# DITA - Do we need a site for just this?

Proposal: DITA

DITA is, for all intents and purposes, an XML file format that has a particular series of tools built around it. It's mainly used as a markup language for documents, with an accompanying set of processing tools.

Is this something we need a site for? It feels like this is really a bigger site than just this. There are a number of document markup languages (DocBook and TeX/LaTeX among them, though the latter also includes presentation information); we shouldn't have separate sites for each. I know TeX/etc already have their own site, but that's not an example we should be following.

HTML is already effectively covered by Stack Overflow, as it interacts with programming languages (and stylesheets) enough to be considered programming. But most of the rest (DITA, Docbook, Markdown, etc) aren't covered. I'm thinking that we should have a unified site for all of these kinds of document markup languages.

Note that we stick all programming languages on the same site (Stack Overflow). So I don't see why DITA is so special in the document markup languages domain to need its own site.

• Was there ever a SE site for Technical Writing? I've seen references to one, but Google doesn't find actual info and SE doesn't seen to list deleted sites. Because that's the site that might generate enough activity: technical writings in DITA, DocBook, S1000D, TEI etc. – jelovirt Jan 20 '13 at 17:23
• I believe the technical-writing proposal was folded into Writers. Questions about technical writing are on-topic there, as are questions about the writing process, including tools. We don't have a lot of tech writers there right now, but as one of them I'd like to encourage y'all to bring questions there (at least while you wait for this site to launch). – Monica Cellio Feb 17 '13 at 1:15

With over 3,000 members tracking the dita-users list, the community around DITA is large and demonstrably growing. That list has been useful for discussions that inform as well as for sharing some specific answers to questions. However, those nuggets of knowledge cannot be easily identified, vetted by community process, and indexed for ease of finding and reusing, so the same questions often get asked over and over on that list.

As an architecture and not just an XML vocabulary, DITA enables building many kinds of applications with it. This sets it apart in a major way from XML vocabularies that are designed around particular data models and those who share only within those circles of usage. With a specific focus on DITA, this site can host questions about the architecture that may have been answered by someone in the tech writing community but be of use to someone using specialization to define business rules or UML data models or content syndication über-manifests.

There is also a sense of a "DITA Way"--a zen of how to work efficiently with the strengths of the design rather than approaching it with conventional predispositions of XML or word processor-informed usage. Some companies may manage their DITA workflows in particular ways for good business reasons, but those are not necessarily best practices for all. The eternal question about "files vs CMS" is a case in point--it all depends. It is a good question, but for most readers on dita-users, trying to argue a good case for a file system may not make sense.

So I'd keep the focus here specifically on DITA so that we can build an accessible, easily accessed and retrieved body of knowledge about the architecture, its peculiarities and subtleties, and its best practices. And hopefully all our angle brackets will stay around!

• Note that we stick all programming languages on the same site (Stack Overflow). This doesn't explain why DITA is so special in the document markup languages domain to need its own site. – Nicol Bolas Jan 20 '13 at 21:04
• Then why does StackExchange need special community categories in the first place? Your argument about sticking all programming languages in one place does not necessarily apply to what I observe as useful categorization of communities of knowledge on StackExchange. DITA is not a DTD or dedicated document markup language, it is an extensible architecture. It has very little to do with languages that do not have similar degrees of extension and application. – Don Day Jan 20 '13 at 21:12
• StackOverflow, as the largest community on StackExchange, is unusual in this sense, most other sites are much more focused. There is no "Web CMS" site, but rather "Wordpress Answers", "Drupal Answers". I think because SO was the first one, it got a very broad definition (basically programming). Maybe they didn't foresee the popularity to come. And now this broadness is not always a good thing. You need to specify the programming language to get relevant answers when you search. The more focused topics of later sites avoid this problem. – Anders Svensson Jan 20 '13 at 21:35
• @DonDay: "It has very little to do with languages that do not have similar degrees of extension and application." You could say something similar for the differences between functional and non-functional languages. – Nicol Bolas Jan 20 '13 at 22:06
• @AndersSvensson: "There is no "Web CMS" site, but rather "Wordpress Answers", "Drupal Answers"." Yes, and I believe this was a mistake. Just as I believe that having Physics, Chemistry, and etc as separate sites was a mistake. Just as I believe that having Ask Ubuntu be separate from U&L was a mistake. The SE model works best to me by having fewer sites with more stuff in each site. – Nicol Bolas Jan 20 '13 at 22:06
• @NicolBolas While it may be easier for some to have fewer, broader sites, I believe that the DITA community does not want to be going through other technologies to discover the answers they seek. For the most part, the community is looking for answers quickly and have little time to weed through other content interspersed through the information they need. This is why the DITA User's Group on Yahoo is so large right now. Grant it, the majority of the folks are more lurkers but the experience base is broad and rarely has time to filter much. – JulioV Jan 21 '13 at 13:31
• +1 Julio. IMHO, the purpose of any question answer site is to connect users with experts. The current stack sites do not aquatically connect DITA users with DITA experts. Because of this, you get no answers, or you get answers which are not relevant, are misleading, or questions get tagged as duplicates when they are really not. Context is very important in providing better help. Ask a physicist and a chemist the same question, and you will get very different explanations. Which one is correct? Well it will depend on whether you are in a chemistry or a physics class... – Casey Jordan Jan 21 '13 at 23:18
• @NicolBolas ... As a practitioner of both DITA and Drupal, I can assure you I would have zero interest in following a generic SE site for web CMS systems. Systems like Drupal, WordPress and Joomla! are so different there's little in common for any practical application. Those interested in the 'pure research' esoteric aspects of how CMS products might be implemented could find such a site useful, but the value for us 'applied research' folks who use these systems in the real world for real applications would be very limited at best. – jjkd Jan 23 '13 at 3:55

DITA definitely needs a site of its own, as the specialization technology that is the core concept of DITA does not exist in any other XML format. The fact that it is XML "just like DocBook and TeX" basically states that all of today's software formats should be handled in the same site.

Various XML standards have emerged exactly because of the vast differences in their application domain. People who have questions about particular meanings or technical backgrounds of tags and mechanisms within an XML standard invariably have those questions from working in a specific domain.

DITA questions have almost nothing in common with DocBook questions, and having one site to fit all will effectively kill its value for most, if not all of its users. If you combine all XML standards into one site, you might as well not have a site at all and keep using Google to find your info.

• I agree whole-heartedly with this answer. I thought about the alternative to propose a more general XML site first, but decided it would be much better with a DITA-specialized one for these very reasons. DITA, as we know, generates a lot of questions on its own. – Anders Svensson Jan 20 '13 at 13:11
• "DITA questions have almost nothing in common with DocBook questions, and having one site to fit all will effectively kill its value for most, if not all of its users." If you substitute JavaScript and C++ for DITA and DocBook, your statement would work just as well as an argument against Stack Overflow. And yet, that works just fine for SO. Why is DITA so special that it needs its own site instead of sharing it like JavaScript and C++ do? We are talking about two technologies that do essentially the same thing. They go about it in different ways, but they're document formats. – Nicol Bolas Jan 20 '13 at 20:39
• @NicolBolas See my comment to Don's answer too. But even though I see some issues with SO with irrelevant answers coming up unless you limit your search to a programming language, there is a difference. An application programmer often needs to know and use several languages, and an answer about C# might be useful to a problem in Java. I have rarely seen specific information about DocBook or S1000D that is relevant to DITA. Things like XSLT would be a potential commonality, but DocBook XSLT questions are likely to be of little use to DITA XSLT, which has its own very particular structure. – Anders Svensson Jan 20 '13 at 21:49
• People on dita-users are used to dealing only with DITA and I can see why people want to make the separation between DITA and rest of the modular documentation languages. However, on StackOverflow tags are an easy way to categorize questions and it works. If you're only interested in dita tags, then the site can more or less hide the other questions. I don't really care if the SE site is DITA-only or documentation markup languages in general, I'm only worried that DITA-only is too narrow a scope and the site will die out. – jelovirt Jan 21 '13 at 8:01
• @jelovirt Im not afraid that DITA-only will be too narrow. There are a lot of questions that are DITA specific. Think of specialization, constraining, creating plugins, ... And we need good tags inside DITA foe these specific questions. – Ben Welman Feb 17 '13 at 16:11

... I would see the same issue with a generalized 'markup languages' SE site. To look at say, as you mention, markdown (RIP: Aaron Swartz) and DITA, and to suggest there isn't a substantive difference between them reflecting the very different intent and implementation concerns me. The target audiences for these are so different it is hard for me to imagine how they would overlap in any significant way.

I would suspect that such a wider site might appeal at best to those interested in the theory of implementation rather than the practice of using. I know that as a practitioner of DITA, I would have no interest in such a generic approach.

I suspect that by combining specialized interests, the assumption that the result is a wider appeal (the 'union' of the interests) is not a good assumption. For sufficiently specialized interests, I think the result would more be limited to that which is totally in common (or the 'intersection' of interests). I think that certainly applies in this example.

I would also venture that the example of programming languages is more of a special case than a useful generalization. With few exceptions, the principles implemented in most commonly used programming languages really aren't that different.

I would have the same question. I would welcome more dita-topics on Stack Exchange, but I worry about doing a micro community around it, that i could mean missing nice exchange on broader topics. Something like a "technicalwriters" or "documentpublishing" would be a better choice in my humble opinion. It would also create opportunity to make dita known to technical writer looking for ways to better organize their work (and not only reach to those looking for Dita as such - I beleive in grouping people by problem more than by solution).

To those saying that dita is different from other tools/standards/format (Docbook, etc), I agree, but Java is different from Ruby, and both live happily on Stack Overflow - I would regret to see them go their separate ways, and there is benefit in the interaction.

This being obviously just my two cents.

Martin

An overriding issue is the need for the DITA community to provide a venue that interconnects developers, implementers, and users. Whether DITA falls into a class of languages that belong on SO is not the issue in this case. SO is not a venue for interconnecting these different roles. Most DITA developers are already members of SO and understand the scope of interaction provided by the site.

As Don mentioned above DITA as an architecture is applicable to, and is being implemented for, a very broad set of use cases. If we were to focus only on applications such as technical documentation and publishing we would not be providing a resource for those implementing DITA in financial services, government, human resources, and other applications that benefit more from DITA architecture than from the base vocabulary itself.