Holes in Microsoft Office XML

By way of that other “someone,” Peter Sefton let me know about Simon St. Laurent, his Blog post, “Holes in Microsoft Office XML.” (Hey, Peter, you see how I openly name Blog writers instead of being concerned about fraternizing with possible undesirables that use words like ‘fat chicks’ and ‘strong African features’?) It would be a healthy academic exercise to respond to his points from my ‘unique’ (and clearly unmentionable) perspective:

Simon: WordprocessingML is a horrible mess.

Bryan: Yes, it is. And I agree with your theory about serialization as my theories about why WordprocessingML is so damn hard to read are more conspiratorial, based on ‘discouraging’ developers to parse the tags. They should provide options to export subsets based on the modularization of schema. In the mean time, I have bypassed WordprocessingML entirely and use the XHTML schemas from the W3C to handle my business.

Simon: Word’s XML functionality isn’t integrated with its existing style functionality.

Bryan: Bingo! You win! As you say, this design flaw is what really makes XML functionality in Word a “1.0 release.” Moreover, I would like to see XML-related Find/Change operations in Word—and redesign of Find/Change itself (with especial regard to handling noncontiguous selections through automation).

Simon: Word’s XML functionality has awful interactivity.

Bryan: I’m not sure at what you are getting at here. But my misunderstanding may be due to my relative command over VBA and C# in VSTO. So any “interactivity” solution in Word is just another development project in my queue. I take it as a compliment that I would be considered “non-traditional” in the most positive sense of the word for those job interviewers who have no doubts about my qualifications.

Simon: InfoPath is its own universe.

Bryan: Yes. And I would not be surprised to hear that at least three people of the other Office Dev’ teams at Microsoft do not want InfoPath to succeed. And with regard to XForms, I look forward to any open source solution that rivals what’s cash for Redmond. I have yet to find any and I am still looking. When, say, Mono technology is used to automate a suite like OpenOffice.org, I will sit up take notice.

Simon: .NET and VBA are two different cultures.

Bryan: Yes, but I am almost holistically required to be able to understand multiple cultures deeply. I write for humans and machines with equal respect for both forms. So I can’t just use any word processor. What I will say, in support of your comment, is that the current state of VSTO makes rapid development in Office a joke—especially for us C# coders. I need VBA to develop Office apps primarily because of its edit-and-continue features. When I finish writing this Blog post in Word, I will run my VBA code that will automatically apply XHTML tags (based on styles and formatting) and export this post to the Clipboard as XHTML to be pasted into WordPress. VBA allows me to drill in almost instantly when any bugs pop up in this process. To have Visual Studio open every time I get ready to write a Blog post about ‘strong African features’ or fat chicks is stupid—even for unmentionable me. Eventually my set of known bugs will disappear and I will write the code all over again in C# under VSTO. Sad but true.

Simon: Who needs this again?

Bryan: This kind of question feels to me like the same surface many skim over when they write for the Web without regard for traditional typography. Who needs em dashes and curly quotes? Microsoft made damn sure to pack all of the compelling features billionaire money can buy into Microsoft Word. I still need these features because I consider myself a seasoned writer. Most open source word processors either can’t or won’t pay attention to a myriad of fine details in the writing experience that Microsoft Word—bloated as it is—provides. For example, in AbiWord, the glyphs are not hinted and smoothed so they sit on the screen in what appears to me a jaggy jumble. So does Microsoft make sure that the developers of this product can’t hint and smooth fonts? Or is there something else?

rasx()