. Niemann: The graphics are SVG?
Mr. Paoli:
No. VML is what we use now. If you didn’t define your own schema, Word
doesn’t know anything about it. Define, then associate, the schema to a
template, then you are creating information belonging to that schema.
Unidentified participant: How do you map a complex type in Word?
Mr. Paoli: I do not claim that we solved all the problems of XML authoring. We think of this as Version 1. We have two modes in Word. In the first one, you design, then you take XSD and apply the tags around in your template, then you go into edit mode and you stop seeing the tags. You have placeholders, and you type in new placeholders. We didn’t take the approach of classic XML authoring tools, or try to solve the problems in helping people see the tags in their documents. Our goal is to let users create documents that they’re already creating today (their business documents) where the tags are created in the background. It’s going to the masses, so for Word, one person wants to create a corporate profile. Someone else creates a template. He fills in the fields that are predefined, or he writes an extension.
So you can look at all those tools as tools that help you create XML documents belonging to your own schema. It will take too long to go into detail. I encourage you go to the XML 2002 conference, where we’ll go into in detail.
Mr. Royal: Your Excel as a user interface has nothing to do with SOAP?
Mr. Paoli: Right. All user interfaces are like the ones you know today.
Mr. Royal: So SOAP packaging, etc. is done elsewhere? It just places a file into a location?
Mr. Paoli: In Office 11, we ship the SOAP toolkit. Excel is a client, and is generating an XML file in memory. Then (we use a DOM) the XML is file sent from a client using SOAP to a server. With Office 11, using XML I’m a client on the desktop. Why? That’s how users get and use their information. If they’re connected, in the middle they can save the information.
Slide 7 [Customer Challenges Today ]: Why do we create XDocs as an application? To help you gather information. Customers are faced with challenges; in data gathering, missing or incorrect information, causes retyping, because we often need the data that is gathered in a different presentation. Also, “forms” are very hard and inflexible to use. For example, HTML and others lack documented editing experience. An example is that people use documents because they’re familiar. Forms don’t grow. They’re good for printing, but not for editing. In general, today, the data isn’t captured using your own schema. It uses some formalization the vendor did, then you have to try transform it to your own schema. XDocs does not work like that.
Mr. Paoli: First, here’s an example of forms created in XDocs. [Mr. Paoli displayed an expense report form separate from the slide presentation.] Here’s a health insurance form [Slide 9], a manufacturing service request [Slide 10], and a purchase request [Slide 11]. The extension is .xml. XDocs don’t have a file format.
It’s an example of interacting—for example, inside a business process, of how to collect sales call information. Say you want to fill in a Sales Call Report. You have an auto-complete for regular input, then tab, and you have a phone number coming from a database.
For each product, you select from a drop-down menu, then input the sales goal, then the actual sales, then unit price, then the total. “Tab” creates a new row.
Ms. Lin: Are these products created using
XDocs?
Mr. Paoli: Yes.
Unidentified Participant: Can you publish these on the Web?
Mr. Paoli: Yes.
Ms. Lin: Can you import using it?
Mr. Paoli: Can you cut and paste HTML? Yes.
Unidentified Participant: Do you do it in MIME [format]?
Mr. Paoli: Yes.
OK, I’ve filled out the form. Now I want comments. I insert text—I have all the Office tools—“Spellcheck,” “highlight,” etc. [The text input box grew according to the input.] So the form is behaving like a document. For this account, I want to list actions I need to do to follow up, so there’s an optional element I click on, and a set of grouped fields are inserted. It gives “Name,” and “Priority Due Date,” in a box that grows. I can change it if I make an error. It flags errors. I can add another action using a drop-down menu. Options come from the XSD schema, so it’s always using valid XML. The fields are grouped together.
I can insert a new action. It looks like a document. Now I can add another customer to the same document. I can highlight whatever portion of the previous one I want, and select an option from the drop-down as to when I put in Name, and other data.
Now it’s all entered, so now I can submit.
Now let’s switch gears. I’m in Outlook. I have a summary I want to review. I can arrange it as I wish—escalate it, etc. The information that’s held is all the information in XML. All I did to change views is use a normal tool. I use XSLT to view and edit. XSLT isn’t just for viewing, it’s for editing.
XSLT is a complex transformation language I needed to give users an easy view of the world, where what they see doesn’t have anything to do with the XML schema. The difficulty was to enable the XSLT views to become live.
Another example—information in an SQL database—we can create reports from that. It’s just showing that what I gathered in my document…
Ms. Lin: How do you display it, if the information needs to be redacted?
Mr. Paoli: I don’t do anything here; you guys can do it. I’m just using the W3C Document Object Model and XSLT.
I’m going to show you the design mode where you decide what security you want…we use digital signatures. We don’t use encoding, because there’s no wide standard yet. XDocs has two core use scenarios: one is at the enterprise level, and one is at the team level.
[Mr. Paoli demonstrated another way of accessing XDocs using Sharepoint Team Server.] You can create document libraries and share them at a team level. We didn’t create anything special—just a way of calling it. The actual format is the same.
Now I’ll create another company on the Sales Report (U. S. Government). [Mr. Paoli created and saved another Sales Report entry] It’s .xml format. I make a new file on the desktop, open it with Notepad. [Mr. Paoli displayed an example of an XML document with many tags, one of which was called “Sales Report.”] It’s all defined by your own XSD. This is all over Office 11. From beginning to end, it’s a well-formed XML file. The way we know whether you opened XDocs is, we added two processing instructions. One is for the icon (the code says “progid=XDocs.document”), and one is a URL to a CAB file. The content of that CAB file is completely neutral. It can be opened in any tool. It’s independent of Microsoft, and 100 percent W3C Standard-compliant (this is where the XSLT is stored with the schema and other XML files). The user doesn’t see any of this. It’s simply the way it’s encoded.
Unidentified participant: What’s important for us as the government to look at, is the output, because 20 years from now we can look at this. It’s structured as a document instance, so it’s frozen as a legal record. It can provide so many views for the information, and that’s the trick.
Mr. Paoli: XDocs doesn’t have a file format.
Ms. Lisa Weber: The hard part is identifying the record and capturing it.
Mr. Paoli: Yes, and that’s in your purview. I don’t know whether to call it “forms” or “documents,” because of the mixture of the types. It’s not a classic form. I’m dealing in semi-structured data models. Not relational, and not just a few fields. To show how it works, just open a document, click on the “Design This Form” icon, and now I’m in design mode. I can change things on the “form;” I can go to the “Layout” tool to add a row, etc. All of this is creating XSLT but in a WYSIWYG way.
Mr. Jim Disbrow: The XML tags in your example aren’t ebXML tags. We’d need to know how to change your example to have tags that have ebXML?
Mr. Paoli: I’ll show you in a moment. It supports all the different types. Continuing, I can drag and drop field names, etc. from the Layout tool menu. I’m going to create a form from scratch. I launch XDocs, and go to “Design a Form.” The menu has three options for Data Source:
¿ XML schema or data file,
¿ database, or
¿ web service.
[Mr. Paoli displayed an XSD file using the W3CXSD standard.] This is where the fields are defined. As long as it supports the XSD standard, we can load it in XDocs. [Mr. Paoli changed a field element name.]
Unidentified participant: Is the validation here?
Mr. Paoli: Yes. 99 percent of them are in the schema. You can add others using scripts.
This is your set of tags; let’s open it. I open it and say, “Finish.” It loaded the XSD on the right hand side, and it includes the new tag called “Government Item” that I just created. So you’ve defined the schema that represents the business process.
How do we then create a form that the end user uses? I go to Design Layout, and make some choices. The “form” will autopopulate based upon some of your choices. Then we can go to Data Source and drag and drop items. You can format it in any way desired. Then go to “Preview Form” and try it. It’s creating XML in the background using your schema. You can save it as XML and have an instance. This is saved in a .CAB file and generates XSLT. If you do “Save” it takes all files needed for this form and puts them in a .CAB file. You can also save to a folder if desired and then we do not use a CAB : the set of files is saved in a folder.
Now I’ll show how we connect to a Web Service directly from within XDocs. Go to “Design New Form,” then “Web Service.” What can we do? Receive and submit data, submit data, or receive data. Then it goes to a data source you define, it fetches the data from your source using the WSDL standard (UDDI), and it finds the WSDL service, and shows the three services that WSDL is showing you.
XDocs finds the XSD schema inside the WSDL coming from the Web Service, and loads it here. It has tags we can click on or drag and drop. I can view or query the information in the “form.” It’s all integrated here.
Back to the original presentation—
Slide 8 [Introducing XDocs]: A Sales Call Demo recap shows graphic of the process used in the demonstration using all the technologies that were used for the data.
Slide 14 [XDocs is a Smart Client for XML Web Services]: This shows the arc of how XDocs interacts with XML. It’s file-in/file out of your file format. Independent of any product.
Slide 15 [XDocs
is built from the ground up on W3C standards]: XDocs is
based completely on open standards. It’s platform independent. It’s based on
XML 1.0 SE, Namespaces in XML, XSD 1.0, XSLT 1.0,
XHTML 1.0, DOM 1.0, XML, DSIG, SOAP 1.1, UDDI 1.0, and WSDL 1.1.
Ms. Lin: How can you extract subsets of information from the document and send them to different destinations, like a database server or different system? Can it be done from within XDocs or without?
Mr. Paoli: We give you the W3C Document Object model so you can go to it and decide what you want to do with it.
Mr. Paoli: No, it’s not relevant. We use a very popular technique to associate a Stylesheet to an XML file. We needed transformation language for the scenarios we’re building here, so we decided to use W3C’s XSLT. The scenarios we’re building had to allow the views to be very different from the structure of the XML data. XSLT is good for the flexibility we need for user views.
Mr. Kevin Williams: Is there “code behind support”, such as C#?
Mr. Paoli: We added scripting languages. The CABs are loaded in a Sandbox model.
Mr. Ken Sall: .CAB files bother me, because what isn’t obvious is that now you have to send the .CAB files as well, right?
Mr. Paoli: You don’t have to, because I’m just exchanging XML. We needed CAB because I had an XSLT and an XSD etc and I needed to send them on the internet so what do I do instead?
Mr. Sall: If people are concerned about storing records, it’s more than the data. There’s the naïve user, the power user, and the XML-aware user, so I’m the naïve home user and…
Mr. Paoli: XDocs is not for them.
Mr. Sall: What do I get as a home user, and how will it interact with others?
Mr. Paoli: XDocs is not for home users. That’s very important.
Mr. Sall: What I mean is, to send a Word or Excel document, how do end users know automatically that a document is XDocs-authored versus a more basic one?
Mr. Paoli: You’re right. First, XDocs is not an end-all to everything. We’re in the business of gathering information. It’s a product that‘s specialized to gather information in the context of the business process—not in the business of the organization of documents; you need someone to do that. We’re trying to push companies to organize stuff.
We envision that someone thinks about schemas, does data models (what’s gathered). Once that’s done, someone creates a form template using XDocs, then decides who uses it and how it’s distributed. You have to have a form template library, but data is the important thing. The goal is to not store presentations, but rather data. Then we have to solve the problem of deployment, once we organize a little bit. For example, you have a web server where you put your documents. All the documents use HTTP and point to them. You’re right—should I have separated the data and the presentation? Or put it in one file—but then what is the format? How do I find the data? Do I compress? Maybe so in Version 2? Is it a good thing? I don’t know. We’re not thinking about the home user here.
Mr. Sall: Say I’m not the home user, but I as an XML developer, see your reference to a .CAB file, etc.. I want to take a straight XML document, and use it with another tool. I can’t validate?
Mr. Paoli: Yes you can. It’s your schema, inside your XML file. You point to your schema. It’s a purely, 100 percent XML file. If I take the XML file, it points to an XML schema, or if you don’t’ want to, and manage it by namespace, fine. I put the schema in the .CAB file because I need it to validate it in XDocs.
Mr. Sall: I encourage you to think about options in how people save the whole package to control the level of…
Mr. Paoli: We have a wizard that lets you change where you are publishing the XML.
Mr. Sall: The XML extension is what you’re using right? You’re required to assign to one or another?
Mr. Paoli: No. You can use whatever XML tool you want to open it with.
Mr. Disbrow: Let’s assume with EGov, with 24 initiatives, we do well, all using XDocs. Every time this is installed, someone has to define the default style sheets. If so, they [the style sheets] start from different places, and nothing goes to NARA that’s consistent. Let’s assume that across the government there’s a default upon installation, and a person doesn’t realize it, and it’s not the one that his organization wants. Is it doable so that when your product hits the street, it can be configured a specific way that certain people want ?
Mr. Paoli: What would you like XDocs to do?
Mr. Disbrow: When installation is done, say “The one installed is ebXML,” for example, so he’s aware of…
Mr. Paoli: We create templates which are associated with a particular schema, so if you send a link to an XSL file, for example, then the user will be using one associated with that schema.
Mr. Disbrow: Can you do it so that all these defaults are to whatever the XML Work Group has said is our standard?
Mr. Paoli: Absolutely! That’s a very important goal. It’s really associated to the form template, so say you create 20 form templates, you send me one, I’m just filling it in and sending it to you, not knowing that I’m using an ebXML schema. That’s very important to us. But, you understand that you send this form to the user or put it on the website.
Mr. Sall: The association will be there.
Mr. Disbrow: But it has to be on the default installation.
Mr. Paoli: You don’t have to install the template on your PC. You install XDocs. Once you do that, you send me “Form 1” and say, “Click on this and fill in.” If I have XDocs, the form is downloaded and installed silently. Then I can use it. I didn’t install anything. I as the end user don’t think of the schema—rather the form template. You decide to use ebXML, XDocs will work fine. The user can’t generate anything wrong.
Mr. Disbrow: Whoever defines the defaults on installation is going to set the direction. I’m saying, is it possible to lock users to a specific template?
Mr. Paoli: You can disallow template creation mode. Then, how do you manage access of which form they can use? It’s a matter of permissions.
Mr. Disbrow: I don’t want to disallow anything that’s in the product. I want to ensure that NARA doesn’t have to set up 500,000 aliases because everyone comes up with their own format. I want to have a uniform starting point for everyone in the government.
Mr. Ambur: That’s the focus of the entire FEA. The DRM is not yet specified, so we can’t provide a default for every element.
Mr. Disbrow: But for certain pieces we can. We aren’t ready. Your product isn’t ready; I think we should think of it that way.
Mr. Paoli: It’s going to open up this management problem. I’d love to have your feedback. I’m not 100 percent sure we shouldn’t have an optional format to put the content and presentation together…
Mr. Ambur: Have you heard of XFDL – eXtensible Forms Definition Language?
Mr. Paoli: Yes.
Mr. Ambur: Its widespread adoption as a standard is down the road aways, but it aims to tie content and presentation together for purposes of records management and legal admissibility.
Mr. Paoli: You have to choose between whether you want to put the data and presentation in one file. We could put the XML file in the .CAB in Version 2 or in a ZIP file. Maybe we could do it—there’s not a big difference. If this is needed, it could be a good feature request for Version 2. No one generates much XML on the client these days; it’s mostly on the server—back-end to back-end. We would like to have XML used on the desktop everywhere, on an everyday basis.
Mr. Prince Billups: Your product is only for clients, not on the server side? Where, in Jim’s scenario, would it let different namespaces choose which schema they’d like to use?
Mr. Paoli: If there’s a server component which is useful, it could be developed, but my opinion is that people on the server who do document management or workflow management systems should have that on the server, and XDocs on the client as a tool that goes to the server. The server tells me what schemas are available; I read in XDocs, I choose, then the user is unaware. The user just picks the forms he might be using.
Mr. Billups: So the package is database independent.
Mr. Paoli: Absolutely.
Mr. Billups: No special middleware?
Mr. Paoli: Absolutely. It’s designed to be a smart client for XML.
Mr. Ambur: Bearing in mind the challenge set forth in the eGov Strategy to make our applications “citizen centered,” that’s a good thought on which to end the meeting – a smart client for XML.
Attendees:
|
Last Name |
First Name |
Organization |
|
Adams |
Susie |
Microsoft |
|
Ambur |
Owen |
FWS |
|
Bargmeyer |
Bruce |
LBNL |
|
Billups |
Prince |
DISA |
|
Campbell |
Richard |
FDIC |
|
Cocos |
Dena |
LMI |
|
Edwards |
Scott |
Corel |
|
Ellis |
Lee |
GSA |
|
Lewis |
Diane |
DOJ |
|
Meng |
Lin |
DOJ |
|
Morrison |
Sylvia |
FDIC |
|
Niemann |
Brand |
EPA |
|
Paoli |
Jean |
Microsoft |
|
Royal |
Marion |
GSA |
|
Sall |
Kenneth |
Silosmashers |
|
Smith |
Jane |
Fenestra |
|
Troutman |
Bruce |
8020Data |
|
Turnbull |
Susan |
GSA |
|
Vineski |
Steve |
EPA |
|
Weber |
Lisa |
NARA |
|
Weiland |
John |
NMIMC |
|
Williams |
Kevin |
BlueOxide |
|
Yee |
Theresa |
LMI |