Orcmid's Lair |
||||
|
|
2004-02-28Authenticating Blog ContributionsMusings: PGP-Signed Comments. Here's a blog entry by Jacques Distler on adding PGP signatures to comments on his blog. It is a neat idea and I am going to try out the business of keeping a link to a key in metadata for my pages. I will have to see how that works with nfoWare when pages are mirrored on CD-ROMs and hard drives.The only thing weird about my visit to this particular archive page, recommended by AnderBill, is that I receive an ActiveX warning (based on my Internet Explorer settings) and although I decline to allow ActiveX there seems to be no difference. It is a peculiar case and I will need to look at the page's source code to see what is going on. Wired 12.02: The New Face of the Silicon Age. The Wired article has more depth and more heart than the newsblurb could convey. My national-internal-economic question remains, and I am still in favor of globalization. I've been lucky here, though I get to worry about Federal Reserve Board chairmen being heard as saying that we must cut Social Security. I would like to see a way to smooth this into a raising of all boats, and the social health that we want for the entire planet. Information Technology EmploymentACM News Service -- The New Face of the Silicon Age. This is about the demographic and global change in IT work and the impact of off-shore outsourcing. I keep hearing this suggestion that outsourcing frees up US firms to invest in the higher-cost creative and marketing end. I have seen that in previous Gartner reports and other analyses. I am a fan of globalization, and it would seem that with economic integration over time, it evens out. The nagging part (ignoring the impact on effected individuals altogether, and that must not be underestimated), has to do with make versus buy when buy is outside the internal economy. The multiplier on taxation and recycling of purchasing within a monetary system is an appreciable thing, so that a dollar spent on an import is actually more expensive, to the economy, than a dollar cycling through the internal economy. I wonder what happened to that economic model and what is it we think will really compensate for that difference -- especially since we operate in a deficit trade position. Just something I remember from economics class and that I wonder about in this picture.Faith in TechnologyThe Semantic WebFeature Article -- The Web within the Web. Here's the February 2004 IEEE Spectrum Online article by Enrique Castro-Leon. Here's the statement that set off my antennae: "But in addition to making users happy, Web services also dramatically drive down costs. Ron Schmelzer, a senior analyst for ZapThink LLC in Waltham, Mass., a Web services consulting firm, says that he's seen software project costs cut by 90 percent because of Web services." It looks like reuse is considered one aspect and the savings in middleware is the sweet spot. It sounds like all of the costs are not being counted.Some simple magical thinking of the kind that shows up as we gee-wiz our way to the semantic web is here too: "The practical impact of delayed binding is enormous. Because of WSDL, a program calling a Web service can check the configuration of the Web service as the program runs, allowing the calling program to adjust for any changes that may have occurred in the Web service. This lets programmers separately develop and test the different components of an application, which will continue to run correctly even if one of its constituent modules is upgraded." This simply doesn't get what it takes to integrate across an interface boundary. The translation of this to semantic-web hype is fascinating: "Because of its pervasiveness, the Web is a subject of intensive research, so, of course, there's a next stage after Web services. It's called the semantic Web [see "Weaving a Web of Ideas," IEEE Spectrum, September 2002, pp. 6569]. Although Web services allow a machine to publish its data, making it available to another machine, the two have to agree on the structure of the data they are publishing. In the semantic Web, this sort of agreement will be largely unnecessary." And last, but not least: "The conceptual work has been done; it's time now for the heavy lifting of application development. It's time for the second Web." ACM News Service -- The Web Within the Web. This blurb is about a review of the Web Services and their deployment. It seems to mangle the roles of some of the Web Services modules. One of the people quoted is an author of the XML and Web Services book that I found so awful in my Web Applications class last summer. Open Access to InformationRemoving Barriers to AccessCommission Press Room. This is the EU Institutions press release that includes the speach by Mr. Erkki Liikanen on Barrier free access, given 2004-02-25. The full text is here, and there are links for downloading too.ACM News Service -- Barrier Free Access to the Information Society. This is about accessibility to the Internet, PCs, and full participation of all. I want to pursue this as a technology spiral for nfoWare and other things that I do. I put this ahead of any device anywhere, although it might be that pursuing accessibility and the accomodation and personalization involved will cover the any device case. Information TechnologyDistributed ComputingPeer-to-Peer OptimizationRelieving Peer-to-peer Pressure. This is the Patric Hadenius 2004-02-25 Technology Review article describing the proposal to do caching of P2P downloads at the ISP level to relieve the net from the request traffic at the ISP edges and the bandwidth glut of the ISP. It has more details and also points out that the problem is the P2P protocols don't play TCP-friendly (a phrase I just learned in the past week). And there is still something wrong with this picture.ACM News Service -- Relieving Peer-to-Peer Pressure. This is a weird article about services from providers of P2P software selling "solutions" for ISPs to put into their subnetworks for caching and locally serving P2P downloads. An interesting problem, not clearly described in the blurb, is how to detect P2P traffic and selectively cache it. Now, what is wrong with this picture? 2004-02-26Computing MilieuxIntellectual Property ModelsOpen Source LicensesSD Times: News & Top Stories - Apache Rewrites License. The David Rubenstein 2004-02-15 aticle on SD Times. The Apache license 2.0 needs to be tracked down directly.ACM News Service -- Apache Rewrites License. It looks like the Apache license will be easier for others to piggyback on, it will have a strong position with regard to patents, and it will encourage sharing without forcing sharing as in the viral element of the GPL. Anti-Competitive ProtectionsACM News Service -- It's About Connectivity Not the Internet. Blurb on the Frankston article captures some of the moves that are going on to curb innovation and how that is done, with discussion of the opportunities that are becoming available left to lurk in the main article. How anti-competitive maneuvers influence regulations is pointed out by Frankston, but he also points out where the real tide may be.It's About Connectivity Not The Internet!. This is a Bob Frankston 2004-02-23 article in CircleID Exploring Frontlines. There are some very provocative observations, including ones that apply to P2P. Finding a way to fry spam |CNET.com. Marguerite Reardon's 2004-02-24 CNET News.com article on John Levine, an Anti-Spam research leader. The article goes much deeper than the blurb, and covers some important areas. p; What is startling for me is that Spam FAX has been handled, but we are resisting empowering the same solutions for Spam e-mail -- Spam is regarded as basically legal. ACM News Service -- Finding a Way to Fry Spam. This blurb turns attention to what doesn't work, who bears the costs, and who is in the best position to intervene effectively -- the ISP. Grid Forum Backs Utility Computing Standards. This Paul Shread 2004-02-24 Datamation article provides links and the overall thrust of the convergence effort between the Grid Forum and work on Utility Computing standards. OASIS, "where most Web services work is happening these days" will be in the act, and there are high hopes for significant convergence. Focus seems to be on the management protocols. This will be interesting to follow. ACM News Service -- Grid Forum Backs Utility Computing Standards. It has been observed that Web Services, Multimedia delivery protocols, and Grid Computing share some common interests around standards. I would say this reflects the growing recognition that the different layers can be deployed across utlity transports of various kinds (yes, HTTP) and are scalable in that regard. The highly optimized cases will remain for highly optimized situations, it seems, and the abstraction of the service models allows great flexibility and ways to achieve interoperability at the edges of coordinated systems. In short, it is a good thing. RSA Panel: Cryptography Can't Foil Human Weakness. Here's the full version of the 2004-02-24 Mark Hachman article cited by TechNews. ACM News Service -- RSA Panel: Cryptography Can't Foil Human Weakness. This news blurb is about how the weak link remains human behavior, and that behavior can undermine any technical solution. There are some concerns about how legislative and governmental approaches have their own down-side, including not always being responsive to privacy and security concerns of the public. One example is electronic voting, and another is the continuing intrusion of Digital Rights Management on IP-holder dictating consumer use, rather than simple preservation of IP. 2004-02-25Computer NetworkingProtocolsInternet Performance AnalysisShivkumar Kalyanaraman. Ah, a cloaked location for Shivkumar Kalyanaraman, with links to papers and other resources about networking.Index of /Homepages/shivkuma/research/papers. This little treasure trove of papers has Yong's paper on TCP Monaco from the 2003 ICCP. It is the PDF here. Measurement Studies of End-to-End Congestion Control in the Internet. Here's a great page turned up by classmate Julian Freeman. The TCP Silly Window SyndromeI started looking into this for a homework assignment (turned in later this date). One thing I have to watch out for is when I publish my blog during homework assignments. I came back and went looking for more information and my search turned up some of my own mentions of SWS in my blog. This is all right for discussions, but for homework assignments, it is like sharing my research with classmates before we have all completed and handed ours in. Ouch.CSIS 4500 Course Notes. Here's another source on SWS. I stopped these notes (captured to Outlook) since I can get to them again. Meanwhile I need to fix Blog This! TCP Illustrated: Silly Windows Syndrome. Here's an article about the silly-window syndrome and the Nagle algorithm too. So there is plenty of material here. The syndrome is actually pretty simple, and so is the remedy. Bill Joy InterviewThe Blog This! link is gone from my browser and I don't know why. Oops, all of my favorites are gone too. Well, let's solve that later. The interview with Bill Joy is a lot of fun. I should find out more about it. Meanwhile, here's a fascinating recollection: "There's this thing called the silly window syndrome which is a very famous technical thing, and it's basically, you know, just a breakdown in communication where the thing just confuses itself into doing really stupid things. And these things arise when the ratio's between different things get out of hand. You know, it's like something that was designed to operate in a range of 1 to 3, suddenly it's got a 1 and 1,000, and some mixture of 1 and 1,000 operating against it. Like somebody's going on a land line and somebody's going on a satellite, and it's trying to measure something and it's always wrong. It's like scissors, rock, paper. You know? And if you ever play people that are really good, they get you every time. And that's kind of what happened. So silly windows syndrome was like the worst case behavior where you'd lose like 99% of your performance. You know, Van Jacobson f'mally came on to Berkeley and figured out a good general algorithm for fixing it so that didn't happen anymore. But that wasn't in the early code." The early code being? "The stuff I did. Yeah, and as long as you were in a normal environment, meaning a pure LAN environment or pure Wan environment, it wasn't an issue. But when you start mixing it up and having satellites and stuff, the thing started behaving really badly. You know, it's kind of a 'Help, my cursor's gone' situation. The system just stops working and wasn't easy to -- the problem had never been documented in the literature. " 2004-02-24Epistemologically Multiple Actor-Centered Systems: or, EMACS at work!. This ACM Ubiquity article by Yuwei Lin is about how when people (actors) from different backgrounds engage in communal sharing and problem-solving, software functionality and innovation are strengthened. The focus is on Stallman's work on EMACS, with some appeal to the motivation and community-building for GNU. ICIR. This is the I.C.S.I Center for Internet Research (and ICSI is the International Computer Science Institute of the University of California at Berkeley. Sally Floyd is at ICIR. Here there is exciting research on a number of Internet protocols, much of it related to congestion control for non-TCP traffic. Equation-Based Congestion Control for Unicast Applications. This is a collection of papers and simulation scripts related to providing TCP-Friendly Unicast Applications. There is a nice list of useful materials, including IETF RFC 3448 and errata. Computer MilieuxTrust and Trustworthy ComputingSecurity ChallengesGates: 'Everything' impacted by security concerns - News - ZDNet. This announcement talk at the RSA Security conference outlines a number of measures. Perhaps the most important is the use of RSA SecureID with Windows.Unlocking Our Future - CSO Magazine - February 2004. This is the Simson Garfinkel article on the grand challenges for security. The individual challenges are described more thoroughly, with practical examples. ACM News Service - Unlocking Our Future. This is about some Grand Challenges for security research. The ones that grip me are 2. development of tools and principles for building large-scale systems for critical and trustworthy applications that also make lucrative targets, such as medical records systems 3. a reliable way to measure risk in information systems, which could allow people to determine how much an organization could save by deploying a specific piece of software, for instance 4. give end users easily understandable security controls as well as privacy they can control for the pervasive, dynamic computing environments of the future; ACM News Service - Computer Security Efforts Intensify. This newsblurb announces the hot topics at this week's RSA Security conference. There are authentication techniques being discussed, including SPF, PassMark, and the Open Authentication Reference Architecture, OATH. Information TechnologyComputer NetworkingMonitoring and AnalysisUsers tap network-monitoring technology. This is the Phil Hochmuth 2004-02-16 Network World article, with an useful diagram.ACM News Service - Users Tap Network-Monitoring Technology. This news blurb is in reference to the IETF sFlow draft standard and its value on high-speed networks. This involves data capture in MIBs on switches and routers and coordination with a collection center that creates a model of network traffic from the snapshots of these important nodes. It appears to be useful in detecting security infractions as well as bad players, something we are talking about in class. Information ProcessingData SystemsImpact of RFID and other monitoringInformationWeek > RFID > Data Avalanche > February 16, 2004. This is the Rick Whiting story. The featured avalanche: "RFID chips have the potential to produce huge amounts of information. How will companies recognize valuable data and avoid getting buried by what they don't need?"ACM News Service - Data Avalanche. Hmm, here's another one on RFIDs and being prepared by the avalanche of data that these babies will produce. "Brian Higgins of BearingPoint comments that RFID systems will boost demand for data-synchronization and transformation tools, and raise the importance of data-quality-management software. New data access issues might also crop up as a result of RFID deployments." Roadblocks could slow RFID | CNET News.com. This is the Matt Hines C|net News.Com article of 2004-02-19. An important situation with RFID tags has to do with ascertaining what is a fact and what is asserted. For example, RFID technology is proposed for securing the integrity of the drug supply chain. It turns out that synchronization also involves normalization of data cods and other features (e.g., how to specify a color), and that disparities in an organizations systems need to be reconciled before RFID descriptive and status data starts trickling through these systems. There are also performance issues and infrastructure questions with regard to the volume of material that will arise with RFID. ACM News Service - Roadblocks Could Slow RFID. I must remember to include RFID (Radio Frequency Identification) technology under Situating Data. This is an important technology, and the gathering of RFID information and of synchronizing data across users is going to require substantial data synchronization provisions. Here it is recommended that RFID technology not be deployed until data synch issues are resolved. Computing MilieuxSocial ComputingGet the geeks out of the way?InfgoWorld: Is social networking just another men's group?. This is the Jon Udell 2004-02-13 article referenced in the ACM Newsblurb. Udell mentions that his book, Practical Internet Groupware, is out of print but available online. (Side note: I love the way O'Reilly does this, and they also leave the "-" characters in the ISBNs. Two attaboys.)I discovered that the "tin ear" is Udell's observation, and that it refers to the absence of social nuances in the questions and categories of current social-networking systems. Along with all of the other testosterone-geek factors. ACM News Service - Tin Ears and the Social Fabric. Comments on the introduction of technologies, in the last five years, that support people working together in information environments: Weblogs, IM, Wikis, comment threads, and Web Services. This kind of loose coupling must also support fluid improvisation, according to John Seely Brown. There is some concern whether the existing software development culture can respond, though the infrastructure seems to be workable. The "tin ears" of Orkut and LinkedIn are an example of programmers not being highly social people. (Hi, What's Your Myers-Briggs? is the tip-off for me.) The notable omission of women in these designs is also felt. 2004-02-23Information TechnologyComputer NetworkingProtocols - Silly Window SyndromeRFC 813 - Windows and Acknowledgment Strategy in TCP. Here's David Clark's paper on the silly window syndrome and how to eliminate it with some useful strategies at the sender and receiver. The current solutions seem more simply expressed, providing a nice indication of how ideas mature and become simpler. The paper is a nice read on what was being thought about at the time, and the first statements of the applicable heuristics.16: TCP Mapping -- Silly Window Syndrome. The IETF was discussing SWS as recently as the 48th meeting. There was a question of providing information about this sort of thing so that the heuristics for efficient usage of the network are better known. RFC 1122 is mentioned. Silly Window Syndrome. Here's a very simple slide that includes the requisite remedies on both the sender and receiver side. There is a reference to an RFC 0813, so that should be interesting too. It's a nice low number. Silly Window Syndrome. Here is an on-line slide presentation that covers the Silly Windows Syndrome. Bill Joy talked about it in an interview. There is good treatment in an online book about TCP. Meanwhile, this is funny for me because I thought it was a failure, but it is a maladaption that leads to a potential glut of tinygrams. There are times when one wants to do tinygrams (as in a Telnet session), but not because of a maladaption of the windowing protocol that leads to a stuck in small packets case. This is a great demonstration of how heuristic these protocol procedures are, and what must be considered with regard to the overall effect. Socket InterfaceWinsock Programmer's FAQ: Winsock Programmer's FAQ. Here's a long FAQ on Winsock, the Windows Sockets implementation. Sections 3.16 - 3.20 discuss aspects of Nagle's algorithm and the Silly Window Syndrome.
|
|||
