Ironcoder is a most excellent idea at its core. The idea is this: we’ve all seen the extraordinarily clever and cool hacks people create at those conferences. It’s a great way for programmers to relax, unwind and work on something completely a) different and b) pointless.
So while Ironcoder has a great idea, it has so far been unable to live up to the potential. Well, what’s wrong then? Too many, and the wrong, constraints.
Ironcoder is a contest for glory and not money. The limits and gradual unveiling of an API and a theme are good ideas in theory, but not in practice. The choices this far have been a bit out of left field: Ironcoder 0 used the Accessibility API + Mardi Gras, and Ironcoder 1 (which just finished) used iTunes visualization + Emily Dickinson. I’m not generally opposed to choices out of left field, but in case 0 it generated a line of entries that felt very forced, and in case 1, no one besides Gus Mueller even submitted an entry (my guess is that five or so others tried, but gave up because of the untenable scenario). Note that conference contests have genres - you generally can cook up whatever you like and submit it in an anything goes pile, or focus on one area (which is still miles broader) if you know exactly what you want to do.
The part that really wrecks it, though, is the 24 hours limit. On a conference, you really do have next to 24 hours at your disposal. Your hack, if you want it to, has your undivided attention, and for 10 hours if you want it to. Even though Ironcoder is held on weekends, I doubt you’ll get above 5 hours of time to do it in.
Again. Constraints are awesome. They help you work better and faster and oftentimes makes the results that much better. But the wrong constraints, and you’re out the window. If the current trend continues, Ironcoder 2’s requirements will read the programming equivalent of “I want a corporate-looking, professional web site. You have five minutes, the colors lime and pink, the fonts Comic Sans and Wingdings, and omnipresent heavy metal umlauts.”
I, as well as anyone, know that Ironcoder draws inspiration from Iron Chef as well, which mandates more than the conferences do. However, I believe that as it stands, the constraints are wrecking the contest completely. Something I believe to be essential in a revised rules set will be for the judge to make a first non-competing entry on his (or her) own, to prove that the constraints work, and that there’s leeway for ideas. Five minutes after the announcement of the subject, I guessed my way to the winning entry, and that simply shouldn’t be the case. This could be so much more awesome if the concept was fine-tuned a bit, and I believe it’s in capable hands.
A final note: this entry means no offense to Jonathan “Wolf” Rentzsch (the judge and consequent “constraints master” of the 0th, inaugural contest), Lucas Eckels (ditto for 1st contest on account of winning 0th contest) or Gus Mueller (ditto for 2nd, 1st contest) or any other contestant or participants. Just writing my thoughts, as always.
Interesting take on the contest. I participated on the 0th one. While I think a good constraint can really do a lot for creativity, I do agree with some of your points. more than the constraints though, I think it needs to be publicized a lot more. Neither of the two contests even had press releases sent out. It’s hard competing for fame and glory, when there’s nobody watching :)
By Andy Kim · 2006.05.23 04:14
I think a press release is overdoing it - the people who are interested would rather find out about it from a link than a press release. It does need to be publicized more, true.
By Jesper · 2006.05.23 06:08
Hi, thanks for taking the time to post this. Feedback is crucial to making further ironcoders a success..
I think Ironcoder v0 was quite successful. Wolf’s choice of the Accessbility API was a stroke of genious that would be hard to live up to. Some very innovative entries were made. In fact one of them is being worked on further and will be released as a product.
I think the second Ironcoder (1) was less succussful not because of the API and theme selection but because of the lack of time spend advertising and spreading the word. The initial ironcoder got a lot of interest amongst Mac bloggers, perhaps because of novelty value. A lot of the #macsb regulars (myself included) happened to be out of town/busy the weekend of the second ironcoder.
I dont think anyone is particularly convinced the 2 x 24 hour format for ironcoder is a winning formula. The 24 hours after API selection isn’t really too useful, especially if the API is known to a developer ahead of time. The 24 hours after theme selection can often be too short for a developer to actually do much (especialy if they’re trying to renovate a house at the same time ;-).
I’ll be posting a post mortem entry on the ironcoder blog about the last contest. Hopefully we can do whatever it takes to make the next ironcoder a succcess.
By Jonathan Wight · 2006.05.23 15:06
I think the killer on the second was the Emily Dickinson theme. Creative and well-intended, yes. But perhaps a little too narrow.
It might have been better to open up the theme to “poetry” in general, which probably would have meshed quite well with the API and led to lots of creative ideas.
By Jon Hendry · 2006.05.29 06:20