Team Murder No Brain No Headache.

15Aug/06Off

Underneath The Drop-cloth

Found via Jeremy Zawodny's always entertaining and time squandering linkblog today, at least for me who reads his aggregator data exploded in a browser while riding the bus, was this article about the hype/marketing versus reality when building application with frameworks. There are a number of really good questions that need to be considered when evaluating these sorts of claims in this post. It's also a pretty realistic summary of the sort of toil you are in for when designing an application that replaces an existing one especially when you have to carry all the data with you. He also made some decisions along the way that cost a considerable amount of time (the decision to switch to PostgreSQL during development) and some tasks that most sane folk wouldn't even consider like rewriting chunks of API (!) that didn't suit his purposes. If you've ever been faced with the choice between rolling your own or using a large framework or library when doing a project and ending up somewhere in between most of this will sound ominously familiar or possibly reopen some old wounds or at least headaches.

The only disagreement I could really find with his methodology is the question raised about using the appropriateness of the toolkit as an issue for debunking when looking at the amount of time that it require to build an application using it. You should be doing most of this before you decide to hand your life over to a framework. This is what becomes so worrisome about addressing all problems with a particular set of tools. Rails seems to be gaining on the Perl abuses of the good old days as it gains popularity and all of the trappings of duh-ness that come with getting religion all involved in your programming languages.

An side order of diversion-jitsu and disclaimer-fu here that needs to be made before the flaming begins: I'm blaming popularity for the abuse and zealous overuse and not the actual language here so drop it before it gets hot. I know it write things in far too many words but I expect you to read and comprehend some of them before I'm going to take any of your ham-fisted 'corrections' to my woeful ineptitude in the properly appraising the One True Flavor Of The Month according to the One True Source Of All Wisdom For Imbecilic Ponderers of Alleged Truth About Things Read About or Otherwise First Considered Ten Minutes Ago terribly seriously. Yes, too many words now please move on before I spill any more in anticipatory defenses of stupid, stupid acts performed by the invisible hordes of idiotic barbarians against my honor. This reminds me that I must get out the CrankyPhone at nearest convenience to demand Southern accent tags from W3C immediately, posthaste, and with damn-the-torpedoes style speeds of delivery... Please choose wisely your hammer oaf-y. Um, so, I kinda lost my train of though in approaching this paragraph or whatever other funny name you would like to give this train wreck of a construction so I will move on immediately before my word count exceeds that of the weblog post I intended to make a brief mention and quick commentary on...

I do think that one of the more enthusiastic positive notes for a first application developed in Django might be the inclusion of a link to the source code for the project which we'll have to shave a eighth of point or so off the overall delivery as it is mentioned as comment appended to post. Most first time developers in a new framework are too busy recovering from their newest ulcer and explaining copious new night sweats to their significant others to bother posting the actual source and/or discussing it in great detail. I think that more explicit discussions like these with some of both good and bad are more valuable as advocacy for a given framework or platform than marketing talk could ever be.  

Comments (0) Trackbacks (0)

Sorry, the comment form is closed at this time.

Trackbacks are disabled.