Christopher.J.Demattio at seagate.com wrote: > > I am currently evaluating DITA as a possible toolset for some of our > documentation. One showstopper that I see with DITA is that it is very > difficult to produce output that shows changes from one version of the > document to another. It is possible, but requires a lot of manual markup > and/or automated script support. > > Obviously, word processing programs, such as Word, have long ago solved > this problem.
This problem must be solved at the XML source level, not at the .docx deliverable level. What if you want a PDF output with change marks? You'll first convert the XML document to .docx and then use a ``PDF printer''? If you just want to *see* the changes, you'll find a number of tools which does this. Example: http://sourceforge.net/projects/diffmk/ If you want to create a XML document where some markup have been added to mark the changes, (at least) DeltaXML -- http://www.deltaxml.com/ -- does this. (Disclaimer: we have not tested any of the above tools. We just trust their authors.) > That's why I was excited to see that your company had > produced a DITA Converter that supports Word 2007 formats. I thought I > would be able to use your converter to generate a 'before' and 'after' > version of the document in docx format and, then, use Word's document > comparison utility to see the differences between the two versions. > > Unfortunately, when I try to compare 'before' and 'after' versions of > some ditac output in Word (using Word 2007's "Compare" function), I get > the following error message: "A table in this document has become > corrupted" and the comparison fails. > > Thinking it might be a problem with my DITA content, I tried the same > experiment using your ditac manual. Here's my experiment: > > 1) Create a version of the ditac manual with no filtering enabled: > "C:\Data\Downloads\DITAConverter_XMLmind\ditac-1_0_0>bin\ditac.bat -toc > -xfc ..\xfc_perso_java-4_3_1\bin\fo2docx.bat out/manual_unfiltered.docx > docsrc\manual\manual.ditamap" > 2) Create another version of the ditac manual with filtering enabled: > "C:\Data\Downloads\DITAConverter_XMLmind\ditac-1_0_0>bin\ditac.bat > -filter docsrc\manual\sreen.ditaval -toc -xf > c ..\xfc_perso_java-4_3_1\bin\fo2docx.bat out/manual_unfiltered.docx > docsrc\manual\manual.ditamap" > > Open both versions of the document in Word 2007 and use Word's Compare > function (not the "Merge") function. During the comparison, I get the > above error message. > > I can recreate this same error with other DITA files that I have > created, even though the DITA files have no tables. Do you have any idea > what might be causing this error? > > Here is the requested debugging info: > The version of ditac. (Execute "ditac -version" to display this version.) > ditac: INFO: ditac version is 1.0.0 > > The name and version of the OS used to run ditac. > Windows XP Professional, Tablet Edition, Service Pack 3 > > The version of the JavaTM runtime used to run ditac. (Execute "java > -version" to display this version.) > java version "1.6.0_14" > Java(TM) SE Runtime Environment (build 1.6.0_14-b08) > Java HotSpot(TM) Client VM (build 14.0-b16, mixed mode, sharing) > > The exact ditac command line which allows to reproduce the problem. > See above > > If relevant, all the input files (DITA files, image files, etc) needed > to run the above command line. > The input files in my example are the DITA source files distributed with > the DITA converter. > > The output files, if such files contain what you consider to be > defects./(See attached file: manual_unfiltered.docx)//(See attached > file: manual_filtered.docx)/ > Your problem is not related to ditac, but rather to our XMLmind XSL-FO Converter product -- http://www.xmlmind.com/foconverter/. You say: [1] I can open manual_unfiltered.docx in Word 2007. [2] I can open manual_filtered.docx in Word 2007. [3] When I use the Word 2007 Compare function, I get error message "A table in this document has become corrupted". For us, the conclusion is simple: the Word 2007 Compare function has a bug (e.g. it depends on data or ``idioms'' which are generated by Word and no other tool, which would be 100% acceptable if Office Open XML were not an ISO standard!). The goal of XMLmind XSL-FO Converter is pretty modest: [1] allow a user to open in MS Word documents generated out of an XML source (may be because the user prefers to work with .docx files rather than with PDF files). [2] In a minority of cases, allow the user to modify using MS Word documents generated out of an XML source. No less, no more.
