§ February 23rd, 2006 § Filed under Web 2.0 Comments Off
Web 2.0 is now a political issue. I thought it was only of interest to the techno-savvy, but apparently it’s bigger than that — it bespeaks narcissism, communism, utopianism and lots of other bad -isms.
Who knew?
Nick Carr has a post titled The New Narcissism referring to on an article by Andrew Keen in the Weekly Standard, the arch-conservative political magazine. In it, Keen says:
SO WHAT, exactly, is the Web 2.0 movement? As an ideology, it is based upon a series of ethical assumptions about media, culture, and technology. It worships the creative amateur: the self-taught filmmaker, the dorm-room musician, the unpublished writer. It suggests that everyone–even the most poorly educated and inarticulate amongst us–can and should use digital media to express and realize themselves. Web 2.0 “empowers” our creativity, it “democratizes” media, it “levels the playing field” between experts and amateurs. The enemy of Web 2.0 is “elitist” traditional media.
Empowered by Web 2.0 technology, we can all become citizen journalists, citizen videographers, citizen musicians. Empowered by this technology, we will be able to write in the morning, direct movies in the afternoon, and make music in the evening.
Sounds familiar? It’s eerily similar to Marx’s seductive promise about individual self-realization in his German Ideology[...]
Wait, it gets worse.
§ Read the rest of this entry…
I have finally come to believe that identities-as-urls isn’t such a bad idea. But as a believer in enterprise-strength directories, I come to this conclusion by a rather circuitous route.
In my leisure reading, I came across an interesting story in The New Republic by Eve Fairbanks:
Frustrated by my persistent failure to call home after I moved to college, my parents pioneered increasingly tech-savvy ways to find out what I was up to. This is how they became the first to discover that, when you Googled me, two or three x-rated links inexplicably appeared at the end of a few pages of legitimate results. The excerpts below these links nestled my name between outrageously obscene phrases involving animals…
[...]
My parents were extremely worried. We consulted family friends, who feared that my racy Google persona would alienate employers and potential suitors. We talked about hiring a lawyer. Desperate to convince my parents I was not actually involved in Internet porn, I suggested that we should be happy for the woman who shared my name, because, evidently, her career was taking off!…
[...]
A year or so [later], I received an e-mail from home. “I’ve been talking to Google,” my mother wrote, “and they got your name off all those sites.” Sure enough, thanks to persistent attention from Google’s user support, the links–byproducts of tricks to drive more traffic to porn sites–were gone. They’ve been replaced by pages of school accolades and a collector’s site featuring the Beanie Babies “Eve” and “Fairbanks.” I’ve been downgraded to a fetishist of stuffed animals: Perhaps I ought to worry about my reputation after all.
§ Read the rest of this entry…
Lately I’ve been investigating algorithms for engineering role definitions for role-based access control implementations. Software vendors such as Eurekify have already solved this problem, but we’ve needed a lightweight tool for defining roles for clients not ready to commit to buying more software.
So, I’ve been learning about data mining algorithms such as clustering, association functions, and the market basket problem. Every time you look at an item at Amazon, it helpfully tells you that other people interested in that item have also been interested in several other items. This is an example of the market basket problem — given that someone has purchased a basket of goods, which additional goods are they most likely to buy?
The reading gets a bit intense:
Two simple similarity functions between transactions are the number of matches and the hamming distance. The number of matches between two transactions is defined by the number of items which were bought in both. In other words, if T1 and T2 are two transactions, then the match function between T1 and T2 is defined by the cardinality of T1∩ T2. The hamming distance between two transactions is defined to be the number of items which were bought in one of the transactions, but not both. Thus, the hamming distance between the transactions T1 and T2 is the sum of the cardinalities of the two sets (T1 − T2) and (T2 − T1). In reality, good similarity functions for transaction data are much more complex than the above mentioned functions. Often the similarity criterion may be a complex function of the number of matches and the hamming distance.
Confused? Me too.
I have to admit, I haven’t gotten too deep into this stuff. Fortunately, you don’t have to for role engineering. The basic approach is to determine, for each permission that is provisioned, what other permissions are always provisioned with it. Each unique collection of permissions coming out of that analysis is a candidate role.
We’ve been playing around with this stuff, coming up with a poor man’s role definition tool using OpenOffice spreadsheets and macros, and it seems to work pretty well. So as it turns out, a small amount of analytic horsepower goes a long way to improving the quality of RBAC role engineering.
§ February 2nd, 2006 § Filed under IT Strategies Comments Off
Geoffrey Moore moderated a panel at the World Economic Forum in Davos on the future of the technology sector with Bill Gates, John Chambers, Eric Schmidt and Skype’s Niklas Zennstrom (podcast available here). Moore has boiled the panel discussion down to three explanatory principles to augment Moore’s Law:
- The core enabling resources of IT–computing, memory, bandwidth–are asymptotically apporaching zero [cost]. For scenario planning purposes, assume they are free. (This is where Moore’s Law continues to have its impact.)
- Expect the digitization of everything. Either digital format will substitute for a prior analog reality, as in media, or it will serve as a proxy for the underlying reality, be that a natural resource, a customer order in transit, or a physical meeting. This allows for the full repertoire of computing to be applied to value creating metamorphoses, whether it be enhancing sensory impact, improving consumer experience, detecting change triggers, or the like.
- Expect the value proposition of IT to migrate from enabling transactions to enabling interactions, whether it be in customer service applications, design collaborations, or self-calibrating sensor-enabled systems. This allows for iterative processes to have shorter and shorter cycle times, resulting in better understanding, faster time to market, more reliable operations, and the like.
§ Read the rest of this entry…