Skip to main content
August 29, 2018

Full transcript of Lightweight DITA podcast: part 1 with guests Carlos Evia and Michael Priestley

Gretyl Kinsey:    Welcome to the Content Strategy Experts Podcast, brought to you by Scriptorium. Since 1997, Scriptorium has helped companies manage, structure, organize, and distribute content in an efficient way. In episode 35, we talk about Lightweight DITA and more with Dr. Carlos Evia and Michael Priestley. This is the first of two podcasts featuring these special guests.

Gretyl Kinsey:    Hello and welcome to the podcast, I’m Gretyl Kinsey, and as I mentioned, I’ve got two special guests with me today, Dr. Carlos Evia and Michael Priestley. Why don’t you both just go a head and introduce yourselves.

Carlos Evia:         Alright, if we go in alphabetically, then that would be me. Hi, I’m Carlos Evia, I’m an associate professor of communication at Virginia Tech.

Michael P.:          I’m Michael Priestley. I work at IBM in the IBM marketing and IBM digital area, supporting marketing systems and marketing organizations with content classification and content tagging systems, but my background that informs this call is also around content standards, and especially the DITA standard and the Lightweight DITA variant of that, and I was one of the original architects of DITA at IBM and the co-editor of the DITA specifications for 1.0 and 1.1 in OASIS.

Carlos E.:              Well, what Michael and I have in common is that we co-chair the Lightweight DITA subcommittee with OASIS.

Gretyl K.:             Fantastic. Michael, I’m going to put my first question to you given that context you just gave us. For those of us in our audience who might not be familiar with DITA, what is DITA?

Michael P.:          Wow. Okay. Yeah, this is the usual caveat that asking the person who’s really close to something is not actually the best person to encapsulate it, but I like to think of it as a standard for content that allows it to be used across systems and reused across organizations and context. There’s really three parts to the standard. There is the definition of chunks of content that are organized around a topic or a set of information that delivers a cohesive point, so it’s not, when I say a chunk of content, I don’t mean a single paragraph or a single word, I mean something more like a meaningful title and the information that goes right underneath it. Maybe it’s an article, maybe it’s a help topic, something more at that level, the unit of use that a user or a client would be looking for and that’s sort of the one part of the standards. We standardized the way to mark up that content in a way that you can then publish it in different media like a pdf, or a web page as well as we use it in different contexts like for example, the same explanation of a concept or the same task might appear in a support context, a troubleshooting context, in the information set for several different products that share a common component, etc.

Michael P.:          It’s that idea of having components of content that you can reuse across contexts, that’s a key part of the DITA’s value. The other two parts of the standard in addition to that definition of the chunk is, one, the definition of a contextual larger collection where you want to say, I want to collect these chunks together and create relationships between them. Those might be related links or link management might turn into the site map for a website or a micro site, or it might become the table of contents and a pdf, and that’s what we call in DITA a map. Now you’ve got these topics is what we call the chunks of content, maps which collect the content.

Michael P.:          Then sort of the third angle on DITA is the fact that we can create more specific types of content and maps and so we might say, we want to have task content and we want to make sure that tasks content always has a set of steps and always starts with a list of prerequisites, if anything. And you can have more specialized types of tasks, so you can say, this is a how we want to structure a task for a manufacturing application. This is how we want to structure a task for developers, this is how we want to structure a task for end users of software.

Michael P.:          You can have types, and types of types, and go as deep as you want and still have them work as part of this larger system and have the content be recognized across different, what would normally be considered silo barriers in an organization. That’s the key piece of–those are the three pillars of DITA, the content, the collections, and the ability to create more specific types within those. The big value proposition for DITA on top of all of this is that we are a standard and that means that if you invest in putting all of your content into a DITA enabled system, you can take it out again and you can move it into another system. That ability to switch systems and not be stuck in your silo forever because of your investment in content that’s really what makes something a useful standard.

Gretyl K.:             Right. We’ve actually had, we’ve had a lot of people have among our client base who we have about 80% or so of our clients are in XML and then within that, about 80% of those are in DITA, so they’ve definitely, I think, seen that value proposition that you talk about, and that, for a lot of them, has been their way of getting out of some of the limitations of being in silos or of being stuck in one single system or tool. It’s given them that flexibility that they’ve been looking for when they come up with a content strategy and say, “Here’s what our goals are for our next steps.”

Michael P.:          Great to hear.

Gretyl K.:             Then, one other thing I want to distinguish is the difference between DITA and Lightweight DITA. I want to hear from both of you about what Lightweight DITA it is and how it’s different from DITA.

Carlos E.:              I think one of the things that Michael was talking about with those stages or steps that DITA allows is that the creation of topics, of chunks of information, the creation of maps that will organize and represent those topics into deliverables that human beings will understand and the expansion of those topics into a specific unit of content that organizations or individuals might need, like you mentioned depends on XML, the Extensible Markup Language and for many years XML has been in a very happy relationship with the field of technical communication, but there have been developments in other markup languages and people might cringe when I say that Markdown is markup language, but HTML with HTML5 has developed some structuring capabilities quite similar to XML and it’s easier to use and easier to implement than XML. Like I mentioned also, Markdown is even easier to write than HTML5 and there are many open source and commercial tools that combine HTML with Markdown and really make something that is robust enough that it can be compared to what you can accomplish with XML.

Carlos E.:              Lightweight DITA is a simpler version of the DITA standard that uses XML in a minimized version of all the available tags for structuring content that has but it also represents a way to structure content, to create those chunks of information, to create those maps, and to customize the way in which people need to see those pieces of information in HTML5 and Markdown.

Carlos E.:              The flavors of Lightweight DITA are three in this initial release that we’re planning, and they are XDITA, HDITA, and MDITA which respectively correspond to XML, HTML5 and Markdown and one of the good things about them is that they are compatible with each other and that means that if you are in your organization, you’re writing content, you don’t have to be 100% committed to writing everything, all of your chunks of information in XML. You can write some in XML, you can write some in HTML5 and you can write some in Markdown, and if you’re following these rules of the proposed Lightweight DITA standard, you can combine them, and they can live together, and they can be exported into information products that for the end user they will look like they came from the same source.

Michael P.:          If I can add, one of the drivers for creating Lightweight DITA was really looking at where DITA as an existing standard was hitting resistance. DITA, the full DITA standard is working great for the group that have adopted it and it’s still getting adopted by new groups today. Where it was encountering resistance, there were two main reasons that we thought we could maybe address. One was, there was a perception that even if you started simple in DITA, the simple starting point was not simple enough. Could we make that starting point even simpler? The second one was when you have groups who would consider DITA and they like the capabilities it brings, but they are really tied to an existing authoring format, and an existing tool stack around that format. We said, “Well, maybe we can kill two birds with one stone here if we simplify DITA to the point where it’s an easier starting point that also makes it easier to map across formats.” That’s how we got to this three flavors, one standard implementation.

Gretyl K.:             Both of you, as I think you mentioned a little earlier, have been really heavily involved in the development of Lightweight DITA. Can you just take us through that a little bit and talk about what that’s been like?

Carlos E.:              Well, Michael can talk more about these because he went to bed one night around 2014 in and he woke up with–he had the dream of Lightweight DITA, and he started writing about it in a blog. And at the time I was just a fan of DITA. I was just a user, and I had been teaching DITA in my classes here at Virginia Tech for a while, but I had never interacted with the standard or the development side of the standard and I had never really interacted with Michael other than liking some of the things that he was tweeting. He wrote that initial blog post presenting, “I have an idea for Lightweight DITA.” He started going to conferences and he was talking about Lightweight DITA as the simplified version of DITA which was still based on XML.

Carlos E.:              One day he started talking about HTML5 as a way to represent Lightweight DITA as well. That’s when I started working with him, and I said, “Let’s just try to expand it and incorporate into Markdown. He created the Lightweight DITA subcommittee with OASIS and I joined the subcommittee and I have been working with him for a few years now on these proposed new standard. Michael, what were you having for dinner before you had a dream about Lightweight DITA? What inspired it?

Michael P.:          Well, I will say that at least part of it was just the challenges that I was facing within IBM, looking at, at that point I was involved in a very cross-silo role looking at enterprise content strategy. And there were certainly some areas that were moving to DITA, and many that had already moved into DITA and we were very happy with it. But there are others that just were not going to move, and it was because of those attachments to existing formats. I was struggling with that idea. My dream has always been that I wanted to get our internal ecosystem within IBM, our content ecosystem, to support the end-to-end customer journey. We’ve got content in marketing, we’ve got content in sales, we’ve got content in training, in documentation, and in support and they’re all coming from different systems and they’re all using different underlying formats. And it’s hard to support a customer who is going to be going through that life cycle of content if you keep having to switch silos and switch standards from an authoring standpoint.

Michael P.:          I wanted to remove a barrier to cross silo collaboration internally by allowing content to have relationships and share common elements, share common classification schemes, and so forth. This format thing was a major blocker, and so I thought, we got to do something. I had been looking at HTML5 in particular and thought, this “article” element that they’ve added to HTML5 is really a reasonable semantic map to a topic in DITA, and where can we go from there? Yeah, once Carlos reached out to me and said, “Hey, let’s see how this works and let’s start making it real and testing it and seeing what flies.” It really started taking it off and making it less theoretical like, “Hey, maybe this will work,” and more, “Hey, this actually does work.”

Gretyl K.:             This concludes part one of our conversation with Dr. Carlos Evia and Michael Priestley. Stay tuned for part two. For now, thank you for listening to the Content Strategy Experts Podcast brought to you by Scriptorium. For more information, visit or check the show notes for relevant links.