Django from the trenchesAugust 21, 2009...... at 9:12 am | Posted in geek | Leave a comment
I walked into the office this morning to an email from one of my colleagues asking me for my opinion on Shabda’s article A response to Dropping Django. The article makes some interesting points and re-affirms my opinion of Shabda as a top-notch python developer. However, I also feel for the points made by Brandon Bloom as they mirror some of my own experiences.
In 2007 I sat down with 2 of the best python developers I am ever likely to work with and over the space of two weeks hashed out a development platform. They were firm TurboGears enthusiasts but I already had working Django code in my repository. What we came up with as an acceptable compromise was Genshi templating (sandboxed), FormEncode validation, TurboGears views and SQLAlchemy ORM. Fourteen months later we launched a kick-ass web application.
In 2009 I sat down with a web developer who had never written a line of python and guess what? I chose Django. Three months later we released JobCrystal. Top flight developers, probably shouldn’t be using any framework other than the one they’ve been involved in rolling themselves, because anybody’s else’s framework is going to itch. For the rest of us mere mortals who aren’t interested in solving authentication again, or url dispatching again, or caching again – Django makes some really sensible choices that in my experience allow you focus on the real issue of delighting your customers a lot quicker than would otherwise be the case. Once you’ve got delighted customers, sure you can go back and re-engineer the parts that don’t feel quite right (I’m dying to play with Django’s contenttypes capabilities to get a higher level of abstraction), but for most web applications the focus has to be on getting to market as quickly as possible and Django provides more of that than any other framework I know.