So very tired as I'm writing this post, so it may not make much sense.

If art is the manifestation of human creativity, is software the most dangerous canvas for creation?

Like other creative forms, software allows you to impact the lives of people you've never met (I remember the best email I ever got from an Audiomatch user was, "I met my girlfriend through your site"). It's also a medium for publicizing your political and social beliefs (look at Tor, or even Napster).

The one way that software differs is that it's never done. When a photograph is exposed, that's it. When a CD ships, it's out there. If an artist doesn't like how a track is mixed, the only alternative is to re-release it. A book is written and printed. Even if an author/artist doesn't like what gets released, there's very little recourse to "fix" it up.

But software allows you to constantly refine and hone. If a particular part of it sucks, you can go back and change it. Software evolves with a series of microevolutions. (And if you're a big fan of RAD, then this is even more pronounced). It evolves with the thinking of its time through its developers. This is dangerous though - the evolution of ideas over time can marginalize and diminish the rawness of an initial idea. The constant nitpicking might take something great and make it just average.

At some point, does niche software reach a point where it doesn't need further improvement? Where it achieves all the goals that it was intended for and not more? I mean, I look at apps like WordPress - sure I think it sucks, but at some point, aren't you just bloating beyond the initial vision and getting marginal returns for the work you do? The natural progression after that seems to be to "reach out" to other related areas.

Commercial software is excluded from this discussion - Facebook has a clear interest in monetizing, so cramming every known feature is a net-positive. Google intends on knowing everything you do, hence expanding the surface area is good. Even MindTouch Deki finds itself spreading itself thin.

What I'm talking about is the pursuit of software projects for the sake of simple creativity. The creation of a project which is beautiful in code or aesthetics; how its one goal is to just make the user happy. I find myself running into less and less of these sites, which is odd, because the notion behind projects like Ruby on Rails is to allow not-as-great programmers create these simple sites easily. Back when all web software was more amateurish, things were a lot more fun.

The closest I've been to this is with listfoo - I created it to make lists pretty quickly and reorganize them. And the final version (which the color/UI scheme isn't what I wanted) is the closest, functionally, to my original dream. Tabulas and Audiomatch were nowhere near what I inteded them to be (I'm not even sure I know what those projects were supposed to be know) There will never be another release of listfoo - it is what it is, and it's limitations and bugs are a part of its character now.

The problem is that people keep creating sites with the intent of monetization, so everything is supposed to scale out - from an architecturual side and from a UI side.

This is why I love sites like longurlplease.com - a simple idea ("short urls are bad for the web"), with a very simple site and a simple task. It doesn't try to be the reverse TinyURL by tracking traffic and pandering itself to a lot of page views so it can place Google Ads on it - it just exists.

We need more of those types of projects out there. Maybe I'll get off my ass one of these days when I'm not pursuing one of my ten thousand pointless hobbies and try to create a short simple project.