Author Archives: Dan Siemon

A common archive format for web forums and email lists?

Here’s a little wish list idea for someone with more time than I to work on.

Since the idea came from the use of Usenet it is probably best to start with a short description of what exactly Usenet is. Usenet is a method for large groups of people to communicate about particular subjects. These discussion groups are divided into hierarchies, similar to how domains are divided. For example, the comp hierarchy contains comp.os.linux.advocracy, comp.os.solaris etc. Whatis.com has a definition of Usenet that may be useful. Anything you can possibly imagine, and more, is discussed on Usenet. In the earlier days of the Internet Usenet was the primary place for technical discussions. Unfortunately, this has changed as more and more people use email lists and web forums.

Google maintains a huge archive of Usenet posts going back many years. They claim to have over 1 billion messages in their archive. Using groups.google.com you can search this archive. Anytime I have a technical question, particularly for programming and networking problems, I always start by searching Usenet. The main reason for this is the fact that all discussions are archived in such a way that you can always see the entire thread and easily move between messages. This is particularly useful when searching for a question. Finding a post that asks the same question is useless if the associated replies that may contain a solution cannot be found. Try a search for “aes vs twofish” at groups.google.com. Clicking on any one of the results will allow you to view the entire discussion thread.

Fast forward to 2005. As the technical abilities of the average Internet user has dropped discussions have moved from Usenet to mailing lists and web forums. This change is happening because users already understand their email client and web browser and have little desire to find a Usenet client or discover the Usenet features of their email client. The problem with this trend is that finding information is now much harder. Try the “aes vs twofish” search with the Google web search. The first result I get is a message called “AES256 vs Twofish performance (Was: twofish keysize)”. This is a email that was sent to the GnuPG users mailing list. Once you follow the link Google can no longer help you. You are limited to whatever features the mailing list archive offers. Some mailing list software provides decent search features but most do not. Web based forums are usually even more difficult to use. Many are ugly, slow and certainly do not present a consistent interface across archives that would make finding information easier.

In order to bring these discussions back into a form where search engines can do what they do best we need a mailing list and web forum archive format. Search engines could pull the archives for each list or forum and present a consistent interface like groups.google.com does.

So that is the task I set out. Define a discussion archive standard and convince all web based forums and email list software providers to support it. The search engines will follow soon after.

LQL 0.7.0 released

Another new version of LQL is available.

0.7.0 changes:

  • Add some new test programs to the tests directory.
  • Fixed a bunch of small bugs that were found with the new test programs.
  • Change the return type of a few _new() functions from GObject to the proper object type.
  • Update docs to match new API.
  • Add some background comments to the documentation.
  • Add support for the DSMark QDisc and corresponding documentation.
  • Add –with-kernel-source=PATH option to configure so alternate kernel include directories can be specified.
  • Add support for the Netem QDisc (everything but distribution tables). This means you now need the headers from a kernel with Netem support in order to compile LQL.
  • Add support for the TCIndex classifier.
  • Put some more time into the classifiers. Still not complete.

LQL 0.6.0 released

I just finished a new release of LQL. This release contains the statistics support I have been working on. The basic statistics are available for each queueing discipline and class but I have not added the extended statistics that are specific to each qdisc/class type yet. Adding support for these statistics will just require modifying a single function in each type. I’m not sure when I will get this done as I don’t need this feature at the moment.

Changes:

  • Add statistics support to all qdiscs and classes. See the get_stats.c example.
  • Add print() method to LQLQDisc and LQLClass so that it’s easy to see their settings.
  • Add printStats() method to LQLQDisc and LQLClass to output all statistics information.
  • Lots of small bug fixes.
  • Clean up and add new documentation.

Lord of the Rings

For the past three years it has become somewhat of a Christmas break tradition to see one of the Lord of the Rings movies in theatre. I managed to be at the opening night of all three releases. Obviously, that was not possible this year. Instead, in what may be a new tradition, Bob, Kier and I sat down to watch all three of the extended editions back to back last Thursday. We started at noon and finished just a few minutes before 1:00AM.

The extended edition of the The Return of the King is about four hours and twenty minutes long. That’s fifty minutes more than the theatrical release. After watching the movie I really don’t know where all of those fifty minutes were added. The additional footage in Return of the King seems much less obvious to me than it did with the Fellowship of the Ring and Two Towers extended editions.

If you can set aside the rather large amount of time required to watch all three movies I highly recommend it. Besides being a relaxing day there is also a great sense of continuity watching the movies together.

Holiday time

Well, another semester is over. I can’t believe I only have 4 months of my undergraduate degree left.

Besides taking a few days off I will also be putting a large amount of time aside for hacking on LQL.