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. No doubt with the new security policy for entering the building, others will join us later, but let’s get started with introductions.
[All participants introduced themselves.]
Mr. Ambur: I have an announcement and a couple other things I’d like to mention. First of all, the XML Registry is among the projects in which Mark Forman has invited joint submission of Exhibit 300s. [Editor’s note: The 300 Exhibit is the Capital Asset Plan and Budget Justification required under OMB Circular A-11 in support of the annual budget process.]
Mr. Marion Royal: It’s been a very educational experience getting an initiative established. NASA has stepped forward and said they’d like to be a partner on it, so we’re already investigating cross-agency partnerships.
Mr. Michael Jacobs: Have you been speaking to Chan Kim from NASA?
Mr. Royal: Yes.
Mr. Jacobs: He spoke to our office recently, discussing cross-agency collaboration on the Federal Registry.
Mr. Ambur: Was there any discussion of the relationship to the DoD Registry?
Mr. Royal: No, we haven’t gotten to that yet.
Mr. Ambur: The “Performance Institute” has scheduled a training session on XML and XBRL in Government for October 15-17. With respect to XBRL, I couldn’t help noticing Marion’s message on the listserv explaining why XBRL was not used in for the A-11 Exhibit 300 but expressing interest in having the Working Group learn more about it. So this training session is a chance for some folks to get up-to-speed on it. [Editor’s note: The training session has been canceled.]
Mr. Royal: I discovered that XBRL is not a simple vocabulary to adopt. It’s an infrastructure, equivalent to ebXML and other infrastructures. It relies heavily on XLink and other technologies. I don’t think we had the proper tools to manage that vocabulary for the I-TIPS [Information Technology Investment Portfolio System]. The I-TIPS schema we developed was used this week for I-TIPS submissions. Over 400 were successfully submitted yesterday using our XML schema.
Mr. Ambur: Marion and I are talking to Judy Spencer about scheduling a forum on SAML. Eve Maler has agreed to conduct the presentation, and it’s up to Judy to determine the date and venue. I’m encouraging consideration of whether to include XNS (Extensible Name Service) in the forum. The XNS.org folks are considering moving under the auspices of OASIS.
The Federal Trade Commission (FTC) asked to see whether we could investigate relationships between the FTC and the XML Work Group (WG), so that’s something that will be interesting to look in to.
Bob Haycock scheduled the XML Conference and Exposition 2002 (XML 2002) for early in December 8-13. We have an invitation for the WG to co-host XML 2002. Have you had any discussions on that, Marion?
Mr. Royal: We have not. I wonder what it means. If it means an outlay of money, where do we get it? If it means using our name, do we clear it with our COO?
Mr. Ambur: I think it means the latter— cross-posting our logos, distributing information on the listserve, etc. I don’t believe it would entail any financial commitment, but there is the issue of propriety and whether our leadership wants to be involved. Personally, I think it makes sense.
Mr. Royal: Does anyone have a problem with attaching our name to XML 2002?
[No opposition.]
Mr. Ambur: Do you need to run it through your leadership?
Mr. Royal: Yes.
Mr. Jacobs: Michael Grimley of the Department of the Navy (DON) has started official designations, with coordination from the DON CIO office, to Technical Committees (TCs) of OASIS. They sign off, saying “You’re now our official representative to this TC, and here are the rules of the road for representing the Navy, how you report back, etc.”
Mr. Ambur: That’s great. You might recall that one of GAO’s criticisms of federal XML efforts was lack of federal involvement in XML standards, so you’re establishing best practices. I think NIST has a role to play with respect to facilitating and helping fill gaps.
Mr. Crawford: The nice thing about what the Navy did is, they formalized participation in OASIS. They plan to follow the same model for managing Navy membership in W3C [World Wide Web Consortium]. It allows the Navy to speak in a single voice, and it’s also a mechanism for coordinating their position on standards as they’re developed. They have a listserve established for work on OASIS announcements, and they’re in the process of identifying technical leads for each area in the Navy that’s affected by standards work in OASIS and W3C. It will allow them to do what GAO articulated—that is, identify their position as a single Department as specifications develop. The standards bodies are excited, because it will allow them to incorporate Navy requirements up front.
Unidentified participant: Is this the first attempt at unifying Departmental participation on XML standards bodies?
Mr. Ambur: I don’t know whether it’s the first attempt, but it’s the first visible effort.
Mr. Crawford: There are other federal agencies doing it.
Unidentified participant: To this extent?
Mr. Crawford: Not to the extent of the Navy.
Mr. Royal: This is not official, like ISO. We haven’t much official movement in the voluntary organizations.
Mr. Ambur: At Standards.gov there’s a listing of government folks involved in standards activities. I don’t know the quality of the data but I do know that it is incomplete, since only a few agencies are contributing to it. After the break, when we talk about scoping potential projects for approval by the AIC, I’d welcome a proposal to extend on a Governmentwide basis what the Navy is doing.
Ms. Susan Turnbull: I brought copies of a magazine from the AIC guide, called “Extending Digital Dividends.” It was put together by several committees. The main findings addressed VoiceXML applications. The Government Printing Office is interested, because it’s the first audio e-book. In the guide, we suggest that Agencies turn to the CIO Council and XML WG for increased XML participation.
Mr. Crawford: There will be an announcement on the OASIS site that they recently approved changes in the way OASIS standards become standards. They’re streamlining the process to shorten it. I think it’s a good, positive move. For those of you involved, you might want to take a look.
Mr. Royal: I’d like to talk briefly about the meeting last week at UN/CEFACT. Theresa [Yee], Mark [Crawford], and I were there. My take is that this was the first meeting of a quorum. It combined a lot of work groups into one big organization. It was a reorganization for CEFACT. I think it was successful. A couple things came out of it:
I took from the meeting, from the Architecture and Methodology group, a very positive feeling in that those people are working on designing how to model e-business processes. We’re getting close to the point where you push a button and the technology does it.
As the week went on a proposal was developed, which went to the UBL group. It proposed that UBL become a part of UN/CEFACT standards work. There are positives and negatives: the positives are to consolidate and cooperate, and reduce the number of meetings that standards people have to go to; the negatives are that the process of developing a standard in UN/CEFACT is very different from OASIS, so the UBL committee may take the position that “We’re making good, rapid progress, and to go to UN/CEFACT may slow us down.” Mark just said that OASIS is streamlining the specification process, so that ‘s a problem that’s going to be vetted in two weeks.
Mr. Crawford: In two weeks there’s the UBL “face-to-face” in Burlington [Massachusetts]. From an XML perspective, I want to add that the way CEFACT has structured themselves, the forum is essentially a virtual group with a coordinating team of chairs for five individual groups. Five groups have their own charters, mandates, and areas of responsibility.
Within the Applied Technology (AT) group, there is an EDI/ Syntax group and an XML group. The XML group has a set of design rules—ebXML, UBL methodology and others—in its responsibilities. The AT group will have a meeting here in Washington in January at LMI, working with XML design rules issues. The week before that, ISO TC154 will be meeting at LMI.
Mr. Royal: The group I participate in, called the “Techniques and Methodology Group,” is basically tasked to work on the specifications themselves that the Applied Technology Group uses to implement, so you take our work and apply it to syntaxes of other groups.
My group out of ebXML (now CEFACT) will meet here in [Washington] D.C. in the second week in December. It’s open to anyone.
Mr. Crawford: It includes Core Components [team].
Mr. Royal: And the Business Process team, and the third major team is the Architecture team. The announcements will be in the minutes.
Mr. Ambur: Mark, is the meeting in January conflicting with the metadata meeting?
Mr. Crawford:
Yes.
Mr. Ambur: We might want to list it in the “What’s New” section on the XML.gov website so that folks are aware of it and the scheduling conflict. Continuing with the agenda, John Kane is here to give us a quick update on the eRecords project.
National Archives and Records Administration
Update on eRecords Management Schema Project
Mr. Kane: The Electronic Records Management Initiative is one of 24 initiatives under the federal e-Government initiative. Records management is an important part of the infrastructure that will make E-Gov work, and we’re working closely with our partners to ensure success.
NARA is the managing partner agency for the ERM Initiative. The projects that are part of the ERM Initiative are organized into four issue areas, each with a lead partner agency: The first one is called “Correspondence Tracking,” in which we address problems with interagency coordination of Congressional inquiries and provide solutions that can be used by Cabinet-level agencies, is being led by the Department of Energy.
The second, “Enterprise-wide ERM,” has as its purpose to provide tools for agencies to use as they plan and implement ERM systems, and is being led by the Environmental Protection Agency.
The third one is “Electronic Information Management Standards,” in which we look at promoting Government-wide use of the DoD 5015.2 Standard for Records Management as the mandatory standard for federal agencies, and it is led by the Department of Defense.
The fourth one is named “Transfer of Permanent E-records to NARA. We’re the lead agency for this one. This area will address an expansion of both the number of formats NARA can accept, and the media and techniques that can be used by Federal agencies when transferring their permanently valuable electronic records to NARA.
We’re currently in the process of looking at three formats (four tasks):
We’re looking at an XML schema to wrap those groups of records when we transfer them to NARA. What records are considered worthy of permanent preservation is predetermined by an Agency. As such, we’re looking at schema to bind the permanent records and metadata to describe the accession.
The schema to facilitate accession is to be done by June, 2003. It’s in the identification phase. There are two tracks: one is made up of records officers who put together the accession package, and the other is made up of NARA folks who process the package.
Internally we’re working with two other NARA programs as well—the Data Administration group (Lisa Weber), and the Lifecycle Coordination staff. The form currently used for accession is the Standard Form (SF) 258. It’s been used for many years, and we expect that the digitized version will not vary a great deal from the paper version.
As the business model for NARA changes with the implementation of the Electronic Records Archives initiative, we expect that the accession process will change. The schema that we create at this point will evolve to reflect those changes.
For now, we’re looking at two aspects of the accession schema. One would be the metadata to describe the accession itself, and the other is the wrapper to package the accession. The first will essentially be a digitization of the SF258 form and for the second we’re looking at METS. METS or the Metadata Encoding and Transmission Standard is supported by the Library of Congress and the Digital Library Federation. It reflects the library business model, and it’s a wrapper for objects going into a digital library environment. The advantage is that it is currently being implemented widely by the library community and by those who follow the Open Archive Information System (OAIS) standard. Because it looks at digital objects as well as behavior, it has lots of potential and flexibility.
We’re enjoying collaboration with LMI through Marion’s office. They’ve begun working with us. LMI will be over at the Archives tomorrow for the first big group meeting. I’m looking forward to something useable for the time being, although it will not be written in stone and will evolve along with NARA’s business model.
How many of you are familiar with METS?
[Few were familiar with METS.]
I’m curious about how widely accepted it is in the archival committee. There are other options—SOAP wrapper, RDF [Resource Description Framework], even a CData scenario. METS has the advantage of a broad-based community developing the standard, that’s looking at digital objects. I don’t know whether it’s the first time it’s being addressed in the archival community, but we’ll find out.
Mr. Royal: I’m not familiar with METS, but we’re looking for an industry standard that we can use without having to create. If it meets our purpose, I’d say, ”Yes, go for it.” Regarding SOAP, I think you’ll end up with a METS wrapper in SOAP, or SOAP in METS, but when it [the message] gets to you, you’ll open it up further to decide what to do. So if METS allows NARA to move records around, great.
Also, this has nothing to do with the format of the record. It’s just the tool that allows them to move it around between Agencies and NARA, and perhaps to move it to its accession process at NARA.
Mr. Ambur: You mentioned the SF 258, but what about the other forms used by NARA?
Mr. Kane: The 258 has the most use.
Mr. Ambur: Meanwhile, NARA is looking at reengineering the process of how records are appraised, which is the focus of Standard Form 115. Regardless of how the process is reengineered, agencies will have to continue to appraise the value of their records. So even though NARA’s current eGov eRecords schema project has been scoped more narrowly, to include only the metadata required to facilitate the transmission of permanent records to NARA, agencies will still need to address the broader scope.
Mr. Marc Le Maitre: Talking about transfer of digital objects from an Agency to NARA, how is the object referred to?
Mr. Kane: The unique identifier, it doesn’t address that. Do you mean a handle?
Mr. Le Maitre: Yes.
Mr. Kane: You could use one; it doesn’t matter. The accession may represent more than one series. Once it’s processed, it may not be required to stay as an object—so it does not need a unique persistent identifier at accession.
Mr. Le Maitre: Transfer protocol.
Mr. Kane: Yes.
Mr. Royal: This is maybe not a single package.
Mr. Le Maitre: You’re not concerned with the records that are in there?
Mr. Kane: Yes.
Mr. Ambur: I’ll be interested to hear more about the other three aspects of the e-Records Project.
Ms. Turnbull: What is METS?
Mr. Kane: Metadata
Encoding and Transmission Standard. [Mr. Kane explained the origin of METS.]
Look for METS under “Library of Congress.” You’ll be directed to its website.
(http://www.loc.gov/standards/mets/)
End presentation.
Mr. Ambur: OK, Ken Sall is going to give us a demo of the wonders of XSLT. Ken?
Mr. Ken Sall
Ken
Sall Consulting
XSLT and XHTML Client-Side Sorting Demo
Mr. Royal: While Ken is getting started, I want to announce that Jim Hunt, who’s been a valuable asset in organizing these meetings, probably won’t be providing support in the future. I want to thank him. Jim, I hope you’re still interested enough in XML to join us at these meetings. Thanks, Jim.
Mr. Sall: I’m Ken Sall, of Ken Sall Consulting and a SiloSmashers solution partner. I’m going to point to an updated version of these slides.
Slide 2 [Goals]: The goal of this is to show you that you can use extensible HTML (XHTML) and Extensible Stylesheet Transformations (XSLT) to show what’s going on in the government and XML itself. It’s also to show that you can use XML on your website today. We’ll look at it in terms of server-side versus client-side; how you can take advantage of particular browsers; and alternatives for cross-browser solutions. The main demo URL is http://kensall.com/gov. Owen has a link to it from the agenda at XML.gov. If there’s anyone on the conference line, I encourage you to go to that link now. At the top of the page is a link to an updated version of the slides.
Slide 3 [Non-Goals]: It’s definitely not a goal at this meeting to dazzle you. That’s not the point. We’re not looking at the broad set of XSLT capabilities; we’re focused on the narrow aspect of sorting and looking at it from different angles. Code Review isn’t what we’re here for either, but rather ways you can add value to XML.gov, should you decide to do so.
The first thing I did was, I took the Home Page of XML.gov (HTML) and converted it to valid XHTML (which as you may know, is HTML in XML syntax). I used an open source tool called “HTML Tidy.” It does most of the work, but its result is not necessarily valid.
Mr. Royal: Tidy is a tool?
Mr. Sall: Yes, by David Raggett of W3C (http://tidy.sourceforge.net/).
Slide 4 [HTML to Valid XHTML 1.0]: W3C has an online HTML Validation Service (http://validator.w3.org/). This service allows you to validate local or Web-reachable files against either HTML or XHTML. That identifies certain errors. You fix them, and that’s the time-intensive part, but it’s not bad because Tidy does a lot. Also, problems are sort of repetitive in nature, so you see a lot of the same problems from page to page. The other thing you can do is brand with the “Valid XHTML” logo. You simply click the logo after editing to revalidate.
[Mr. Sall clicked on the “Valid XHTML Demo” link on Slide 3, which returned the “XML.Gov Demos: XHTML” web page. He then clicked on the “Valid XHTML 1.0” link to validate.]
Here is a converted XML.gov page, now in XHTML format. It’s similar to the one we’re used to. It has the special icons on it, so if we click on the “W3C XHTML 1.0” icon, it sends the page to the W3C HTML Validation Service, validates it, and in this case results in the message, “No errors found.” [Mr. Sall followed the process he had just described.]
As a footnote, the Cascading Style Sheet (CSS) can be validated the same way—but that’s not the point of the demo.
Slides 5 & 6 [Sorting Government XML Efforts]: The next thing we did was, we looked at the registered XML efforts related to the WG. We started with the “Efforts” page at http://xml.gov/efforts.htm.
Slide 7 [Sorting Government XML Efforts, continued]: We cared about two columns of data in a table, but not about anything else on the page. We wanted to sort the efforts by organization and also by effort name. The original HTML contained nested tables. We wanted to sort only the innermost one. There was an added benefit to using XSLT. I discovered, when writing the XSLT, that there were inconsistencies in path names; some were relative paths, and some were absolute paths with a full URL. I could identify the inconsistencies and fix them with XSLT. By the way, Netscape 7 [shown here] just became non-beta.
The main point here is that you’re effectively mining XHTML to extract tabular data using XSLT, and you can preserve links.
Slide 8 [XML WG Participants Sort]: It might be interesting to define what it means to be a WG participant. What does the WG want to track? What would you want to extract? We made a strawman DTD (Document Type Definition) of what we might want to collect. We made it such that “Name” and “Organization” are required, and all other information was optional—so “Title,” “URL,” “ExtraURL,” “SpecialInterests,” and “ContactInfo” are all optional. I want to stress that this strawman is merely my opinion of a starting point, selected strictly for demo purposes.
Mr. Ambur: I’d like to interject that one of the things I’m interested in is the possibility of working with the Knowledge Management Working Group to implement an XML schema for people to document their skills and expertise. John Andre expressed interest in working with us toward that end, but we haven’t gone beyond that point yet.
Mr. Sall: This could be a starting point; the DTD is easily modified.
Slide 9 [DTD for XML WG Participants]: Here’s two thirds of the DTD. It took about an hour for me to write this. It was easy to create; it could be extended, or it could be replaced with an XML schema, if desired.
Slide 10 [XML WG Participants Sort]: The Result is another slide that shows the ability to extract information. It’s kind of like a text-based database; there’s lots of information here, but you only want to view certain subsets at any given time. On the left, we pulled out four columns; on the right, five. They’re sorted differently. The left sort is by organization; the right sort is by last name of the participant.
Slide 11 [Server-side vs. Client-side]: We also looked at server-side/client-side transformations. Server-side transformations mean that the transformations occur on the server—in real time if it’s dynamic data, or at a predetermined time if the data doesn’t often change, or one time only if the data is static. The information is sent to the browser as HTML or XHTML. The advantage is, you can guarantee it works in all browsers. The disadvantage is, if it’s dynamic data, or if it’s a popular website, you’ll have a heavy load on the server as a result.
In contrast, in a client-side transformation, the browser is sent raw XML (rather than HTML or XHTML) and an associated XSLT stylesheet. The browser (client) has to apply the stylesheet real time to the XML instance data, and then render it as HTML. This lightens the load on the server. The potential benefit is, you have the data already downloaded to the client. If you can do in-place manipulation of data, you can get lots of views without going back to the server.
Mr. Royal: This is very exciting. The style sheet has to come from the server?
Mr. Sall: Yes—or reachable in some way from a URL, meaning it could come from a different server.
Mr. Royal: This allows a custom view of standard data if you’ve established what the data looks like?
Mr. Sall: Exactly. For this, we created a little DTD, just for a model.
Mr. Brand Niemann: This is part of separating data from presentation. You can create presentation of data independent from Web Services, etc.
Mr. Lex Poot: This is not new. It’s been there for years.
Mr. Royal: We know how, but not who is doing it.
Mr. Donald Zacherl: This is what we’ve been trying to get people to do for years.
Mr. Poot: It’s like a cover, or table of contents, all from XML. You use different style sheets to render the document.
Mr. Royal: Yes, but my point is, how do I customize my own XSLT to view standard documents the way I want to, whether by VoiceXML, or larger font, or whatever?
Mr. Niemann: You can buy XML Spy 5.0, which does this. There are additional tools that can migrate sites to XML, do customization of stylesheets, or find errors in style sheets…
Mr. Royal: Still—XML Spy is not for the end user. It’s for the developer.
Mr. Niemann: Actually it is. It’s part of the integrated environment. It’s like Corel; they’re trying to offer it to a whole spectrum of users.
Mr. Royal: We have a way to go.
Mr. Niemann: You’re right, but the strategy for XML Spy, Corel, and XMetal is to deliver “XML offering” and content management to an enterprise at various levels, so the user might not even know he’s delivering XML. Or you could take a standard tool and purposely create it, or have a very sophisticated team working on it, so there are three levels.
Mr. Kevin Williams: It’s important to distinguish between instances and design. I see what Marion is saying. You have something sitting on your client, and you style it the way you want to.
Mr. Royal: Where we are now is that the producer of the XML document provides alternative choices.
Mr. Sall: The next step is providing the ability to create an alternative at the user level and associate it with an XML instance. That’s tricky, because the URL of the XSLT stylesheet is referenced from the XML data by means of a processing instruction. If you can transform independent of the browser (i.e., the server-side solution I mentioned earlier), it’s not an issue. The good news is that the possibility is there; the bad news is that you have to have an XSLT processor built into your browser. That means that, at a bare minimum, you need the MSXML 4 parser with Internet Explorer (IE) 5.05 or above, or you need Netscape 6 or above.
Slide 12 [Client-Side Sorting: xml.house.gov]: Client-side sorting: there’s a very interesting case of this on the xml.house.gov website that I believe many of you have already seen. They are delivering raw XML to your browser. It uses a CSS, as well as XSLT, and renders on the fly as HTML, on the client side. The initial view is a sort of Congressmen in alphabetical order. Each column heading is active, so I click on the State column heading, for example, and the browser sorts a lot of data. Quite a lot of data.
Mr. Jacobs: By the stylesheet?
Mr. Sall: XSLT stylesheet driven by JavaScript. Actually, Microsoft’s JScript in this particular example.
Mr. Niemann: Another thing to mention is that the information didn’t originally exist in a table. When it’s marked up with XML, you can apply a query and a stylesheet transformation that does this.
Mr. Sall: This is very impressive, especially in IE 6.0, but in Netscape 7 or Mozilla, it looks the same but doesn’t work the same. This is not to criticize the developers. Their XSLT stylesheet makes clear what they’re doing, and they’re aware that they are departing from standards in order to gain functionality, and that they’re going to leverage true standards at some point in time. However, it made me think that I want a real cross-browser solution. It got me thinking about other possibilities, especially about accessibility. The most basic thing would simply be a text message to the user on the page saying, “You need IE”, etc.
Slide 13 [Client Side: Browser Detection]: But a better approach would be to dynamically detect what browser the user has, and to display a JavaScript alert if they are going to be sacrificing functionality due to their choice of browser. For me it’s a question of accessibility, but not in the common Section 508 sense. I prefer Netscape. If I can’t access something because of that, there’s a question of accessibility. If it’s done this way, you can still use IE and leverage its proprietary features, but we Netscape users at least get something useful.
Slide 14 [Client-Side Sorting: IE Only]: I next explored IE. I wanted to get familiar with its proprietary features related to XML, XSLT, and the DOM (Document Object Model). There was an “XML Journal” article in early 2002 that talked about an IE-specific implementation. I used the techniques discussed in the article to manipulate the same participant information in an IE-specific manner. [To illustrate the difference, Mr. Sall first displayed the page in the inappropriate browser and then in the correct browser.] This is an XML.gov participant sort table. The data is sorted by last name, and I can readily switch to any other sort I want to. It’s still client-side, in-place sorting. As long as I’m using IE 5.5 or higher with the MSXML 4 parser.
Mr. Royal: Does the browser cache that data?
Mr. Sall: I think it does. Yes, it uses the DOM, and you’re really manipulating that.
Mr. Zacherl: It doesn’t cache the sort, but you wouldn’t want it to. If you want to save it specifically, do it the way you would any other browser.
Slide 15 [Client-Side Sorting: JavaScript]: I thought, “I want this functionality,” but I want something that’s cross-browser, so I posted a message on the XMLdev listserve. I got one response. A gentleman from Greece named Manos Batsis. I’ve been working with him since September 2. He’s been exploring the same problem from the JavaScript perspective. He’s donated his code with a General Public License (GPL). [Mr. Sall clicked on a link to the page displayed on the slide.] There it is. It’s a table of sorted dates, names, etc. This is it in IE. The question is, what about in Netscape? It looks the same, with one subtle difference—the cursor displayed in the column headings. But more important, the sort works fine, just like in the IE case!
Mr. Jacobs: Is this HTML?
Mr. Sall: There is XML data that’s being fed. It’s in a shell that lets us do the table in HTML.
Slide 16 & 17 [Client-Side: Cross-Browser API]: Manos Batsis is developing a browser that implements the DOM to give us the sorting across browsers. It doesn’t work in all of them, since they don’t all have the same DOM implementation, especially older browsers. It allows us, with extensions, to transform nodes. It makes the sorting work smoothly. It’s implemented in JavaScript, but it’s based on the standard W3C DOM.
Mr. Poot: Does it need MSXML extensions to use the data?
Mr. Sall: Probably only IE, because it doesn’t apply in Netscape. Whether you really do, I think not, because it’s cross-browser.
Unidentified participant: Does anything need to be installed on the client?
Mr. Sall: No. The necessary JavaScript is delivered with the data.
Mr. Williams: How big are the documents?
Mr. Sall: I don’t know off hand.
Mr. Williams: In JavaScript, they might be huge.
Mr. Sall: On the order of 1,000 lines? [Update since meeting: less than 400 lines for the example shown, including the data and the JavaScript.]
Mr. Williams: Kevin how far along is he?
Mr. Sall: Well, we received an update yesterday. He has gotten further. He’s very interested in working with us and communicating results to the WG, so stay tuned, especially since he’s willing to do a GPL.
In order to implement this, he had to know considerable details about both browsers. From the developer’s perspective, however, it’s just a simple Application Programming Interface (API) that Manos is creating.
[Mr. Sall demonstrated that it did not work in Netscape 4.7x since that is an older browser.]
But things are active. We can sort in IE. It looks identical, and it works identically in Netscape. I think this is a major accomplishment. I should mention that Manos is on the OASIS TC for Human Mark-up Language.
Unidentified participant: At what point does going back not work?
Mr. Sall: 4.x. It works with 6.0 and above. [Note: There was never a Netscape 5.x.]
Slide 18 [Multi-Page Sorting of Brainstorming Ideas]: A week ago, Owen said it would be nice if some of this work could be applied to the WG’s [GAO Report] brainstorming ideas. We discussed this a bit, and, two days later, we had something working.
[Mr. Sall displayed the HTML brainstorming table as provided by Owen Ambur—Slide 19.]
This is Owen’s original HTML table that we’ve seen. First I converted the page to XHTML. Since it was all row/column oriented, I thought it was confusing for me to manipulate it that way (by referring to row/column numbers, for example). It makes more sense for me to wrap the data in custom XML so I could think about it in terms of element names. I developed a simple DTD and XSLT for a one-time conversion from XHTML to my custom XML. Then I wrote additional XSLT stylesheets—one for each column sort. If I were doing this for something other than a demonstration, I’d develop modular code, since each stylesheet had more in common than not. This was just quick and dirty solution, however.
I also wanted to consider both server- and client-side versions of this. If you look at the server-side examples, they should work in every browser.
Slide 20 & 21 [Brainstorming: XML Data Rep. and In XML Spy.]: Here’s the XML data I was sorting, shown in IE and then in XML Spy.
Slide 22 [Sorting of Brainstorming Ideas]: So here’s the table sorted after the conversion steps. I used a red outline in the table headings to highlight the current sort criteria. If I click on “med,” or any other column heading, you get a different view. This is server-side, in which I’ve already done the transformation from XML to HTML. I also have the client-side solution implemented. In this case, I’m not using in-place sorting, however. When I click on a client-side case, my browser is sent the XML and XSLT, the browser performs the sorting transformation by whatever column criteria I select, and produces HTML fairly quickly. Here I can switch to other column headers. It does everything live. It works in the most recent versions of both browsers, so on the brainstorming structure slide we have an idea; it has a description and rankings; other data, and an idea of a focus group and volunteers. I can update the data, and all should be fine.
Slide 23 [Sorting of Brainstorming Ideas]: The client-side solution took about three times as long as the server-side solution. I wouldn’t draw too much from that because it’s my server, and there’s not too much traffic. But remember, the client-side solution involves more work at the browser end, off-loading the work from the server.
Slide 24 [Recommendations for XML.gov]: In conclusion, I’d like to suggest what I believe are the best ways to add value to XML.gov at this time.
Mr. Ambur: Your recommendations are applicable to websites generally, aren’t they?
Mr. Sall: Definitely.
Mr. Ambur: According to press reports, at IRMCO Brooke Dixon talked about OMB issuing guidelines for agency websites. It strikes me that it might be an opportunity to advance the kind of best practices you’re suggesting. In addition, GSA is considering content management capabilities that might be offered Governmentwide through FirstGov. That might be an opportunity as well.
Mr. Sall: Yes, it’s sort of funny, because it’s XHTML. It hasn’t caught on, with the exception of the small device community. It’s not as fast as we thought, but it is a good leg up. Once converted, it’s easy to make it valid. Once you have valid (or at least well-formed) XHTML, you can use XSLT and other XML technology to manipulate the content.
Mr. Royal: I like good ideas, but not ideas without an action plan, so how would you make this recommendation extend to other communities—like FirstGov?
Mr. Ambur: First make them aware. I attended one of their focus group meetings and try to stay in touch with Tony Byrne, who has provided consulting services to them. That’s one avenue. If someone would like to propose a task team along these lines, we can get into more detail after the break.
Mr. Poot: Have you looked at IIS [Internet Information Services] for the server? That’s what we’re using.
Mr. Sall: Yes. It can do all the transformations. It doesn’t have to be IIS. It could be JSP [Java Server Pages]—any technology that can do the transformations.
Unidentified participant: Is there a down side to number 1?
Mr. Sall: Not as long as you make it valid XHTML. I had that problem on a website I do for a ski club. I tried to explain it to some people. They sometimes don’t check for it, so it can be a problem. Yes, it’s an issue that someone has to monitor—do that little validity check.
Mr. Niemann: The down side—Elizabeth Castro has written “Quickstart XML.” It doesn’t recognize XTML. I think they should go all the way to XML, and not use the interim bridge of XHTML that OASIS provided. Also, the FEDWEB conference this month offers a full day on the theme of changing government websites from HTML to XML. The program includes three-four talks. If you read her comments, it implies to me the question, “How would you do these things without using XML?”
Mr. Sall: The last thing is to pay attention to browser differences, and pay attention to the server side initially, because it’s more predictable. Then move toward client-side, where it makes sense. That’s it. Thanks to the WG and Owen for the opportunity to speak.
Mr. Williams: Do you know what the backward compatibility is for XHTML?
Mr. Sall: Yes. It’s one of the reasons I recommend using HTML Tidy and the validation service. XHTML 1.0 works with all major browsers [version] 4 up, and I believe [version] 3, provided you follow XML/XHTML syntax rules. Tidy takes care of all that—quoting It has to do with the elements for the image, etc. Tidy takes care of all that—quoting attributes, properly identifying empty elements (<img />, <br />, <hr /> ,etc.) using XML syntax, all that.
End presentation.
Mr. Ambur: Thanks, Ken. The next item is a discussion of the WG charter, which lapses at the end of this month. Coincident with the lapse of our charter, the AIC committee, to which we report, is reorganizing, so we’re shooting at a moving target. Generally speaking, the thought expressed by John Gilligan and Debra Stouffer, who chair the AIC, is that the committees subgroups should be more project/task oriented. In our first charter, we were explicitly directed not to take on operational roles, so this is a bit of a paradigm shift.
The AIC hasn’t worked out the process it’ll use to determine priorities, so there’s no template in that regard. After the break, I’d like to engage in discussion of what the high-priority tasks from this WG should be, and encourage folks to volunteer to work on them in such a way that we can advance them to the AIC for approval.
[Mr. Ambur passed out copies of the proposed charter.]
As far as changes to the charter, they’re pretty minimal. I changed a little of the background. The name of the AIC has changed since our original charter. In the Purpose section, there is a minor wording change in the first bullet with respect to developing our own best practices. The major change is in the fifth bullet—to propose to identify products or projects that will help multiple federal agencies, consistent with the general thrust of the new direction we have received from the AIC co-chairs.
Our exiting charter outlined a voting process, but we haven’t taken any votes and I don’t foresee the need for any. The working group operates by consensus, and if there is any point on which consensus is not possible, it is Marion’s and my responsibility to determine whether to elevate it to the AIC. So I have eliminated the voting process in the draft extension of our charter.
Mr. Royal: Is there anything else from the AIC?
Mr. Ambur: Only that none of us knows what the outcome of the AIC reorganization is going to be. Under the original proposal, they proposed a structure comprised of three subcommittees:
All the existing WGs, except for the Architecture WG, would fall under the Emerging Technologies Technical Committee. I question the value that structure adds. For example, the issue of Emerging Technology is much broader than I personally could address, and another way to view the issue is from the perspective of the value to be added by the proposed Emerging Technologies subcommittee. Nevertheless, since our charter lapses at the end of the month, this group should come to consensus on how we think we can best add value in the future.
Mr. Royal: We already have a recommendation for extension that came out of the strategy meeting. I concur with this charter you’ve proposed.
Mr. Ambur: Has any one else had a chance to scan it, or does anyone have any points of disagreement? Let’s take a break. When we come back, we can discuss any points you think of over the break.
Break
Amending
and Extending the XML WG Charter
Scoping
Potentially Near-Term Deliverables
(XSLT Demo)
Mr. Ambur: Let’s get started again. Ken is displaying some of the brainstorming ideas in his demo sort.
This part of the meeting is unstructured. We just passed out a listing of brainstorming ideas we had following the briefing on the GAO report. The importance of each of those ideas was rated in Raosoft’s survey, and this tabulation is displaying those in descending order based on total weighted total score. Since not all respondents rated all ideas, there are some discrepancies in the rankings based upon total score versus average score. Those differences can be revealed more clearly using the XML sorting capability that Ken has devised. In addition, so other interesting results can be observed. For example, as sort on the “Lo” importance column reveals that the idea ranked 48th in the total rating score was considered by very few people not to be at least somewhat important. [Editor’s note: It’s total weighted score was low because most people rated it of medium, rather than high importance, and relatively fewer people (18) rated it.] The sort capability helps reveal things that are not otherwise apparent.
My objective for this meeting is to further focus on those ideas and others that haven’t been identified yet, and to move the ball forward to propose actual tasks for approval by the AIC. We’re looking for people to volunteer to scope out the tasks, if not also to carry them out. We don’t yet have the template by which the AIC will determine priorities, so we’re shooting in the dark a little, but I hope we can flesh these out a little into potential projects and tasks.
Ken added a column called “Project” because some of the ideas might be parts of the same project, and others might need to be broken down to two or more projects. With that, I’m open to comments, suggestions, and volunteers.
Mr. Poot: Section 508 isn’t on the list. It’s probably something to look into. It’s more complicated than most people realize.
Mr. Ambur: David Heiser of the IRS can speak to that. David?
Mr. David Heiser: Yes—a little background. I’m in charge of electronic authoring; not XML per se. Our office puts out a lot of tax products. We also put out a 7,000-page Internal Revenue Manual. Our issue is that we still have to do it in paper; however, everything we do is posted to the Web, and converted to HTML and PDF. The problem is that HTML and PDF are somewhat inaccessible (though we’re now implementing about thirty talking tax forms). But we still have the issue that all public IRS documents are converted to HTML, and they’re graphic-intensive—anywhere from tax forms to icons. I can’t imagine we’re alone. Today we’ve converted as many graphics as possible. It’s a problem because our vendor has a format on the Web that doesn’t work well. Even if we use “alt tag” or “long desc,” we’re still limited as to how much we can describe.
Mr. Poot: Yesterday, we actually described a figure inside the text. As long as it’s described, it doesn’t matter whether it’s in long desc or alt text.
Mr. Ambur: I understand what you’re saying, but there’s still a requirement to describe figures that may not already be described.
Mr. Heiser: We’re one of the largest Braille users in the country. We’re working extensively with that. Adobe introduced a new feature in Adobe version 5.0 that helps, but we still have a black hole where you’re not giving the browser the same capability that the sighted user has.
We also contracted with an entity here in town to use “Arbortext” on the authoring side. The goal is to develop code that can be used across DTDs to describe visual products. Hopefully we can work with this. I’m interested in defining a proper tool. We looked at SMIL [Synchronized Multimedia Integration Language] and others. None are where they should be to address this problem.
Mr. Niemann: Did you look at SVG [Scalable Vector Graphics]?
Mr. Heiser: Yes we did, and we’ve had lots of suggestions…
Mr. Niemann: I want to respond to your email in detail, because there are a number of alternatives. [Mr. Niemann specified several.] My son works for a non-profit that’s repurposed a lot of your content into XML. Lots of alternatives.
Mr. Ambur: David has made an offer of meeting space.
Mr. Heiser: I have some resources…
Mr. Ambur: It sounds like Brand has made some resources available.
Susan chairs the Universal Access work group. Susan, would you like to participate?
Ms. Turnbull: Yes, we will participate. I'm aware of a number of research activities that are related to this issue. Regarding the reference made to XHTML and SMIL, the audio e-book version of the AIC guide "Extending Digital Dividends: Public Goods and Services that Work for All" uses these standards. I have heard concerns about the IRS approach because it supports only one screen reader product. I’ve also looked at SVG.
Mr. Heiser: Yes.
Mr. Ambur: So David is taking the lead on scooping, and Susan, Lex, and Brand are volunteering to work with him. Anyone else? [Mr. Sall volunteered.]
Mr. Ambur: Presuming we’re still in business next month, I’d like to consider a specific project proposal to the meeting. If we have direction from the AIC as to the format for the proposal, we can move forward.
Lex?
Mr. Poot: Have you had any contact from the Access Board, or not?
Mr. Ambur: I’ve had some.
Mr. Poot: Doug Wakefield is in charge of the Section 508 work there.
Mr. Ambur: The issue is what the AIC should be doing to facilitate and advance the cause from where it stands today.
Mr. Niemann: I work with the American Foundation for the Blind. I can help with testing.
Mr. Heiser: We have resources there as well. I don’t know whether you heard my twist. I’m authoring in SGML, and I plan to stay there.
Mr. Ambur: Does anybody else want to volunteer to lead a project or task?
Mr. Steve Vineski: A number of things fit into EPA’s agenda for XML. We’re working on schema design guidelines, then we need to deliver on schema best practices. We’re finding we have to deal with namespace architecture. What they look like for the EPA and partner States, probably will flow into the federal namespace.
We have another group, which doesn’t part in this WG. They’re working on Web Services in standards, protocols, and compatibility for servers to provide services. For some, like schema design guidelines, I can step up. I do have Agency deadlines I have to meet. At some point, I have to deliver to the Agency. For the Registry/Repository group, I defer to Bruce [Bargmeyer] and Larry. They’re already working with them on that. I can commit to the namespace. Our pace might be quicker than what this group is looking at, because we have to deliver in November.
Mr. Le Maitre: I’d like to participate in the namespace work, and I’d like to throw the discussion to a more generalized one of identity—which namespace takes different components?
Mr. Royal: I’ve been bouncing that around as well. My own history, as far as namespaces is, I was involved with GSA’s registry for XML.gov, X.00, and NSAP. The problem with delegating names—for example “.gov,” is that for each step you have, each delegation, there needs to be a body that manages that part of the registry. So we had looked at (and I’d be happy to make it in the form of a proposal or white paper) the idea that the government should adopt namespaces using DNS naming authorities, but different from just a URL. Start with “US” at the top tier. The second tier is “gov.” The third tier would typically occur on the left side—for example, GSA. My namespace would be US, gov, gsa, then whatever GSA extends beyond. The authority of GSA is gsa.gov inherently in the U.S.
Mr. Le Maitre: People in my company are working on XNS. It can sit on top of URLs. It’s written in EBNF, on URN services. We proposed it to an alliance suffering similar problems. The URN service is both machine- and human-readable, so you can name many times, so you have globally unique identifiers. I’d like to apply some of our thinking to this space.
Mr. Royal: Maybe set up a special meeting to get together and put these issues on the table, and come up with resolution quickly, so it meets your [Mr. Vineski’s] guidelines, and move on it.
Mr. Ambur: If we can deliver usable output, that’s great, but if all we can do is scope out, and determine the required resources, then that’s good too. In some cases we may need to go to the AIC and say, “We’ve identified this need but we don’t have the resources to address it.” We should try to identify and highlight crosscutting needs, and formulate project/task proposals so that the CIO Council can consider and add value to them.
Mr. Royal: We have one we can go forward on. I’ll take the lead, with Steve [Vineski] and Marc [Le Maitre].
Mr. Niemann: Having attended a week-long W3C meeting last week, they identified three emphases. One is the Architecture group. It’s on the outside, on the website. The bottom line is that URIs are it on the identifier. It’s important to get the W3C input.
Mr. Ambur: So you’re volunteering?
Mr. Niemann: I would, but I need to get it from the “horse’s mouth.” They’re bringing in David Booth. They want to be more involved. He’s very active in W3C committees. He’s a W3C fellow—he could be the interface, so if we want to get him here we could do that. It’s important because that’s their most important activity, now that Tim Berners-Lee is discussing future work of Web architecture—working on identifiers, architectures, and protocols. On identifiers, URIs are it.
Mr. Royal: Without going into the substance of what we discussed, it’s important to recognize that the standards evolve. Last year’s are not the standards for this year or the next, but on something like namespace, it’s important that as agencies use them, they recognize a way to uniquely identify within a global. There are a limited number of registered methodologies. We don’t want to go down the road of X55 or NSAP. It makes sense to use DNS. It’s scalable, it exists, and there are no new naming authorities.
Mr. Le Maitre: Once you’ve given an object an identifier, the next problem is authenticating that identity. So it has a name; now make sure the name is representative of the owner of the object. Then you have to describe that object’s relationship with other objects. RDF does a good job, but not completely. We started off with the problem of identity, and we ended up with how to describe relationships that control the flow between objects and provided the bounds for the plumbing between two objects. It’s not just authorized access to data, but it’s about the usage of information around it—how you can control it in a peer-to-peer environment. That’s what XNS looked at. They’ve done a good effort. They’re putting it into OASIS this week. They put it in there because it needs to be matured. It’s a 370-page specification as it stands including the URN. It needs to be matured.
Mr. Ambur: I’ll rely on Marion, as the leader of the namespace task team, to spell out the scope of the task. However, another, related potential task is the forum on SAML that we’ve been discussing with Judy Spencer of the PKI Steering Committee. Part of the scoping of that task is to determine whether to include XNS in the forum as well.
Mr. Le Maitre: XNS embraces SAML. We bind the assertion into a contract that allows it to go forward after the initial session.
Mr. Ambur: Judy is not here. The task for Marion is to decide whether we recommend that XNS be included in the forum.
Mr. Le Maitre: There is potential duplication of effort. That’s an area we’re cognizant of, so how does it fit into the security standards? There’s not enough detail to know. Assuming my colleagues at Verisign promote those at OASIS, it will be an interesting debate.
Mr. Ambur: I’ll leave it up to Marion to scope his task team’s effort. If the decision is that it’s outside the scope of the task he has agreed to lead, we could consider it as a separate effort.
Mr. Vineski: On the schema design I mentioned, it was posted to the XML.gov website last spring. I can set up a weekly or biweekly call if anyone in this WG wants to comment on that.
Mr. Crawford: There exists on XML.gov a set of federal design rules, based on the DON design rules document. The DON’s has been updated. The federal one hasn’t, and there’s been a lot of harmonization between the EPA’s and Navy’s design rules, and UBL. Hopefully, since I’m the chair of the group developing them for CEFACT, CEFACT’s will also be harmonized.
Mr. Vineski: How were the federal ones reviewed?
Mr. Crawford: They haven’t been. They’re based extensively on the Navy’s, which had a great deal of internal review.
Mr. Vineski: How were they named federal?
Mr. Crawford: They were put out as “draft” on XML.gov.
Mr. Royal: They’ve been before this group for a long time now. We need to come to conclusion. If we approve them, they could go through the OMB [Office of Management and Budget] architecture group, or CIO Council AIC.
Unidentified participant: They’ve evolved a lot at EPA.
Mr. Royal: So incorporate, and work on, the federal guidelines.
Mr. Niemann: We’ve encountered a practical problem of EPA working with the States, etc. If you don’t have all the people at the table during schema work, it brings about the need for subschema or components for harmonization. It’s important for people to know that, and that they’re making a conscious choice in that direction. You either need a larger group at the table and develop sooner, or down the road. It has implications on the Registry and the collaborative tools you use. It’s important to acknowledge. We need that in the document sooner, rather than later, because lots of people are making choices, and need to know their strategy.
Mr. Royal: That brings me to the point that, if you don’t have anything, you build it yourself. If you have something, even if it’s not perfect, the more who adopt it, the greater the chance it has to be interoperable.
Mr. Niemann: If you have a collaborative tool you see overlaps on vocabularies being developed, possibilities for reusability, etc. sooner, rather than later. People tend to be isolated and look around later, but sooner, rather than later, has advantages. We need to reflect that on the design documents, so people realize the ramifications of that. In our case, we went from two schemas to 50 or more, and with subcomponents, we’re still not done.
Mr. Vineski: I hear you Brand. We want the right people. We can’t decide who joins the calls.
Mr. Niemann: Lesson learned.
Unidentified participant: It’s important when you scale to the full government. You have the early adopters who harmonize with, say, 10 people, and consider it done. The 11th guy says, “I wasn’t a part of it.” Harmonization tends to require more time.
Mr. Royal: But if the early adopters are not meeting at first, then you have 11 implementations, so even the 11th guy can then say, “Here’s my contribution.” But if you don’t do anything…
Unidentified participant: You agree to do the harmonization and have a migration path, which means these things will evolve
Mr. Vineski: They’ll evolve, but I have a mandate to have a draft in eight weeks. I’d at least like to have it compatible with federal guidelines.
Mr. Ambur: What’s the actionable task? Is it to make sure the draft on the XML.gov site reflects the latest and best thinking? Do we have the resources to do that?
Mr. Crawford: The draft federal guidelines are based on the DON version 1.0. Some things are missing. The differences are negligible. One thing missing is that the Navy (and this is a difference between the Navy, federal, and EPA) focuses on design rules for schemas and transactional exchanges, and doesn’t broach the issue of content. The EPA does—whether rules are applicable for one type versus another.
Mr. Niemann: My recollection is (a month ago, I saw that Mitre had completed a one-year assessment of the DoD registration process), as far as I can tell, in the government they have the most schemas—90 schemas and DTDs, and 15,000 data elements…
Mr. Crawford: There’s no requirement to harmonize objects that are stored, or reuse what’s there. It’s split up into namespaces, with no attempt to harmonize.
Mr. Niemann: I couldn’t agree more, but they didn’t have the foresight to harmonize. The registry doesn’t do it automatically, without other processes in place.
Mr. Crawford: The model intended was if people reuse them, fine. It’s not up to us to control it.
Unidentified participant: I believe the model went beyond that. It was that way in the beginning, but there was an attempt to make it part of the COE [Common Operating Environment]. “Lower the barrier of entry to get it up there, and set it up so people don’t see all that stuff. Set it up so that people see that there’s community agreement.”
Mr. Roy Morgan: Lex, I think the amount of DoD use is too big to manage one registry. With all these diff communities of interest, you’ll never find one COE. That’s the issue of enterprise namespace. There won’t be agreement on tags
Mr. Ambur: I’d like to solicit someone to lead and others to volunteer to participate in scoping this.
Mr. Crawford: I’d just like to say that the DoD recognizes that, and they are involved in scoping their requirements put in place registry requirements and management that will do that harmonization. I submit that DoD is not too big to have a standardized way of tagging. They’ve had if for 40 years.
Mr. Ambur: If the Navy is willing to step forward and try, we’ll see. I encourage and laud them for their courage. Back to the Developers Guide, it seems the minimum step is identifying whether we need additional resources to update the Guide. If so, we should scope a proposal to the AIC to obtain those resources and, beyond that, think of a marketing plan and outreach strategy as well.
Mr. Royal: I don’t mind volunteering, but I take more pleasure out of volunteering others. If the EPA has it under way, and it’s a supplement to the existing federal guide, getting the latest from the federal, and combining it with the latest EPA one, that would be great.
Mr. Vineski: I have the resources to push the EPA one forward. I think there’s a mapping effort to look at discussing the federal one, the DON’s, and the EPA version.
Mr. Ambur: That’s the minimum: Look at the scope and tell us what we need to be doing.
Mr. Vineski: Sure, but that mapping’s not one of our deliverables.
Mr. Ambur: Even if it all that it means is saying, “We need “x” dollars to do that,” I want to have a list of potential projects as well scoped as possible to provide to the AIC.
Mr. Vineski: If people are interested, please contact me.
Mr. Royal: As a FY 04 budget item, the XML Registry/Repository is funded by GSA, but we need things done before 04. We need things done in FY 03 to move us to a scalable government-wide registry/repository. We need, as a project a team, a group of people that can help us develop a strategy of what to do in 03, to move us to deployment in 04. The business case analysis is finalized now. It’s supportive of a government-wide registry/repository, and it gives several options:
The direction in moving forward needs to be determined by this committee.
Mr. Morgan: The topic for this afternoon with remarks from the Bureau of Public Debt and from Bruce, is, “What are the processes that have to go on for people to do their work in what has to be built?” The committee needs to think about the processes of what happens: deciding what XML has to do in the business world; what the government needs; what should be brought into being to do the government’s work. I propose that the committee take on that task—to define briefly a sequence of capabilities that will one day lead to the point where the 04 budget line item procurement is turned on, and a light shines across the world. We need incremental capabilities, and a picture of what they are. I concur and am willing to work on it. A team is in place. The degree of participation is not known, but we’ll find out.
Mr. Bruce Bargmeyer: There are real design pieces of work to do. “Federation” implies that they communicate. There is a variety of standards being worked on to accomplish that.
Mr. Ambur: Relative to reorganization of the AIC, they’re planning to have an architectural component subcommittee, to consider the logical separation of components. Identifying what “federation” means could be a task for that subcommittee. We’ll see if the AIC has the resources to make such determinations. If not, at least we can identify the needs.
Mr. Royal: We have to talk about money. We have planned money to take us to 03, but I’ve heard rumors of year-end money as well on the AIC, and I think this issue is one we should lobby, because it’s important.
Mr. Crawford: Bruce, how many of these issues can be put to OASIS for resolution?
Mr. Bargmeyer: Many. UDDI…
Mr. Crawford: Ask the UDDI people to establish a joint technical committee. We’ve populated into there a little to try to encourage that convergence.
Mr. Bargmeyer: There could be a big convergence of UBL and UDDI activities.
Mr. Ambur: That may be something with which the CIO Council could help. The first task is to convince the higher-ups that it’s something that should be done.
Mr. Bargmeyer: I’m not sure they all want to be in the same room. It could be thorny.
Mr. Crawford: The door exists for discussion.
Mr. Royal: I’m not sure we have to go that high initially. Say we draft a short statement saying, “We have a concern, and what is OASIS’s position?”
Mr. Crawford: And Patrick [Gannon] would be responsive.
Mr. Royal: And if that’s not satisfactory, we could raise it up to the appropriate level.
Mr. Ambur: So is someone volunteering to draft a memo to Patrick?
[Mr. Bargmeyer volunteered.]
Mr. Niemann: At W3C, David Booth talked with Mark Forman. Forman recommended, and they accepted, a broadening of the definition of Web Services, to build on the WSDL specification. A question was, “How does it relate to OASIS’s? Is it larger or smaller?” The consensus was that it was a broader definition than that of OASIS. The Web Services description group definition is a quite general one of capturing all relationships that exist now, using words like “choreographing Web Services components across a global network.”
Mr. Ambur: So as head of the new Web Services group, you’ll keep us abreast?
Mr. Niemann: It’ll be interesting to watch these two standards bodies move forward in this area.
Mr. Ambur: Any last comments or volunteers?
Ms. Turnbull: I’ve never been part of an OASIS Technical Committee (TC), but you mentioned streamlining. I was wondering whether it’s ordinary that the OASIS process entails a collaboration tool for members to vet; a discipline, hierarchy, or platform on which to build without breaking each other’s contributions? My question is, is the XML arena routinely using these, or should we learn how to use them to bring things that work together, so they’re not isolated blocks?
Mr. Crawford: I can’t speak for all TCs, but I’m not aware of collaborative tools being used. OASIS doesn’t provide collaborative tools for their work. OASIS has assigned editors responsible for the products, and work done on teleconferences and email. The editors are responsible for the input of team members and turning it into a product of the group.
Mr. Le Maitre: I spoke with Simon Nicholson, and asked him the same question. He said there’s no mandate for groups to work together. One might be to confederate to this cause as a mandate from Cisco.
Mr. Crawford: OASIS does have a process for putting in place harmonization between various TCs, called establishment of a Joint Coordination Committee (JCC). There are specific guidelines for how they’re formed and what they provide. Regarding Bruce’s idea, maybe we should push for, as a first step, the creation of one of these JCCs for UDDI and maybe ebXML.
Mr. Niemann: There is some of this collaboration. We’re seeing some of these TCs overlapping in their meeting, coordinating these standards to be harmonized and consistent. We’re seeing it with Tim Berners-Lee and his committee.
Ms. Turnbull: I’ve been exposed to software collaboration use on WebDav, coming from a founder of Apache, who’ll be in town on October 11. I’d be interested in setting up a meeting with him for people interested in hearing about Apache.
Mr. Ambur: At the Registry/Repository meeting this afternoon we’ll hear a presentation from Kevin on his XML collaborative tool?
Mr. Williams: I’ll talk about the role of collaboration: “How do you maximize value,” combined with a white paper on XML.gov.
Mr. Ambur: Thank you all.
End meeting.
Attendees:
|
Last Name |
First Name |
Organization |
|
Ambur |
Owen |
FWS |
|
Bargmeyer |
Bruce |
|
|
Crawford |
Mark |
LMI |
|
Duliere |
William |
Bureau of Public Debt |
|
Ellis |
Lee |
GSA |
|
Glace |
Jessica |
LMI |
|
Heiser |
David |
IRS |
|
Kane |
John |
NARA |
|
Le Maitre |
Marc |
|
|
Morgan |
Roy |
NIST |
|
Niemann |
Brand |
EPA |
|
Royal |
Marion |
GSA |
|
Sall |
Kenneth |
Ken Sall Consulting |
|
Troutman |
Bruce |
8020Data |
|
Turnbull |
Susan |
GSA |
|
Vineski |
Steve |
EPA |
|
Weber |
Lisa |
NARA |
|
Yee |
Theresa |
LMI |
|
Zacherl |
Donald |
SiloSmashers |