<?xml version="1.0" encoding="UTF-8"?>
<opml version="1.0">
  <head>
    <title>cmdln.net_2008-06-08</title>
    <expansionState>1,2,19,37,38,48,68,92,96,109,133,142,166,167,176,187,188,196</expansionState>
  </head>
  <body>
    <outline text="Intro" Offset="00:17"/>
    <outline text="Security Alerts" Offset="01:56">
      <outline text="New counter measure against worms" Offset="02:15">
        <outline text="http://rss.slashdot.org/~r/slashdot/eqWf/~3/304871205/article.pl"/>
        <outline text="New research from Ohio State University"/>
        <outline text="Infected machines start scanning a wide variety of other systems"/>
        <outline text="Looking for likely new targets to infect"/>
        <outline text="Common sense indicates unusual volume of scans indicates an infection"/>
        <outline text="A good heuristic for taking such machines off the net and diagnosing"/>
        <outline text="Researchers looked at relationship between how long an infected system was allowed to actively scan and probability a virus would spread"/>
        <outline text="Scans and scan-like activity are also part of normal traffic"/>
        <outline text="What is the best cut off point, especially for automated response?"/>
        <outline text="Researchers developed a model and tested against Code Red and SQL Slammer"/>
        <outline text="Turns out key value was well beyond what occurs on a normal network"/>
        <outline text="But this threshold was very quickly reached by infected machines"/>
        <outline text="Achieved some impressive results"/>
        <outline text="Recommend it as a complementary measure alongside other protections"/>
        <outline text="Similar to work also done at Pennsylvania State University"/>
        <outline text="Sounds like this is a network level technique, so more effective for ISPs, companies and schools"/>
      </outline>
      <outline text="Mozilla contemplates successor to same origin policy, Site Security Policy" Offset="04:56">
        <outline text="http://rss.slashdot.org/~r/slashdot/eqWf/~3/306307630/article.pl"/>
        <outline text="Idea is in the very earliest of stages, discussion"/>
        <outline text="No specification, yet, though there is some experimental code"/>
        <outline text="Code is in the form of a Firefox extension"/>
        <outline text="Intended to spur discussion and lead to a full proof of concept"/>
        <outline text="Effort is a response to the rise of XSS and XSRF attacks"/>
        <outline text="Intended to complement, supplement application level security, not replace it"/>
        <outline text="Looks like an expansion of same site origination policy"/>
        <outline text="Server gets to declare white lists for different kinds of sources and targets"/>
        <outline text="So intend for a safe way to allow two sites/services to call each other's scripts, content"/>
        <outline text="Uses simple HTTP headers, means non-capable browsers would just ignore"/>
        <outline text="Requires browsers to correctly adhere to policies"/>
        <outline text="Unclear yet how they will deal with spoofing of policy information itself"/>
        <outline text="Also unclear whether a system admin or an application developer would be responsible for setting policy"/>
        <outline text="Not sure how to popularize this to end users"/>
        <outline text="Will really rely on uptick by end users"/>
        <outline text="Hopefully we can get to proof of concept stage quickly as a live demonstration may make the operation of SSP more clear"/>
      </outline>
    </outline>
    <outline text="News" Offset="09:50">
      <outline text="Princeton paper on government transparency" Offset="10:03">
        <outline text="http://feeds.feedburner.com/~r/techliberation/~3/303046023/"/>
        <outline text="David Robinson, Harlan Yu, William Zeller, Ed Felten"/>
        <outline text="To be published as a final draft in Yale Journal of Law and Technology this Fall"/>
        <outline text="The policy recommendation is simple"/>
        <outline text="Current policy focuses on governmental web sites"/>
        <outline text="These are expensive and prone to falling out of date"/>
        <outline text="Nettle of regulatory compliance makes developing such sites difficult if not impossible"/>
        <outline text="Paper suggests instead that government provide open data feeds"/>
        <outline text="Third parties could take over building the human facing sites"/>
        <outline text="Argues that competition would see many different, better ways into the data">
          <outline text="Some of this is happening, anyway, just at higher cost"/>
          <outline text="Govtrack.us, Maplight, Sunlight"/>
        </outline>
        <outline text="Realistic enough to recognize not all areas of government would benefit"/>
        <outline text="Cost savings in most popular could offset burden of maintaining feed and site for less popular"/>
        <outline text="The engineer in my likes this idea"/>
        <outline text="I even like that they extrapolate out to a browser or browser feature for perusing this data directly"/>
        <outline text="I am concerned that some requirements are glossed over"/>
        <outline text="Accessibility, at least, seems to be a big one"/>
        <outline text="Would third parties have to comply, which raises their cost?"/>
        <outline text="If not, then would that not disenfranchise some?"/>
        <outline text="Unique issues to private parties are overlooked"/>
        <outline text="Common carriage versus private forum"/>
        <outline text="This is confusing enough for purely private plays"/>
        <outline text="Detecting bias in this scenario could be difficult"/>
        <outline text="If end user choice is limited, how can a consumer really use Free Speech to deal with any issues?"/>
        <outline text="I think this is an earnest beginning, but it will clearly need more thought"/>
        <outline text="At a minimum, open data feeds alone will not be enough in the short term or for some aspects of government"/>
        <outline text="Subsidizing a browser or browser plugin could be an interesting compromise to traditional web sites"/>
        <outline text="This is some of early thinking from Robinson, Felten; look forward to more"/>
      </outline>
      <outline text="The real ACTA threat" Offset="16:02">
        <outline text="http://feeds.arstechnica.com/~r/arstechnica/BAaf/~3/302908867/20080602-the-real-acta-threat-its-not-ipod-scanning-border-guards.html"/>
        <outline text="ACTA is a trade agreement, negotiated in private"/>
        <outline text="Nominally it is an anti-counterfeiting trade agreement"/>
        <outline text="Would apparently include piracy measures in its scope"/>
        <outline text="Would be between US, Canada, the EU, and Japan among others"/>
        <outline text="Recently a document was leaked outlining the draft"/>
        <outline text="There are implications that its scope would be broader"/>
        <outline text="I have heard of ACTA before though not very familiar"/>
        <outline text="Even earlier references seemed to be low key"/>
        <outline text="Ars confirms this, that the treaty has been public knowledge since last year"/>
        <outline text="Information has been scant"/>
        <outline text="Agreement is bypassing the World Trade Organization and the World Intellectual Property Organization"/>
        <outline text="This would eliminate some of the leeway that has been allowed with past WIPO treaties"/>
        <outline text="There is also the suspicion that later signees would not get any negotiating room"/>
        <outline text="There apparently is no actual draft as of yet"/>
        <outline text="Leaked document is a &quot;discussion paper&quot;"/>
        <outline text="Much hand wringing in the press, theories about border search and seizure for copyright infringement on digital devices"/>
        <outline text="Discussion paper is not a draft, suggest provisions that might make it into a draft"/>
        <outline text="Ars analysis yields much blander customs language, though it would amount to new powers"/>
        <outline text="Really trouble provisions target hubs like Pirate Bay and seek to spread DMCA safe harbor under takedown provisions more broadly"/>
        <outline text="The piracy hub language takes out consideration of commercial gain, seems to speak directly to pirate trackers, other non-profit entities"/>
        <outline text="Under the ISP provisions, would also perhaps erode judicial oversight in acquiring subscribe information when infringement is suspected"/>
        <outline text="At its heart, this seems like another trade play to spread maximal control we've been seeing in the US globally"/>
        <outline text="EFF on ACTA" Offset="6/2/08">
          <outline text="http://www.eff.org/action/sunlight-acta"/>
          <outline text="Some more history"/>
          <outline text="Also a link to take action"/>
        </outline>
        <outline text="Patry on ACTA" Offset="6/3/08">
          <outline text="http://williampatry.blogspot.com/2008/06/acta-call-to-arms-no-more-secret.html"/>
          <outline text="Outlines why making copyright a trade issue s a problem"/>
          <outline text="US Trade Representative historically has always pushed for more copyright"/>
          <outline text="Has not always had the power to enact, however"/>
          <outline text="Under the US Constitution, that power is meant to reside solely with Congress"/>
          <outline text="ACTA would reshape IP law under the USTR and do so in private, without the scrutiny a Congressional action would involve"/>
          <outline text="Biggest objection is to closed proceeding, rather than any specific provision"/>
          <outline text="Does mention another provision supposedly in the draft, from a trusted source"/>
          <outline text="Will apparently include ISP filtering measures"/>
          <outline text="Seems consistent with the Ars discussion of changes to information gathering, ISP involvement"/>
          <outline text="Patry doesn't dwell on it, preferring to stick to the main issue"/>
          <outline text="Other sites are already picking up on it, though"/>
        </outline>
      </outline>
      <outline text="Ease of spoofing copyright infringement online" Offset="22:37">
        <outline text="http://bits.blogs.nytimes.com/2008/06/05/the-inexact-science-behind-dmca-takedown-notices/"/>
        <outline text="Abuse of DMCA takedown has been speculated before"/>
        <outline text="The investigative techniques of the RIAA has repeatedly been called into question"/>
        <outline text="This is the first systematic study of the problem"/>
        <outline text="Tadayoshi Kohno, Michael Piatek and Arvind Krishnamurthy undertook a detailed study"/>
        <outline text="Two studies, actually, in May and August of last year"/>
        <outline text="Originally were looking at who participates in BitTorrent traffic"/>
        <outline text="Received so many false takedown notices, launched this study, including data from the fist"/>
        <outline text="Launched monitoring agents, were not downloading anything"/>
        <outline text="Received more than 400 takedown requests"/>
        <outline text="Seem to confirm poor practices and even intentional framing"/>
        <outline text="Concluded enforcers are only considering IP addresses, not actual content"/>
        <outline text="Also speculate enforcers and ISPs are being driven more and more to automation to deal with volume"/>
        <outline text="Did not even have to use IP spoofing"/>
        <outline text="Explain a false positive based on timing between tracker requests and DHCP leases"/>
        <outline text="Also, demonstrate an attack using an extension to BitTorrent meant to help with proxies and NATing firewalls"/>
        <outline text="Their demonstration involved framing print servers, which received multiple take down notices"/>
        <outline text="Researchers hope their paper will help bring about more open-ness in what, exactly, enforcers are doing"/>
        <outline text="Give feedback to P2P users, too, suggest more effective ways to identify monitors"/>
        <outline text="Prove black lists are no good but explain simple analysis for spotting monitors"/>
        <outline text="I am encouraged that this is a disciplined study"/>
        <outline text="Also that it was picked up by mainstream press, although it is NYT's tech blog"/>
        <outline text="Trends in paper may also push enforcers more to 10x-100x more expensive content analysis or, more likely, to keep coopting ISPs for DPI"/>
        <outline text="Felten on DMCA take downs based on inconclusive evidence">
          <outline text="http://www.freedom-to-tinker.com/?p=1298"/>
          <outline text="Suggests that a soft &quot;warning&quot; might be warranted by the lightweight monitoring detailed in the paper"/>
          <outline text="Might even be an effective deterrent with the understanding of the threat of more conclusive investigation"/>
          <outline text="Deconstructs the letters, too"/>
          <outline text="Points out how they do not jibe with the studies findings"/>
          <outline text="Speculates a little beyond cost as to why rights holders don't do more in depth verification"/>
          <outline text="Claim positive infringement in cases where that is impossible"/>
          <outline text="Means that takedown notices are effectively little more than warnings"/>
        </outline>
      </outline>
      <outline text="Reverse engineering the brain" Offset="30:17">
        <outline text="http://www.spectrum.ieee.org/print/6268"/>
        <outline text="Profile of David Adler's work"/>
        <outline text="Using advanced imaging to map out smallest neurological structures"/>
        <outline text="Project at a campus called Janelia Farm, part of Howard Hughes Medical Institute, is courting Adler"/>
        <outline text="Their goal is to understand the human brain"/>
        <outline text="To answer that core question of neuroscience, how does the brain do what it does"/>
        <outline text="Starting with a fruit fly brain"/>
        <outline text="Proceeding from the assumption that the difference to a human brain is quantitative, not qualitative"/>
        <outline text="Very similar to reverse engineering an integrated circuit"/>
        <outline text="Fruit fly brain is more complex than an IC but not by a huge amount"/>
        <outline text="Neurons are more analog, though, unlike an IC, hence the need for Adler's imaging"/>
        <outline text="Highly multidisciplinary team to bridge that difference between simply logical circuits and neural wiring"/>
        <outline text="Article details the challenges, continuing to contrast to IC reverse engineering"/>
        <outline text="Sample preparation is tricky, given the organic nature"/>
        <outline text="Amount of raw data is immense, especially with pushing the cutting edge of imaging with Adler's help"/>
        <outline text="Looking to use machine learning to deal with data glut"/>
        <outline text="Characterize the images faster to cut down on what needs to be stored"/>
        <outline text="Article entertains speculation of building thinking machines with these &quot;wiring diagrams&quot;"/>
        <outline text="Regardless, this is some pretty hefty basic science"/>
        <outline text="Benefits are not hard to appreciate, from understanding disease to various aspects of human cognition"/>
        <outline text="Wouldn't necessarily need to fully understand the human mind to realize these benefits"/>
        <outline text="Also the technological innovations, like improvements in imaging and data processing"/>
        <outline text="Sort of a Manhattan project for neuroscience"/>
      </outline>
    </outline>
    <outline text="tail -f" Offset="34:49">
      <outline text="Media Defender defends syn attack on Rev3" Offset="35:09">
        <outline text="http://feeds.wired.com/~r/wired/topheadlines/~3/303096647/mediadefender-d.html"/>
        <outline text="CEO claims it didn't realize who they were targeting"/>
        <outline text="Claims the tracker in question was serving a large volume of pirated content"/>
        <outline text="Basically doesn't sound like MD is owning up to anything"/>
        <outline text="Louderback's point that MD should have investigated and tried contact is still valid"/>
        <outline text="His hyperbole about air traffic control is a bit unwarranted"/>
        <outline text="It is enough that MD took out the operations of a legitimate business, beyond the scope of taking out a pirate tracker"/>
        <outline text="With the FBI definitely investigating, there may be more to come"/>
      </outline>
      <outline text="DTrace improved in latest OS X update" Offset="36:19">
        <outline text="http://blogs.sun.com/ahl/entry/apple_updates_dtrace"/>
        <outline text="Original post described experiments that showed some DTrace code failed"/>
        <outline text="Tracked it down to happening when certain applications, like iTunes, were running"/>
        <outline text="DTrace is a development tool original developed by Sun but adopted by others"/>
        <outline text="Apple shipped it with Leopard prompting Adam Leventhal's investigation"/>
        <outline text="Recent update, 10.5.3, improves DTrace behavior, Adam shares some empirical results"/>
        <outline text="There are still some issues, though Adam doesn't lay these at Apple's feet, per se"/>
        <outline text="One that he does is that application names are still redacted"/>
        <outline text="Correctly points out that other tools, like Activity Monitor, show these applications by name"/>
        <outline text="Asks whether revealing just application names in DTrace would be any different"/>
      </outline>
    </outline>
    <outline text="Outro" Offset="38:40">
      <outline text="Contact me">
        <outline text="Email to feedback@thecommandline.net"/>
        <outline text="Web site at http://thecommandline.net/"/>
        <outline text="IM to command.line@skype"/>
        <outline text="Listener comment line is 240-949-2638"/>
        <outline text="del.icio.us tag is &quot;for:cmdln&quot;"/>
        <outline text="http://twitter.com/cmdln"/>
      </outline>
      <outline text="I'd like to thank libsyn.com for AAC hosting and Wouter de Bie for MP3 hosting"/>
      <outline text="These notes and the show audio and music are covered by a Creative Commons license">
        <outline text="http://creativecommons.org/licenses/by-nc-sa/3.0/us/"/>
        <outline text="Attribution, non-commercial, share alike"/>
      </outline>
    </outline>
  </body>
</opml>
