I work full time, so my projects are done in evenings, on the train etc.
I have an A4 sketch book and pencil I carry with me pretty much everywhere. Ages before i start coding, I'll be sketching out idea after idea for the project; everything from screen layouts, data flow, bits of text (If a good idea for some working comes up), names, db design etc. I've tried doing this stuff in software, but nothing comes close to pencil+paper for me
These drawings/notes aren't means as permanent artefacts, but rather as an aid to helping me think through the idea and flesh it out. As I build up the idea I spend a lot of time playing with it in my head (I'm very visual and seem to have a good ability to mentally model things)
When I have a few hours free, I'll start coding, but then often only until I find the next thing I hadn't considered and then I'll head back to the sketchbook.
When coding, I start with the back end process first. I also use Rails, but I'll build my models first, and I make them so I can control them completely from the console. I don't start building the UI until I'm happy with the way everything sits together.
Then, once things are nicely underway and most of the major problems are sorted out I'll lose interest and start on something new (actually, at any one time I've usually got a minimum of 3-4 project somewhere in their lifecycle between initial enthusiasm to abandonment)
It's interesting that you build your models first. This has been the way I've worked, but has felt clumsy so far on this project - hence my asking of this question. I do like the idea of controlling the models entirely from the console before building any UI though; that's definitely interesting to me as I can continue to develop the models to fit use cases that I find (I will be able to speak with a potential customer in the coming week so I'm hoping to glean more insight then) and nothing at all is wasted.
I am also not that visual, I do find it easier to fit the jigsaw together mentally, but I like to get this down somewhere solid where it can be reviewed/altered. Also, my memory isn't so great and in the past I have forgotten realisations that I'd made regarding faults in my design or changes that I should make.
Thanks for your insight; I'm edging towards the idea of modelling directly in Rails migrations.
Engineering the project is the hard part, but for a lucky some of us, it's the fun part. "Polishing" your project, on the other hand, blows. Who wants to spend the day building flow for user account cancellation or tweaking the registration page, when you could be optimizing code?
The win-win situation here is clear: Do the engineering that you love & pays the bills, hire cheap talent to do the cheap, boring parts. Because, at least speaking for myself, if I'm being honest a $12 hour designer would do these "menial" jobs BETTER than me.
I have an A4 sketch book and pencil I carry with me pretty much everywhere. Ages before i start coding, I'll be sketching out idea after idea for the project; everything from screen layouts, data flow, bits of text (If a good idea for some working comes up), names, db design etc. I've tried doing this stuff in software, but nothing comes close to pencil+paper for me
These drawings/notes aren't means as permanent artefacts, but rather as an aid to helping me think through the idea and flesh it out. As I build up the idea I spend a lot of time playing with it in my head (I'm very visual and seem to have a good ability to mentally model things)
When I have a few hours free, I'll start coding, but then often only until I find the next thing I hadn't considered and then I'll head back to the sketchbook.
When coding, I start with the back end process first. I also use Rails, but I'll build my models first, and I make them so I can control them completely from the console. I don't start building the UI until I'm happy with the way everything sits together.
Then, once things are nicely underway and most of the major problems are sorted out I'll lose interest and start on something new (actually, at any one time I've usually got a minimum of 3-4 project somewhere in their lifecycle between initial enthusiasm to abandonment)