Push Button Paradise

I made live a few minor tweaks to the examploforms stylesheet: the XHTML2 namespace, minimal layout, and correctly populating the itmes inside the select1 controls. Enjoy: http://examploforms.org -m

If you point your browser at http://examploforms.org you'll find the first public release of Examploforms. Think of Schematron and Examplotron, applied to XForms.

What motivates the development of a tool like this? You can think of Examploforms as the antithesis of graphical drag and drop designers. There are two ends to the spectrum: really complicated, data-intensive forms can be painful to work with in a touchy-feely WYSIWYG tool. On the other side, sometimes in really simple forms modeling and getting the back-end integration is the bigger job; actually designing the form is essentially trivial. In both of those cases, Examploforms can help.

This initial release focuses on ideas and technology, not on optimization or making the output look beautiful. (By the way, any stylesheet designers out there who would like to contribute some CSS to make the generated forms look nicer--I'm happy to accept contributions) It includes an XSLT stylesheet that transforms an Examploforms document into XHTML+XForms.

In includes a few useful datatype libraries/naming conventions that prove highly useful for form design, and will look familiar to regular readers of my blog.

I've set up a mailing list for questions and comments. Details within the specification document. -m

Things I hate about register.com:

1) Clueless tech support

2) Expiring-domain-warning message consistently get flagged as spam

3) Mind-numbingly-dense tech support

4) Their "Renewal Manager" has terrible support for .info

5) Tech support personnel dumber than a bag of hammers

6) Expensive

7) Stupefyingly obtuse, deci-witted, turing-challenged tech support staff

So I'm bitter. But I'm also switching away ASAP. -m

My home page of http://dubinko.info was embarrasingly out of date. Fixed. -m

I'm getting sore thumbs from excessive trackpad use, so I ordered a mischief (or "hoard" or "whole bunch of", depending on your upbringing) of optical mice. Might as well get rid of all these clunky, dust-attracting mechanical things.

Logitech has a nice model for about 11 bucks. No more do you have to choose between separate 'mouse port' vs. USB models. They come with a USB connector, and an adaptor to the round mouse connector. (The electronics for this has to be onboard the mouse itself, because the adapter is way too small to be a general-purpose USB converter.)

On OS X and Windows, it was literally plug-and-play, not even a click needed. On my SUSE 9.1 server, though, I was surprised that it didn't Just Work. It detected it allright. It lights up, and shows up as "optical mouse" in the device list. It just doesn't have any influence on moving the pointer or actually registering a click. Sigh. Link: http://www.amazon.com/exec/obidos/tg/detail/-/B0000659A7/dubinkoinfo-20 -m

Some interesting comments posted anonymously on Groklaw. I want this in my personal, searchable archive, so here it is...

About the cost of ownership stuff. TCO should include cost of exit - how much does it cost you to junk the software when it's obsolete and switch to a different vendor. That's one of the most important factors but is usually ignored.

The textbook economic theory of software pricing is as follows:

The marginal cost of production is zero (more or less) - it costs nothing at all to burn another CD. So the price in a perfectly competitive market will be zero, and in a real market determined by the degree to which the market fails to be competitive - primarily the incidental costs incurred by a purchaser switching from one vendor to another.

So the major factor in predicting your future software costs is the cost of exit - how much will you have to fork out to stop using your current software to switch to something else.

Windows has high exit costs. Proprietary protocols, file formats everything. Breaking the M$ addiction means going cold turkey to a completely unrelated system.

Not satisfied with your Linux vendor? Start installing a different distro on new hardware, migrate your data as usual - cost of switching is minimal - properly planned, it's just part of your usual upgrade cycle.

A good argument for standards. -m

Lots of "congrats" messages. One bite on actual new work (not bad for day 0). The most interesting part, though, was several bits of information (or an offer for more information under NDA). I love getting little pieces like this. Keeps the challenge of assembling the overall-strategy puzzle more interesting. :) -m

The amazing Ben Nolan sent me the source for the nForms XForms engine (another one of my Ten Favorites), under a GPL-license. It's smaller than I would have thought--about 400 lines of XSLT and 1000 lines of JavaScript. I'm putting together a small group to help me decide what to do with this, possibly including further development. Contact me if you're interested. -m

If you like the crisp look of http://brainattic.info you have Derek Hansen to thank. He not only came up with the excellent basic design (as seen on CSS Zen Garden), but also gave me permission to base my web site, with new images and minor tweaks, on it. Thanks! To see more of his work, follow the links at the bottom of the Brain Attic page. -m

Article out on IBM developerWorks. Full link: http://www-106.ibm.com/developerworks/library/x-xfvalid.html This is a new approach to opening up code...

I got the idea from my excellent laywer. He has a nice set of web pages up that give detailed instructions on how to form a corporation, limited liability company, nonprofit, etc. in Arizona. It would be possible for someone who didn't value their own time or just wanted an intense, hands-on learning experience, to follow the instructions and get everything done eventually. Or you can pay a small, fixed price and have the professional take care of it. Laywer link: http://www.keytlaw.com/az/formllc.htm

I have taken a similar approach with the XForms Validator. If you look in this blog, or my slides for the XForms Tutorial, or my book, or this new article, you'll find all kinds of technical details on the validator; enough to roll your own validator if you have tons of free time or are looking for an intense learning (and testing) experience. Or you can pay a small, fixed price an have a professional take care of it.

No matter what happens, this will be an interesting experiment in combining business and open source. -m

Basically, I've started my own consulting company. I'm now available to help you with projects large and small, especially if they deal with XForms/web forms, standards, open source, or XML. The marketing message (i.e. what kind of impression you want to make on people who haven't otherwise heard of you) revolves around managing information overload. Link: http://brainattic.info

The name, of course, comes from the great Sir Arthur Conan Doyle, specifically from A Study in Scarlet, and has ties all the way back to one of my definitive blog posts. Lots has changed, however, since then.

For one, I'm not quite pumped towards XML. It certainly is useful in the right places, but then again the same can be said for non-XML. XML is a tool, not a religion.

Also, the wonderful Chandler project from OSAF has made lots of steady progress. I expect this to become a major component in computing, and I plan to be ready for it.

So overall, I'm going into this without any key software projects in play. I plan to contribute significantly to at least one open source project, and perhaps start at least one, but on day zero this is completely a consulting path I'm on.

So, contact me and let's get started. -m

I'll be travelling for the next week or thereabouts, mostly without connectivity. I promise interesting news upon my return. -m

The Keeping Things Found project from University of Washington has found that, not too surprisingly, electronic notes and records are easier to manage and find things in than paper. Link: http://kftf.ischool.washington.edu/index.asp

It gets worse. According to a study, at least half of the participants admit to losing a paper document at least once per week. So it's not just a matter of paper taking slower, manual searches. It's a matter of finding your old info or not; a concrete difference in value of your accrued notes and snippets.

There's still work to be done, though: Just 10 percent of those surveyed were extremely satisfied with their ability to keep track of computerized records.

Research link: http://www.uwnews.org/article.asp?articleID=5405 -m

Often in forms, instead of some particular data type, the value allowed somewhere is just a simple enumeration. XML Schema has a way to define these, but often you want something shorter, without having a pile of simpleType restrictions lying around. Thus, the dt:enum family of datatypes.

This works only for enum values that are simple tokens, specifically those that conform to the xforms:listItem datatype. The datatypes are defined as the string "enum", a hypen, and a hypen-separated list of tokens. Thus, if the allowd values of element content are "full", "compact", and "minimal", the datatype definition would be dt:enum-full-compact-minimal. I haven't yet defined a namespace for this library.

I need to decide whether this is preferable, in an examplotron context, to just including each possible elements, and using eg:occurs="-" on all but the first. -m

Nice little bit of JavaScript from Daniel Glazman. I wonder how many people are running with JavaScript disabled these days... (I use server-side PHP to accomplish the same thing on some of my sites) Link: http://disruptive-innovations.com/zoo/20040830/HTMLoverlays.html -m

The SVG Working Group has just released the first public draft of an officially-standardized XBL. http://www.w3.org/TR/2004/WD-sXBL-20040901/ This first draft claims to be close to a 'last call', so now would be a good time to get busy reading and commenting on this. -m