Keep Reading!

You're not just sitting around doing nothing, right? You're reading all these fascinating documents on the Internet, aren't you?

• Last week was the EuroSys 2012 conference in Bern, Switzerland. Happily, the folks at the University of Cambridge Computer Laboratory's Systems Research Group put together some great summaries of the research that was presented:
  • Day One

    There is also some stuff to do with constrained execution, to avoid variable-length instruction sequences being misinterpreted and doing dangerous things (I don’t fully understand why this is relevant, but I missed a bit of the talk being distracted).

  • Day Two

    To reduce DRAM latency, they constructed a lock-free, unbalanced 4-way tree with the same concurrency properties as a binary tree, but is only half as deep for the same amount of data -- plus each node fits into a cache line.

  • Day Three

    Of course, the migration business is somewhat challenging with regards to atomicity and consistency. They address this by moving to a fine-grainedly locked (per row) hash table for requests, which is only 2% slower than the coarse-grained version in stock Linux.

  Overall, it looks like there were some great papers presented at the conference, with lots of interesting stuff to follow up on.
• Phrack 68 is out, and XORL, a blogger from Greece, gives his usual great synopsis of the contents:

  We all know that such rootkits are backdoring Androids in the wild for quite sometime and h0h0 has even made a presentation on it at DefCon in 2010, but it is always good to have some technical documentation to get started with.

• Scott Hanselman writes a stellar article on Facebook's complex privacy settings: Facebook's privacy settings are too complex for ANYONE to use - Change these settings today

  Under tagging you can choose what happens when someone tags you and tags that friends add to your own posts or photos. You can also control tag suggestions.

• Kenneth Iverson's book on the APL programming language: A Programming Language, has reached its 50th anniversary, and the book is online.

  The systematic treatment of complex algorithms requires a suitable programming language for their description, and such a programming language should be concise, precise, consistent over a wide area of application, mnemonic, and economical of symbols; it should exhibit clearly the constraints on the sequence in which operations are performed; and it should permit the description of a process to be independent of the particular representation chosen for the data.

Happy reading!