T he concept of ‘megacommunity’ is another way to perceive ‘power to the edges’ thinking about emergent organization and leadership. Megacommunities are large multi-organizational systems that are oriented to multilateral action. The megacommunity approach acknowledges that no single organization or entity can make the decisions required by today’s complex social and infrastructural systems. The challenges facing community leaders are based on complexity: the growing density of linkages among people, organizations, and issues all across the world.”CMT: There are data sources and RSS feeds, but until recently we bloggers had no pipes! Aggregating and integrating disparate data sources using JavaScript isn’t for those of us with regular day-jobs. But Yahoo! Pipes is a new object-oriented data manipulation and data mashup environment that can make it easy to do this. Other mashup environments are DabbleDB and Dapper.
— Mark Gerencser, Megacommunity Manifesto, 2006
DSM: Yahoo!Pipes is a free hosted service that lets you mix RSS feeds and create data mashups in a visual programming environment. The name of the service derives from mimicking Unix pipes, which let programmers push output from one command to the input required for the next command and so on ad lib, to create simple utility programs from a series of daisy-chained procedure-calls on the command line. Yahoo!Pipes is not strictly serial, though. The input for one function can come from more than one upstream source. The Yahoo!Pipes drag-and-drop user interface makes it easy to set up your own pipes to deliver specialized remixed feeds to address special interests that you have. You can stuff the pipe output to your RSS reader or publish it to a website or to a file. You can set up your pipe applet to solicit interactive input from your keyboard or pointing device. Or you can set up your pipe applet to harvest and filter and sort and pre-process or slice and dice what it finds without any interactive input from you. If you want to find and sort all recent feeds concerning violas you can do that. If you want to build a very specific news clippings robot to capture all articles that concern a specific composer or composition, you can do that. The flexibility is really pretty remarkable. And all very simple and easy. You don’t have to have any programming experience. You just have to (1) know what you want to do, (2) know where the information will come from on the web [the URLs], and (3) have a rational, logical mind.
CMT: What about error-trapping? Does Yahoo!Pipes have a compiler that checks that what you’ve dragged-and-dropped and hooked together is syntactically correct or semantically and logically feasible? Does Yahoo!Pipes have a performance optimizer or rule compiler to recast pipes that the user assembled naively? Can it make slow-performing pipes faster without the user having to know why her/his syntactically-correct but inelegant pipe is sucking so badly?
DSM: No. In this initial version, Yahoo!Pipes doesn’t seem to have any significant syntax-checker or compiler functions or error-trapping. You can build a pipe that may not work. But a non-functional pipe will not crash your machine or cause any harm. The Yahoo!Pipes service at least has that much bullet-proofing. And Yahoo!Pipes does not [yet] have any optimizer built in, so far as I can tell. If you put your filters upstream of your sorts, then the sort operations will run faster. But Yahoo!Pipes lets you chain things together in whatever sequence you wish, and it doesn’t complain or override the structures you build or replace it with run-time syntax or dispatcher sequences that reflect an optimized query plan. Likewise, there is no ‘plan’ command or compiler ‘hint’ capability such as you may be used to in Oracle or DB2 or other full-function relational DBMS systems.
CMT: The user interface, to me, resembles Insightful Corp’s iMiner datamining environment.
DSM: Yes. That’s a good analogy or resemblance to note. iMiner basically enables a statistics-naïve user to drag-and-drop various dataretrieval and statistical analysis functions onto a workspace within the iMiner application and then “connect” them to visually specify the processing that you want the software to perform. You do that instead of writing hundreds of lines of S-PLUS code. You can then ‘save’ your iMiner workspace with all of the process objects and connections and report templates and so on, so that it can be re-run in exactly the same way whenever you want. That’s the same way Yahoo!Pipes works. Once you’re satisfied with the output your pipe is producing you can save the pipe in your Yahoo!Pipes workspace, or you can publish it and make it available to others, or you can embed it in the HTML or CSS code for your website, or do other things with it.
CMT: What if I don’t quite know what I want to do? Or what if I just want to get a sense for the range of things that can be done with Yahoo!Pipes?
DSM: Well, the main Yahoo!Pipes page includes a gallery of ‘clonable’ examples of pipes that have been built by other people. Many examples involve remixes and filters of newsfeeds and iTunes and concert-listings and other kinds of applications that would enable you to quickly get a fairly comprehensive and practical notion of what you can do and how you go about doing it. You can ‘copy’ a generic example and change a few things in your copy to customize it to do just what you want. You don’t have to start totally from scratch . . .
CMT: Matthew Hurst has been producing some really cool computer-generated maps of the interconnections in the blogosphere, and his latest image is a beauty. He’s centered it on DailyKos (the bright white dot in the center) and says that the dense area of interlinked sites below it represents the socio/political community while the less densely populated but also highly interlinked area in the northeast quadrant of the image represents the heart of the ‘technical blogosphere’ with BoingBoing as the brightest node there. The size and brightness are proportional to the average number of hits per day that a blog receives. One of his blogosphere maps appears on page 22 of the May issue of Discover magazine, with green links representing one-way links (that is, blog A links to blog B) and purple links denoting reciprocal links (blog B returns the favor). Surely Yahoo!Pipes remix activity is not going to be counted in that sort of blogosphere mapping and social networking analytics.
DSM: You’re correct. Meta-processing and dynamic runtime inter-blog referencing is not (currently) being captured. Possibly in the future those types of dynamic “pulls” would be metered. But right now it’s just static blog-blog references that are counted.
CMT: This Yahoo!Pipes thing is pretty cool! I just set up a simple pipe for aggregating a number of classical music blogs’ feeds together, sorting them by date, then by blog name within date, then by blog post title within blog: CMT Classical Music Pipe. I bet our friends here may enjoy playing with this or creating custom ones of their own . . .
DSM: Because people communicate so easily across national and organizational boundaries, the conventional managerial decision-making style — in which a boss dominates decisions or delegates them to subordinates — is no longer adequate. A single classical music critic or powerful editor dominating decisions about what music events get coverage and how they get covered – is no longer adequate. A few non-profit commissioning organizations dominating decisions about what kinds of composing and performing get funding – is no longer adequate. Solutions require multi-organizational systems that are larger and more oriented to multilateral action than conventional cross-sector approaches are. In those future regimes, the most successful leaders are not those with the best technical solutions, the most compelling vision, the most impressive portfolio or ticket sales, or the most sensationalistic or charismatic style. “Winners” are those who manage – in whatever ways they can – to influence others in a larger system that they don’t control. Gerencser calls this type of non-hierarchical system a “megacommunity.”
CMT: So this is the kind of thinking that motivates the Megacommunity Manifesto! It resembles Hugh Brody’s study of indigenous cultures, The Other Side of Eden. Indigenous cultures are profoundly complex-adaptive. Rather than attempt to dominate the environment, they adapt themselves to its demands and become an integral part of it. Indigenous assistive technologies for classical music and new music, implemented in Yahoo!Pipes !
- Barlow A. The Rise of the Blogosphere. Praeger, 2007.
- Brody H. The Other Side of Eden. Faber & Faber, 2002.
- Gerencser M, Napolitano F, van Lee R. Megacommunity Manifesto. August, 2006.
- Ho E. Remixing the Web with Yahoo! Pipes.
- Hurst M. Datamining/Visualization/Social Media blog.
- Lebowsky J. Complexity, Megacommunity, and Adaptation. 21-AUG-2006.
- iCommons.org
No comments:
Post a Comment