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: We might as well get started. I’m Owen Ambur, co-chair of this Working Group. My co-chair, Marion Royal, has been ill, so he won’t make it to today’s meeting. We usually start with introductions, including your name and interest XML.
[Participants introduced themselves.]
Mr. Ambur:
Our first speakers are John Dodd of CSC and Davis Roberts of SAIC. John and Davis are regular participants in
this Working Group and are here today in their capacities as representatives of
the Industry Advisory Council (IAC) to talk about their work on white papers
relating to the Federal Enterprise Architecture (FEA). John?
Mr.
John Dodd, CSC
Ms.
Davis Roberts, SAIC
Leveraging
XML Standards and Technologies Linked to Enterprise Architecture: Key to
Government Transformation!
Slide 1 [Title Slide]: I’d like to tell you first what the Industry Association is. The Industry Advisory Council is made up of over 400 companies that are part of the federal IT community. It’s also tied with the federal side of the CIO Council, so there’s a federal part and a commercial part. Part of our charter is to bridge that gap and address challenges around the architecture. We formed the SIG in April of 2002. Both Davis and I got involved with the meetings and our names somehow bubbled to the surface. We didn’t initially intend them to be bubbled so high.
The Federal Enterprise Architecture is putting out a set of reference models. There had been an earlier enterprise architecture framework document in 1999. We set about looking at companies who used it, then we were given a set of challenges. One is how to leverage enterprise technology. Today’s is to leverage XML standards and leverage them to the enterprise architecture. The key reason to know about that is that this is the perspective from which high level decision makers in the government want to look at their enterprise technology.
Slide 2 [New Challenges Facing Agencies]: Our goal is to pull organizations, both in industry and federal agencies, and provide best practices, support, and outreach. We want to work with the CIO Council and government customers to ensure that our papers make sense. At the end, we’ll do a welcome for industry and government and ask representatives from both to participate and review our work.
The big challenge of the enterprise architecture concept is business needs and their focusing on business lines. We’re strongly recommending to all that they let this cross industry lines. We think we should align our process around the reference model and influence how the reference model is done.
This started with Bob Haycock. One of his first meetings was here with this Working Group, and he challenged us. We wanted to look at that challenge, use those reference models, and see how those technologies fit into the reference model.
I’ve given you a little about the IAC. As a start, we’re looking at a set of standards, so that every time you create interfaces, you don’t have to reinvent the wheel.
Right now, XML technology is crossing a chasm and becoming mainstream. It’s not there yet, but it will be mainstream. There’s a lot to be done, because some pieces are mature, some are just now emerging, and some are really off in the nether land or just great thoughts.
Slide 3 [Agenda]: [Mr. Dodd did not discuss the agenda, however the contents of the slide were the following:]
· IAC Role and Status
· FEA Reference Model and New Approach
· Enterprise Models- Linked to XML and Web Services
· Focus on “Cross Government Information Sharing”- Business Line Architecture
· Value and Leveraging XML: Challenges of Infusion
· Recommendation for XML-based Technologies becoming “Mainstream” and Pervasive
Slide 4 [Context Setting]: I don’t know how many of you have heard of the reference models, but I want to give you a little on that. They are target architectures that agencies need to look at, to see if they fit. “Are they useful and usable by eGov projects and agencies, and do they provide a reference so companies can map technologies to them?”
Mr. Ambur: They’re also accessible through a link at XML.gov. [Editor’s Note: See http://xml.gov/index.htm#fea]
Mr. Dodd: One thing that will change is that the business case is now important in the government. It’s also represented by joint agency business cases. If you have a joint case for one agency, that’s good. If you have one that is joint between agencies, across multiple needs, that’s very good. If you have a neat technology, you have to show how it’s useful.
Slide 5 [Business Operational View]: One of the focuses of our papers is to bridge the operational view of the enterprise with the technical view. They’ve been in two ethers up until now; there must be something to bridge the gap. They must be business-goal driven—show the transformation process, where you are today—and look at leveraging inherent features of enabling technology. You now have to explain this to senior managers, because these days they’re more and more aware of security. At the very tops of agencies, they’re talking about security and information sharing. I saw an article in the [Washington] Post today about how the Governor of Utah is talking about interoperability. It’s “dangerous” when political officials talk about interoperability. I was afraid to even say it before, and they’re talking about it now.
New technologies—you have to introduce and explain the technologies, because people are scared of new technology. One of the things is also acting in business analyst roles. At the same time we’re advocating a federated approach, it’s not necessarily a new technology—centralization versus decentralization—but you can have a community at a central point or a few nodes, and have the other parts be distributed. Maybe a few zones, with other areas. You may have .NET, or you may have J2EE, and the business needs reign supreme.
We’re also looking at supply chain. This has to be managed. This has been done informally; Global Justice [Network] has a process. EPA has one. Is there a way to do it so that the federal government doesn’t have to deal with 35?
Is there a structure or an infrastructure that would define and help security and higher layers of reuse?
Slide 6 [Overall Context Diagram: IAC Papers Linked to FEA/RM and FEAF- IAC]: So we looked at the reference model and talked to Gerry Wethington of NASCIO [National Association of State Chief Information Officers] and Bob Greeves [DOJ Global Justice Network]. We said one problem is information sharing. Everyone “wants to.” Gerry is CIO of the state of Missouri, and Vice President of NASCIO. Now, everyone wants to do it differently. We ‘re talking about defining a government-wide framework. First, we have to look at model usage. We have to define the business model.
This
is the wedding cake. The top layers talk about services to citizens; then
there’s a set of internal services; then below them are the infrastructure
services. So we say, “The Department of Education, Homeland Security,
etc.—different agencies—are trying to use this reference model and define
service components,” so we developed two white papers. The next thing people wanted was the
Performance Reference Model. It’s not yet released. We did a white paper and
developed a balanced scorecard on how to do the architecture, and linked it to
the Business Reference Model. Our whole recommendation is, link them together
and provide focus around value management, and provide tools that are business
tools and business analyst tools.
There has to be a vision for how all the authentication/ auditing/ access control/ privacy concerns are done. Even if it’s not complete, it would tie the work of NIST with the Federal Enterprise Architecture. We said, “One of the big issues is, this will create a lot of process change. What are the cross-cutting security considerations? How do you divide the organization into zones and responsibilities? How do you move your organization up in this enterprise? How do you get yourself started?”
We did a set of papers on the Data Reference Model. “What data do you need to share?” Not everything, but there have to be common approaches. There are standards that are facilitators for those kinds of federated data, and that help with that kind of federated trust management. Those are the things you have to share. You’ll see diagrams on that—SOAP with a security approach—but there are some deficiencies. That’s leading to, “How do you govern this across business lines, so that people working Homeland Security, who have two needs—security and a set of customers…” They need to have a vision of where to go. They need security, and they need to have the people aspect.
The business line needed to have very much what Susan Turnbull has done. It has to be able to form communities. It has to have a model aspect. There has to be an agreement aspect to let the people agree. It needs some kind of starter kit so you‘re not defining the standards each time; and then it needs an ability to extend, so it’s a core set of infrastructure.
Slide 7
[Transformation= New (Thoughts + Models+ Process) with Value-Focused
Actions that are Managed!]: So this is my math picture of
how to link together. There are a lot of problems at the top about the way of
thinking. You hear a lot of rhetoric, but if no one takes any action, then
it’ll be just that—rhetoric. We talk about it being citizen-centered, along
cross-government business lines. Every vendor has to be linked in the services
they can provide. One of the things you have to do in government is provide a
way of focusing on which technologies are valuable today—using SOAP, any of
these—how do you present it and convince people that it’s a new way of managing
that’s worthwhile? There are lots of technology models that people are putting
out. We think there’s another component: the process model. Each of our white
papers tries to define 6-8 actionable steps, to be put in the blueprint of how
government can change, so we’re trying to get them to be more transformational.
Slide 8 [The FEA Business Reference Model (BRM), Version 1.0]: This is the business model in terms of the citizen, so all these have to be linked.
Slide 9 [XML Standards and Web Services Alignment with FEAF and FEA RMs]: This is, “Where do all the technologies fit?” There are technologies that fit…the architecture has a bus architecture, which before the enterprise architecture was an IT architecture. Now, it’s a business architecture—how the organization runs the business. There is a new set of emerging standards involved with business process management—BPMS, ebXML/BPSS, BTP, UML/EDOC, UEML, WSDL. A bunch are related to data and information modeling, some to security architecture, and some are just emerging related to components. It’s a complicated situation. We need to relate the technology to the audience. There are multiple audiences. There’s a technology audience at the bottom, which is interested in many aspects, but at the top are business analysts—so as you introduce IT you have to be able to talk those different languages.
Slide 10 [Standards Mapped to the Reference Model…..Work with Standards Organizations to Meet e-Government and EA/BLA Needs]: So we recommend that here’s another cut of where the technologies are. So we wanted you to think…when you talk about performance, there’s not a clear set of standards or initiatives, except XBRL, which is really financial. Many are moving, but you’ll find by looking at the reference model that pieces are missing. What the IAC is recommending in general is a technology management process as we look at Web Services.
Slide 11 [Recommend Technology Management Activities]: I’ll go through a couple areas. We’re emphasizing that technologies have to be leveraged. They are maturing, going through phases. Some are going through a quick migration from incubation to mature; some take forever; and some won’t make it out of the lab, so in general, high level government mission people start to think and understand the high technology, so OMB has to select key emerging technologies that have high leverage value. So the CIO Council has formed groups that discuss that. The XML Working Group is one.
We’re looking at business technology; asking questions about proving the technology. We want the benefits, the value propositions. Those have to come from you guys. You’re the technology leaders. You have to provide that value proposition and technical education to all levels of government, and have to provide roadmaps. I’ll show you some examples of where XML technology is today. It’s not one thing; it’s a set, and it’s a complicated roadmap. You have to talk about how mature it is. You have to look at where it is today, and be involved in open standards. We’ve taken this to the Mark Formans and Norman Lorentzes—that they have to have strong participation in W3C, OASIS, etc.—have a government point of view, such as with Roy [Morgan] and NIST.
There has to be another vehicle. If this idea of infrastructure happens, and has pieces, you are the competency center for the technology. For instance, Karen Brown at NSA has a competency center. It could provide benefit across other areas of government. Have a technology integration timeline and integrate into a sequencing plan—so you expect a technology to come out, and standards to occur. If it doesn’t come out, you put up a red light that “this is hung up.” We know these things hang up. We know there’s a risk factor. We have to examine that.
We’re also advocating some “solution labs.” You can’t make every project take on the risk of introducing a new technology. There has to be some shared experimentation, especially around integration. If we go around business lines and integrate, we have to share the work.
Slide 12 [Strategic Selection of Key Enabling Technologies]: So here’s the process of piecing out the technology. It’s really just a process, that has to be in there, of technology assessment and cultural awareness. “Is your organization aware?” Sometimes we can collectively share, because some organizations are reluctant to change.
Slide 13 [TRM and
XML: Web Services—Self Describing Components]: One of the
key technologies government is looking at is componentizing technology.
Components can’t sit without components between them, so we’re looking at a
reference model of components, of characteristics. WSDL is one of the things
that gets to characteristics.
Slide 14 [Self-Describing SOAP-based information Exchange Package- Key Part of Infostructure]: We’re looking at an approach to messaging tied to security headers, where XML encryption and signatures are linked to agency policy, so we can work on stuff that SOAP and LegalXML have done. We need a common approach to things like auditing and authorization; minor details we don’t all need to invent. It could be part of the infrastructure.
Slide 15 [Leveraging Technology Work Groups should define Technology Roadmaps: Keeping Options Open]: We also recommend that the organization, as it’s updating, look at technology. You might have some tools, and say, “What’s currently available and mature?” Maybe we need to use some emerging technology because it has a high benefit. Maybe there will be some experimenting with what I call “incubating technology.” Some government organizations are doing stuff like LegalXML or Topic Maps. They sponsored some of this over there at DARPA and NIST. We have to move it from emerging to mature, so we could have workshops.
Slide 16 [XML Technology Map and Maturity Assessment-01/2003]: Here’s my diagram. The green is mature. The center has the core XML components. Some are mature, some are emerging, and some new ones are incubating. Some of the benefits are, you have to have a “trust and security” aspect to it. There’s also a movement about federated trust management, so we all could have a debate on this. Right now I don’t think XForms is mature. I think Owen said that whether something is perceived as emerging depends upon the company in the room. If you have a product in one of those categories, then it’s mature. If you’re talking about a rival, then it’s incubating—not that we would do those debates. So I’ve put together some of that, but one vehicle would be to use this as a talking point. The thing we haven’t done in our thought process is, “What happens when those technologies mature and we start using them?”
Some are simple add-ons, some are a significant change-—XLink, XPointer, Xpath, and XQuery can change the way you’re structuring your website, the portal. It gets to content management.
Some companies think they have, say, 85 websites. They might really have 1290 if they were to examine it. How do you go from “As is” to where you want to be, and understand what you have? We haven’t wrestled with it as a group. It’s something you may want to think about.
Slide 17 [Technology Readiness Timeline]: Another way to do it is with a timeline. You link a timeline to an initiative. We tried here to link department and agency initiatives to emerging and incubating technologies to look at how they could be fostered over time. The whole purpose is, if you can find a value proposition, that’s critical. For the government, I think it’s the whole emphasis on information sharing. How do you manage interfaces? It’s “dangerous” when governors talk about it. I heard Senator Liebermann talk about it, and it shocked me. The problem is integration. How do you put it together?
Slide 18 [Focus
on Information Sharing: Interfaces, Interoperability and Integration]: If we focus around a business-line architecture, it’s a core for
a model that can be extended; a way to exchange packets of information using
common components. How do you work with other organizations? It’s not just
federal; it’s government (states). How do you get involved in task forces;
encourage participation by government; create a standard profile?
Ultimately it has to impact the acquisition process. If you’re not asking for it, you won’t get it. You want to foster an open standards approach, or it’s not going to have an impact. We don’t advocate having just one process that fits every government agency, but we also don’t want thousands. The key is to take a few actions. Some say, “This is all good, but if we could get one or two changes every year, it would make a significant impact.” One of the messages I’ve been hearing is, “How do we get that business line to make changes?”
Slide 19
[Business Service Framework: Federates Data and Links Business Services
across Business Lines]: Here’s one on federating data. We have
zones. It could be a way to have business-service data federated, shared, a map
of the common infrastructure. Maybe Homeland Security, Department of
Education—affected States—they’re all business lines. Some data have to cross
between those entities. You’d think Homeland Security and the EPA have no
commonality, but maybe hazardous material, for example, has a common area
between the two.
So
there are integration points, a context or zone, a level of security, with
roles, integration points, and firewalls. If we define a common infrastructure,
we then have a framework, so not everyone has to define it each time. There
would have to be extensions.
Slide 21
[Cross-Government Enterprise Architecture Management]: There are some channel descriptions. There has been some work on
defining channels, but it’s pretty weak—just papers. It would be a help if
these were in a common model, or format, so they could be shared. We could
reuse and integrate around those models.
Slide 22 [Next Steps]: So I tried to leave time for questions. Each paper has elements of XML, elements of technology in them. We’re putting sidebars on of Web Services and XML. There are nine papers and an overview. We’ve reviewed six so far with the government. The overview changes each time we review, so it evolves. We want to encourage external reviewers. We’re working on three more papers—Security and Privacy Service Framework, Web Services to EA with Action, and Business Line Infostructure and Aligned Process with NASCIO. XML security is somewhat embedded in these presentations. We intentionally review these with Mark Forman. We bounced ideas off of him for the next set of papers in the overview, so I can determine whether we’re headed in the right direction.
Hopefully the IAC Enterprise Architecture SIG won’t be totally divorced of these white papers and thought problems, but we’re going to shift our focus to how we shop our outreach and deployment of ideas to the broader community. I put Davis’s email address as the POC. We’d like more feedback. We’d like these either expanded by ideas we haven’t thought of, or someone could tell us, “This is dumb. It won’t work ‘because.’” We want to hear it. The cynics always have something of value. We have to listen.
Mr. Ambur: We’re right on time. We have time for questions. With respect to IT architecture, I would just say I like to focus on information. So many technology folks want to focus just on the technology. It’s important to focus on what we’re aiming for, which is high-quality, high-value information to use in decision-making processes. I think you made an important point with the graphic about actionable items. A common failing of IT architecture is that the planning is too abstract and doesn’t lead to anything that can be implemented.
I believe we should start with the bottom-up approach, focusing on communities of interest. Those starting from the top down should be careful to scope their efforts narrowly and focus on very important drivers, like Homeland Security. Also, if we’re going to express a vision like citizen centricity, it should be made explicit in terms of XML data elements. Otherwise it may be little more than rhetoric. My personal interest is the Data Reference Model rather than the Business Reference Model because I don’t think we really know whether we have common lines of business until we get to the elements that actually describe the business. I encourage a bottom-up approach, stimulated by strong inspiration, vision, and support from the leadership. I’m trying to remember the name of the fellow who was Deputy Director of OMB, who went to NASA.
Mr. Dodd: Shawn O’Keefe?
Mr. Ambur: Yes. When he was at OMB, he said the creation of a Federal CIO could have the inintended consequence of absolving agencies from doing what they should be doing anyway, without being told to do so. I’m afraid that the top-down approach can have that effect. People may wait to be told what to do. That’s unfortunate. Nor am I a big believer in modeling. In my view, we have all the models we need. They are the forms we use every day in our business processes. We should inventory the data elements on our existing forms. If we can harmonize the elements as we are registering them, great. If not, at least we’ve taken the step of inventorying our data elements and rendering them in XML.
Mr. Jim Disbrow: You [Mr. Dodd] represent 400 companies, all of whom have a business model to sell a product to us. If you could change it to sell one copy to GSA, that’s then available to others. Then we have a way to work together. For example, Topic Maps. If I want to use one, at this point I can go to one company only. My agency won’t spring for that, but if a COI [Community of Interest] pays for and gets one, then we can use it. I can develop a wonderful ontology, but I can’t hand it off. No one can exchange.
Mr. Dodd: That’s the idea of not just infrastructure--of standards—but of a competency center. It could be a government agency, or it could be a person. If you want this technology, the technology would be there, it could provide the technology and support. I’m not saying it’s a contractor, but it would be a place together.
Mr. Disbrow: But that one company would sign a contract with the competency center, and would then make further agreements with others when used further. That model doesn’t work well. I like SOAP, but for everything else, you have to pay that $400,000 dollars to get that one piece of software.
Mr. Bill Morgan: You have a proposed evolutionary target architecture, and you have different agencies at different levels of meeting the target. In terms of a transition plan, what’s happening is that even if the reference model is 100 percent fine, different agencies move at different rates. The implications are that ultimately government operates by the least common denominator, so it’s critical that there’s a transition plan that moves agencies at the same rate. It’s politically difficult. It’s also constrained by funding. For example, we develop a government direction for contracts. Agencies will be saying, “Why work with you, because it’s not in our business process. There’s no benefit.” That’s an issue. No matter how well the architecture is made, if there’s not some plan to move at the same rate, then we’re constrained…
Mr. Dodd: “No agency left behind?” You’re saying move forward at the same rate so we all get some benefit.
Mr. Disbrow: If you’re in the private sector--Cisco transformed their organization in a short period of time. If the window is too long, it will never happen. You’ll always have legacy systems--some that don’t meet the architecture, so it’s critical…
Mr. Dodd: Say we took an infrastructure, with three levels of sophistication, and we want everyone to be at a minimum level--say, “By 2004 have this…”
Ms. Davis Roberts: The budget process has a lot to do with that.
Mr. Dodd: You have to link the enterprise architecture and business plan to the budget. It all comes down to dollars. The big thing is, government manpower resources are really going to be hurting due to key retirements in the next few years. There are a lot of inhibitors. I think they inherently know that. I think you’re all saying we need to make those inhibitors more visible. We have challenges defined in each paper, which we’re trying to address, but there are remaining inhibitors.
Mr. Bob Greeves: There’s danger in what the IAC is doing. You can’t hold the IAC responsible for every negotiated government relationship. I think you’re trying to help Bob Haycock with a meaningful approach to the enterprise architecture. I think the missing piece is, “What are the nine papers--how do they relate to the reference models? Are they drivers? Are they implementations on the back end? What is the purpose in the flow of getting the enterprise architecture out?”
Mr. Dodd: In the Business Reference Model, it was already out, so we looked at usage. In the Data Reference Model it was early, so we looked at the concept, and we’re trying to influence it.
Mr. Greeves: You understand that, but the audience doesn’t necessarily understand it. It’s important that they understand the frame of reference.
Ms. Roberts: We’re looking for a place to post the current draft. Maybe we can get a summary to Owen or Roy and you can post it.
Mr. Roy Morgan: Sure.
Mr. Dodd: Everyone has some areas that they’re more interested in than others.
Mr. Ambur: You’re saying you need linkage to the FEA models and their status or maturity?
Ms. Roberts: We need your insight on them. We’ll get them to you.
Mr. Dodd: Every time we talk to a group, we get new insights. These are the titles [Of the papers. Mr. Dodd pointed out titles contained in the material in Slide 6 on the screen.] There are two on model usage; one on PRM; one on the Security Service Framework; one on enterprise architecture; one on Business Line Infostructure; one on interfaces, interoperability, and integration; one on the Data Reference Model; and one that is overall.
Mr. Greeves: You need to take this and reduce it to one or two pages on how these relate.
Mr. Ambur: First, just a 2- or 3-column tabular display.
Ms. Roberts: We’re working on version control.
Mr. Dodd: When you have 60 companies…talk about cat-herding!
Mr. Brand Niemann: I heard from Norm Lorentz at the Federal Enterprise Architecture SAWG [Solution Architects working group], that “leveraging technology groups” such as the XML Working group and the Web Services Working Group are going to be operationalized within the AIC [CIO Council Architecture and Infrastructure Committee]. Agencies are asked to play an active role in that. It’s not just top-down. It may look that way now, but it will become more bottom-up as a result of including the agencies in the committee. They’re trying to set the infrastructure to bring a bottom-up aspect to it.
With regard to Topic maps, we have that as part of the February 18 agenda. Richard Brown and his contractor (Coolheads)are talking about how they’ve implemented the best example of Topic Map Web Services in government (80,000 topics for 95 IRS publications, with 800 FAQs), and how the rest of us might use this to address the E-Gov act requirement to categorize and taxonomize our government content. There are about a half dozen contractors in that space. They’re piloting it in a number of offices to get feedback. We’re also asking him to tell us how he’s working with his contractors. We can gain a lot from Rich Brown at IRS. It seems to satisfy E-Gov on taxonomy.
Lastly, I couldn’t agree more with Owen. We have a new pilot of e-forms for E-Gov that OMB has asked us to do. Rick Rogers of FENESTRA is leading the E-Gov effort on e-forms. He’ll report on what they’re doing. They have commitments from 10 major players to take a small selection of forms and mark them up with XML. The vendors have agreed to show us their implementations. Hopefully that can jump-start this use of forms, and do what Owen has asked us to do for years. We welcome anyone else with a challenging form to send it to us. The IRS has done that. David Heiser has given us his most challenging form. It had a flow chart in it. The experts can’t do it with SVG, but SMIL and XForms can easily handle it. We’d like to get the most challenging examples of forms you have to include in this pilot.
Mr. Dodd: I think I started with people pulling each other. I think the thing we can do is pull each other up. I hope you can help us improve.
Mr. Greeves: Be careful about standardizing on terms. I notice you’re using terms like ‘applications’, etc., so be careful about confusing people. Lastly, I’d rather see you, instead of hammering on XML, then relating it to Web Services, shift that around. Most executives get the message on XML now, but they haven’t gotten it on Web Services yet.
Mr. Dodd: When I talk to the Web Services group, I will do that. I like to talk to the audience. There’s a whole set of sidebars. It’s really XML-based Web Services.
Mr. Greeves: Executives haven’t gotten the message that the way of the future is Web Services. It’s a very important message.
Mr. Dodd: There has to be a bridge. “Service component” means application…I was at a breakfast December 18 with Mark Forman and Norm Lorentz, and I asked whether emerging technology could be considered in XML standards in making them adoptable. Mark said “Definitely.” We’re addressing verbosity, complexity, and most important—ambiguity. We have capability and security built in from the ground up, and with 20 moving parts of what’s necessary to implement Web Services across tiers, if we can cut down to what’s evolving, we would like to see that considered. We need help in how to do it.
Mr. Ambur: I see the wheels turning in Davis’s mind… Maybe we’ll help her recruit some reviewers. Let’s take a break, and come back at 10:35 AM.
End Presentation
Mr. Ambur: I just want to mention that we have an attendance roster if you’d like to register your attendance. We had a few folks who came in late if you’d like to introduce yourselves.
[Introductions]
Mr. Ambur: With that, I think we’re ready for the next presentation. Chuck Myers is here with a contingent of people from Adobe. Chuck?
Mr.
Chuck Myers
Adobe
Systems
PDF
and Acrobat in the Federal Enterprise Architecture
Mr. Myers: As Owen said, I’m Chuck Myers from Adobe Systems. I’m part of the PDF group, and I’m a product manager for forms products. [Mr. Myers introduced his Adobe colleagues.] Between the seven of us, you have Adobe’s representatives to Idealliance. I’m on the board of that, PDF-A, W3C, OASIS, and some others—so we’re strongly committed to these processes. I have 60 slides. I’ll only do 40. I want to talk about PDF and Acrobat, and how they fit into the FEA.
Slide 2 [Presentation Goals]: I’m going to do a review of E-Gov initiatives and Adobe’s fit in that regard, do a review of where we are today, and talk about a couple of distinctions you really don’t hear much about from Adobe—PDF and Acrobat. While they’re related, they have completely separate lots. Then I’ll talk a little about XML documents and PDF use, and talk about two worlds; the data world and the document or publishing world. They’re distinct worlds. Mostly you hear us talk about publishing. Today a lot is on the data end. We’ll talk a little about the next version of Acrobat and give you some backup on what’s going on with other Adobe products if time permits.
Slide 3 [Adobe Systems]: They make us put this one in. This gives you an indication of our size and mission. We have a major emphasis on getting business to integrate and communicate better, and in so doing we’re getting to the transaction side of things.
Slide 4 [E-Government Definition and Goals]: We’re looking closely at e-government initiatives now. We did a talk and met Owen at the XML 2003 conference. As a matter of fact, I missed Bob Haycock’s presentation to meet Owen. This is what we see as the meaning and goals of e-government—minimizing paper, and making information flow in a friction-free way, which would make life better for real people.
Slide 5 [E-Government Initiatives (24)]: You probably know this list by heart now. [The list identified the following functional areas:
· Internal Efficiency and Effectiveness
· Government to Citizen
· Government to Business
· Government to Government
Slide 6 [E-Government Initiatives (continued)]: I’ll show that Adobe is already active in these initiatives. The ones in red are e-records. U.S. Courts have been using PDF for a while. NARA now has a pilot; there’s tax filing in government—citizen-to-government. A lot is going on in the IRS.
Slide 7 [Why Acrobat and PDF?]: Agencies are using PDF and Acrobat for a lot of compliance information, whether they be FDA [Food and Drug Administration] or USDA [U.S. Department of Agriculture] or others. Many are getting such that their information comes in through Acrobat. There’s the question of why you should use Acrobat—why care about PDF? it’s a reliable way of sending information around. You know it’s always going to look the same. Many applications don’t care about the way data looks, but if you care about it for perpetuity, it’s wonderful. It’s also a great way to integrate paper processes into data processes, which are more and more XML-based, so you can bridge between the two.
We’re hearing more abut e-paper. It fits the paradigm of being the way people think. OMB comes out with new forms, so people think that way. “How do I make the transition from paper to e-forms, and at the same time, work offline rather than online.”
Last but not least, Acrobat is virtually everywhere, so if you choose you can count on getting information out in a variety of ways.
Slid 8 [PDF Example Use Cases]: A way to look at this on where we’re going and the way people take these efforts on PDF and Acrobat is to look at use cases. We have three use cases; document/bill presentment (“How does the Social Security Administration send out information to you?”); offline form applications, and database/Web Service-connected form applications.
Slide 9 [Document/Bill Presentment]: So in bill presentment you have an ERP system. Whether it’s SAP or another is unclear. You’re going to make XML data, or it could be more retro stuff too. We have products that take this XML data, format it into PDF, stick metadata into it, and send it out as PDF. You can archive it. People can go in to see how it was sent out.
Slide 10 [Document/Bill Presentment (continued)]: The key thing is, I can have one kind of document and it goes to three constituencies. An example is your phone bill. I send it to my mother. She looks at it, prints it, and sticks it in the “to be paid” box. The only thing is the presentation.
The next level is small business, except the bill is electronically enabled. You can click on the bill, export the data, and have the transaction go through. It’s not automated. It starts with the presentation and a link to electronic capability. But on a large scale, you’re not going to have someone look at them. You’re going to have a system that sucks the XML data out and does the transaction—maybe take the PDF part and store it as an archive. Adobe has entered into a relationship with SAP to integrate these kinds of things.
Slide 11 [Form Entry (Compliance/Applications)]: The next one is forms. I don’t have to tell you about applications for forms. The typical interaction today is, a person queries for forms. Right now with IRS, you get a single form. I’d like to get forms with my information on them. Get a package back. Data integration with that really works. I work with that offline, then when I’m done, I send it to a server that sucks the information out, makes a PDF, and sends it on its way.
Mr. Ambur: I’m delighted to see this slide, because typically database designers and administrators ignore the record-keeping requirements, and that is a prescription for waste, fraud, and abuse. It’s rampant. So I’m delighted to see that Adobe recognizes the need.
Mr. Myers: It’s sort of built in to our assumptions. For a while, Acrobat could deal with online databases—could populate, validate, and prefill. We see people going with this to use Web Services, with SOAP as the main transaction and WSDL for the data. People go to Web Services to invoke procedures to make transactions kick off, so the form is the front end to applications and databases.
Slide 14 [PDF and Acrobat Map to Significant Portions of FEA Component Architecture]: With use cases, we try to explain the split of PDF and Acrobat—and which parts of our technology relate to the Federal Enterprise Architecture. The buzzwords are still unclear to me. If you look at the right part of the slide, it’s the architecture with five layers. We do X.509 these days. PDF has support for that. Acrobat supports the SSL in the interface part. Most of these are Acrobat-based, supporting XSL-FO, SVG, and others. PDF and XFT (our XML-based forms template) are PDF-based. PDF has Java or ActiveX script built in, so you can take business logic and embed it in a PDF file.
Data interchange: XML is built into all pieces of PDF internally, and interfaces are built out of Acrobat. We were an early participant of RDF. We’ve built an RDF platform called XMP. It’s going into all Adobe products, so they have a consistent interface across all products. For data management, that’s something that Acrobat does.
Mr. Walt Houser: Are you following the XML digital signature standard from W3C?
Mr. Myers: Yes. You’ll see some of it today.
Slide 15 [PDF and Acrobat Today]: I’m going to talk a bit about PDF and Acrobat today, and give the succession of what’s going on with Acrobat and our technology in support.
· [PDF Adoption
· PDF Standardization
· Acrobat Deployment]
Slide 16 [1999: PDF is a Dynamic Content Container]: If we go back to four years ago, the message was, take good data, sock it into PDF, and you might be able to do something with it some day. It wasn’t clear. You had content, images, forms, forms data, transaction data, metadata, and rich media.
Slide 17 [Technology Foundation (PDF Adoption)]: With that, PDF has been used (governments don’t adopt this type of thing as a standard, but suggest that certain things be used). Many government agencies are using PDF as a submission format—FDA, DOJ—50-100 agencies use PDF quite actively. The most fun numbers are, there are over 2 million PDFs on .gov sites today. The IRS is still our poster child for PDF forms. We still see a spike in April as people download forms.
Slide 18 [Technology Foundation (PDF Adoption), (continued)]: This is a publicly available specification that anyone can import or use—there’s no “Adobe” or copyright on the specification. It’s used around the world. These are a few examples within that.
Slide 19 [PDF Progress on Standards Tracks]: Adobe has a history of publicly available specifications. When we started 20 years ago, we took postscript and said “Anyone can create or implement this.” In today’s market, 80% of the printers have postscript implemented with no Adobe code in them. I worked with the SPDL group initially. Our main route is to develop internally, then go to an ISO specification or put it on a track for internal use.
There are three right now that are interesting. One has gone up through ANSI to the ISO level, called PDF/X. It never goes to paper or film anymore. It drives a whole industry. The courts were a major influencer, as well as FDA. PDF/A is coming up with a profile of PDF usage applicable for archives. If you’re going to turn something into an archive, there’s lots of stuff you don’t want in there—for example, Java scripts.
Ms. Melonie Warfel (Adobe): Not too many of you participate in PDF/A right now. Those who aren’t, I encourage you to do so. The URL is on the side.
Mr. Myers: The last on the list is the earliest in the cycle. It’s PDF/is, which is an Internet fax standard. The fax technology has not improved in 15 years. There’s no way to put metadata with it, so there’s a whole effort around that. We looked at a specification of a subset of PDF use to go forward.
The key thing to remember is that PDF is a container architecture that can hold a lot of stuff—XML, graphical stuff, SVG, raster graphics, JPEG, GIF, etc.. Various compression schemes can go into one common architecture.
Mr. Houser: Is it specified for 822 SMIME compliance?
Mr. Myers: All we have is, we’re a registered MIME type. Maybe there are a few things I need to learn. So lets talk afterwards.
Slide 20 [And PDF Doesn’t Come Only From Adobe]: The other thing is, PDF is not just from Adobe. We license it to others; we’re for-profit, as are others. Some in the pharmaceutical space license from us—Metavante is the largest bill and statement maker. If any of you get a Zinio file, they’re licensed from us. The file is a PDF file.
Besides people to whom we license, there’s a whole set of independent implementations, so if you load Red Hat Linux, and view a PDF, you’re viewing it with XPDF. It’s actually written by an electrical engineer who did it for the fun of it. There’s also RenderX, Apache:FOP, Digital Applications; Apple in OS10 has their own implementation, so a bunch of people are doing it—so PDF is not an Adobe thing.
Slide 21 [Acrobat on the Desktop]: But we do build Acrobat. We have 1800 partners, with over 8 million copies sold. The Reader has gone to over a half billion people. It’s basically on every new machine—Palm, Symbiant, Unix. You can count on it being just about everywhere.
Slide 22 [Roles
of XML and PDF Today]: [Skipped.]
Slide 23 [Document Formats in Enterprise Workflows]: Why do XML? Why do PDF? PDF is for presentation, and a long term static form. Other than printing it, there’s no way to work on it in the form. We want to be able to work on it from both worlds. Going back to two years ago, we integrated XML into PDF and Acrobat.
Slide 24 [XML Within PDF and Acrobat 5]: We took forms, and went from PDF syntax to XML-based syntax. It was a logical progression. XML-based—that is, one schema, which goes to an XML space. Through the acquisition of Accelio (which Adobe purchased in April 2002) we got XML forms templates. You have embedded data objects (stealth feature). You can take any data object, stick it in, and get it back later. Mostly…
Mr. Houser: Have you cleaned up the printing?
Mr. Myers: In the latest release just now coming out, there was a good deal of work done.
Mr. Houser: When you read PDF output from JetForms it’s just a little dot in the corner.
Mr. Myers: The next one I talked about earlier—about the metadata. Now Photoshop Illustrator and others have gotten on line. The main three are really data oriented. The bullet about content structure is really publishing (tagged PDF). I’d like to take tagged information and keep it within a PDF file to use for several things. Context-sensitive searches. I want a content-accessible screen reader. A classic case is columned text mushed together. If you have XML in there it knows where it is. I could talk more about it, but I want you to get a taste.
Unidentified Participant: Do you do CSS [Cascading Style Sheets] sand XSL [Extensible Stylesheet Language]?
Mr. Myers: These are Acrobat functionality. We have Web capture that takes Web pages and brings them into PDF. We built our first XSL formatting objects to take XSL formatting content and bring it into PDF.
Mr. Ambur: Regarding metadata, do search engines index it from PDF?
Mr. Myers: Search engines are not available yet. They tend to be laggers. At least we have Altavista and Google searching the content. We have major content managers supporting XML as a way to get metadata about Adobe components. It’s falling more into content management.
Mr. Houser: The ugly secret is that no one updates metadata, so if we get it right the first time, we’re better off. We typically lose that when we do conversions.
Mr. Myers: Our effort was to see how the metadata cold be kept through it.
Mr. Houser: Not taking out what was put in…
Mr. Peter Kacandes (Adobe): You can write a Javascript to populate metadata for documents. Since it’s scriptable, you can pull data out of the documents.
Mr. Ambur: GSA, through FirstGov, is contemplating content management capabilities that might be made available on a Governmentwide basis. I have proposed that GSA and NARA get together to construct a process that will gather information quality data about agency Web records and report such information via FirstGov, so that stakeholders can decide what, if anything to do about it. It is necessary to make reality transparent before people can get to a point to do anything about it.
Mr. Myers: We always had an extensible metadata scheme in PDF. We did some blocking things, and we never made a user interface so anyone could see it. We had two pushers: one was to make an external hierarchical framework, so we chose RDF. We had multiple domains. The other was to get a user interface built into the product so people could build things around us. People said, “Yeah, this is neat, but there’s no infrastructure,” so we built the infrastructure. It may take a version or two to get the kinks out of it. I hear stories from Eric Miller of W3C of putting metadata files into and out of Photoshop and PDF that we never imagined, so yes, it’s working.
Slide 25 [Publishing: Abstract Model]: Where we see publishing, we see it as just an abstract model for what makes publishing. If you look at it, you take a neutrally-coded structured markup, put it through styling or a template, and get a flowable presentation out of it. Content that’s not bound to anything. Not pages and screens from which you then take the information and format; you get the final presentation, internally represented in the program or display list, and externally represented as postscript. It doesn’t matter. Then it’s outwardly rendered to something people can perceive. This is where the specifications live. The styling is XSLT or XLT. You get HTML, CSS, or XSLFO out of that. You either get internal or external views out of that, then it’s rendered to a screen. So it’s a Web publishing scheme, with the first part on a server, and the last part on a Web browser. If you look at the publication flow, you take the first two levels of the circle on the back end, then distribute as PDF. The process is still the same.
Slide 26 [Publishing: Abstract Model-Standards Mapping]: People are now building 3-tiered schemes of this. [The slide depicted a 3-tiered communication hierarchy of style processor, formatter, and renderer.
Mr. Kevin Williams: Is XFT a published format?
Mr. Myers: Yes.
Mr. Williams: XML based?
Mr. Myers: Totally XML based. The early version was Adobe’s contribution to the working group.
Mr. Houser: So one can obtain middleware to apply styling, and then present the document for a client to render?
Mr. Myers: Yes.
Slide 27 [2001: PDF is a Dynamic Content Container]: Now we see the specifications, where FO and PDF and SVG fit. Now we can do more cool things with the picture. We can still take the content and package it in PDF, but we can split it back out again. Now with all the metadata and forms data, we can do lot of cool things with it. So this is where we were a couple years ago. Now I’ll talk about where we’re going. We haven’t announced anything yet; we’re still talking about it. I think the slides here we won’t distribute.
Slide 28 [PDF/XML/Acrobat Directions]: [Title slide.]
Slide 29 [The Following Slides On Futures]:
· [Are nondisclosure information
· Will be discussed at session
· But are not available for dissemination prior to meeting.]
Slide 30 [PDF With XML Directions]: Here’s the next round of Acrobat. The key things are to work with PDF—do things to take legacy transaction systems and integrate them into XML environments; plug into the Web servers of the world, whether they be UBL, or whatever, so people can automate document processes.
This next list is not a comprehensive list. We show three things—allow Arbitrary XML in forms data and templates; integration of PDF into XML workflows, and allow intelligence, data, and process access to forms and integration into Web Services. I have six or seven slides to look at, then Gavin will do a demonstration.
Mr. Niemann: What do you mean by “arbitrary?”
Mr. Myers: You pick, rather than we.
Mr. Niemann: Create your own schema?
Mr. Myers: Yes.
Mr. Ambur: That relates to my question, regarding interaction with the XML Registry. Theoretically, can you go to the registry and import a schema into Acrobat?
Mr. Myers: Yes. You could get WSDL to do it, or pick up an XML schema and bring it in. I’m a long-time SGML head. Long ago, arbitrary schemas were a good thing.
Mr. John Brinkema: We want to take a text document, mark it up with certain information as text, then cause it to become PDF for post-processing and XML-tag the information processed. Are you including that as a functionality of the Arbitrary XML in forms data and templates.?
Mr. Myers: The ability for content went in in 2001. Here, we’re talking about forms data.
Mr. Brinkema: Will namespace references be dynamic? One of the problems is that namespace definition is static. It’s there, but meaningless. If it’s not typed in right, the browser might not render XML correctly, but…
Mr. Myers: We did namespaces right, so if you do them right, you’re not binding to the string, but to the URI from the namespace reference.
Mr. Gavin McKenzie: There’s a transition period from supporting the syntax to doing it right. All that matters is the URI.
Slide 31 [Allow Arbitrary XML Forms Data]: This is how we define allowing Arbitrary XML Forms data.
· [User-defined XML as forms data
· XML-based form definition template (XFT)
· Embedded calculations and business logic in form through JavaScript
· XML data mapping and transformation
· Adobe Forms Designer]
Mr. Niemann: The thrust of the XForms standard is to eliminate the need for scripting. How do you position yourself for that?
Mr. McKenzie: On one hand, the message is there abut limiting scripting; on the other hand, the reality is there and you do it through XPath. Business logic in XML forms is expressed in XPath, and I see that as a script.
Mr. Niemann: Are you a supporter of the standard?
Mr. Myers: We’re a participant. We don’t yet have everything working that way. If you do it right, you can get the information in the form, then out and back, and not drop the information. Then you’re compliant.
Mr. McKenzie: I was there at the beginning of XForms, so we’re still there.
Mr. Houser: Is Larry Masinter involved in the project?
Mr. McKenzie: Larry participates. We occasionally bring him in as an expert.
Mr. Ken Sall: How would you relate that to Microsoft XDocs?
Mr. Myers: I can’t compare us as a vendor to another vendor, especially when I don’t have their product. Ours has the data capabilities. We have the ability to really go and have a solid presentation. The key thing we’re both going for is, “Lets use Web Services, get the data in, and let you decide what works. Adobe says Microsoft’s entry validates the market we’ve been in for a long time.
Mr. Ambur: My assumption would be that a long-time form automation vendor like JetForms would have a better forms designer than Microsoft, but when Jean Paoli was here, I asked him a similar question and his answer was more direct.
Mr. Myers: Microsoft has not released XDocs as yet…
Mr. Niemann: They have. I have it, and I’m challenging them to illustrate its use with the registry.
Mr. Kacandes: All the registry is doing is giving us a way to classify and do searches on schemas, then a way to do a query. Once you pull it out, it’s just a document, and the cap is to pull a document in. The question is, do you want a registry interface to be part of a product? It’s a different discussion.
Mr. Myers: The triad is WSDL, UDDI, and SOAP. We’re doing SOAP and WSDL schemas. UDDI we’re on the fence with.
Mr. Kacandes: UDDI is only a way to register URIs. This discussion is with the registry as a repository of not just XML data, but of any type of data—for example, binary data, text documents etc.. Similarly, you can store PDF documents in the registry/repository and you can use the XML meta-data about the PDF documents to do custom classifications of those documents. You can define your own meta-data system for the PDF documents and you can write scripts to populate and extract that meta-data. There is an open source API for manipulating the meta-data of PDF documents.
Mr. Morgan: Or intelligent registry processes, like a button that says, “Please register this,” and it’s done.
Mr. Houser: So to get this into an answer for my management, when they ask, “So what?” if we use this, we can update a form at a registry, and when a citizen uses the form, he can get the updated one, and things like that…
Mr. Kacandes: Once you get his completed form back, you can take the complete form with the user data and store it back into the registry….or you could use SOAP capabilities. You could put a PDF document as a MIME attachment to the SOAP message, so maybe the business process requires that the data be sent to a Web Service that is implemented as a SOAP endpoint. The service consumes the SOAP message and extracts the attached PDF document.
Then you have some business logic that pulls them out, extracts the XML data from the PDF document, and then does whatever you need. For example, it could be processed by an EJB component or a .NET component.
Mr. Houser: Wouldn’t you need some presentation services with it?
Mr. Myers: If someone needs more depth in this, we can set it up.
Mr. Jim Disbrow: On workflow—this morning I pulled up Internet Explorer and pulled up a leave slip, which went to an OPM SF71 leave slip. I filled it in, gave it to my boss, and it was eventually rekeyed. When can we do this so that I forward it with an electronic signature to my boss, and she does the same?
Mr. Greg Pasaki (Adobe): We can do that today. Your instance just isn’t implemented. It happens at Adobe and in the private sector today. Some agencies, e.g., USDA, have implemented it today.
Mr. Myers: System integrators can do it. We’re integrating into Documentum—an open text. Adobe sells workflow products. Development is the main issue.
Mr. Disbrow: That would represent an excellent opportunity to expose government employees to Web Services, XML, the work flow…
Ms. Warfel: Lets’ set something up for later and get a live connection to Adobe.
Mr. Pasaki: Lots of folks are experiencing phased implementations, but there is a progression of functionalities that various agencies and private enterprise are implementing.
Mr. Myers: Level One is: print, then fill in. Level Two is: fill in on-screen. The key is the implementations and integration, and how people are going through it.
Slide 32 [Adobe
Form Designer 5.0: PDF Form Creation
Method]: [Skipped.]
Slide 33 [PDF Forms in XML]: This is about trying to make PDF work in XML. PDF is a first class package. It does everything in the world. Some people just believe in XML though. They want XML databases. That’s fine too. The issue is, instead of taking XML stuff and stuffing it inside PDF, what if there’s also a representation where the XML is on the outside, and the PDF is on the inside?
Mr. Houser: Like ebXML did with EDI?
Mr. Myers: No. So we have an XDP extensible
data package, which is wearing the stuff on the outside.
Slide 34 [XML Container for PDF Forms (XML Data Package (XDP)]: This slide simply shows what XDP contains and the benefits.
Slide 35 [Interchangeable Data Containers: XML Data Package (XDP), PDF]: Is it a particle, or is it a wave? I have the XML part and the PDF part. Both are usable by Acrobat. It doesn’t matter where or what.
Mr. Robert Proceed: So the specification for XDP is published?
Mr. Myers: It’s not published today. It’s basically an issue of how it should be released. The intent is to publish.
Mr. Niemann: Will Acrobat become a native XML authoring tool like XMetal? Where will the XML come from ?
Mr. Myers: I’ll let you see the demo, and you can make that call for yourself.
Mr. Niemann: The XML comes from where?
Mr. Myers: From wherever it comes from—a database, another process, an XML service…
Ms. Ann Zimmerman: What are the interfaces?
Unidentified participant: Let’s do the demo.
Slide 36 [Acrobat as Intelligent Front End to Data and Processes]: The form is an Acrobat front end for data. Before, you’d just fill it out. Now, we’re adding Web Services to that, so you can do SOAP calls—either document-oriented, or using an RPC model to invoke functions externally—so this is the other part of where we’re going with this.
Demonstration:
Mr. Myers: This is a non-released product (pre-beta).
Mr. McKenzie: The chances are, something will go wrong.
Ms. Zimmerman: Can you define the products as you go along?
Mr. McKenzie: I’ll probably be fuzzy about versions and names, but I’ll try. The first one is the Form Designer—for example, the Thrift Savings Plan Interfund Transfer Request form. I refer to them as templates. We have a lot of boilerplate text and field descriptions. In some cases, say it needs to be formatted in a certain way, expressed declaratively without script, etc.. From this, what I’ll do then, is create an XML descriptions template by saving and create a PDF-fillable form. So I’ll go to Acrobat and show you the form. First I’ll fill in a little of the form.
Behind the scenes is a PDF file with an XML description of the template. So I fill out a few things, such as my name. First I want to export data out of this form into an XML format. [Mr. McKenzie displayed a pull-down menu.] I can select an XML file or a data package. In this case, I export data content from the form into an XML file. So then, I head over to XML Spy and look at the XML. The XML we write is very terse, to minimize white space. I’m going to beautify it.
You can see my name filled out. There are a lot of empty tags corresponding to fields I haven’t filled out, but the names of the tags and the descriptions are dictated by the form I created in the designer. This is rather simple. No namespaces. It uses elements rather than attributes. Despite the simplicity, you could use attributes or namespaces—no matter.
Mr. Houser: It looks cleaner than what comes out of [Microsoft] Word.
Mr. Niemann: Can you change the XML file and have it show up in the form?
Mr. McKenzie: Let’s do that. [Mr. McKenzie input a date of birth and several allocations for different funds. By rule the allocations were required to sum to 100%, however Mr. McKenzie purposely chose a non-compliant sum. He then returned to the form, where a validation message appeared indicating that the percentages were wrong. The date of birth he had entered appeared as expected. Mr. McKenzie then went to the appropriate form fields and corrected the percentages.
Unidentified participant: What if you export out and it doesn’t make the change?
Mr. McKenzie: The difference is between people saving work in progress and the needs of the form—so the answer goes both ways. So now I re-export to the same file, and go back to XMLSpy. Note it’s changed. Now I re-beautify the code. Now I’ll take some information and go in at the personal information in the fund. I’ll add some stuff that doesn’t correspond to fields in the form. The template didn’t predict or anticipate it. Then I go back into Acrobat and re-import. The information looks the same. [Mr. McKenzie then input a pre-configured digital ID.] Then I sign the document, and I’m going to export the form’s data. [Mr. McKenzie returned to XMLSpy.]
So there’s an XML DSIG [digital signature] at the top. I collapse the signatures, and the information from the form is still there. It didn’t disturb it.
Mr. Sall: Where is an XML schema fitting in here? I thought with the validation there was a schema, but with the things that don’t fit, how does that work?
Mr. McKenzie: There are two answers to that. When I designed the template, I didn’t provide a sample data file or understanding of what it should look like. Here I can’t show a schema for it.
Mr. Myers: The schema is in the form-design end of things, but within the form client, it works at the well-formed level, because a parse error the user can’t work with.
Mr. Philip Levy (Adobe): It didn’t require the XML to be validated.
Mr. Sall: We’re talking about extensibility, but if anything goes, what have you extended?
Mr. McKenzie: This is intended to be a feature on this demo. We’re not taking advantage of schema validation.
Mr. Myers: Not transforming and throwing away is the idea.
Mr. Levy: It’s not required for the form. The form doesn’t mess it up.
Mr. Disbrow: Following up on Ken’s point, let’s say the government is using an XML registry standard across government (ebXML). In your second part, those are ebXML standard, but the first part isn’t. Can you click a button to make those standard?
Mr. Myers: For the form things, those can be anything you want.
Mr. McKenzie: Any namespace or schema you choose.
Mr. Disbrow: But you can’t link to the existing registry with the data and have it do it?
Mr. Myers: You’re starting from yours, that you got from the registry.
Mr. McKenzie: Let me jump ahead to show you two things. The document says it’s one page, but I’ve imported two records here. It’s not traditional PDF, but a metaphor for the underlying data. It came from a file, but could have come from the data source.
[Mr. McKenzie continued the demonstration.] This shows that if you need some information in the clear so you can work on it with your XML files, it allows that. So you can exchange between the form and the XML as you like.
Mr. Williams: The base 64-encoded will actually be pulled in?
Ms. Zimmerman: Is the data also stored independently of the XML so you have the PDF and the pure XML?
Mr. McKenzie: Yes, and we maintain the linkages. There’s going to be more content than what shows up in the fields. XML is not know for brevity. The message is one of choice. You choose the workflows. When the value of using XML outweighs …
Ms. Zimmerman: XML compresses very well.
Mr. McKenzie: Yes. If you zip it or use other compression, it compresses well.
Unidentified participant: What is the file size increase between simple XML and the PDF/X?.
Mr. McKenzie: I just don’t know the answer. You have to choose whether the value of direct access to the XML content outweighs the file source and size.
Slide 38 [Deployment of Intelligent Acrobat Forms Without Constituent Cost]: You can talk about technology, but you need to discuss how to deploy it to people. When you look at forms, the person who approves the value is the one who runs the data process. You need to get it to the user without a cost, so you look at a way of deploying Acrobat that has key functionality, that isn’t default-enabled, such as digital signature, etc.. When you turn them on on the server and license the sever product, anyone who has them will be able to use the functions. The IRS was a pioneer customer with that, and they made a CD for small businesses initially. Now you can license on a per-form basis from Adobe.
Unidentified Participant: Why did H&R Block take the form and define their own service?
Mr. Myers: The form is politics rather than technology. Let’s take a non-political one. USDA has wildlife reports that are required of people with farms or open lands. People have to send them in. Right now it’s paper based, and manually input by the government. How can a citizen fill it in and send it electronically, where you suck data out, and go about your business? USDA is doing a pilot with this, where people save forms locally, offline, then mail them in to USDA.
Ms. Zimmerman: The real ROI comes from automation on the back end. What tools do you have now—or are you providing—to facilitate building that business process?
Mr. Myers: Our thing is, other than what we’re building for the workflow server, which is oriented toward small-medium business, the main route is for system implementers to build.
Ms. Zimmerman: Not even API?
Mr. Myers: Massive API sets, and document servers , so we’re building tools for the back end, but the idea is people would integrate.
Ms. Warfel: We have people doing this today, so those aren’t necessarily tools that aren’t available today.
Mr. Myers: When I think of the FDA doing submissions, the example is always new drug applications, but we dealt with people in FDA’s veterinary medicine arena doing their routine filings, and went through a process. It was an incredible hit. People are achieving significant business gains with the technology as it stands today. We’re adding the XML to the end of that.
So you have PDF everywhere, with reliable presentation and archiving. We’ve done work to integrate, with XML and WSD being part of it. It’s part of the transition from paper systems—whether removing them or using forms to download and fill out, etc.—giving capabilities to folks without having to load software on their machines. They don’t have to worry about licensing etc.
Mr. Ambur: If there are no more questions or comments, I thank you all and we’ll conclude with that.
End meeting.
Attendees:
|
Last Name |
First Name |
Organization |
|
Ambur |
Owen |
FWS |
|
Brinkema |
John |
US Courts |
|
Conlee |
John |
Clear Methods |
|
Cox |
Bruce |
US PTO |
|
Disbrow |
Jim |
DOE |
|
Dodd |
John |
CSC |
|
Dodge |
Doyle |
OnLine Forms |
|
Edwards |
Scott |
Corel |
|
Ellis |
Lee |
GSA |
|
Evans |
Pat |
Global Systems |
|
Greenwood |
Michael |
US Courts |
|
Greeves |
Bob |
DOJ/OJP |
|
Houser |
Walt |
VA |
|
Kacandes |
Peter |
Adobe |
|
Kanaan |
Muhan |
DynCorp |
|
Kane |
John |
NARA |
|
Kantor |
Bohdan |
LOC |
|
Karecki |
Philip |
CSC |
|
Le |
Dyung |
NARA |
|
Levenson |
Stephen |
AOUSC |
|
Lewis |
Diane |
DOJ |
|
Mason |
Pamela |
NARA |
|
McKenzie |
Gavin |
Adobe |
|
Morgan |
Roy |
NIST |
|
Morgan |
Bill |
GSA |
|
Myers |
Chuck |
Adobe |
|
Niemann |
Brand |
EPA |
|
Roberts |
Davis |
SAIC |
|
Sall |
Kenneth |
Silosmashers |
|
Turnbull |
Susan |
GSA |
|
Van Feenburg |
Michael |
NASA |
|
Warfel |
Melanie |
Adobe |
|
Weber |
Lisa |
NARA |
|
Williams |
Kevin |
BlueOxide |
|
Windley |
Phil |
|