While doing doing a project prop­erly from the start is a fine goal, there’s some­thing to be said for just ship­ping. It’s pretty rare that that you have all of the require­ments from the get-go and you cer­tainly can’t pre­dict every prob­lem that’s going to arise while your project is in pro­duc­tion. If you attempt to antic­i­pate every fea­ture you could need it’s likely that the project will never be fin­ished, it’s cer­tain they you’ll waste a lot of time on uneeded fea­tures, and it’s still pos­si­ble that you’ll miss fea­tures and bugs that are impor­tant. In fact, it’s more likely that you’ll miss things as many issues don’t arrise or make them­selves appar­ent until you reach production

It’s best to ship early and iter­ate reg­u­lar­ly. Remain light and don’t invest too much in any par­tic­u­lar iter­a­tion and you’ll be much more able to address issues as they arrise. Ship ear­ly, ship often, adjust as you go. Sim­ple. So sim­ple, it’s com­mon knowl­edge, really.

I used to have a bad habit where I gen­er­ally would­n’t take this advice. I’d think I was tak­ing it but I’d even­tu­ally real­ize that my 1.0 was really way more com­pli­cated than it needed to be. This was a mis­take because I ended up wast­ing time but it was a mis­take dou­bly so because some­times the project when nowhere. Wast­ing two or three weeks on a dead end project is a hor­ri­ble waste, espe­cially when a two day mockup would be enough to test the fea­si­bil­ity of the pro­ject.

I remem­ber one project I worked on for months. Tech­ni­cal­ly, I’m still work­ing on it but it’s not any closer to release than it was a year ago. It could have been a lot sim­pler than I made it but my ambi­tion got in the way and now it’s noth­ing. If you look at the projects listed here on my site how­ev­er, you’ll notice that they all are very sim­ple, some are even half-baked, but they’ve proven more suc­cess­ful than any overly ambi­tious pro­ject.

So, my goal these days is to keep things sim­ple and iter­ate if nec­es­sary. The sim­plest thing that could pos­si­bly work, works at least and most of the time that’s all that’s nec­es­sary.