Software Development from Ryan

There's lots of great development tools coming out and a lot of interest in building services for people. There is also the "build it and they will come" mentality everywhere.

More often than not it doesn't seem to work. Fold.com has recently closed its doors because it couldn't keep up with competion. Micheal Arrington's comment on fold.com's closing has a good point about building applications:

"Well, the inevitable is starting to happen - a few new web startups are starting to close up shop as they find that building an application is a lot easier than getting users to try it out, and keep coming back" - Micheal Arrington

Innovation in development tools and techniques has a value ceiling. I had a discussion on Joshua Wehners blog this week about the excitement behind Ruby on Rails and why people should not be so excited. Micheal Arringtons point clarifies this again. Technology can get very good, allowing you to make better apps, but everyone is going to be making better apps. There are a lot of areas beyond the tool that you use that you could use to differentiate yourself from the competition to attract people and retain members.

 I've been playing around with different techniques over the  last month. On whirlpad.com, a free travel blogging service I built earlier this year, I introduced a Travel IQ test called the TQ test. It measures your travel knowledge by having you match photos of toilets from around the world to their location. With a bit of seeding this test consistently brings 1000 uniques per day. This is quite a bit more than the homepage of whirlpad.com.

The interesting thing here is that the bottleneck to creating more tests, and getting more traffic is not the technology, RoR or LAMP. They really don't matter. The important work is in the creative effort to develop the test. The travel blogging service, however, has a lot more development work than marking/PR/people-attracting work. 

Being a good developer and usuing the greatest newest technology is not enough anymore. Besides, that type of work can be easily outsourced. Just because its easy for me to grow tomatoes in my backyard doesn't mean I should start a business around it.

If I had to do large projects with a LAMP framework I would go crazy quickly. There is a lack of organization. Or, maybe its just that I've never worked with any professionals who use this and haven't picked up any trade skills. PHP sucks. Its a half baked language. Why do scripting languages get popular and think they can take on the world?

Over the past six months of a working vacation I've had the oportunity to ditch C++ software development and dive head first into making large systems with LAMP.  I've learned that doing large scale systems with lamp is a bad idea. PHP alone is not enough to build a solid structure in a large system. People, including myself, use LAMP because its fast and cheap. And I can run in on a $10/month 1and1 account. So what do you do to take advantage of the cheapness and spead and still build manageable large systems? You build a language on top of php.

Everyone does it apparently, and most people say its a bad idea: to build a template framework using php. But I did it and its been nice to work with. So far I've built these sites with a template framwork with almost zero php or sql code:

1. whirlpad.com
2. canpost.org
3. focalpost.com (canpost.org 2.0)

Number 4 is in the works.