Ask HN: How do you start your hobby web app?

[benregn]

I've been wondering how people start out their hobby projects as I'm having some difficulties myself.

Do you guys just dive right in after you get your initial idea? Or do you plan it out, right down requirements, wireframe etc.?

Maybe you even have some specific process if you have done a lot of projects?

[imp]

Dive right in. Write the bare minimum that you need to demonstrate the uniqueness of your project and then get feedback on it immediately. It's good to have some high-level ideas of what it could be, but don't waste time with formal documents.

Wireframes are pretty useful for making any web page, not just a hobby web app. My process is to sketch out wireframes, make the HTML and UI, then build the backend.

Also, I try to come up with at least three ways to implement a UI for a new app. That way you don't get emotionally tied to the first one just because you thought of it first. Usually, I find that my first one is crap, and the second one is good enough that I never can think of a third one that's better than the second.

[benregn]

I like your way of coming up with design ideas, makes a lot of sense to me. Not sure that I could do it though... I suffer very much from the common "tech guy" problem - my design skills suck! :P

[dgunn]

I try to tackle most of my project by just jumping in. It's definitely the best way to learn. I don't let any decision take more than a few minutes unless it's really essential to the project then I would do maybe 10 minutes of cursory research. The goal is to not get bogged down with detail for me because I will if I don't watch myself.

As for design, for a tech guy, I think I'm pretty good, but one really nice thing I'm noticing is that web design seems to be getting simpler. Or at the very least, you don't get points taken away for being sparse. (Usability is different tho, I'll definitely take points away for that.)

One great example is stackoverflow. It represents a design I would have never made. I would have called it ugly. But I really like it for some reason. :)

[user24]

I have a checklist I go through before I invest significant time into my own side projects: http://www.puremango.co.uk/2011/05/why-your-web-app-wont-wor...

Then I plan everything to the last detail, then I build it. When I was younger I just built it, and that's a great way to learn lots about both social economics and code. Nowadays I feel my time is better invested in producing something with a high chance of success than just a messing-about project. Unless the concept really grabs me, like http://thingsinbooks.com, in which case I just build it.

[ayers]

I dove straight into the code for my current project. Since it is something that I do not have exposure to at work, I was more concerned about learning the language and environment than developing the concept out. Next time I will definitely spend more time planning and on specs. I have found that if I had spent more time developing out the concept and defined initial core features at the very least, I would have less coding interruptions and some set goals that I could gauge my progress.

[cbs]

Sometimes with reading RFCs, design, graphical work, database layout, looking at new tools to try out while building the project or just digging in.

As long as you start somewhere, it doesn't matter, it is a hobby project after all.