Internet Alchemy

In this interview, Aaron makes reference to the problem that Pepys (my wysiwiki) sets out to solve: the disconnect between reading a wiki and editing it.

While wikis have achieved some of this, there?s still a lot to be done. Tim calls them the ?poor man?s? equivalent. For example, wikis aren?t WYSIWYG and make it too difficult to use links and other advanced features. And you can?t just see a typo and correct it, in situ, you have to find the edit link, then find the typo again, then correct it, then find the save button. You can?t use images or spreadsheets or any of the things that have been common in word processors forever.

These are all things on Pepys’ upcoming feature list and right now it solves the disconnect by allowing live editing, linking and formatting. The original wiki solved the problem of stale documentationby making it easy to edit, correct and create new pages. Pepys takes that a step further and makes it insanely easy to do it – just type and the links appear, ready to click on.

Having a wiki be that much more live, makes a big diference how you use it. I keep it open all the time and jot down things as I think of them. It’s all in one place and searchable. In fact, I just switched to it and played with the Find Pages function. I noticed a bug which stopped the results window from being closed properly. I just clicked on the home button, then on BugList and added it to my list of bugs that need looking at. It took about 30 seconds and I have a permanent record of it. This has got to be the killer development philosophy: develop software that you would want too use yourself.

Martin Fowler writes about Very Low Defect Projects:

I was particularly happy to run into the second case I saw – since it involved several of my old friends from C3. They are now building some portal software at Chrysler. Although they started off at the shocking defect rate of one a month, during 2002 they recorded exactly one bug against their system. During this time they were releasing new version of their software every week or two.

He’s careful to point out that adopting XP won’t automatically lower your bug count rate to barely detectable levels, but maybe it helps.

This is such a neat solution to a problem that’s been nagging me for ages: how to get a table-free page layout with a right-hand sidebar but with the sidebar content being after the main content in the HTML. The last point is important for search engines which may rank the importance of content by its position in the document, and also for accessibility – don’t you just hate those ’skip to content’ hacks? This solution was provided by Ryan Brill in response to a question thrown out over at Mezzoblue

Here’s another social networking service: Orkut which has been designed by one of Google’s employees. I didn’t know this, but apparently Google allows employees to spend up to 20% of their time on personal interest projects. That’s pretty cool and is definitately going to foster some amazing spin offs. I think, perhaps, that we’re going to look back on Google in a few decades and compare it very favourably with Xerox Parc. More at Search Engine Watch

I’ve just come across Grunk, a Java based toolkit for scraping semi-structured text. The concept is a bit like having regular expressions that work in terms of words rather than characters. The patterns are specified in an XML format so conceivably a library of common ones could be built up. Output is also in XML so coupled with some XSLT this could be a significant addition to the scrapers toolset.

After a 12-month hiatus, I’m back hacking on Pepys, my desktop wysiwiki. I’m just about done on a number of outstanding bug fixes, clearing the decks for a series of new features. I’m also going to start writing about it more here and at Semantic Planet because there are some interesting (IMHO) ideas that are driving this. More soon…

I’ve just come across a this list of
Ingredients for Serious Thought by Lance Fortnow, a researcher in computational complexity. Although I don’t claim to be capable of proving even the simplest theorems in complexity theory, these are exactly the conditions I need to get in the ‘zone’, especially 8 and 12: I get fantastically productive on the train and late at night. Now I wonder what a late night train journey might conjure up…

Interesting weblog by the way. One to add alongside Jacques Distler

This looks very useful: Turtle – Terse RDF Triple Language (formerly N-Triples Plus) by Dave Beckett. It’s an extension to N-Triples adding some of the neat features from Notation3 which makes it a pretty sensible alternative to RDF/XML. For comparison, here are the RDF/XML and RDF/Turtle versions of my RSS 1.0 feed and the RDF/XML and RDF/Turtle versions of my FOAF. It’s certainly easier to read and write but you have to be careful with your punctuation – it’s very easy to forget the (.) at the end of each set of triples. One niggle which I think also applies to Notation3 is that no punctuation (. , or is allowed in the last triple of an blank node. It would be nice to allow a semicolon so that you’re less likely to forget to add it when aappending new triples. (This is a common idiom when writing lists in Perl). Another clear advantage over the RDF/XML is that this style lends itself to appending triples. Appending to an XML document, let alone an RDF/XML one, is a non-trivial task making it harder to build logging applications that write XML logs. Nice work.

I’ve been trying out the new Opera 7.5 preview which now supports RSS 0.91, 1.0 and 2.0! Best of all it’s fully integrated into the M2 mail client which means that you can use all the mail filters to organise the entries in any way you like. Fantastic!

Jon Hanna has launched himself into webspace with a true tour de force: Entry-Level Unicode for XML. This article covers just about all you need to know about using Unicode with XML, either as an application or parser developer.
My only nit is that I wish Jon had put the long appendices on separate pages – many people will want to print out the article and they won’t necessarily need the code samples. Apart from that, this is definately a document to point newbie XMLers to in the future.