Pushing String

In making the case for the Universial Business Language (UBL) over the last few years, Jon Bosak has often talked about the potential of modern B2B technologies to empower small businesses — as in really small, like a mom-and-pop shop in a developing country making fabrics that could be used in automobile manufacture. If the biggest technical barriers to entry are removed and the power of simple, universal, powerful technologies such as XML (with vocabularies that have legally rigorous semantics), HTTP/SMTP, and so on can be harnessed, even the smaller and less influential suppliers can play. And that can make their lives dramatically better.

We’re not all the way there yet, but barriers continue to fall. Last year I came across a cool website called Worldstock.com, an offshoot of Overstock.com. Worldstock specializes in “handcrafted goods from around the world”; to quote from their self-description:

Around the world there are artisans who know how to make exquisite centerpiece items. Yet they have trouble accessing the US market because they are small-lot producers in an age of mass distribution. Often there is no way to get goods from their remote villages to here, and when there is, too many layers of mark-up make them unaffordable. The tragedy is that if we bought their goods, the artisans could prosper without abandoning their native crafts and culture, and without depending on charity.

Overstock’s main business is bringing small lots to consumers at affordable prices. Three years ago we realized that that this capacity is exactly what artisans need. Thus was born Worldstock.

(I don’t know whether XML is involved in making this site a success, or whether individual artisans are using web forms that submit UBL-like purchase order responses to the Worldstock servers! But it’s neat to imagine. And use of such technologies could increase the opportunities for more small businesses — and more retailers. On the other hand, before the full flowering of the Bosak B2B vision, small-lot producers would probably have to increase the consistency of their output quite a bit, such that goods from multiple sources all conformed with tiny tolerances to really detailed specs. And there would probably still be a need for jobbers, which perhaps Worldstock-like companies could fill.)

As one example of what Worldstock has accomplished, it’s helping many hundreds of women in Afghanistan to make a living and feed themselves and their families. Their goods can be browsed by country of origin; in addition to Afghanistan, they offer items from Bali, Brazil, China, Egypt, India, Indonesia, Ghana, Iran, Israel, Lebanon, Mexico, Morocco, Nepal, Pakistan, Paraguay, Peru, Thailand, and Tibet.

Watching the tragic tsunami news from southeast Asia, I was put in mind of Worldstock. In addition to donating money and other resources to help the survivors, consider supporting businesspeople in the affected countries. Economic growth fueled by such purchases could, indeed, help prepare them for future catastrophic events by making them more resilient.

One of the LRF units on my laptop just gave out. Actually, it fell out, which was quite disconcerting. It must be time to get a new laptop.

All right, all right, I’m panicking over nothing. Surely LRF support is a minor feature and I can use the laptop just fine if one of them is missing.

I’ll just have to use this as a learning opportunity (what my old boss and mentor calls “AFGO”, for “another growth opportunity”).

But I still want a new laptop.

While shopping in the Broad Canvas crafts store in Oxford in July 2003, Lauren Wood and I cooked up the idea to do some stitching projects for the XML conference that year. While I was grateful to see that the selection of ready-made kits in the store wasn’t limited to pictures of puppies and kittens doing unbearably cute things — I ended up buying a kit for stitching Stonehenge — I couldn’t help noticing the relative paucity of, shall we say, technically oriented patterns. We had to design our own.

My XML 2003 artwork exhibit entry: The Infoset is a Unicorn - It doesn’t exist

I ended up acquiring a program called PCStitch to put together my first design, and I’ve used it for several others since.

Snippet of the PCStitch pattern for Infoset Unicorn

Now, when I say “design,” I’m really talking more about piecing together elements found elsewhere than invention from the ground up, though sometimes I do need to do work from scratch. There are quite a few free patterns online, sometimes in PCStitch format or the format of another program, but often in badly scanned PDF form. (We need XSML — Cross-Stitch Markup Language…)

I usually input raw stitch data into PCStitch by hand because I need to assemble a whole bunch of pieces from disparate sources, play around with colors, print out large-size color versions for the tricky bits, etc. The unicorn head came from a free pattern that was a bit more extensive (and very hard to read in its native form).

I found the “fancy alphabet” used here online as well. The input process pays off particularly well with alphabets, which PCStitch can treat as a cool “stitched font” that lets you literally type your text into the design, center it, and so on. Victorian sampler-stitching maidens never had it so good.

Of course, I have to design nonalphanumerics such as ampersands, angle brackets, and semicolons myself since they’re never included. (I wonder why that is?)

Who am I and why am I here?

I was an old-time SGMLer who specialized in information modeling, and was one of the folks whom Jon Bosak invited to be involved in the creation of a technology we eventually called XML. I’m writing here for the same reason I co-wrote a book once about SGML DTDs (more about that anon, probably). No, not for the oodles of money or fame… Sometimes I discover that I have something to say, and saying it seems a better idea than the alternative. The profundity quotient will, no doubt, vary quite a bit.

My current day job is the care and feeding of various identity/web services initiatives on behalf of Sun Microsystems, and I’m involved in the OASIS SAML committee and the Liberty Alliance, among others. (N.B.: The opinions expressed here are my own, and neither Sun nor any other party necessarily agrees with them.)

What kind of musings will you find here? This may be the world’s first (and last?) XML-and-cross-stitching blog. For the last two years I have been designing my own stitching patterns with XML themes in order to display the results at the North American XML conference series’ artwork exhibits. I run the exhibit just to show the XML world that structured markup geeks have many talents, or at least interests…

Since I had designed a rather large project for the 2004 conference, I snuck in a lot of stitching during Liberty Alliance standards meetings. At some point Nick Ragouzis caught me in the act. His encouragement to blog about this bizarre hybrid habit really stuck with me.

Others who egged me on, by way of thanks: Tim Bray really turned on the pressure. Lauren Wood offered advice on stylistic matters. Norm Walsh suggested a number of topics (which may surface here someday). And Ben Hyde not only encouraged me to blog, but suggested a design for one of my eventual stitching projects and even blogged about it when it was done. And my husband Eli Israel offered lots of support, both IT and editorial in nature.

Why Pushing String? While stitching away during another lecturer’s presentation on XML schema design at CSW’s wonderful XML Summer School in Oxford, U.K., I commented that some aspect or other of choosing a schema design style was really difficult, “like pushing string.” Looking down, I realized that that’s what cross-stitching is all about.

It turns out to be easy — if your implement is pointy enough.

UPDATE: The post above has more thoughts on the many meanings of “pushing string.”