This post originated from an RSS feed registered with Ruby Buzz by Jim Weirich. Original Post: No Fluff, Just Stuff (Sunday) Feed Title: { | one, step, back | } Feed URL: http://onestepback.org/index.cgi/synopsis.rss Feed Description: Jim Weirich's Blog on Software Development, Ruby, and whatever else sparks his interest. Latest Ruby Buzz Posts Latest Ruby Buzz Posts by Jim Weirich Latest Posts From { | one, step, back | }

A Word on Organization

The Saturday writeup mentioned a mixup with Hotel venues before the conference and may have left an impression that the conference was unorganized. Quite the contrary. Everything was very well put together. Seminars started on time, ended (more or less) on time, and the break times always had cookies available. As far as I could tell, Jay Zimmerman ran the whole show with the help of one assistant, and did a very fine job. I was very impressed with the little touches. For example, the binders everyone got when they registered was personalized with their name on the front cover, and the summary of the conference schedule on the back.

You should also see ChadFowler's comments on the conference.

Building a Data Persistence Tier — Leveraging Object Relational Bridge (OJB) (John Carnell)

Sunday morning started out with a double session on the OJB Object/Relational mapping tool. OJB uses an XML mapping file to define the relationships between the object model and the relational database. I see the value in the O/R mapping, but it seems to me that the mapping file can be quite complex, and that worries me to some degree. XDoclet can probably help here, but then that’s yet another tool in the mix. (update: Ryan tells me he has download OJB and started looking at the tutorials and was very impressed with them).

John started out the session with an example of code that had a subtle JDBC bug. The code forget to close a prepared statement and the program would receive an "open cursor" error under heavy load. At the break, I couldn’t resist pointing out to John over the break that Dave’s Ruby talk showed how to address that problem in Ruby so it can’t happen by accident.

• Cool Quote: Beware of refactoring through search and replace.

Lunch Time Quiz (Jay Zimmerman)

After lunch time, we played a little question and answer time with Jay giving prizes. Our table got a J2EE question, but unfortunately we had no J2EE experts sitting with us and we were clueless. Jay was nice enough to give us a second chance with a simpler question on ant. This time we got it right and won a pen.

The next game one where we guessed a phrase that was presented in a Wheel of Fortune style format. We tried to answer as Jay slowly filled in the letters. I actually won a Compuware polo shirt in this game. Great fun.

Applied Ant (Erik Hatcher)

I missed Erik’s intro to Ant, but did catch the follow on session. This talk mainly presented a bunch of different targets and goals that ant was capabile of handling (some with additional jar files).

Some things I came away with …

• I really dislike reading XML, and especially dislike having to manually maintain XML files.
• I use a small Ruby script called rake to do a lot of my builds. Although rake is not nearly as mature as ant, it is interesting to compare the design choices. Needless to say, rake does not use XML as input (rakefiles contain normal Ruby code).
• Ant has a lot of really cool build targets that are well suited for Java development. It would be worth my time to study these (for rake enhancements).
• Someone said that the native C/C++ compilation in ant is very nicely done. Ok, here’s another thing to look at in ant.

Agile Practices for Code Reuse (Maciej Zaawadzki)

Its getting late and I must confess that my information overload meter was in the red for the past two sessions. I only heard about half of what Maciej had to say. He has some good ideas in this area. I especially like his idea for a code station, a binary artifact repository. It sounds a bit like source control applied at a binary file (e.g. jar file) level.

The Conference Closes

Somehow for me, the end of a conference is rather anti-climatic. The last session is over and everyone just … walks out the door. It is a bit of a let down. But by Sunday afternoon, everyone was dead tired, and I know a couple folk left before the start of the final session.

Was the weekend worthwhile? Absolutely. Few technical conferences ever make it to Cincinnati, so I applaud Jay Zimmerman (the conference organizer) for bringing his road show to the "heartland". It really makes this kind of training more available to your average programmer. If a NoFluffJustStuff conference comes to a city near you, I strongly recommend that you grab everyone in your development organization and take advantage of the situation.