Let's Go to the Tape: Q & A with a Microsoft XML Guru

Charles Heinemann
Program Manager, XML
Microsoft Corporation

May 11, 1998

The following article was originally published in the Site Builder Network Magazine "Extreme XML" column (now MSDN Online Voices Extreme XML).

Many of you have written me since this column debuted in March, asking for the scoop on just what we're up to here at Microsoft concerning XML. How would I know? I just work here. So I took a stroll down the hall asking if anyone would be willing to come to my office and give me the scoop -- on tape.

Adam Denning, the group program manager for XML and, incidentally, my boss, agreed not only to come to my office, but also to come clean. Now, for those of you who think that I chose this format for this month's column just to lessen my workload, let me tell you that getting this interview was no cakewalk.

First, I discovered that Mr. Denning, a native of the biggest of the British Isles, dislikes Marmite sandwiches, making for a very uncomfortable moment when the Microsoft Catering cart arrived. (How was I to know? I thought they all loved that stuff.) Second, he is often difficult to understand. (They told me he would be speaking English, but that sure ain't the kinda' English they speak where I'm from.) Third, he kept making obscure references to some show called Beyond the Fringe. (He later told me it was some comedy show with that short guy from Arthur -- who, just between you and me, I never thought was all that funny. Now, Benny Hill -- he's funny.)

All my troubles aside, enjoy:

Q. How did Microsoft get involved in XML?

Adam Denning: We were in the vanguard, so to speak. There were a few visionaries at this company who, along with a number of other people outside this company, believed XML to be far more than simply a document markup language. So we fostered it and supported it as a language for data. We firmly believe that XML as a universal text-based language for structured data will fundamentally change the Web, and our involvement in it from the beginning has reflected this.

Q. What role do you see XML playing on the World Wide Web?

Adam Denning: The primary role would be moving information from one place to another. Right now, the average work server provides that information as pages, which means that what you get is basically a visual rendition of information that you might actually want to process. What you really want to do is receive the information in such a way that you can understand it computationally, do things with it, and maybe then ultimately display it. Presentation will be merely a facet of the Web, rather than the whole thing.

Q. How does the average user fit into this vision?

Adam Denning: "User" being the key word here. Home shopping. Imagine you want to go and buy tape recorders to do interviews. How do you do that now? You go to a number of sites, look at the things available on those sites, find out prices, products, and then purchase the item, or maybe even go to a physical location to buy the product. But let's say you wanted to look at different tape recorders, made by different manufacturers, side by side so that you could compare their features and prices. You could then choose the best one at the best price and purchase it. XML will make this possible because information will be available and accessible, as well as more easily searched.

Q. What role do you see XML playing in terms of future desktop applications?

Adam Denning: The thing about XML is that it is totally broad. So you could use it for virtually anything. For instance, it could be used to describe documents produced by a word processor, components within a spreadsheet, or other structured information.

As an aside, I do believe that over time, and I'm not sure when, Microsoft Office will become one of the main tools you will be able to use to create marked-up information: data, documents, or whatever you want to call it.

Q. Besides Internet Explorer, what other groups at Microsoft are supporting XML?

Adam Denning: I think it's more a question of which groups aren't. Literally, everyone that we speak with has some interest in XML, and for a variety of different purposes.

Q. What groups are not just supporting XML but actually using XML?

Adam Denning: Internet Explorer is already using it for various purposes, such as Channel Definition Format, OSD, and so forth. Commerce Server is using it as the data-transfer format within the Commerce Interchange Pipeline. Microsoft Office is using XML within documents to hold information about the document. These, and a host of other groups. Within the next year, you'll see XML being utilized just about everywhere.

Q. Why would these groups choose XML over some other format?

Adam Denning: If the system is totally closed, then it's a good question. Why use XML when you could use a potentially more efficient format? But in the case of most of the groups that I speak with here at Microsoft and elsewhere, in fact, the system is not entirely closed. Therefore, you need a universal format that doesn't require that you communicate only with certain systems.

Q. Why not use existing standard formats, such as EDI?

Adam Denning: The problem with EDI is that it takes a long time for people to agree on a standard, because everybody wants their individual customizations, and so forth. And it's a format that is not particularly human readable. So XML can, in many instances, replace EDI, because it allows for extensibility without requiring any underlying changes in the format.

Q. Are you concerned at all about the compatibility of Microsoft's XML technology with, for instance, the XML support in other browsers?

Adam Denning: I'm concerned because we have to make sure that it happens. But I believe that there's enough industry momentum behind it to ensure that, at least at the macro level, compatibility exists.

Q. How important do you see such compatibility as being?

Adam Denning: Critical.

Q. Why critical?

Adam Denning: Because if people want to use information from a wide variety of sources, there will be no guarantee that others will be using a specific browser. There's no guarantee that anything will be constant, apart from the information being transferred. Therefore, to allow for the broadest reach of applications using XML, you need to ensure that browsers have the same support.

Q. Why won't we have the same problem with XML that we did and do with HTML?

Adam Denning: The problem with HTML is that HTML is a defined standard where not only the syntax is defined, but the tags within the syntax are defined. There are a definitive number of tags within the syntax that can be used. Each of those tags has a defined behavior. So therefore, if you want to extend HTML in any meaningful way -- for example, add data binding capability or enhance it to use ActiveX controls or other objects, or if you want to use Dynamic HTML -- you essentially have to change HTML. You either have to add new tags or extend the behavior of existing tags. Doing so, in both cases, means that what you've now created is proprietary.

With XML, the syntax is defined, but not one single tag is defined. Consequently, any semantics that a tag has are entirely up to the author. There is no predefined tag saying, "Display me in a paragraph," or "Put me in bold," or anything else. All tags are open, and therefore you can't have the same problem as with HTML, because the standard is deliberately open. Now, you may argue that that situation therefore introduces a different problem -- and that is that we may both define <TITLE> tags, but have different meanings associated with those tags. But this is solved through the use of namespaces . With namespaces, both <TITLE> tags can be used at the same time, because they are both made unique.

Q. What is Microsoft's main focus, in terms of XML and XML support, over the next six months?

Adam Denning: For the short term, our plan is to provide the functionality that we've been evangelizing for so long, and to make sure that people can actually fulfill our vision and their vision using XML.

Q. What about over the next couple years?

Adam Denning: Over the next year or two years, we see people across the Web universe beginning to understand how XML can be leveraged by their customers. Therefore, we see people beginning to expose the types of services we spoke about earlier. So you will begin to see booksellers exposing information, their list of books if you like, in such a way that you could query it, that you could find the best price among them. We'll see search engines that can understand XML tags, and can therefore not only find documents that contain the words Jane Smith, but contain information about the "author" Jane Smith.

Q. What, in your opinion, is the key to XML becoming widely adopted, to it enjoying the success that HTML currently enjoys?

Adam Denning: In fact, there are three keys. First, we have to see that it has value. People have to understand why, exactly, XML is interesting to them and why it is valuable. They won't simply adopt it because it's a new wave of technology; they have to see that it has value. Second, we have to provide them with the capability to utilize that value. So that means having browsers that support XML, some standard parsing and rendering, but also all the other bits: querying, transactioning, updating. Third, industry bodies have to decide that they can agree upon XML vocabularies, that booksellers can come together to agree upon a vocabulary for books. So that everyone knows that if you want to speak books, there's a vocabulary for doing so.

Q. Why would two booksellers want to cooperate like that?

Adam Denning: Well, inevitably consumers are going to demand that they have access to data marked up in XML.

Q. What you're saying, then, is if people are going to get involved with business on the Web, they're going to have to get on board with XML, because that's going to be . . .

Adam Denning: The lingua franca of commerce.

Q. So in a certain sense, XML will enable the customer, the quite-average customer, to actually have access to not only the products, but also the data that surrounds the products?

Adam Denning: Oh, yeah.

And with that, the Englishman checked his watch, apologized in a most gracious fashion for having to go, and headed off down the hall, no doubt to find some "real food" to eat.

Charles Heinemann is a program manager for Microsoft's Weblications team. Coming from Texas, he knows how to think big.

Extreme Mail

Dear Charles:
I'd basically like to display a virtual comics page so I don't have to chase down each site every day. However, this seems like it would violate their copyright to the strips. Is XML mostly for use with non-proprietary type info, like stock quotes, or would it be okay to use it for comic strips?
Thanks.

Charles answers:
XML is a data format intended for universal use and access. This means that in building XML support into Internet Explorer, it was assumed that XML would be used to describe both proprietary and non-proprietary information. We do, however, understand that the transfer of information, regardless of the format, gives rise to copyright and security concerns, and we are working very hard to ensure that these concerns are dealt with in an appropriate manner.

For technical how-to questions, check in with the Web Men Talking, the Web Workshop's answer pair.