[dita] Summary of Key Scoping Telcon

["Nitchie, Chris" <cnitchie@ptc.com>]

Here's a summary of the key scope discussion between Kris Eberlein, Eliot Kimber, Michael Priestly, Dave Helfinstine (who was mostly silent), and myself.

The primary focus of the discussion was around the relative merits of adding the ability to explicitly reference a defining scope when referencing a key, vs. requiring all inter-scope key references to have their own unique, global key. Michael took an action item to provide an illustration of the former, and the latter was illustrated here: http://lists.oasis-open.org/archives/dita/201201/msg00058.html.

We all agreed that there was merit to both approaches, but we differed on how to measure each ones' costs. Citing real-world examples from his own experience, Michael worried that the requirement for unique global keys for cross-scope key references would require an unworkably onerous workload for some map authors/publication managers. Eliot and I worried that adding key scope addressing, with all the additional provisions that might be needed to make it work, would add too much complexity into the spec, and suggested that some of the work that Michael is worried about could be managed by tools. Michael was unconvinced that it was appropriate to offload obvious work saving provisions to tools. And that's more or less how we left things.

In addition to providing an example of how scope-specifying key references might look, Michael took an action item to ask around at IBM about whether his concerns were well-founded. Eliot is going to spend some time thinking through some of the complexity issues related to key space addressing to see if our concerns are well-founded.

Chris Nitchie
Senior Software Engineer

T 734.352.2879   F 734.997.0201
E cnitchie@ptc.com

PTC.com