Boise Open Spaces Recap

Boise’s first Open Spaces event was a raging success (IMHO). Thanks to Scott for organizing the event, and Jarod for guiding us through the (non) process. I hope that we can repeat the event soon with a wider audience and longer duration.

I sat in on 3 groups, the first as a full member and the last two as a bumblebee. Here are my rough notes from each session. Most of these are just the questions that we hashed through without the answers, I guess you had to be there for that.

• Domain Driven Domain
  • Do you choose Repository or Active Record?
    • Can a dynamic language give you the quick ramp up of Active Record and the long-term maintainability of Repository?
  • How do Domain Specific Languages relate to DDD?
    • What if you create a DSL to define your business rules?
• Teaching Object Relational Mapping to junior developers
  • Which ORM do you choose?
    • • In the .NET world it seemed to boil down to a choice between
        • • NHibernate
          • LINQ for SQL
          • Castle Active Record (apparently built on top of NHibernate)
      • Java
        • Hibernate like mapping (EJB, JPA, etc…)
          • Start the junior developer off learning the domain model and try to isolate them from the object-relational impedance through HSQL
        • iBatis approach
          • Expose them to the internals of the relational database and make the mapping to the domain process manual
• Technical leaders moving into management
  • Two different tracks that exist at a lot of companies
    • Technical track
    • Management track
    • Both are intended to have similar pay and promotional opportunities
    • At smaller companies, this dual track doesn’t always exist and technical people are forced into management to chase promotions
  • As a technical leader (ie Architect) how do you stay “in the code enough” to make competent long-term technical decisions?
    • Make sure your code is architected so you can recover from poor mistakes
      • For example be able replace your persistence layer in 2 weeks, mitigating a poor persistence layer technology decision.
    • Will your developers feel like you are the non-coding architect passing down edicts from above without knowing the details or impact?
• The event closed with a great pissing match of who has worked on the largest system, I think Chris and Jarod tied for first place.