Old News

New news is on the class home page.

Week 4

• Project grading notes posted (see sidebar). It's a good idea to go through your own project as if you were grading it. Have you forgotten anything? It's surprising how often groups lose points because they simply omit something important.
• Meet in lab Thursday?
  With the due date Friday, I know there aren't going to a lot of extra mental cycles to spend on lecture Thursday. So, why don't we be realistic about it and just meet in the room 100 lab. If things are going great, you can give me an informal demo and maybe show it off to classmates. If you're struggling with something or have a question, it's an opportunity to ask about it (or maybe get a little help from a classmate ... that's perfectly ok in this class, as long as you properly credit those who help).
  OK?
• Some previous projects:
  • Spring 2002
    • Group meeting time scheduling (web-based app found free times in common for group members). 2002.
    • Simple web-based chat application. Even in 2002 there were more fully-featured chat applications, but this one had the advantage of not requiring any software installation at all, just a web browser with Java applet support. I used it for a while for my regular meetings with colleagues in Italy and Taiwan.
    • An action game. I don't recall much about it, except that the illustrations were nice (one of the members was quite skilled in Manga-style drawing).
  • Winter 2003
    • A 3D pong game ... crash and burn.
    • A skinnable pac-man game, in pure Java, with an OSU beavers vs UO ducks skin as demo. Quite impressive, but they had to work very hard to get acceptable performance from Java.
    • Linear algebra calculator. Like yet another calculator simulator, but instead of numbers it manipulated symbolic formulas. Pretty cool.
  • Fall 2006
    • Sudoku. Yeah, there are zillions, but there were some innovative touches in the interface, which was inspired by a much older computer game called Sherlock.
    • Library stripper --- Given a Java program and all its libraries (.jar files), produce a single .jar file that contains just the library modules it really needs. Although Java reflection turned out to be a bigger obstacle than expected, this was very well executed, and I hope to see it turn into a widely distributed utility program in the future.
    • Trip mapping / sharing based on GPS data. Think Flickr, but with mapped routes instead of photos. One of the team members had been cooking this idea up well before bringing it to 422, and is continuing to work on it. I'm looking forward to getting a GPS unit so I can use it.
    • Interest-based group formation and event notification. Social networking definitely seems to be a hot topic currently.

Week 3

• Lecture slides available. These will be linked from another portion of the course web site, but here they are for now:
  • Course Intro
  • Process
  • Teamwork
  • Herbsleb & Grinter paper

Script fixed (Friday 19 Jan)

Erik Brown spotted a bug in the script log_normal.awk and provided a fix. His fixed version is now the one you'll get if you download the script again. Thanks, Erik!

Week 2

• Snow again! The roads are slippery. There will be class if you are able to make it, but if you do not feel safe traveling to school you should skip it today. With luck we'll have at least one representative of each project group for the user interview.
• We will meet at our regular classroom and then walk together to Deschutes, where we will interview Prof. Lo in room 260.

• Tuesday: Requirements interview with Prof. Lo.
  (Initially I wrote "Monday" but I hope you all figured out that was a mistake, since we don't meet on Mondays.)

  We will meet at our regular class time, 10am, and have half an hour to prepare. Prof. Lo will join us at 10:30, and we'll take 30 minutes to interview her regarding Visigoth. When the interview is over, we will have 20 minutes to discuss it before the end of our class period.

  Please think about what you want to ask. It may be useful to divide each question into two parts: the general issue you are trying to learn more about, and a sharp, concrete question designed to uncover the information you want.

  More info: We may walk to Deschutes to meet Prof. Lo. We'll still meet initially in our classroom; I'll know by then whether we are walking to Deschutes or Prof. Lo is walking to Gerlinger. Either way we'll have 30 minutes of interview time, and we can talk during the walk time if we are doing the walking. [Update: It's definite, we will interview Prof. Lo in Deschutes.]

• The schedule called for presentations from groups on Tuesday. The group presentations are canceled in favor of the requirements interview. This is too good an opportunity to pass up, and we really need to do it early in the project.

Week 1

• I'm not too worried about having missed a lecture today --- I'll be able to fill that in later --- but I am concerned that we missed a chance to talk about the project and your design ideas. I encourage you to email me with ideas, questions, whatever, and I'll respond as promptly as I can. It's not a replacement for a face-to-face discussion in class, with everyone there, but it's better than nothing.

• Here is a link to a sample source file for visualization, and a small (quick and dirty) Awk script that parses the source file and produces a less compact but easier to parse format. Before you grab them, here are a couple of caveats:

  • The log file is 25 megabytes in its compressed form. When you unpack it, it will be 110 megabytes.
  • It is compressed with the bzip2 (bz2) format, because the more common gzip format only gets it down to 35 megabytes. If you run linux, you've probably already got a bzip2 decompressor program; likewise if you run MacOS X (just use the "open" command on the file to invoke the Mac archive decompressor). If you run Windows, you may need to obtain and install a bz2 decompressor (but there should surely be free programs available on the net).
  • Documentation of the file format, copied from the Mirage web site where I got the data, is copied into a comment in the Awk program. If you run any kind of Unix, including Mac OS X and Linux, you already have Awk (a simple scripting language). If you run Windows, you may need to obtain it.
  • The output of the Awk program is simpler than the input file format, but it's also much more verbose ... if you saved the data after Awk processing it might be half a gigabyte or so.

  OK, you've been warned. So here is the source data (25M) and the simple Awk script to massage it into an easier to use format.

• No class Thursday due to unsafe road conditions. I'll have more information here, including a reading assignment, later this morning.

• Reading:

  Herbsleb, J. D. and Grinter, R. E. 1999. Splitting the organization and integrating the code: Conway's law revisited. In Proceedings of the 21st international Conference on Software Engineering (Los Angeles, California, United States, May 16 - 22, 1999). IEEE Computer Society Press, Los Alamitos, CA, 85-95.

  It is best to download the paper directly from the ACM Digital Library if possible --- UO has a subscription so you should be able to do this either directly from campus or through the UO Library proxy. However, since ACM doesn't provide a URL for this paper (probably because it is published by IEEE), I have also made a local copy.

  Please read by Tuesday. We will discuss it Tuesday or Thursday, depending on whether I can arrange a requirements elicitation meeting on Tuesday.

• Our first project will be support for visualizing portions of the Gnutella overlay network, using data collected by the Mirage network group.

  A project description is now available: [PDF]. The version passed out in class was 1.426, the correct current version is 1.5. It has a couple of pictures and some links to visualization engines.