Ideas For Dozens: September 2005 Archives

« August 2005 | Main | October 2005 »

September 14, 2005

An Upload Manager for the User-Writable Web

Uploading a bunch of photos to my Flickr account as well a live performance art recording to MFDZ in the last couple of days, I began to get frustrated with how the file upload process as a whole works on the web.

There's basically no support for uploading at all in any browser I've ever seen. While your file's heading over to the site's server, the browser just waits to hear back. Often times, even though the browser's given you no feedback to this effect, closing the window will disrupt the file transfer, preventing the file from actually uploading.

Sites that take user uploads all the time have had to build nasty hacks around this problem: reminders not to close the window, tiny little windows that open up and hide in the background, crazy ad hoc progress bars with a million page refreshes, whole helper applications that have to run on the desktop, etc. While some sites, Flickr comes strongly to mind, have managed to minimize the pain, the process never seems natural, like you're forcing the browser to do something it doesn't want to.

Contrast this with what happens when you download a file from the web. The browser does all the work for you. A nice little window pops up with a progress bar and a record of your previous downloads. This Download Manager makes it convenient to find files on your local drive once you've downloaded them or to kill any uploads while they're still in progress.

As the web becomes ever more 'two-way writeable', as we start to spend more of our time online uploading our own creations rather than just passviely consuming those of other people, we're going to start to need Upload Managers. I'm sure there's all kinds of technical details I don't understand what it would actually take to implement something like this, but I've got some suggestions for what I'd like it to do from a user's point of view. I even dummied up a little visualization:

Here's the way I picture this working:

Fill out an upload form (like Flickr's Upload Manager) and click the "Send" button.
Your browser's "Upload" window pops up with a progress bar for your new item (like the bottom entry in my mockup). Simultaneously, the browser loads a return page on the site that sent you.
While the upload is in progress, you can use the Upload window to keep track of its progress or to cancel it. You can also keep browsing around the web as normal without having to keep a particular window open.
When the upload finishes, the browser sends an alert just like on the completion of a download (Safari, in OS X, hops up and down in the dock). Also, the progress bar disappears and the Upload window display links both to the original file on your drive (the folded paper icon) and to the appropriate url on the destination site (that would have to be supplied by the site to which you were uploading). This history would stick around until you manually cleared it, giving you easy access to all your most recent uploads around the web.

It seems like once this process became widely adopted, sites might want to be able to trigger JavaScript on the completion of a download or be able to access the Upload manager for other reasons (or bookmarklets might want to use that event for something so cool that it would be hard to think of now). It could be a platform for all the new types of user interactions we're going to need for this new kind of user-writable web.

Technorati Tags: Upload, Manager, Flickr, Web 2.0, Safari

Posted by Greg at 5:27 PM | Comments (3)

September 13, 2005

The Struggle To Right Oneself: On Artists' Statements

Art critics, especially good ones, are fond of saying that there are few forms of writing less creative than art criticism, that groping struggle to hang heavy, awkward words on the ethereal and ineffable. Each time I've tried my hand at writing critically about art I've empathized with this despair. But one part of the process always bucks me up and gives me faith in the necessity of the writing: reading artists' statements.

Below even badly translated technical manuals, arcane government regulations, and most poetry, artists' statements may be the most oppressively awful literary genre in existence. With a uniformity that is almost shocking, they follow a forkless path on the way to pretensiousness and inscrutability: invocation of an obscure, usually gloomy, Philiosopher followed by a fuzzy-headed description of the themes that infuse both the Philosopher's work and the artist's that takes the form of broad platitudes and near meaningless phrases which sound more like the titles of interdisciplinary graduate school classes than the description of any actual physical objects: Embodiedness and Globalization in the Decline of Nature and Aboriginal Identity, etc. Next comes the self-mythologizing section: an explanation of how they gained a profound personal undertanding of all these broad themes only when they travlled to rural Borneo to the deathbed of their long lost twin. Finally, they finish up with an actual, if vague, description of the work ('by taking pictures of cats, I. . .') and a precisely described, if imaginary, projection of its effect on the world ('. . .will inevitably bring greater attention to the crisis in the region').

The truly maddening thing is that the excreableness of these statements is completely uncoupled from the quality of the work they describe. Less horrible statements can describe inane work and the most baroquely kafka-esque of them can describe work with a simple and direct power.

A great, maybe quintessential, example of this is Kerry Skarbakka's artist's statement about his series titled The Struggle to Right Oneself. I will try not to put you through too much of Skarbakka's statement, but to give you a taste, it starts:

Philosopher Martin Heideggar described human existence as a process of perpetual falling, and it is the responsibility of each individual to catch ourselves from our own uncertainty. This unsettling prognosis of life informs my present body of work

And ends with this:

The images stand as ominous messages and reminders that we are all vulnerable to losing our footing and grasp. Moreover, they convey the primal qualities of the human condition as a precarious balancing act between the struggle against our desire to survive and our fantasy to transcend our humanness.

Now, like you (and Skarbakka himself, most probably), I have no idea what this means. I don't know much about Heidegger (except that his name is supposed to be spelled with an 'e') and I can't even parse all of the dangling prepositions in that second excerpt into a gramatically meaningful unit (the balancing act is between the struggle and the fantasy but the struggle is also against the desire and all of this is somehow the primal qualities of the human condition as the balancing act?).

Beyond this question of bad writing, though, Skarbakka's statement completely misses the point of his work, which is, largely, light and playful in tone. Take, for example, Naked:

At first, you see the captured action itself: bouncing on the bed. His raised left leg and back-tilted head emphasize the arc of his flight and give you a strong feeling of downward motion back towards the bed.

But then something strange happens. The body seems to settle into place in mid-air. It becomes a kind of classical sculpture. Rather than a single moment stolen out of an ongoing process, the poze is eternal, general. The shadows it throws on the wall and bed from the direct lighting make the body pop off of the rest of the background increasing this illusion of its three dimensional solidity. Like a bronze, it seems heavy, but at rest.

Often times, Skarbakka makes this joke of the flying body as sculpture all but literal, like in Fence, where the figure balances impossibly on the tip of one toe like a plastic pink flamingo in a garden pushed almost all the way over by the wind:

or in Studio which catches the artist in a permament Wiley Coyote moment, after the edge of the cliff, but before the fall. Where, in true cartoon logic, the only thing that can cause the fall is the realization, which will never come to this frozen figure, that you're walking (or, here, lying) on air:

Skarbakka does also have a dark side, which is equally misrepresented by the existential moanings of his statement. Images like Sarajevo (left) and Hopkins (Belize), of bodies falling from rundown buildings in third world countries, in a post September 11th world:

These pictures resonate strongly with the television news footage of people jumping to their deaths to escape fires on the high floors of the World Trade Center towers -- one of the central images of our often apocalyptic-seeming times. Although Skarbakka has, under political pressure, explicitly foresworn any such connection, this seems like an especially rich area for him. If he can bring his understanding of frozen moments of falling, and even his sense of humor about them, to bear on this horrific possibility of contemporary life he would really have acheived something of the depth and profundity at which the muddled words of his artist's statement can only hint.

Technorati Tags: Skarbakka, photography, contemporary, art, Wiley Coyote, artists statement

Posted by Greg at 4:21 PM | Comments (10)

September 10, 2005

The Linux Long Tail Clarified

So, in addition to posting my previous entry on "the Long Tail of Linux developers" to this blog, I also emailed a version of it to Chris Anderson himself. Much to my surprise and delight, he responded within a day and indulged me by participating in a little bit of back-and-forth in which I tried to clarify and make explicit some of the (it became rapidly clear to me) mudled ideas in the original post. I think that by the end I convinced him that there was something to my idea, though it may not have turned out to be exactly what I thought about the situation that was important, but rather something more basic.

What I had been taking for granted was the idea that the development of the Linux kernel as a whole represented an example of the Long Tail. That is, if you put lines of kernel code on the y-axis and contributors on the x-axis, you'd get a power law distribution where the portion of code supplied by vast numbers of small contributors added up to equal that of the small number of developers who lead the movement. What really makes it an example of the Long Tail is how this distribution of programmers was enabled by new tools for online communcation and collaboration amongst all those developers. The email discussion lists, universally downloadable source code, and open acceptance of contributions that make up the Linux developement process (Andrew Morton's role is much like the Public Editor of a newspaper, soliciting input, culling through it for quality and appropriateness and then properly placing it within the public enterprise as a whole) served to make it economically feasible for so many isolated indivduals to participate in a single large scale project of such great public importance. In other words, these new communication technologies and the lowered cost of participation they created (both for the central authority trying to do the organizing and the individual contributors who want to participate) allowed the Linux movement to effectively harness (or market to) a large and diverse wealth of programming power that would otherwise have been unavailable to it.

Further, the effectiveness of the ad-hoc Linux programming team as a whole was a product of the fact that so many of its members occupied niches. This fact ensured a diversity of talent and interest which meant that the right person would be there to take on each of the "thousand tiny projects" which Morton describes as making up the larger project of building an operating system.

Thinking about the story of Linux in this way leads me to wonder about the web programming/scripting skills I've acquired in the last couple of years. In terms of a power law distribution of programmers as a whole, as a Liberal Arts graduate with a degree in Art History currently spending his time as a musician and pastry slinger, I am way down the tail of productive programmers (on a graph where the y-axis is rate of effective software production and the x-axis represents individual programmers I am somewhere out of frame to the right). But the heart of my experience as a nascent web programmer has been the discovery of the new tools and platforms which are emerging for web programming -- PHP, mySQL, and Ruby on Rails for example -- tools which make it possible for me, with my limited knowledge and abilities, to complete more and more sophisticated and complex programming projects. These tools lower the barriers to entry and the opportunity cost of creating businesses and services of every stripe on the web. They are why I, along with a Philiosophy graduate/recording engineer and a history professor, have been able to create a Flickr-like grassroots music distribution site, which uses all the hip technologies such as tagging, RSS feeds with enclosures, and an open API to allow anyone to upload and sell their own music, make comments on music made by other people, and form social network links to other artists and fans.

When fools like us begin to be able to make sites like this one, not because we are expert web 2.0 programmers, but because we love and make music, the whole ballgame is wide open.

Technorati Tags: Long Tail, Linux, Chris Anderson, Open Source, OSS, PHP, MFDZ

Posted by Greg at 3:51 AM | Comments (1)

September 9, 2005

The Long Tail of Linux Kernel Developers

Along with many others, I've been following the development of a set of ideas about the rising importance of niche markets in a world with ever-decreasing costs of production, storage, and communication called The Long Tail by Wired Magazine editor Chris Anderson (or as we call him around Music For Dozens, "the other Chris Anderson").

Yesterday, I thought of his ideas while listening to an IT Conversations podcast of Andrew Morton's presentation to SD Forum. Morton is the lead maintainer of the Linux kernel (the technical heart of the operating system and in many ways the spiritual one of the open source movement as a whole). Morton is second in command only to the eponymous Linus Torvald himself.

In his talk, detailing the process through which development of the kernel takes place, Morton gave an interesting example of the Long Tail which I had not yet come across elsewhere. While Anderson and other commenters have outlined many ways in which the open source software world plays a part in many different Long Tails, Morton described what is, essentially, a Long Tail that exists amongst kernel developers themselves.

Morton explained that while patches have been contributed to the Linux kernel by a couple thousand programmers, its core constituency, which is responsible for the majority of development consists of about thirty people. Above that, of course the head narrows ever further down to the movement's leadership at the Open Source Development Labs, and eventually down to Torvald himself (there's an interesting discussion in the Q&A section about what would happen if Linus got hit by a bus -- according to Morton, it wouldn't be that big a deal -- which is an interesting question for Long Tails more generally [what happens if the head suddenly disappears?], but I digress).

Morton's argument is that the maturation of Linux and its rapid and successful recent growth comes from the emergence of this head, of a class of professional kernel developers who are employed by companies making products that interact with the kernel for whatever reason (whether it be Red Hat and the other big commercial Linux distribution companies or just some hardware manufacturer that needs drivers for its products merged into the Linux tree so that their products will work with Linux systems adn they won't lose Linux users as potential customers). At a certain point it became economically necessary for these companies to actively fund kernel development and the strucutre of the Linux project puts in place incentives for these companies to merge their improvements back into the public version of the Linux source code so they can benefit from the massive quality assurance and testing efforts continuously undertaken by Linux's constituency of early adopters and so that their features will be built upon, expanded, and maintained by other future Linux developers.

So, I guess my question is this: is Linux development the first example of a Long Tail to emerge tail first?

Technorati Tags: Linux, open source, oss, Andrew Morton, SD Forum, IT Conversations.

Posted by Greg at 3:13 AM | Comments (0)

September 5, 2005

Flickr's Interestingness and Images of Katrina

Recently, Flickr released a new feature they call "interestingness":

There are lots of things that make a photo 'interesting' (or not) in the Flickr. Where the clickthroughs are coming from; who comments on it and when; who marks it as a favorite; its tags and many more things which are constantly changing. Interestingness changes over time, as more and more fantastic photos and stories are added to Flickr.

This is a functional definition. A photo's interestingness quotient is determined by the behavior of Flickr users, it mechanically records, aggregates, and expresses their tastes as demonstrated by their use of the site.

Over a million strong and from all over the world, the Flickr user base is a relatively diverse bunch. It is surprising, therefore, to look at the interesting photos from the last 24 hours and see such strong visual coherence. These photos look less like the result of an automated algorithm than a traditional hand-picked curatorial process. They have a coherent aesthetic. What's more, it's a surprising one for a twenty-first century photo-sharing internet website: high formalism. All the formalist genres and cliches are here: perfect exposures, a prevalence of black and white, surrealist collage, macros, urban emptiness, poor strangers figured as types, careful compositions that emphasize the geometric qualities of the subjects, etc.

Here's a sampling of Interesting images from 31 August 2005, so you can get a sense of what I'm talking about:

Now, what's surprising about looking at this grid of images is how much it differs from the aesthetic more commonly found throughout the wider Flickr photostream, which is made up, in large part, of flash-flattened snapshots of family and friends, low-res cameraphone snaps, cute pictures of pets, hand-made art and collages, screen captures from the web, inspirational imagery decorated with stirring slogans, etc. A glance at the all time most-used tags will quickly confirm this: baby, beach, cameraphone, cat, christmas, family, party, trip, vacation, and wedding are a representative sample.

So you can get a sense of the contrast with the Interesting photos, here's a recent image labeled with each of those tags:

Rather than the slickness of Modern high formalism, this looks a lot more like the pre- or post-modern aesthetic of the snapshot: images that gain their meaning not from their visual power, but from their personal associations and the context of their use. Their takers treasure them because of their power to make present loved people and cherished events. They are aids to memory, not fuel for the aesthetic eye.

So what does the divergence of these two aesthetics tell us, either about the tastes of Flickr's users or, if they exist, the company's human curators? One conclusion to draw might be that the professional and semi-professional photographers who make up a minority of Flickr's users are having a disproportionate influence on the metrics that go into Interestingness because they are more active. They make more comments, mark more photos as favorites, look at more pictures not by their current contacts, and therefore their activity has a greater weight in the algorithms that choose the Interesting photos. Similarly, to the extent to which Interestingness is curated, it seems that Flickr is choosing to align itself with these people's pro-am aesthetic, predictably in that they are the company's most active and vocal customers.

Right now, in the midst of the aftermath of Hurricaine Katrina, it's an especially interesting time to look at Flickr in the light of this contrast. When it comes to images of Katrina, these two aesthetics are colliding in interesting and surprising ways.

The web has been saturated with powerful professional journalistic photos of the human misery and physical devastation caused by the flood. And a lot of those images have found their way onto Flickr:

In their captions for these pictures and their comments on them, people tend to record their own shock and horror on seeing these images for the first time. Re-posting the photos on Flickr is a kind of outgrowth of these original reactions. The users want to assert their own powerful emotional experiences in front of these images as real and to share those feelings with others. This usage reminds me of the omni-presence of pictures of the burning World Trade Center towers after September 11th. While Katrina has not produced such a singular iconic image, the desire for images which can act as avatars of our feelings about such an event is still very strong.

Another somewhat surprising use of Flickr in response to Katrina is the posting of pictures taken of TVs and screen shots of news websites displaying coverage of the event:

This is a kind of contemporary version of the long held practice of saving newspapers with headlines anouncing shocking, important, or historical events. They are (in this case virtual) tokens which connect those of us witnessing events solely through mass media with those events' physical reality. Just as the physical presence of a saved newspaper filled the void of the absence of first hand knowledge of, say, the Kennedy assasination, these captured images provide substitute mememento mori for distant tragedies, things we can show our grandchildren as proof that we were there.

Secondarily these images, along with those directly reposted from mainstream sources, provide a vehicle for people to critique the media's coverage of events. For example, the person who posted the third image above (showing an announcement aimed at people in federal disaster areas that was broadcast on MTV2), captioned his photo: "something tells me that anyone in a federal disaster area is not going to be sitting around watching a Gorillaz video on MTV2, nor are they going to have access to the internet...". And the third photo in the previous group, of the man holding a baby stadning in waist-deep water, drew the following comment from a user called djune: "The look on his face says it all... 'won't you help me instead of shooting photos?!'"

The original images (collages, drawings, photo alterations, etc.) that people have posted in response to Katrina are some of the strangest and most disturbing representations of the disaster that I've come across:

They have aspects of nineteenth century Folk art in the way they struggle, often not wholly successfully, to transform powerfully felt private imagery into universal expressions. Often they evince extreme points of view, whether religious or political, which, while they may seem uncouth in the context of traditional mid-twentieth century journalistic values, make up a big part of people's real world reaction to historic events.

The Katrina images on Flickr that have most touched me personally are the ones most in line with the second of Flickr's two aesthetics that I outlined above, that of the snapshot:

These are pictures people take in the same mode as any other they place on Flickr: to capture something they come across in their own lives, most often of familiar and beloved people and places. The power in these photos doesn't come from their visual surfaces, but from the stories and feelings they embody: the care package ready to be mailed off to the Astrodome, addressed to "Any Katrina Refugee"; the soaked bird rescued from Miami beach which bit the photographer's finger and flew off the next day; kids playing at running a lemonade stand dedicating its profits to "Hurricaine Relief" on a beautiful sunny late-summer day.

Technorati Tags: Flickr, Katrina, interestingness, tag, hurricaine, modern, formalist, photography, snapshot, photojournalism, amateur

Posted by Greg at 2:40 PM | Comments (9)

September 1, 2005

Back with a New Strategy

Recently, I've been off on one of my all-too-frequent blogging vacations. There are many causes for them: busy periods in my various projects, a lack of motivation to write, large scale posts that become overly ambitious and that I therefore overwork, etc. No reasons that are really ever very good.

In an attempt to put an end to this current vacation, I've come up with a strategy. I'm going to set myself a quota of three posts a week, but give myself the freedom (even the commission) to let them range more widely in topic than I normally would. Though it may be less than totally apparent, I do imagine some kind of connective theme to the things I write here. They all involve "ideas" of various kinds, as silly as that may sound.

So, some of these posts will surely be of less interest to you than others. Some may be more technical, some may be more personal, many will be more random. Your job is to flow freely with the feedback and I'll hone the topics. It's a throw it up against the wall and see what sticks kind of thing.

Also: Thanks to the very helpful Curt Merrill, I just managed to upload all my old posts from when I used to host this blog on my own. Eventually, I'll go through and tag them all into categories, but in the meantime, they're something to read.

Posted by Greg at 12:41 AM | Comments (4)