Eight Births

Related link: http://www.jtc1sc34.org/repository/0524.pdf

I celebrated eight births by the end of this May!

First, my own birthday. I went to Sydney's best Japanese restaurant with some friends who happen to have birthdays the successive three days to mine: 18, 19, 20, 21. Apparantly we are all in the third decan of Taurus, however I suspect the decans need re-factoring, because while I indisputably fit most of the profile, the woman with the birthday the next day to me is an always-on organizer of celebrity drag bake-offs for AIDS charities who didn't seem to fit any of it; and the guy in the next position perpetually turns down promotions to stay where he is happy-just last week he turned down a job for twice the salary. On the other hand, another site says that Taureans are good with Capricorn partners, because we are content to maintain the status quo. Err, but that is the opposite of the first site's ancient wisdom,that we have a potent desire to succeed: shades of St Augustine's astral twins! Oh well, at least we are all in tune with pleasure-loving venus, which sounds fine.

Azuma make a chrysanthemum of thinnest raw white fish, then splash it with hot sesame oil and chilli, so that the top side is slightly cooked while the under side is raw, which is a great feat. (For any other Sydneysiders, I think the other best Japanesey food in town is Jazushi for eel pie and Ichiban Boshi for Tokyo ramen in Bondi, at successively lower price breaks. )

The fifth birth is that I finally delivered what I hope is the final draft of ISO Schematron. Daniel Cazzulino, who implemented Schematron.NET, had a couple of late good ideas: allow variables on patterns and phases, so it is much more regular. A PDF version is here. ISO Schematron is really a framework of elements into which you can slot your own query languages, so it does not require you limit your self to XSLT or commit yourself to XQuery. I will put out another blog shortly with some of the features.

The last three births were the releases of three new Topologi products, which I designed. They are really all about reporting as the key to quality in industrial publishing. How do we know that there has not been tag abuse? How do we know that all the graphics have been converted successfully? How do we know how much time a programmer will take to create a stylesheet for a particular document type? How do we know that a new person is marking up documents in the same way as other people are? How do we know that all the instances of some term that is supposed to have been marked up has in fact been marked up? How does a post-technical manager (you know what I mean) verify that documents are in the state that they are supposed to be in? How do you support an efficient division of labour? You get the idea.

In programming, these kind of issues used to be called "programming in the large", but industrial markup is no different: the "markup in the large" issues can be just as vital as the nitty gritty issues of how many keystrokes it takes to enter data. There is a PDF of annotated screenshots here.