Open Source Alternative for DITA Open Toolkit

Question

I am working on a web application that should process DITA documents from persistent storage (probably JCR). The DITA Open Toolkit is the only DITA implementation I know of, but it requires all your documents to exist on the file system. Ideally, I would like something that works like DITA OT but allows you to provide a resolver (like XSLT URIResolver) to pull link content from other sources.

If people have other ideas, such as using a virtual file system to trick DITA OT, I would also like to hear them. Thanks!

Edit: I forgot to mention in the original publication that I am looking for an open source solution, as this is a project released under the license of the educational community.

Answer 1

After some evaluation, the latest version of XMLMind Dita Converter (ditac) is really working. Performance is at least twice that of the Open Toolkit for creating identical projects: http://www.xmlmind.com/ditac/

Answer 2

One comment about XMLMind Dita Converter (ditac) is that it is released under the Mozilla Public License, which, according to http://www.gnu.org/licenses/license-list.html#GPLIncompatibleLicenses, is not compatible with the GPL.

Answer 3

Take a look at Arbortext (specifically Arbortext Content Manager). Arbortext supports xinclude, catalog files, and it also has a publishing tool for publishing in PDF and digital media that you don’t get with OTK. OTK is not really intended for production.

Yes, I’m a seller (now), but I started working as a developer more than ten years ago. I answer many community questions and sponsor two dozen resources to answer people's questions. The best of which is the user group SF Bay PTC Arbortext (Virtual).

Do you want to do something like what Juniper does? (I can publish only one link, so it will be mine.) Go to the juniper network, select support, technical documentation, ex-series platforms, any of the ex-series documents. They show topics on the Internet directly (this is also inside the source code on the router and in pdf books). This will help if I understand what you are trying to do.

Feel free to contact me offline.

Answer 4

Perhaps DITA2Go can help:

http://www.dita2go.com/

DITA2Go allows your files to be anywhere as you requested. It also has many extensions beyond what it provides for OT, such as key keys and dinats, which are covered by DITA 1.3. It was created with the intense collaboration of two TC members working on large live projects, and is currently being used by hundreds of people. It is also about ten times faster than OT, thanks to C ++ and does not require programming skills at all.

It is free, but it is not Open Source. It is fully supported, and developers fix bugs immediately and often add new features in a day or two upon request. According to his recent review, he shares most of his code with the commercial product Mif2Go, which is a tool used by approximately 25% of FrameMaker users who switch to DITA.

I do not see the requirements for the tools used to create a freely licensed document, which should be Open Source itself. There are no restrictions on the use of output, which obviously belongs to the user, not Omni Systems.

NTN!

Answer 5

This new set of DITA XProc pipelines on the EMC developer network may be worth exploring. It can be downloaded free of charge for development (and there is an XProc engine).

This package appeared at the end of October 2010. Quote: “The goal of the project is to provide an alternative to the DITA Open Toolkit (DITA-OT), which does not rely on file system-based processing, is not directly dependent on Java and Ant, and uses XProc XML processing capabilities to provide greater flexibility, extensibility, portability and ultimately also better performance. Conveyors use as much of XProc's standard features as possible, so users can use them with any compatible XProc implementation at the lowest cost. Pipelines have been tested with EMC D ocumentum XProc Engine (Calumet) version 1.0.12.

Answer 6

My colleague will just talk to me about the DITA compiler. This is apparently part of the xml mind. According to him, the implementation is not quite completed.