As Joel Spolsky points out, every programmer should have a basic understanding of character sets. Of course one key factor is knowing how your language(s) deals with character sets. For Python, Uche Ogbuji has a nice collection of Python and UNICODE wisdom and nuggets.

And trust me, anyone who claims to be seriously developing webfeed apps and Web 2.0 more generally, needs to know a bit about UNICODE.

Ericka Menchen, University of Illinois, Chicago communications graduate student, recently conducted a survey of del.icio.us users. The results, entitled "Feedback, Motivation and Collectivity in a Social Bookmarking System" are now available.

Ericka also maintains an outstanding blog, regarding her adventures in graduate school.

Merrill Brown, former MSNBC top honcho, has been tapped to head up the Carnegie Foundation's, News in the 21st Century project, one element of Carnegie's overall program on revitalizing journalism. Fabulous!

At the risk of biting the hand that feeds me, when the original Carnegie announcement was made, I was struck by how little the issue of changing audience behavior, aided and abetted by technology, was a component. It read to me as revitalization through creating folks who are really, really good at reporting as opposed to just really good at reporting. Readers are just twiddling their thumbs waiting for more better investigative reporting!!

Maybe with his background and credibility, Brown will be able to give a little push in the audience facing direction.

O'Reillys, Nat Torkington transcribed some of the Long Tail Panel at Supernova 2005. The thoughts were interesting (a rock radio station dies everyday), but it tweaked one of my not quite pet peeves.

Long tail disciples almost uniformly refer to the need for tools to "filter" extraordinary amounts of niche content. A better analogy is probably "navigation", avoiding the term search as search is highly conflated with information retrieval techniques. My guess is that if you look at how people deal actually deal with long tail ecologies (e.g. electronica DJs, comic book collectors, etc.) they navigate social networks surrounding and embedding the niche media.

If you look at Amazon, I would bet that the "People who bought...", Listmania, and review features work a lot better than the recommendation engine in terms of encouraging sales. These could only loosely be construed as filtering.

Instead of filtering tools, how about "expansion" tools? Tools that work from a kernel of known preferential content, and expand the horizon outward. You don't need to sell folks on the entirety of the long tail at once.

Link parkin:' plasq's Comic Life looks like an elegant tool for people to author new media objects out of those pile of digital photos they have lying around.

TagCloud is a meta-webfeed service that takes bundles of webfeeds and gives back weighted lists of frequently occurring terms. Unclear on how the chosen terms are selected and what the time window is for analysis, but it's a well done first cut.

Disclaimer, I don't listen to podcasts and I'm only peripherally interested in them as another form of sociable media on the Web. This post was inspired by some observations by Ernest Miller on the scannability of podcasts.

Linkblogging is the equivalent of television soundbites for the Web. An exclusive diet of soundbites is obviously not good, but they do serve a useful purpose when done well, transmitting useful information in a condensed fashion. It's when they get completely divorced from their original context that soundbites get dangerous.

Mechanisms to support what I'm terming podbytes (probably too cute by half but hang with me), could help amplify the impact of podcasts. If users had a cheap and easy way to select, snip, and link to chunks of podcasts a.ka. podbytes, serious podcast listeners could more easily spread the word about the highlights of good podcasts. Less hardcore users could ride the collective intelligence of the core podcast audiences by keeping track of podbyte blogs (podbyting?) through webfeeds or WebJay playlists. This would probably be a much more effective means of scanning than having to play the podcast at double time. Podbytes would also make podcasts much more of the Web than they are now.

There are two thorny issues as I see it. One, is convenient interfaces for selecting and snipping podbytes. On the desktop, I don't see this as insurmountable, but on other podcast listening devices such as iPods, there'd need to be careful physical design. I'm not holding my breath waiting for a soundbite marking mechanism to come to these devices. Plus, who needs to be podbyting when they're driving or running on the treadmill!

Two, linking into regions of MP3s, the common podcast file format as far as I know, isn't particularly natural. Jon Udell has been exploring many of these issues, but as recently as January of this year, linking to podcast samples wasn't peachy and I haven't heard anything to change that picture.

About a week ago I got a note from Jury Gerasimov, a developer on Surfpack. Gerasimov was excited by my wistful dreams of an "aggregator as platform". Aparently Surfpack is aiming to fit that bill.

There are two elements that I think are crucial for extensiblity, and missing in just about every aggregator. One is the ability to hook into the fetching/crawling mechanism. Inside every aggregator is a little personal Web crawler. As far as I know, no aggregator makes it easy to monitor and extend what that crawler does.

Second, once the crawler has done it's work retrieving all of that web content on the user's behalf, presuming the aggregator persistently stores what it finds, you'd like to have a nice interface for interrogating that data. A little feed item specific query and manipulation language would be my ideal, but then again I'm a geek. Along with an extensible user interface this is one of the key aspects of "aggregator as platform".

Again, making an analogy to Emacs, text navigation, querying, region marking, etc. are fabulously well supported in Emacs lisp. What would be the appropriate language mechanisms for working with a database of webfeed items?

Meaty. Alex Halavais makes a case for social informatics growing up as a field, a.k.a. getting beyond emergence.

Still pondering but looks challenging to say the least.

Nothing collaborative going here, just coincidence. Flickr has revamped their API a bit to deal with 3rd party Flickr apps needing authentication. In essence, Flickr will hand out time limited tokens to the app instead of the app asking for user's login/password combo.

Meanwhile, FeedBurner has launched an Awareness API. FeedBurner provides a level of indirection on webfeeds, which allows them to do readership calculations among other things. You proxy delivery of your webfeed and they give you interesting stats. The Awareness API lets folks get at their data through web services, so further applications can be developed.

My guess is that for any individual feed this is pretty boring, but for decentralized networks of sites this could be pretty important. FeedBurner could turn into the Audit Bureau of Circulations for webfeeds! Of course ABC is a relatively transparent non-profit and FeedBurner is a privately held company so there's a bit of a difference.

pyc is a full blown compiler from Python source to Python VM bytecode, written by Stelios Xanthakis. Yet another vehicle for doing interesting programming language design, implementation, and pedagogy.

Now here's a whacky idea, taking off from Andrew Appel's compiler text, Modern Compiler Implementation in Python.

Link parkin': The World Association of Newspapers' 100 largest newspapers by circulation. All I've got to say is that there are hell of a lot of papers being sold in Japan.

One mild irritant with AudioScrobbler is that you can't seem to get your listening history data out in any reasonable way. This is mitigated by the fact that you could write a custom plugin to record the information in your own local data store, or configure a prebuilt plug-in to go through a proxy that does the archiving.

Hmmm, I wonder if you could build a fully distributed P2P version of AudioScrobbler with easy to manage social boundaries, backed by good security and privacy, nice global statistical services and a well designed query interface?

I can dream can't I!

del.icio.us now has gotten a bit smarter about media types. URLs ending in particular extensions automatically get tagged with a few special labels. For example, .mp3 will automatically be labeled system:filetype:mp3 and system:media:audio. Apparently you can also use these special labels yourself on any old URL.

Bonus podcasting boosting effect. Any RSS feeds that include one of these magic tags become enclosure containing. I'd have to work through the math, but I think with some creative tagging this can be an instant podcast creation tool, presuming of course all the media is already on the Web.

Recently, I've taken to characterizing my research interests as being centered on sociable media on the Web, especially from an empirical perspective. Jyri Engestr

If you don't want to shell out the cash for John Scott's "Social Network Analysis: A Handbook", a quick scan of Hanneman & Riddle's "Introduction to social network methods" seems to indicate the free, open, and online text might be a reasonable substitute.

Thinking differently, the Apple on Intel announcement could serve as a benchmark of sorts for blog analysis tools. Anything that purports to provide insight into what's being written or discussed in the blogosphere, and Web in general, should be able to do some retrospective analysis on the timeframe from June 1 to June 10 or so, and determine whether they're doing better than random.

SVS, or Social Versioning System, is Simon Yuill's neat take on version control and collaborative development. I haven't quite figured out exactly what SVS is, but it looks like a toolkit for building collaborative development tools.

Even cooler, SVS is being put to use for the development of spring_alpha, a networked game designed to explore development and experimentation with social practices, in both software production and game design. One of my (many!) rejected NSF grant proposals essentially had these concepts as the kernel of a research program. But the software in the proposal was a bit vaporous. I'll have to keep svs/spring_alpha in the back pocket for future attempts.

Oh yeah, Python inside!!

Aggregator++. Getting a bit crowded in the marketplace I'd say. For Alex King and Scott Sander's FeedLounge it looks like a nice UI and tagging (feeds and posts), from the ground up, will be the claims to fame.

Name in lights moment. The culmination of 2 course instantiations and a summer's worth of work went live about a month ago. Carl Smith did the heavy lifting on our Interpretive Digital Essay on The 1909 Plan of Chicago, (scroll down for credits) but I'm glad to have done my small part to make it happen. At the very least, our co-taught class helped wrangle research assistant talent for that big summer push!

If you really want to know about the birth of modern Chicago, dig in. Besides the interactive features are just dang cool!!

Eric Zorn, of The Chicago Tribune, wrote a moving column regarding Alicia Frantz, a local woman who passed away when she fell off her bike and was hit by a truck. Pure chance.

So what's the big deal? Zorn also maintains a Tribune branded weblog and in connection with his column he also wrote a moving post commemorating Frantz's life. I never met Frantz and I don't think Zorn did either, but as part of the Chicago area blogging community he demonstrates that he "gets it" by including amateur pictures of her, referencing her audio work, linking to her site, and pointing to other places to get more of Alicia's story, highlighting her place within the blogging community. I'm sure The Trib is picking up quite a few reads on its site and major whuffie with Chicagoland bloggers.

This is what I was talking about before when I said newspapers might be better off achieving A-List blogger status. Just actively listening and fully participating can go a long way.

Blentwell is del.icio.us for mix-DJ's. Where have you been all of my life!?

Sniff, it's just so beautiful. Can a brother get a tissue?

Digging through the logs for beep I ran across at least one satisfied customer. At 16 years of age, K. C. Cassimally might be a bit below the targeted demographic but the enthusiasm for the project is encouraging.

And it's always neat to get shout outs from Mauritius!!

PubSub is now in the business of creating specialized searches on various aspects of government. Seeing the web's political reactions in realtime will be a major impact of large scale watch engines.

Link parkin': Darren Hobbs notes a major scientific project almost losing 5 years of work due to range limits in C integers. There's one for the CS 213 anecdote book.

One of the reasons I generally like Vin Crosbie's writings on the media industry is he usually has nicely contextualized statistics to back up his thoughts. Witness the fact that if you look at the raw numbers of newspaper circulation declines, you start to see whole towns disappearing. Extrapolate to major metro areas and there are significant communities (Anaheim, Syracuse, Fort Lauderdale) conceptually going by the wayside. Definitely a different perspective than 1% circulation drop which is my quick estimate of the lossage from Vin's numbers.

If daily news trends at these circulation scales continue on their current harrowing trend, citizens journalism might not just be an alternative, but the only option.

Over a year ago, Marc Canter planted his flag on the term "digital lifestyle aggregation" or DLA. When I first heard of DLA, the concept rang a bit hollow but thinking about del.icio.us, Flickr, allconsuming, and audioscrobbler, the pieces are there. All of these applications generate webfeeds so they could be slurped into your generic RSS aggregator, but I think most of these are subtly incorrect, and point to where I'd tweak DLA.

What we really want is Sociable Lifestyle Aggregators for doing DLA. Not only do we want to monitor and aggregate our personal digital lifestyles, we want to have good mechanisms for selective and controlled revelation. Aggregators with buddy lists could be a baby step, but aggregators equally or priimarily designed for social activities are a whole different beast. The issues are a little deeper than "river of news" versus "e-mail".

Regardless of what you think of Stanley Fish as an intellectual, academic, educator, or adminstrator, this New York Times op-ed piece on designing new languages should resonate with anyone who calls themselves a computer scientist.

If you could get the infrastructure right, Fish's exercise would be fun to pull on a bunch of unsuspecting super neophyte programmers. Students who didn't know enough to "know" that designing a language is supposed to be hard. Combine with Stroustrup's maxim "library design is language design" for maximum effect.

I have no idea what MOVES stands for, but it's an organization connected with the Naval Postgraduate Institute. Apparently MOVES is the driving force behind Delta3D, an open source game engine with the backing of the US military for development support. Runs Windows or Linux.

Python inside(able): wOOt!

I've mentioned before that I thought the Yahoo! Music Engine looks cool simply because it's pluggable. Although a bit stale, when yTunes shipped, Ian Rogers posted an entertaining overview of why adopting the Yahoo! Music Engine might be worth your while.

COMMAND LINE SHELL PLUGIN make me drool!

Link parkin: Elgg is an open source "personal learning landscape", a.k.a. a social blogging and media tool. Might be worth digging into.

Speaking of journalists vs bloggers, I'm coming to a somewhat contrarian position regarding traditional news organizations and citizen journalists. The old organizations should run away from hosting citizen's content like the plague. Comments, blogs, photos? Don't bother. Everyone can have their own press, why let them use yours.

Instead, these organizations should become the A1 A-list blogger of the community. Set up a continuous web beat desk. Staff a small number of folks to aggregate and point to the best web material of your community. Use your reporting skills, acumen, and connections to provide raw material for the audience, e.g. put in the time to make public records information more usable, or have the web desk work on providing on-line back story material from reporters.

Don't bother herding cats. Make them chase your tail. Be the first place locals want a link from, and the first place they link to.

Once an avowed Wired hater, I've taken a shine to the recent vintage. The magazine's not so heavy on The Long Now boosterism, although some might argue its just been replaced by The Long Tail boosterism. In any event, I took a particular liking to Noah Shachtman's article on unmanned aviation in the military. Enough to attempt a tortured analogy.

Shachtman's piece captures a cultural divide between the Air Force, which views UAVs as planes requiring professional operation, and the Army which treats these devices as populist tools for more effective tactical ops. Both sides of the divide are rapidly adopting the technology, and adapting their cultures but in different means and ways, each choosing different shapings of the tools.

Lesson learned? Not either/or but and.

I'm not big on the journalism vs blogging debate, but could a similar thing be happening with blogging packages and other social media tools? The pros are cautiously figuring out how to integrate and adapt, while the grunts are charging forward. Out of both efforts will come a variety of adapted tools, molded appropriately for their particular strategic and social contexts.

I told you it was a tortured analogy!

I'm not dead yet.

Pietro Speroni has written a long piece on the difference between tag sets and tag clouds. The former is just a list of labels applied to an item (e.g. url), the latter is the same list with label frequency factored in. Tag clouds essentially carry more information. Speroni also speculates on the rank/frequency distribution in del.icio.us tag clouds, invoking power law distributions, but at least retaining some skepticism that they really behave that way.

The interesting leap is that tag clouds can be considered feature vectors, which are subsequently amenable to well known clustering and information retrieveal techniques. I'm not sure sets versus clouds actually make that much of a difference. I'd have to do the math, but if you assume something approaching a power law distribution, almost all of the information is in the first couple of labels anyway. The rest will be sufficiently non-discriminating that a good clustering algorithm will ignore them anyway.

Based upon these thoughts, Terrell Russell has cooked up a charting engine to see tag set behavior for del.icio.us URLs. However, it strikes me that some kind of rigorous study is in order.

Stan James' Outfoxed is a Firefox extension to support rating pages and distributing the metadata across social networks. James built Outfoxed in less than six months, as part of a Masters' thesis, and the site has a detailed narration of the system's design, implementation, and consequences.

I'm not a huge fan of explicit rating systems, but Outfoxed is a good case study of how to build new functionality directly into an extensible browser.

Hey, isn't about time for someone to start a new commercial effort around Firefox to make it "safe for business"? Do some heavy duty engineering to make the browser much more efficient (memory consumption is ridiculous), provide some cleaned up well maintained groupware extensions, tie the whole bundle into readily available backend systems (Exchange, wiki & weblog tools, del.icio.us clone), and call it a low end Groove. Might be worth somebody's effort.

Flickr is ditching Flash and going to DHTML for displaying photos in many places. Can't find any details anywhere, but apparently this new scheme also supports active links in notes on photos. Gustavo G has an example photo with embedded links.

The photo mesh is slowly creeping up on us.

It was only a matter of time, but this little blog that could, finally succumbed to Adam Penenberg's, Media Hack column in the Google rankings. Could be my general lack of posting activity, but I'm guessing Penenberg's work fact checking Michelle Delio garnered a number of high quality links.

Sniff

Still the number one New Media Hack though.

Detailed instructions on how to trick out Firefox using the Greasemonkey extension. From Mark Pilgrim. Nuff said.

Dubbed yTunes, Yahoo! has released Yahoo! Music Engine, a desktop digital media application in competion with iTunes, Winamp, Windows Media Player, et. al. Being something of a duffer in writing plug-ins for such players, the first thing that lept to mind was "is it extensible?".

Sho nuff, you can write plug-ins for yTunes.

Yeah, I'm still alive. One of the things sucking up my time has been beep, a joint project between Medill's Media Managment capstone project and Chicago's suburban Daily Herald newspaper. Rich Gordon gives better background over at E-Media Tidbits.

Working on this project served as a nice, real world introduction to Drupal. I've mentioned before that Drupal is a next step blogging tool. Actually, the built in blogging module is quite marginal from my perspective, especially in comparison to something like Movable Type. Example, getting a solid multi-user blog is fairly painful. (Yes, I know about taxonomies, but that's not the built-in blogging module now is it?)

HOWEVER! Drupal has a very nice extension framework. The system supports the creation of conceptually new types of content much better than Movable Type. And the ability to escape hatch and create nodes of Php code is huge, especially since such pages will automatically be integrated with the rest of Drupal's permission, logging, theming, etc. mechanisms. Theming and URL management is also well done in Drupal.

In short, I think Drupal is a great tool for breaking out of the blog model and thinking more of microcontent creation.