Hacker News Reader

Comment by sova | Hacker News Reader

Some very practical, down-to-earth, solid advice.

Lines I found insightful: "... make it functionally complete in the first place.

... If it can't be made functionally complete it's too big. You've tried to bite off a piece that's bigger than you can swallow. Just get back to the drawing board and split off a piece that can be made functionally complete."

^_^

I really like the idea of splitting a project into phases. You might have Three phases, each one building on the last, but the very first complete phase should be a complete and usable thing in itself.

First the skateboard, then the scooter, then the bike, the motorcycle, and finally the car.

View on HN

> First the skateboard, then the scooter, then the bike, the motorcycle, and finally the car.

I've always found this metaphor incredibly amusing, because, to me, it actually highlights the _problems_, rather than advantages, of the iterative approach. At every step of the process, you're not only increasing complexity, you're also pivoting your product to the point where:

1. your previous clients are unlikely to still be well served by the new product, and are likely to churn as a result

2. the engineering challenges are different enough that most of the components aren't really reusable, and trying to shoehorn one stage into becoming the next will give you a sub-par product.

It's not about incremental development, it's about passing the baton. The spirit is motion/movement/transport, and the corresponding vehicle should be a natural expression of that with your available tools. Who said that the goal should be product-oriented? If I want to reach Alaska I will gladly take whatever transportation suits me best at the time. If I were insistent that my tank was clearly the most capable vehicle, although I would eventually get there, it would not be as enjoyable, imho.

If I ever buy a car with skateboard as a chassis, I'll be very disappointed!

Sorry, just had to =)

Probably it should be build the tyre first. This is a functionally complete thing that can, along with many other functionally complete things be composed into the car.

Car built with scooter tires... Scooter built with car tires...

Just not seeing it.

When Java was first becoming popular, I recall emphasizing how important it was to use the standard, well-documented, well-tested java.lang.String, rather than a custom String class as had been done in C++. Application domain experts started asking about strings that didn't fit in RAM (documents).

A great, tested component in one domain may require a significant redesign in a domain with other requirements.

Coding is a conceptual art-form, it's not about building a car with scooter tires, it's about upgrading vehicles to get the end result in a workable fashion. Try using spark plugs and pistons to get somewhere!

Car built with car tyres.

Analogy wears thing because a tyre is a standard off the shelf thing. In software everything that is interesting (and requires employees) is custom to some degree. So it is like a tyre for a rover on the moon.

However no need to get that to also work for a scooter on pluto! Unless the sales guy has already promised it!

step 1 "invent the wheel"

No, you miss so much when you try to make a component you think will be useful. The first piece must be something that makes someone's life a little bit easier - that's the only way to be sure you're making something actually useful.

Aside of it being convenient for users, keeping the components small is good for the developer as it tends to tame the complexity to the acceptable level.

Reply on news.ycombinator.com