« March 22, 2004 | Main | March 24, 2004 »
March 23, 2004
Blogs as Vehicles of Scholarly Influence
Crooked Timber has a great discussion of why academics do or don't blog. Brian Weatherson made some points that I think echo my feelings on the subject to some degree:
First, having to get my thoughts into a state where it's not embarrassing to have other people read them is a real spur to clarify what I'm doing and sort out bugs. Second, when I go to write the paper, I've got first drafts of some of the trickier sections already written, so I can cut-and-paste them in and start editing. Third, whenever I have an idea that isn't going to go anywhere, my readers tell me about it fairly quickly. That can save a lot of time, because there's nothing so bad for work output as spending lots of resources on a dead-end project. I've been saved from a few disasters this way.From Crooked Timber: Academics and blogging
Referenced Tue Mar 23 2004 14:32:56 GMT-0700
Of course, this is true whether you're an academic or not. The larger question is do blogs matter academically, not just as a writer. Another comment says: "I've thought about it repeatedly and I keep coming up with the answer `I'd like to, but no.' I (very) occasionally contribute to a big blog. It's lots of fun. But the press of time - specifically [with respect to] tenure and promotion schedules - makes it impossible to commit that amount of time every day."
This last comment makes me sad because I feel its an abandonment of scholarship to a system that may or may not promote scholarship. Specifically, one of the chief goals of scholarship, particularly in the realm of science, is influence. Good academic work is likely as not to get accepted in a scholarly journal, which meets the needs of the tenure committee, but is only read by a handful of people. An entry on my blog is likely read by several thousand people. Which has more influence? That's not an easy question to answer.
Of course, my blog is missing something that academic journals have in spades: the trust of tenure committees. So while I can't say for certain which has more influence on scholarship, I can tell you with certainty which has more influence on my career.
One of my chief complaints about traditional scholarly journals is that they are mostly for profit enterprises. Even those run by non-profits frequently have to make money to support other activities. The result is a vast system with high overhead that cannot make scholarly work freely available without collapsing in on itself. So, when I publish something in a scholarly journal, it almost inevitably goes behind a pay curtain that makes the New York Times and Wall Street Journal seem positively freewheeling by comparison. This is in direct conflict with what scholarly publishing is supposed to be about.
Publishing in journals used to be the best means of promoting one's ideas and exerting influence in a field of thought; it has become one of the worst. Technology has moved beyond the high overhead of print publishing, but since the judgment of peers, in a very narrowly defined sense, has become the way we judge the quality of a scholar's output, the old system continues. I'm not thinking I can change that system, but it doesn't mean, as someone who values scholarship, that I shouldn't do that which is best even as I do that which is required.
For a more lengthy treatment of this topic, see Where Do I List This on My CV? Considering the Values of Self-Published Web Sites by Steven D. Krause.
02:49 PM | Recommend This | Print This
Step Forward and Be Identified!
"A name is now no longer a simple identifier; it is the key to a vast, cross-referenced system of public and private databases, which lay bare the most intimate features of an individual's life." Are you required to identify yourself?
02:28 PM | Recommend This | Print This
Passport to Nowhere
A c|net news article gives Passport's eulogy.
02:24 PM | Recommend This | Print This
Investing in Underserved Markets
If you are going to launch a new venture outside the technology hotbeds of Silicon Valley or Boston, you have to be an overacheiver...or at least, believe you are capable of beating the odds. What do you have against you? Well, all the things that the Valley has in its favor in multiples: a large and extremely educated and experienced workforce, a deep capital base, experienced service providers (attorneys, accountants, bankers, PR firms, and so on), a history of innovation and success (local heroes like Cisco, Sun, Seibel, Oracle, Google, Yahoo, and others), and the passion to create.From Investing in Underserved Markets :: AO
Referenced Tue Mar 23 2004 09:15:33 GMT-0700
So starts an article at AlwaysOn by Nick Efstratis of the Wasatch Venture Fund on why Wasatch believes in investing outside SV and Boston.
09:18 AM | Recommend This | Print This
The Need for Identity Management
Information Week has a good article on identity management. The article talks about the benefits and barriers to achieving those benefits. Identity management promises to "improve security, boost worker productivity, cut costs, and reduce the integration friction usually connected with giving employees, business partners, customers, and suppliers access to internal systems."
ID-management vendors such as BMC Software, Computer Associates, IBM Tivoli, Netegrity, Novell, Oblix, and RSA Security have promised for years that their software would deliver those benefits. However, there are few industry-wide standards and most applications are proprietary. This forces companies to install a hodgepodge of software and devote a great deal of time to getting the apps to work together--even before making them work among businesses. It can take as long as nine months for two companies to integrate separate ID-management apps well enough to allow employee authentication and authorization across company borders, says Michael Barrett, VP of Internet systems at American Express Co. That's why most businesses are focused on improving their internal ID-management controls to make it easier to identify and authenticate employees and customers seeking to access internal information.From The Need For Identity Management
Referenced Tue Mar 23 2004 08:21:20 GMT-0700
The article makes a case for too many standards, SAML, Liberty, and WS-Federation, being one of the barriers, but doesn't mention that there are systems that can seamlessly handle all three. One such system is even open source: SourceID.
Michael Barrett talks about some of the things AMEX is doing internally.
Until there's a single ID-management standard, businesses are making do with the tools available today. "There are a lot of companies doing what we're doing," says American Express' Barrett. "They're kicking the tires and deploying it in an internal way." American Express is working on an ID-management initiative designed to deliver its business credit-card and travel services directly to the customers' intranets. "We're getting pressure from our corporate clients to be able to use our services in such a way that they can link [them] into their identity-management systems without having to create and manage a separate user name and password for each service," Barrett says. And while only a few companies have deployed the technology, there will be a significant number of deployments by the end of the year, Barrett predicts. "This stuff is real," he says.From The Need For Identity Management
Referenced Tue Mar 23 2004 08:24:57 GMT-0700
This is a pretty interesting project that I've heard about in other venues. AMEX is making this stuff real, right now. From what I've heard, I think they're going to be way out in front on this front in comparison to their competition. There is considerable detail in the article concerning AMEX's plans and approach:
The potential cost savings and productivity gains are so large that it's important to move forward, even if standards are still being developed, Barrett says. American Express' internal ID-management architecture includes a homegrown Internet-authentication system, a mainframe-based access-control system known as RACF, Active Directory and LDAP databases, and Netegrity and Oblix tools. Almost all the applications are deployed as part of small departmental apps. Some are compatible with the Liberty Alliance identity-management specs. It would be too expensive for American Express to implement a proprietary single-sign-on application that covers every network, system, and application company-wide, Barrett says. "It just isn't cost justifiable, and the effort would be enormous," he says. So American Express will move gradually, consolidating some identity databases and making other apps compliant with Liberty Alliance specifications to reduce departmental "islands of identity," Barrett says. About 30% of employee calls to the financial-services company's help desk are because of forgotten passwords. So one of the first implementations of new standards is to reduce the number of user names and passwords employees must create, he says. Implementing Liberty Alliance standards will "allow seamless flow of identity among our internal systems," he says. But good ID management does more than cut down on calls to the help desk and save money, Barrett says. Simplified access can improve employee productivity and reduce aggravation. "Multiple logons just drive employees crazy," he says. It's quite common for employees to be working on more than one application or in more than one system and have to switch between them. If they've been off one system too long, they have to log back on. "That's an employee irritant and a major productivity sink," Barrett says.From The Need For Identity Management
Referenced Tue Mar 23 2004 08:29:06 GMT-0700
This approach reflects some tried and true principles. First, a single benefit, reduced internal help desk calls about passwords and identity issues, can offer enough benefit to justify many of the identity management system changes that lead to increased efficiency and new opportunities elsewhere. Since these latter benefits can be hard to quantify, its nice to have the upfront ROI. According to the Information Week article, research firm Gartner says a company with 10,000 employees that automates provisioning for 12 applications can save about $3.5 million over three years and see a 295% return on investment.
Second, identifying and then consolidating "islands of identity" is important. Equally important is to not feel like it has to be done all at once or separately. Metadirectories and federation are two solutions that allows identity islands to be consolidated without divisions giving up autonomy over identities, at least at first.
Third, reduce the number of identity credentials that a single employee must maintain (in most cases username and password) and automate, as much as practicable the provisioning of access information for existing identity credentials.
Basically, this strategy comes down to
- Justify the expense. Plan and sell the ID management project.
- Enable future ID initiatives. Federate identities across organizational boundaries.
- Offer increased benefit and value. Make it easy for legacy and new applications to use the consolidated identity credentials.