I built delicious while working a fairly grueling day job.

- I had a deck of index cards and a binder clip. Have an idea, write it down.

- I'd sit down and try to do just one thing. Look through the cards, find something easy. Even a little thing. After that it was okay to go to bed.

- Keep the code chunks small. Make it easy to page in all the state for working on that bit of code. Everything but the file that actually rendered a single bookmark fit on a single screen.

- Understand that some days you just won't get anything done. Go with it.

- Get something out there. Feedback is incredibly motivating.

- Relentlessly cut things back. What's the minimal amount you have to do. What can you remove? What can you get rid of? What's the minimal design that has conceptual integrity? If it's not necessary, get rid of it. I find that lopping off chunks of the ideas mean that other parts of the code and interactions with other parts of the system become less and less complicated.

- Don't start working on something until it's reasonably complete in your head. It's ok to doodle in code, but don't spend hours building something that is conceptually fragmentary. Sometimes I think about an idea for months. There are ideas I came up with 5+ years ago that I'm still noodling on.

- Keep an idea log. If you aren't working you aren't generating real ideas.

- Be ready to abandon an idea for another one if you find yourself thinking about it more.

I'm so undisciplined sometimes I wish I had my own personal drill-sergeant (akin to that guy from Full Metal Jacket) to kick my ass as soon I start losing my train of thought. . .does anyone else here consider their wandering train of thought a problem?

Usually I spend so much time trying to figure wtf I should do that I either end-up micro-managing some code snippet that doesn't matter or day-dreaming about how awesome my hypothetical startup is, which frankly doesn't do anything at all!

How do you guys manage to do work that really matters on your own time without anyone else there to keep you on task, implement a set constraints for you to work within, or tell you what the problem is? i.e. gimme some tips to self-manage my train of thought.

Since building something is only 10% of the battle (if that) you're going to want to apply the same general techniques to the other things -- web design, marketing, support, etc.

My favorite three words are "Automate, outsource, and eliminate". If neither you nor your customers require your personal attention to X, you should ideally not be X-ing, because you and your customers require your personal attention on enough things as it is.

[Incidentally: Telling your parents and friends about your internal deadlines as a motivational tool is pretty useful. Telling your customers is a great way to get unproductively stressed when life gets in the way. Your parents and friends will still be there later if your schedule slips a week because your girlfriend needed more face time. Your customers, on the other hand, tend to be a little more insistent that when you tell them 2.0 is coming out on July 1st that it is actually out on July 1st.]

> Unfortunately, [basic project management principles] often seem to go straight out the window when programmers sit down to work on their own time.

Well stated. When you are working on your side project and not being paid/accountable to others, its easy to mismanage your resource (yourself).

Does anyone use the free 1-project version of Basecamp for hobby project management?

This sounds like it would turn my hobbies into yet another job. I understand where it's coming from, but sometimes I like working on things that I don't have to finish.

At first, some of these items seem to conflict (like "something every day" and "let life intrude"), but these are great rules of thumb. From what I've seen, too many developers just set out on a path without really mapping it out first.

In my own experience I have often used the excuse that my project is so small that I don't need to write anything down for it, but that always comes back to bite me when I have to "let life intrude" for a week or so - I come back and I'm totally lost as to where to go next. Having lots of little milestones makes things go much more smoothly.

I'd say the number one bit of advice would be "Get people to use it".

Looking back at side projects that are sitting on a disk somewhere, abandoned, they went that way because no one was using them.

Once you manage to get some people using your side project, it's a lot harder to just shelve it, and a lot easier to iterate and grow.

... and leave an easy task to pick up when you resume.

You know how I motivate myself? I would subscribe to an oversized slice @ slicehost to remind myself everyday, I'd better get this thing out as soon as possible, otherwise I would end up paying those un-utilized resources as everyday passed by. It works magic, I ended up releasing my beta version within 4 months instead of the planned 8-12 months.

For me and my company, my ability to finish things has had a lot more to do with me believing in the project.

In my limited experience, you don't invest your time or your money into a startup. You invest yourself. You sacrifice your health, your relationships, and your political capital to make it work. And you don't do that because you have to or you want to. You do that because you only have one goal, and that's getting your company off the ground.

There's nothing that I've ever invested myself in so fully as my company, aside perhaps from my marriage.

First, know when "your spare time" is. After "completing" something (or just calling it quits for the day), go for it. Doesn't matter what gets "built". Just knowing you've done something enjoyable for yourself, whether it's tangible or not, is an accomplishment in itself. Second, if serious about building something others can benefit from, create that list...a plan in action...whatever you want to call it. Then you can measure your progress until it's DONE. Good luck.

I think the golden rule here is: build something that is going to be useful for yourself too, for the next years. If you are not an user of your own programs it's hard to get motivated. For example this is how I abandoned the development of hping. Because I stopped doing security for work I started using hping rarely, and I was no longer motivated to continue the development.

The "something real" that was created in the article is a mind mapping web app. Clearly there is a need for a really good mind mapping and idea tracking software, since so many people are taking a stab at the problem (myself included).

Personally, what I lack the most is that "small amount of external pressure", but I've been trying to fix that, trying to get out of my vacuum and start collaborating with people.

Nice post. Great reminder for those of us who tend to put big, imaginary barriers between ourselves and finishing anything!

It seems like so-called project management is mostly ideas adapted from the the great eastern traditions - yoga, buddhism, zen and eastern way to live in general. You can get almost any old book and you will find something called 'the way' - 'do'. Aikido is a good example, Bushido, and related teachings.

btw, I admire those genius who drawing users-guides and promote projects in anime style. =)