What’s Next for Dev Derby?
By all accounts, our first go at running a Dev Derby was a success. While we didn’t emerge from the day with a production-ready application, no one involved thought that would happen. What we did do is meet a number of new developers (most local to Bloomington), highlight a need for technological intervention into the non-profit sector, and lay the foundation for doing this all again. (FYI, the 2010 Dev Derby Developer Challenge (PDF) can be found online.)
My thinking right now is that Dev Derby could happen 2-4 times a year, not just in conjunction with The Combine. It would intentionally make use of the artifacts from past events—submitted code, design documents, visual design, and marketing plans—to seed future events. In addition to generating and iterating code, we could have similar challenges for designers (to create research and grounded concepts) and marketers (to suggest plans for effectively implementing the creative and technical work).
Rather than getting everyone together in person to do a post-mortem, we’re going to attempt to solicit feedback here, through the comments attached to this post. The developers panel Saturday afternoon was a great start, revealing some preferences for more lead time with the challenge and some team building prior to the coding event. We are very interested in hearing from: Participants (Developers and Designers), Judges, Organizers and Volunteers, Fans and Observers, Businesses and potential sponsors, and Non-profit organizations and volunteers.
In particular, we are hoping to get some discussion going on any and all of these topics:
- Should we do this again?
- How often should this occur?
- What is the ideal length for a similar coding challenge?
- How much lead time is needed to prepare for the challenge?
- How would design and business best be included in future challenges?
- Are the monetary incentives needed to get participation?
- How do we promote iteration of the resulting code?
- How do we improve gender and cultural balance in participation?
We welcome open discussion here, as well as any additional input emailed to info -at- devderby.com.
What do you think we should do to improve Dev Derby?
jfalkenthal said:
Sep 20, 10 at 2:28 amI guess I will begin this conversation.
We should definitely do this again and I think that as we gain experience during each event, it will get better and better. How often the events are held should reflect the interest in the events.
I would be particularly interested in scheduling a “design derby” a month prior to each Dev Derby. The format would need to be a little different than a 6 hour design session if the intention is to produce a quality challenge for an upcoming Dev Derby. But as we saw from our experience this time, it is possible to produce a design document in a week in under 10 hours of work that is in a state ready for the Dev Derby competition to use. We have enough designers in the area that we should be able to find more people interested in it than we did when recruiting for the design team this time, if it is during a different time of year. Being able to show the final design documents that were used should help in recruitment.
It would help both the design derby and dev derby if there was more recruitment and partnership with the School of Informatics and Computing in recruiting their students to participate. Specifically if faculty could assist in promoting this opportunity. As a student at SOIC, I was very disappointed in the level of student participation. There were quite a few students on the Java team from the cognitive science program. These students didn’t have experience in developing web applications and yet were interested in learning from this experience. The students that I tried to recruit to the competition didn’t feel like they had the experience to compete, when in fact, they probably knew enough. So either we have a confidence issue or our recruiting efforts need to be revised to emphasize it as a learning experience. All of the professional developers involved talked about what a great learning experience it was for them. I’d like to instill the value of service-learning and lifelong learning within our students at SOIC. Perhaps partner with the SERVE IT Community IT Clinic to.
Front-end teams. It appears that each team could have accomplished more if they had people dedicated solely to working on the front-end development. I think this would be easy to recruit for if there was an application process specific to front-end developers and then the dev team leads got to bid on their choices. Front-end developers could also emphasize if they had an interest in working with a specific team. Students should feel more confident and experienced in front-end development in XHTML, CSS, etc.
An alternate that I heard a few developers request was for the front-end to already be completed prior to the contest. I understand that this wasn’t the intent of the contest because we wanted to allow freedom for the teams to decide one of many paths to go, but I think the developers might prefer this.
Design/research consultants or team members. We had discussed this prior to the competition. Again, it would help the dev teams do what they do best: backend.
Questions for the developers:
What kind of challenge are you looking for?
Is being able to have freedom to choose between suggested features to include something you care about?
Honestly, I understand that it isn’t possible to complete an application in 6 hours, but I would find the competition more appealing if it was possible to create a more finished product during the competition. Perhaps that could be accomplished by decreasing the scope or increasing the time.
As far as diversity goes, I haven’t met a local female developer that isn’t affiliated with the university. I would love to be corrected on this and for someone to point out that we have an abundance of female developers in the area. If there are, we probably need to think about what we can do to increase their participation in all of our community technology events such as Geek Dinners and Tech Lunches before we start trying to recruit them into more formal events like Dev Derby. We might want to consider how we can strengthen this community while be inclusive of minorities.
As is common for other events that are trying to increase female participation, offering free childcare services for the day might help.
It would be easier to recruit amongst college women if the deadlines for applying were announced a little farther in advance of the communication pieces going out for recruitment.
I hope that’s enough thoughts to start a conversation.
maxbeatty said:
Sep 20, 10 at 4:11 pmI wrote some thoughts on my Dev Derby experience here http://maxbeatty.com/blog/2010/09/dev-derby-2010/
Beyond that, I think monetary incentives and the application process give the competition credibility. Participants shouldn’t get caught up in the exact monetary value since building tools for the non-profit sector should have an intrinsic value.
Having a dedicated front-end team member may not make sense for every team. Our PHP team identified two people to develop the front-end because we were using a MVC framework. Someone who only knew HTML, CSS, and JavaScript wouldn’t have been much help without knowing anything about the framework we were using.
Ultimately, each team should be responsible for limiting the scope of the project. A team of 8 is going to produce more than a team of 2. The requirements shouldn’t be catered so that the smallest team can complete them. The smaller teams (and even the bigger teams) should cut out what they know they won’t be able to complete so they have something to present. All of the teams, including mine, did’t focus enough on having something to present at the end of the 6 hours.
Like I said in my blog post, I’m really glad I participated and would do it again.
Kevin, is the code going to be released on GitHub? I think creating an organization on there for participants would be a great way to continue development in our free time.
Kevin Makice said:
Sep 20, 10 at 4:32 pmYes, it is definitely going to be released. I’m waiting for the legal team for Humanetrix to give me the appropriate license text and a green light. I’ll be speaking with the Team Leads to make sure I’m clear on what I’m releasing to the world. I hope to have this done by the end of this week.
Ditto on the awards.
lawrence said:
Sep 23, 10 at 11:38 amI would recommend:
- having the design team do more, including generating basic HTML templates/CSS
- have the programming teams do less; have them build out a fully functional CRUD app IN ADVANCE of the competition
Reveal the challenge on the day of the contest, and expose a set of more limited, and purely functional challenges:
- implement full text search “by hand” without using plugins or preexisting solutions
- create a batch processing workflow queue for arbitrary work items with complex precondition/postcondition requirements
In other words make it more of a programming competition not a contest to slap together a pretty GUI and a bunch of CRUD screens. Because that’s the type of thing that there are plenty of tools to assist with. That’s probably why I won: I planned in advance what tools we could use that would allow us to generate a basic CRUD app as rapidly as possible. http://www.lmcalpin.com/post/1109835129/dev-derby-2010
Limiting the scope of the contest would also be beneficial since it would allow us to make the judging more objective:
- Selenium test scripts could be used to verify that the GUI behaves to spec in response to changes in back end state
- expecting teams to create web services that expose internal state could be used to verify other system behaviors