Older blog entries for simonstl (starting at number 60)

Land of Abandoned Dreams - Dan York asks "What made people abandon their dreams?" and their farms in upstate NY.

I'm not sure that people precisely abandoned their dreams here - some moved on to follow those dreams someplace else, and in other cases dreams weren't shared across generations. Upstate New York's farms have been emptying out since about 1820, when there were massive migrations to the west. Better land called, and New York's own massive internal improvement, the Erie Canal, reduced the cost of getting goods from those farms to markets in the east.

There are still farms up there, as there are all over the state, but it's not easy to keep them going. Even here in central New York (where the weather and land are slightly more encouraging), figuring out ways to keep farming an active local business is difficult.

There are some bright spots, though. Organic farming has very different economics and is generally less conducive to enormous scale than industrial farming. (Much New York terrain tends to encourage smaller-scale farms.) The higher prices for organic food offer better prospects for some farmers. I also feel like there's more active farming north of the border, though that's just based on what little I've seen from my car. There's no simple answer to it.

If you can't stand XML, it's possible that you just don't like markup, have problems that need other answers (like RDF, perhaps), or just don't like pointy brackets.

It's also possible that markup is a good solution to the problems you have, but the many spiky tools that have been sold as "XML" are driving you away. If that's the case, my piece on "Sane XML" might be helpful.

4 Jan 2004 (updated 4 Jan 2004 at 02:03 UTC) »

It's a new year, and I've been thinking about where to put my energy.

XML, great stuff though I still think it is, is pretty much complete. I think we've learned over the past few years that the stuff that was actually a simplification of existing practice was good, and the rest should be used cautiously or (in the case of W3C XML Schema and specs it's infiltrating) ignored. The RELAX NG folks have created a sane schema language, the interesting action in the space has largely moved away from the W3C, and we're now at the point where everyone can create whatever vocabulary they like.

Not that they create particularly good vocabularies, especially if they focus on W3C XML Schema as the path to new vocabularies, but there's only so much I can do to keep people from banging their heads against the wall.

So if XML is no longer my main technical focus, what's next?

I have two main areas of interest at the moment, one even more abstract than XML and one more (well, mostly) concrete.

On the abstract side, I've been reading Christopher Alexander's The Timeless Way of Building and A Pattern Language, as well as the first volume of The Nature of Order. It's a lot more exciting than the Gang of Four Design Patterns work that derives from it, and I think I'll write a lot this year on how programmers can learn to do better work by taking aesthetics seriously, on a lot of different levels.

On the more concrete side, I've been looking into mapping and computing. I've been writing a blog that focuses strictly on one 96 square mile town, and maps are an important part of that, especially given the planning process that's currently in motion here. It's been interesting to see how most of the road network was in place by 1900, but a few key changes have had dramatic impact. I'm also sorting through the avalanche of census data that's available from the 2000 census, examining it both through GIS tools and through databases.

Between those two things (and a healthy continuing dose of XML, I'm sure), it should be a good 2004.

I just sealed my driveway. Two days of cleaning, prep work, and finally applying the sealer.

I should be so careful in my programming.

(Now let's just hope it doesn't rain before the sealer sets.)

the web is no good unless it can be a sound foundation for the semantic web and web services too.

Gah. And my house's foundation is no good unless it can be a foundation for a skyscraper and a gas station too. Comments like these from people I think should know better occasionally drive me to rant.

In more exciting news, I'll be ranting next week at the Extreme Markup Languages conference in Montreal. It's always been one of my favorite shows, and this year they gave me a Daily Polemic. The original version of What can you do with half a parser? is pretty good (I think), but for the polemic it's getting a cast of Playmobil figures and some toxic black goo.

18 Jul 2003 (updated 18 Jul 2003 at 02:21 UTC) »

Last week was a blur of two conferences - OSCON and the Sells Brothers Applied XML Conference, both in the Portland, Oregon area. The two conferences were very different for a few fundamental reasons:

  • OSCON is completely dominated by open source cultures and values, despite the Microsoft-paid lunch, while Applied XML was largely a .NET fest, with a host presently employed by Microsoft and lots of Microsoft-oriented content.

  • While OSCON had an largely tutorial XML track (which I thought went very well, though I'm biased as I chaired it), Applied XML was a single-track conference where every session had something to do with, and generally focused on, XML.

Applied XML was smaller and more tightly focused, but it had a similar energy level to OSCON, at least in the hallway conversations. OSCON was amazing as usual for bringing together developers from a variety of different communities, and letting them explore sessions they found interesting. A lot of tracks benefited from crossovers between different kinds of developers.

One especially interesting bit was the crossover between the conferences. There was a BOF session Wednesday night on what it would take to implement dynamic languages (Perl, Python, Ruby, etc.) efficiently in the .NET environment, which was built with statically typed languages (think C#). That conversation's probably just getting started, but there were some fascinating bits. Peter Drayton and Brad Merrill of Microsoft, who hosted the BOF, were at both shows, and seemed happy in both contexts.

(There was also a surprising amount of RDF in conversations at both conferences, continuing a trend I noticed at OSCOM. It wasn't on either conference program, but it was in the bars and hallways.)

Though I'm really tired of traveling, I'm looking forward to one last conference this summer - one that combines the small size and XML focus of the Applied XML conference with the community approach of OSCON. Extreme Markup Languages, running from August 4-8, has seen fit to give me a Daily Polemic. I suppose it's appropriate given my usual style, but it's still a pretty scary responsibility.

To live up to that challenge, I've turned to the wonderful world of Playmobil. Playmobil figures make excellent computer consultants, especially now that Playmobil offers office equipment. I'm hoping to use SMIL for the slides, though maybe it'll be SVG. We'll see.

I'll be taking next week off - I decided it was time to spend some time working on my own projects and remember why I like doing this stuff. (Maybe I won't - we'll see!) In addition to the Playmobil photo shoot, I'm hoping to put some work in on my Java tools for processing XML. I haven't been able to spend more than one day at a time on them lately, with months between sessions.

A week of training

I spent last week in Ottawa, taking five days of training from Ken Holman on XSLT and XSL-FO. I had plenty of job-related reasons to take the training, from growing use of XSLT in my work to re-connecting with users in the field, but I'll admit it was a pleasant luxury to take a week to look at a complex technology up close.

I blogged each day over on my O'Reilly blog (1 2 3 4 5), but it's interesting to look back at the course as a whole. Three days of XSLT and two days of XSL-FO is a lot, but even that's kind of a forced march through the technologies, really only scratching the surface. Ken did a great job of covering the overall story, making us think our way through exercises on key features, and pointing out potential problems. Actually having worked through the details (and coming back with a handout covered in post-it notes) should help me remember what I did.

I haven't taken a formal training course in years, though. I've been mostly self-taught, with the occasional conference tutoral supplementing books, specs, and email. Face-to-face has some huge advantages, I have to say. Immediate interactions and immediate gratification are wonderful, as is having a pre-built set of exercises really intended for this kind of back-and-forth interaction.

I already knew (and heck, disliked) a good deal of XSLT before I went in, but this was an opportunity to check and expand on what I knew. Fortunately, the course was structured so that different users could get what they needed - including those of us who'd already been over a lot of it. (The XSL-FO class included at least one attendee who knew a lot more than I did about any of the material.)

Training isn't always an option for everyone, but after the past week, I'm thinking it's something I'll be encouraging a lot more.

I seem to have a lot of conferences on my schedule at the moment, just when I'm busiest with "real work". In any case, if anyone's interested in XML-related conferences, I wrote up what's on my radar. I may or may not make it to everything here, but suspect there may be a few folks here who will.

Content-rich keynote presentations seem to be pretty rare, though they've improved some since the bubble faded. I was lucky last week to go to a conference (XML Europe) which had two excellent keynotes back to back, both exploring the synergy of open source and XML.

I've written both of them up for xmlhack. Jon Bosak explored how he hopes open standards and open source can work together, while Daniel Veillard examined how XML is used by open source projects. They were very different presentations, kind of like looking through opposite ends of a telescope, but they complemented each other beautifully.

I've been thinking a lot lately about computing cultures. XML culture, for instance, feels very different from Java culture. Though I do most of my programming in Java, the work I do leads me into creating XML-oriented interfaces that are far removed from the suggestions in Effective Java, for instance.

While I program in Java, I don't think I'm part of Java culture - I even find some aspects of it profoundly disturbing. I've concluded over time that Python is probably a more appropriate medium for what I want to do, but I've got all this easily-mined work in Java...

I think similar issues arise in information modeling and storage. I wrote a short piece on it yesterday, "The (data) medium is the message". The bit I quoted from McLuhan, which I think is pretty much at the heart of the matter, is:

"Environments are not passive wrappings but active processes."

Programmers tend to think of ourselves as active and the environments we program as passive, but it's definitely a two-way street, even before you get into the environment-changing possibilities of open source.

51 older entries...

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!