The topicmapCafe is turning into
The topicmapCafe is turning into a who's who in topicmapland.
XML.com: What Are Topic Maps?
XML.com: What Are Topic Maps? [Sep. 11, 2002]: a new and excellent article explaining topicmaps. Still, practical topicmap application seems to take time - there are some non-technical hurdles that seem even harder than the technical ones.
LIFT for Dreamweaver - Nielsen
LIFT for Dreamweaver - Nielsen Norman Group edition. NN continue to be the number one brand in usability - much like adaptivepath is becoming the leading brand in information architecture (or is it the bear?). I wonder when they'll start doing research and charging for reports? I don't think that's a bad evolution.
I was told by a
I was told by a librarian (a real one!) that research has repeatedly shown that when indexing information - even with a controlled vocabulary and trained indexers - the maximum rate of agreement hovers around 25%. So for 75% of items, the indexers don't agree on which terms to use to describe them. Anyone want to (dis)confirm this?
I also recently discovered the existence of language geeks. Compared with the language geeks, your basic Slashdot-type technology geek is nothing but an amateur in geekness, a pathetic wannabe. A friend of mine is cataloging the different words used in New York state legal texts to describe places, and their meaning - as a hobby. He was also translating 18 century German texts into English, learning German to do it. German is harder than Perl, believe me. No debuggers available. Wordnet looks like a typical language geek hangout.
Finally, I was reading Information Anxiety 2 (at amazon.com), and thus learned that Richard Saul Wurman coined the term "information architect" back in 1975. What he meant leans more towards what today we would call an "information designer" though - many information architects wouldn't meet his definition.
That's all folks. I'm off to Belgium for a few weeks. Updates will be rare and probably not worth reading.
My MS interview: tie master.
MS contacted me after seeing the XFML site I've been working on. I wouldn't have applied for the position based on the profile they wrote, it was way of - too much experience needed. They wanted to interview me anyway for a Program Manager position. A program manager is basically someone who determines functionality of a piece of software - and MS set it up in such an ingenious yet obvious way (coders do not report to program managers) that leads to quality code (yes, I just said that) that I'm surprised it hasn't been copied more in the industry. It really is quite brilliant.
Microsoft are famous for their gruelling interview practices - and I have to agree: it was fun but my brain was fried at the end of the day. But again - it just works. I was impressed. I talked to 6 people, and each gave me a design problem, something they were actually working on. I was then supposed to come up with solutions. It is a great way of interviewing, but it really fries your brain fast. By noon I couldn't think straight anymore. (They didn't offer me the job.) Another thing they were trying to find out is if I could be an advocate: had I done talks at conferences? Did I know a lot of people in the industry?
I wore a suit and tie but really shouldn't have bothered. Microsoft are hiring agressively at the moment, I saw at least 10 other people waiting for interviews. None wore ties. Nobody I met wore ties. I was the sole tie master. For the interview, I was flown to Seattle, MS paid. MS got me a nice hotelroom for two nights so I could check out Seattle day two. It looks like a great place. MS are famous for treating their employees well, I would have liked to work there.
FacetMap is getting really interesting.
FacetMap is getting really interesting. Read his statement for this month: damn cool stuff.
Meanwhile, with the design of XFML, I'm facing this question: allow for controlled vocabulary features like related, broader, narrower and alternative terms, or just stick with the a categorization using one term for each topic. Mmmm... Any ideas?
Christina's book is out! Information
Christina's book is out! Information Architecture: Blueprints for the Web.
Faceted Classification Goodness
Some posts by people a lot more cleverder than me on the XFML list: anyone interested in faceted classification and thesauri should read these.
Travis from Facetmap makes some great points:
"(about topics having multiple parents) [...] The example given in ISO2788 is that "organs (musical instruments)" can have the parents "wind instruments" and "keyboard instruments".
[...]
Many readers of this list probably see where I'm going with this. "wind instruments" and "keyboard instruments" don't belong in the same taxonomy. They are separate facets of "organs". "Wind" goes in the "sound production" facet taxonomy along with "percussion", "string", etc. "Keyboard" belongs in "input devices", which would be a
different facet taxonomy because it describes a different aspect of the instrument.
[...]
Having written an XFML implementation from scratch, I would also like to add this: The relational database graph theory involved is much more elegant, and more efficient to run, if you limit facet nodes to a single parent. The practical upshot of such a limitation is, of course, that a facet mapping engine can handle many more facet nodes and resources before it starts to get swamped."
... and gets responses from the library camp:
"For many years the literature on thesaurus construction has been emphasising the need for facet analysis as an underlying tool to ensure that valid thesaurus relationships are created. It has been good practice to present a controlled vocabulary both as a faceted classification and a thesaurus, since Jean Aitchison "Thesaurofacet" was published in 1969.
[...]
I think we may be talking slightly at cross-purposes here. It may be helpful if I set out the definitions of some of the terms used in the library / information science community, as people coming to this problem from an information technology or "topic map" background have adopted different terms for the same things, which leads to some misunderstandings. [...]"
I updated my personal site
I updated my personal site since I'm no longer looking for a job.
I am stealing ideas from
I am stealing ideas from the RSS people for XFML: I am now using copyrightHolder and copyrightLicense attributes instead of just one copyright attribute: thanks. And to address the scaling problem that occurs when an XFML file gets popular (files can get large), I have added skipDays and skipHours elements as in RSS 0.94. Thanks.
Information architecture: learning how to
Information architecture: learning how to classify - ia/: "[...] when times change, terms will change." Very true. When your view of the world changes, your classification changes.
New book:Beyond Borders: Web Globalization
New book:Beyond Borders: Web Globalization Strategies. Looks interesting - I'll be ordering it. There are few books that deal with non-technical globalisation issues such as workflow and culture.
The BlogMD Initiative (via The
The BlogMD Initiative (via The Noisy One): "The BlogMD initiative, by creating a standard ping API supported by multiple CMS developers, will remove this barrier to entry from the weblog metadata application space."
What it basically means is that you can make your metadata available through an API. It's similar yet different from XFML:
- BlogMD wants to provide a centrally imposed set of metadata, mostly focussing on things like title, creation date, author and such. "[...] think creatively about what a good set of standard metadata to track around blogs would be." XFML wants each author to create it's own metadata structure, and is more focussed on topics without values (a date has a value, a topic doesn't).
- BlogMD shares its data through an API (using pings, like TrackBack), XFML shares its data by publishing an XML file (like RSS). Pings will scale better and offer some possible cool advantages, XML files offer simplicity and ease of developing for (everyone knows how to parse an XML file).
- XFML is in version 0.8 (not published yet), BlogMD is just starting. We're ahead ;)
- Ease of implementation: XFML exporting is very easy to implement, full XFML functionality (importing, merging, ...) is a lot harder, but BlogMD is even harder.
Overall, they sound like two complementary technologies. Go check them out and if you're technically inclined, give them a hand!
Cities of Text: Some Notes
Cities of Text: Some Notes On Some Notes on Intranets, Knowledge Management And Urban Planning: "Instead of file servers - byzantine hierarchical mazes into which we dump inscrutable containers of chartjunk called blarvitz.doc and blarvitz.ppt and blarvitz.wks - we have Web servers: byzantine, hierarchical mazes into which we dump now-scrutable containers of chartjunk called blarvitz.html.
In short, we've traded one generation of junk-making tools for another, one generation of data junkyard technology for another. "
NUblog: Dubliners, where the Joe-man
NUblog: Dubliners, where the Joe-man laments the slow adoption of RDF: "What we need is for Web pages to categorize themselves, which categorizations could then be computer-read and -collected. It's already possible, but it ain't happening."
He's missing the point: there is no incentive for me to try to understand RDF or add it to my website. Langauges like RSS or XFML adresses the incentive problem by adding value to a website with additional incoming and outgoing links. People can add metadata to other people's sites (a core concept that makes topicmaps (and XFML) really powerful): not everyone has to be a librarian.
RDF is dead as a popular metadata format. If it was gonna happen, it would have already. Forget about it.
[topicmapmail] occurrence abuse?: a thread
[topicmapmail] occurrence abuse?: a thread on the topicmap mailinglist that nicely illustrates why topicmaps are too complex for use by non-experts: even they can't seem to agree on the definition of something as fundamental as an occurrence.
But that's ok - we don't need topicmaps to be understood by the user. Compatible child formats like XFML can be developed for specific purposes - and UI's should remove the last level of complexity. People will mess up metadata: they are lazy and metadata is hard to understand. That's a given we just have to work with.
I have a strong feeling
I have a strong feeling culture (and how it interacts with technology) is the next thing the web people will discover. The signs are there. It's such a rich topic, with a huge influence on how we design interactions.
Learning about culture is a logical next step: start with visual design, learn that interaction matters more so do interaction design and information architecture, learn more about social sciences, cognitive psychology, then read about cultural changes associated with disruptive technologies. Start (self-promotional) reading up.
Design disolving in behaviour (via
Design disolving in behaviour (via Matt): I was gonna quote a large part of this but an evil side effect of the CSS used on the site means I can't select (copy and paste) text.
Scripting News: "The designers of
Scripting News: "The designers of 1.0 wanted to forget that 0.91 happened. 0.91 had the version number, 0.90 didn't. There ought to be a law taught in Format Design 101. Include a version number. Rule number two. If version n-1 has a version number, version n must also have a version number. Rule number three. You can't ignore previous versions."
I'm trying to avoid the whole RSS versioning mess and all the RSS politics with XFML. Dictatorship is a good thing when designing a format.
I came to my senses
I came to my senses and am taking in the sun this week. Seeya later.
gladwell dot com-- Designs for
gladwell dot com-- Designs for Working: " [...] the Ford Motor Company, along with a group of researchers from the University of Michigan, recently conducted a pilot project on the effectiveness of "war rooms" in software development. Previously, someone inside the company who needed a new piece of software written would have a series of meetings with the company's programmers, and the client and the programmers would send messages back and forth. In the war-room study, the company moved the client, the programmers, and a manager into a dedicated conference room, and made them stay there until the project was done. Using the war room cut the software-development time by two-thirds, in part because there was far less time wasted on formal meetings or calls outside the building: the people who ought to have been bumping into each other were now sitting next to each other. "
Imparting knowledge through storytelling: this
Imparting knowledge through storytelling: this is the kind of wide-eyed knowledge management blab that you shouldn't let within 10 miles of your company. "For example, a hyperbolic tree representation of the relationships between elements of the story could be combined with sound and video effects as the reader/user explores the story." Ha!
I'm looking for an appartment
I'm looking for an appartment in Hoboken (NYC). Any ideas where to look? The Village Voice isn't being very helpful, Craigslist is better, but I could still use some more pointers.
Outlook is really slowing me
Outlook is really slowing me down. I estimate I loose at least 20 minutes a day on it. My 128M of ram just isn't enough for it. Bloatware! Help!
Interesting screenshots of the next
Interesting screenshots of the next generation MSN.
The top ten usability problems
The top ten usability problems in Mozilla. Developers are slowly getting into the usability game, and that's a Good Thing for open source. I believe OS will make it if they follow (or stay ahead of) the curve from functional to usable to invisible software. We've done functional, right now the challenge is to make it usable.
I had an insight that
I had an insight that let me simplify the XFML DTD a lot: now the only things in an XFML document are facets, topics and pages. Nothing else. XFML DTD 0.7 and example file.
Mind your phraseology! - Using
Mind your phraseology! - Using controlled vocabularies to improve findability.
Switching to Outlook: why is
Switching to Outlook: why is it so slow getting email from my server? Anything I should know? (Screenshot)
Use language to design a mental model.
Designing a mental model: Six Degrees - Why messages, files and people?.
I am in the process of designing a mental model for a metadata app I'm working on, and it is fascinating to think about the choices you have, and the way a mental model is closely related to language: the nouns and verbs you use to describe what is going on. I think I need to find some good books on linguistics. Any recommendations?
What do I call my "metadata"? "XFML map" or "metadata document"? Metaphors abound. "Connections" or "links" between topics? One of my best inventions I think so far on this is that each map has a "network". Just saying that: ...
"A map has a network of connected maps around it"
... explains a lot of the philosophy and the technical details of XFML. No further explanation. One word! Before I said that, you may not have known this thing, after I say it it is obvious. Questions will arise:
"How is a map connected to other maps in its network? - Through connections between individual topics."
The word "network" conjures images of connected maps, exactly how I want users to visualise this. Use language to design a mental model. Language, especially the categories and metaphors you choose, guides thinking. I've been reading Frank Herbert (again) - gotta be careful with that.
I have used Eudora for
I have used Eudora for years as my email app, mainly because I was worried about virusses with Outlook and didn't want to succumb to MS. The last few weeks it started crashing on my machine, so now I'm moving to Outlook after all. It also means I can finally try out SixDegrees. Oh wait, darn, my trial installation expired and I never even used it (because I was using Eudora - SixDegrees doesn't support Eudora yet). What now?
Robert Barta wrote an excellent
Robert Barta wrote an excellent topicmap tutorial this time using AsTMa, another way of expressing topicmaps (usually XTM is used). AsTMa= Authoring Tutorial. Check it out!
A thoughtful post by DonnaM:
A thoughtful post by DonnaM: Content Inventory - when is it enough?
Simon did a great job
Simon did a great job on email archives, one of the last bastions of techie-induced design where nobody seemed to have heard of usability or design. Finally, that is changing. css-discuss archive.
As you may know, I
As you may know, I am coding some software that works with XFML. I am going to do some testing end August, through September. I want to find out if the philosophy of XFML (= "Distributed, loosely connected metadata can solve the problems of badly entered or very specific metadata.") holds up in real life.
If you have a weblog or content site and are interested in this exercise, get in touch. People joining will be expected to write about their experiences good and bad with authoring faceted metadata, I'll set up a central weblog for that.
Sticky test
This post should only show up at the top of the working without a job category.
Four thousand two hundred pages
Four thousand two hundred pages driving Donna nuts. Is a content inventory really worth the work?
Joshua sees the light. My
Joshua sees the light. My take on this: don't size text. I've given up on it. What's wrong with default text sizes anyway?
Victor points to this good
Victor points to this good looking faceted browsing interface and notes exactly how clever it is in its focuss on its audience on the result page.
Cell Biology (washingtonpost.com) (via Peterme):
Cell Biology (washingtonpost.com) (via Peterme): "When somebody turns off their cell phone for you, it's true love."
The whole metadata thing I'm
The whole metadata thing I'm working on with Simon Willison is becoming more and more obvious. I'm really surprised nobody has done this: the more I think about it and now that it's taking shape, the more absolutely obvious it seems: metadata should be syndicated and connected. Somewhat like news items (RSS). I can't even remember why exactly it took me so long to get my head around it all.
Thanks to Simon's Secret XML Powers, it's also turning out to be surprisingly easy to code. I wouldn't be surprised if we have working code in about 20 hours of work. I'm interviewing this week, so I don't have a lot of time for coding, but even so it's not as much work as I'd thought. I think we're succeeding well avoiding the political pitfalls that make the RSS specifications such a mess - the advantage of dictatorship I think.
Coding together while living in different countries seems to be working out as well - mostly because we divide our work clearly. We'll see how it goes, I'm curious how we'll deal with coding together without using CVS or some versioning system.
I think the first real life applications will be on my colombia website, and then I want to make it work with a mailing list archive. Maybe the Sigia-L list? We'll see how it goes...
Techie stuff IA's need to
Techie stuff IA's need to know. (I know, I was going to stop posting. I'm trying, really!)