dita-lightweight-dita — archive

[Stanley Doherty]

Greetings -- I had hoped to get these testing results to you all by Friday, but I ran into some new-job orientation tasks that required attention. I have completed the testing that I had planned and posted results to my GitGub repo https://github.com/StanDoherty/LightweightDITATesting . Feel free to clone and test it for yourself. If you have updates/fixes, please let me know and I'll add you as a contributor. I have to add some additional detail to the testing report, but otherwise I am done. Here are a few highlights: 1. When we remove HDITA and PDF builds from the text matrix that I worked on last year, we are left with 24 tests that exercise xrefs, keyrefs, and transcludes between XDITA and MDITA. 18 of the 24 tests pass. The four that fail involve conditions and MDITA conref validation. This is pretty good news for anyone wanting to exercise interoperability between XDITA and MDITA. 2. Block elements: I tested each block element against Oxygen validation (XDITA only) and DITA-OT HTML5 build fidelity (XDITA and MDITA). With 35 elements, the test matrix involved 105 test cases. 33 of 35 validate in Oxygen XDITA. 20 XDITA elements build successfully in DITA-OT HTML5; the failures are mostly in the multimedia elements. To the extent that there is not a 1:1 correspondence between XDITA and MDITA elements, there are numerous "Not Applicable" test cases. 3. Inline elements: Other than superscript, subscript, and phrase, XDITA and MDITA maintain parity. 4. XDITA: The more I work with it, the less like it. If there are no commercial applications that read/write XDITA, then I assume that 100% of the people working with XDITA already know DITA 1.3. Stan