Round-Tripping versus Single-Source

[orcmid]

Document Engineering projects seem to be more amenable to single-source publishing in multiple formats than round-tripping of (editable) documents via an editable intermediary that exercise provisions of a platform like GitHub.

If the intermediary is a vehicle for single-source publishing, then having it in the middle of a round-tripping arrangement might be effective. The challenge there is that this becomes an editor and production case and that writers might not want anything to do with it. There is also the matter of making translations between natural languages and going from documents to on-line flavors, such as Help systems.

Twenty years ago, there was considerable interest in single-source of on-line help and documentation for applications delivered on Microsoft Windows. This was somehow aggravated by the fact that Microsoft was evolving from one application help system to another, and the means of using them outside of Microsoft was diminishing. I notice that the top-of-the heap single-source products and their producers in that era seem to have vanished.

We will make some round-trip cases in OpenDocument Format (ODF) and in Office Open XML Format (OOXML). This will identify feasibility and also the constraints under which othering must adhere to in relying on software that delivers those formats. The first choice intermediary is the pandoc "native" derivative of Markdown. An additional challenge is how this might not be successfully viewable as GitHub Markdown :) or whether it is close enough.

What are your experiences and considerations over these cases, especially with regard to roles and how well authors are supported?