I’ve been poking around the Web looking for GPS gadgets that connect to a PDA - Palm, iPAQ, Zaurus, etc. Most of what I’m finding is GPS with piles of navigation software that’s actually not that interesting to me.

Anyone know a good place for info on GPS hardware for PDAs with programming in mind? Java programming would be ideal, but I can find my way through other environments if I need to do so.

Basically I’d like to write a program that starts by establishing a location through GPS, and then has the user record a variety of information about objects (mostly trees) within about a 60′ radius. I’ll be taking that information and turning it into SVG later to produce a simple map, as well as storing the information for later use. Eventually it might be nice to use the GPS to collect the locations of the objects, but it’s still a simple function call asking “where am I?”

This isn’t GIS rocket science by any means, but I also don’t have much of a budget, and the GPS unit has to be able to do a decent job under forest canopy, another big problem for navigation-oriented units.

All suggestions are welcome.

Any ideas?

Related link: http://www.cnn.com/2002/TECH/ptech/06/24/diagnosing.cars.ap/index.html

An Associated Press story reports that car manufacturers are keeping repair codes a secret, cutting independent mechanics out of the loop.

Here I thought Sony’s repair policies for PlayStation 2 systems (mail back to manufacturer) were annoying, but it’s depressing to hear that this kind of business brilliance is spreading to other industries that traditionally had a more open approach.

[Slashdot also has a story, which links to a Minneapolis Star-Tribune piece with more details.]

What’s next? My lawnmower?

You’ve found yourself an open Wireless Lan - and you want to share the knowledge…do you slap it on the web?
Well, that’s a good idea, except that to find the node, you have to be on the net in the first place. Not much good if you’re just off a plane, in a strange city and need to grab your mail.

So why not WarChalk - Matt Jones’ hobo-inspired idea for spreading the wireless word.

You find a node, and draw the correct symbol on a nearby piece of public furniture - a wall, the pavement, the side of a lamppost. Anyone knowing in the ways of the WarChalking will recognise what it means, and get online. No more wandering around bandwidthless, and no more struggling with online maps.

Genius.

Is this genius, or what?

Sudden moments of realisation are what many of us live for. Give me a fizzle of my frontal lobe every few hours, and I’ll be yours forever: ideas, inspiration, moments of flow, of poetry, of elegance, whether in our work or in our play, are what gives many people a reason to live.

(It may actually be, and I digress at this point, the reason that many of us congregate here: the thrill of a elegantly rendered hack is so much more potent than simple bread and circuses. This is your brain on drugs, this is *mine* on Knuth)

So, anyway, you find me on the come-down of a idea-hit. My wife and I have decided to emigrate. Ok, so that might seem a little inwardly thrilling (I mean you’re thinking, well done Ben, but why should I care, right?), but bear with me.

You see, I just realised that no one - at least no one who pays me a wage - has any real idea where I am. I’ve never met my editor here at O’Reilly, Simon St Laurent, but given that I’m English he could possibly guess to within a few hundred miles, and my boss at The Guardian could perhaps narrow it down to within 10, but at the end of the day, my address is my URL, my email and my Instant Messenger accounts. Add in a mobile phone, and I’ve worked seamlessly from Tehran to Rangoon with no one noticing.

So, we’re off - we finally decided today - and we’re moving to Florence, Italy. For the price of living in London, we can double the size of our appartment, and get a balcony over looking the same view as this live cam, and as long as I’ve got bandwidth - no one would ever know the difference.

Don’t get me wrong, this is not some smugness in the making. This whole idea fascinates me - perhaps, for the first time, after years of prophecy, we can now truly declare the death of distance.

Is this madness, or can one really work from anywhere with a dial tone?

Related link: http://www.xml.com/pub/a/2002/06/19/vdv-wxs.html

While this article by Eric van der Vlist doesn’t explicitly discuss SOAP, it indirectly addresses an important problem with SOAP today: SOAP RPC Encoding is misaligned with XML Schema. SOAP Encoding provides an object-graph view of data, which allows for references but doesn’t provide strict typing for complex types. XML Schema, on the other hand, provides a tree view which doesn’t support references but does provide strict typing for both simple and complex types. Depending on the data model used (Encoding vs. XML Schema) our view of the data changes. This is an issue I’m grappling with while writing my next book on J2EE Web Services. Anyway, I recommend that you read the article, which doesn’t discuss the SOAP issue just mentioned but does discuss how schemas influence our perceptions of data.

Related link: http://www.theserverside.com/resources/article.jsp?l=RMI-IIOP

I have always held IIOP interoperability in Enterprise JavaBeans suspect. Having worked with CORBA before EJB, I know firsthand how difficult it is to make two different CORBA platforms interoperate. This article by Billy Newport exposes some of the problems with Java RMI-IIOP and explains why IIOP interoperability is difficult.

Related link: http://www.learnthenet.com/english/html/70alan.htm

This web page on LearnTheNet.com is like a quick guide to networking and Internet technologies. Its fun and easy to read.

It’s somewhat LAN centric, which is actually a refreshing change of perspective. Reading it reminded me of a stint I did at Northwestern Mutual Life in the early 90’s working with their Networking Team. At that time Novell was king of business networking, and we were only beginning to experiment with TCP/IP as an alternative to Novell’s native protocol. The World Wide Web had just been invented and was unknown to most of the world. How times have changed.

I’ve been doing some volunteer work for a forestry group whose main office is effectively off-the-grid. There’s a phone line, but that’s it. They have an ancient diesel generator, but while it’s effective for circular saws, I don’t think running a computer off it would be wise or especially convenient. It’s louder than the saws.

The inside of the building is lit with solar power, generated by a small array of panels. Right now there are only four 12V car batteries for storage, but that’s been sufficient for lighting. This place is in upstate NY, which is notoriously cloudy.

They’re hoping to put a computer there, and I’ve suggested an inkjet printer, possibly with a scanner/copier all-in-one thing. Laser printers are much hungrier for power, so hopefully that’ll avoid a problem.

My larger question is the computer. They have an ancient Gateway sitting in a corner, but if my past experience is at all accurate, it’ll be very hungry for power, as will the monitor for it. An LCD might help, but I’m also wondering more generally about the computer.

Anyone have suggestions for energy-efficient computing? Laptops seem like the place to be in general, and most of the solar info I can find is panels for laptops. Something that runs on 12V would be nice, but most of what I find in my searches for 12V is specialty auto/marine stuff that isn’t cheap. Cheap inverters are also not especially efficient.

Any thoughts on how to run computers off the sun?

Related link: http://www.devx.com/free/press/2000/110100.asp

I was surprised to discover that the XML Schema date types are tied to a single calendar system, the Gregorian (Western) calendar. Its funny that XML Schema uses an international character set, Unicode, but not an international calendar system. Although Java and C# both use Gregorian as their default, they also support other calendar systems. The article was written in November of 2000, so I contacted the author, Lee Anne Phillips, to discover if plans for the next version of XML Schema included support for an international calendar system. She said no. Apparently, people can’t agree on how to implement an international calendar system so its been shelved. Just so you know, people do use calendars other than the Gregorian. While its true that calendars are not as varied as languages, they deserve the same consideration as non-English languages. Its not a matter of being politically correct, it’s simply logical and practical.

Related link: http://www.sys-con.com/xml/article.cfm?id=423

While I’ve criticized (some would say whined about) the quality of the W3C XML Schema specifications for a long while, it seems like a lot more dissent is rising to the surface lately, especially this week.

Tom Gaven’s Desperately Seeking… Help for XML Schema article arrived in the XML-Journal that appeared in my mailbox yesterday. It starts with a bang:

“There’s been a lot of interest in XML Schema, and well there should be. A lot of very smart people put the XML Schema spec together, and I’m sure it took an amazing amount of effort. But if I can be frank for a moment, XML Schema is rocket science. From noNamespaceSchemaLocation to block attributes and substitutionGroups, the XML Schema syntax is simply too complex….STOP the MADNESS and just say NO!“

Gaven goes on to explain how complex type inheritance is broken, detailing one small but critical piece.

Meanwhile, on the ietf-xml-use mailing list, James Clark explained at length why RELAX NG is preferable to W3C XML Schema, noting that:

“There seems to be a tendency for people to suspend their technical judgment when it comes to W3C XML Schema. The attitude seems to be “It’s a W3C Recommendation; everybody is using it, so we should too, regardless of its technical merits.” I don’t think this
attitude serves the best long-term interests of the Internet.“

On the xml-dev mailing list, Clark’s comments led to an outpouring of mostly similar comments (Threads: 1 2 3 4 5). The xml-dev discussion also appears to have inspired Dorothea Salo to write a blog entry called “Maybe I’m Not Stupid,” reflecting that:

“Confession: I don’t get XML Schema. I just don’t get it. When one is shoved under my nose, I can kinda-sorta follow what’s happening, but to get a real idea I have to have a reference handy. I get lost…. I haven’t studied RELAX NG much, but the few examples I’ve run across were aha moments. Aha! I get this! Aha! This is pretty keen!Aha! I always did want to be able to say that in a DTD!“

On XML-L, G. Ken Holman also described W3C XML Schema as:

“too complex and, in
some places, just simply broken and unable to express some of the constraints we often need to place on structured information. If it meets the needs of data-heads then they can go ahead and use it, but for doc-heads and web-heads I’m of the opinion RELAX-NG[1] (pronounced “relaxing”) is part of the future and is already here for us to use today.
“

Back on ietf-xml-use, John Cowan went beyond Holman’s claims for RELAX NG’s applicability to documents:

“With the exception of identity constraints, RNG provides a superset of XML Schema facilities. It works equally well for documents and data, with no compromises required.“

Developers interested in exploring RELAX NG, the most frequently proposed alternative to W3C XML Schema (and now an ISO Draft International Standard) may want to explore:

• RELAX NG tutorial
• RELAX NG specification
• RELAX NG Technical Committee at OASIS
• Document Schema Definition Language (DSDL) at ISO, including RELAX NG
• Eric van der Vlist on XML Schema (XML.com)
• RelaxNGCC — Bridging the Gap Between Schemas and Programs (XML.com)
• RELAX NG, Compared (XML.com)
• Comparing XML Schema Languages (XML.com)
• Schemarama (XML.com)
• RELAX NG Gets “Feasible Validation” Support (xmlhack.com)
• RELAX NG Becomes ISO Draft Standard (xmlhack.com)

Do these rebels have a chance against the W3C empire?

Related link: http://www.ics.uci.edu/~fielding/pubs/dissertation/top.htm

Thomas Fielding famous dissertation about REST (Representational State Transfer) has giving rise to a grass roots movement, which competes head-to-head with SOAP/WSDL-based Web Services. The two concepts are really very different. SOAP/WSDL is operation driven, where data is exchanged according to the operation (method) invoked. There are many operations and few endpoint addresses (URLs). REST, on the other hand, is URI driven, where data is exchanged according to its address and only a few standard methods are offered (GET, POST, PUT, DELETE). In addition to Thomas Fielding’s dissertation you can also read a couple articles by Paul Prescod which summarize the concepts nicely: Second Generation Web Services and REST and the Real World.

Related link: http://www.xml.com/pub/a/2002/05/29/perry.html

Walter Perry makes an interesting argument about the possibility of fraud with standard XML vocabularies (a.k.a. Standard Data Vocabularies [SDVs]). The premise is that SDVs leads to automation based on SDVs. This automation is data driven, so the contents of the documents affect the outcome of operations. People will eventually figure out how to submit SDV documents in such a way as to give themselves an unfair advantage over a particular system. It reminds me tricks you can use with search engine based on the key words you place in the header - in both cases, you are submitting data in a way that manipulates the system to provide you with an advantage. I believe that ebXML is particularly vulnerable to this kind of manipulation, since that organization is establishing SDVs for many different industries.

Related link: http://www.chappellassoc.com/articles/article_who_cares_UDDI.html

David Chappell manages to sum up my own misgivings about UDDI. It seems far too complex and abstract to be practical. Something simpler is needed.

Brewster Kahle sent this to the archivists-talk@yahoogroups.com news group. Interesting stuff, via Aaron Swartz.

Subject: Public Access to the Public Domain: finding the list of out of copyright books and getting them online
From: Brewster Kahle

To: Steve Harris, Gregory B. Newby, David Wolber, Raj Reddy, Robert Thibadeau, Michael Shamos, Jim Fruchterman, James Michalko, Judith Bush, Lawrence Lessig, archivists-talk
Date: Wed, 29 May 2002 01:46:02 -0700

This is a note to a bunch of we technical and lawyerly doers in the world of Public Domain (PD) books to propose a short term strategy in order to get feedback and coordinate efforts.

Steve Harris is a guttenberg person and lawyer, Greg Newby is a guttenberg person and tech guru, Dave Wolber is a prof at USF and is working with the Internet Archive this summer with 12 of his undergrads, Raj Reddy is “universal access to human knowledge”, Robert Thibadeau is a tech master of the Universal Library at CMU, Michael I. Shamos is leader of Univeral Library at CMU and lawyer, Jim Fruchterman started bookshare.org: 20k scanned online books for the blind, Jim Michalko runs RLG (a union catalog), Judith Bush is running the RLG project to get the catalog onto the net, Larry Lessig has started the Creative Commons. archivists-talk@yahoogroups.com is an unmoderated list that discusses this kind of thing. Sorry if I left key people off– it is late.

If we want to help people put a pile of books online here is a strategy:

1. take a large catalog of books in libraries,
2. tag each entry with its US copyright status,
3. prioritize those that are out of copyright,
4. try to inspire the world to digitize the out-of-copyright books,
5. format the books for online distribution,
6. organize the resulting digitized books,
7. cause enlightenment in all corners of the globe.

Status:
1. Get catalog: Research Library Group (rlg.org) is up for it and will have their catalog ready in July. Maybe OCLC would be up for it too.

2. Tag PD works: Tagging Needs to be done.
for US PD rules see: The Public Domain: How to Find and Use Copyright-Free Writings, Music, Art & More (Public Domain, 1st Ed)

steps to tag:

A. get the registration records in dirty OCR’ed form: Newby has started this. (since most are in page image form right now)

B. maybe hand clean reg records up, (guttenberg started see Distributed Proofreaders) maybe with money from IA or CMU for offshore help, maybe dont clean it up but use smart computer algorithms to get it right enough. (harris pointed out US Catalog of Copyright Entries is a start at a hand cleaned version)

C. do a “join” to match them up with the records and find the books that have been registered in the catalog: the rest are then Public domain.

D. Publicize what we have done and the methods in a series of meetings with lawyers, publishers, and authors groups so that we can refine a “best faith effort” to do large scale copyright clearance. This step is key so that we are not surprising anyone.Maybe Creative Commons can sponsor these? IA can help.

3. Prioritize: we need good ideas here. Maybe we could get checkout counts, or how many libraries have a book to know which ones should be digitizing first. Otherwise we could just leave it up to individuals that would scan, and they do it in any order they want to.

4. Scanning: There are a number of ways to inspire this. We need a repeatable and inexpensive process that builds in QA. Universal Library is now cranking on a centralized approach, can we augment this with a napster-like version? bookshare.org is an example where blind people are scanning in books.

5. Format: archival and access formats are a problem. there is no “MP3″ of online books yet. Of course keep the high rez scans (IA will provide free storage for any needers) and then have some meetings where we try to get the list of supported access standards down to a manageable number. I suggest a gritty meeting in August in San Francisco that, again, the IA can sponsor.

6. Organize: I think we can build an amazon.com-like site with a catalog that starts with RLG’s records (they have verbally agreed). Internet Archive is building such a site this summer, hopefully Dave Wolber’s army will help populate it with the 20k or so existing online books–we could use help.

7. Educate and enlighten. We are now in touch with many world wide literacy programs. I am sure you guys are in touch with them as well. If we deliver a free, massive, and relevant library they can do alot of the hard work of getting it to the many many people.

Please comment, destroy, or whatever this strategy.

Then pick a hunk to take charge of.

The Internet Archive is best suited to #5 and #6 (Format and Organize) but can help or do any pieces.

“Public Access to the Public Domain” is a step towards Raj’s “Universal Access to Human Knowledge”. How hard can it be?

-brewster

After four years of running my home network with a Windows NT Server at the core, I’ve shut down the server and disconnected the domain.

For about the last three years, I’ve promised myself “this will be the last year on Windows.” I came to Windows from the Macintosh, and stayed in Windows largely (ironically) because of Java. I’ve had a network of Windows, Macintosh, and Linux systems in various forms since 1995, and finally got an NT Server license in 1998 (I’d run Workstation before that), adding to my slowly growing collection of Windows software.

The notion of a server that could handle Windows and Macintosh systems was appealing, as was the fact that I could run Windows applications of various stripes on the server. (Don’t try this at work, but it does very well at home.)

Over time, though, the appeal dwindled. The Macintosh printing worked well for the PostScript laser printer I had, but poorly for the inkjet printing - NT insisted that the Mac send it PostScript, which it then mashed into something resembling what I wanted on the printout. Storing Macintosh files produced all kinds of weird filename issues that made daily work and backups extra-complicated. Maybe Microsoft got around to fixing these, but I can’t say the upgrade prices to 2000 or XP looked attractive.

Meanwhile, I was shifting a lot of my network functionality to independent devices. After a bad trial period with the Microsoft Proxy Server, I bought a WebRamp and life was so much easier. (I still use it as a DHCP server, even though the modem is long gone.) I put the laser printer and then the inkjet on their own Ethernet print servers, and the Macs were finally full partners and the Linux system was finally a partner. As I ran out of disk space on the NT server, I decided to buy a SnapServer instead, and it’s worked well for all of my systems, as has a tiny router.

I’m not completely out of Windows. My wife’s embroidery software is Windows-only, so there’ll be a few around for a while. As an editor and writer, I still need access to Windows software on a regular basis for testing and exploration.

The “everything on one server, with centralized authentication” approach may have been appealing to me a few years ago, but now I’m much happier with a loosely-connected set of specialized devices. While I don’t think this approach is necessarily appropriate for offices, it’ll do just fine for a home network.

How has your network evolved?