Check out this video. I think that this guy, Sean Kelly. He talks about developing applications and the difference between UI development and otherwise.

In the video he compares & contrasts some web frameworks (concluding that J2EE sucks… even with hibernate). The main take away for me is that the most important property of a UI framework is it’s ability to facilitate rapid deployment and, more importantly, rapid *re-deployment*.

I think this is important.

Should python/ruby/php/java code be written well always? Yes, but that doesn’t contradict fast development.
Is python/ruby/php less robust than java? Yes, but it probably doesn’t matter.

I have more to say, but this is probably too much for a comment. With the blessing of the congregation I’ve been inspired to continue my response in my *first post* in my blog…

I suppose my main hope was to get people thinking about time consumption and the reality of certain projects. Presumably funded companies are going to last a while, making good code much more valuable. Presumably personal projects probably won’t last a while, which means one of two things: either you learn how to write good code from the project or you get a prototype out fast so you don’t waste your time writing code that you’re not interested in learning from.

Thanks for the feedback!

I don’t believe that funding is the most critical issue in the development method that should be chosen for an individual project. There are too many factors to list, but I believe the most critical criteria differentiating the two projects you have listed here is scope; helpd is a far less complex app than cellarspot as far as I can tell, and I don’t imagine cellarspot would have done well coded in inline php. Poor coding or design choices will come back to bite at exponentially increasing degrees as the time spent on a project increases.

In addition, the framework used in constructing the project is a critical factor; in what little I’ve seen of JSP, it’s far more time consuming to write and far easier to create terrible code for than a framework like Ruby on Rails, which is great for far more than prototyping (barring a number of performance issues).

Overall, I don’t think sacrificing code quality in favor of super-agile development is ever worth it unless the project in question is extremely small-scale.