I admit that I have been very vague, but that's partially because we don't really know what to do either.

I talked with our commissioner and I have a bit to show you.

This is output of the program they have. They know how to get a most optimal solution, but it also shows 5 suboptimal solutions. We don't have to concern ourselves with picking one of these options. What you see here is sort of a breeding plan, in this case for plants. You see genetic traits in red or blue, blue being desirable, and you want to end up with a completely blue one. There are different paths to take and different ancestor plants to get to the same goal.

The problem here isn't determining what the 'best' (most economical or w/e) path is, they already do that themselves. Our goal is to find out what other information we can extract from comparing the different options. Like, the ancestor plants here are marked with colors, and as you can see one of them (the orange one) is used in all possible paths, but either in the first or second cross. That is very useful information to have.
So what we need to do/think of is different things like the one I just mentioned that we can distill out of the different options. What intermediates are more common, what ancestors, how many possibilities are there to go to the end goal without adding more intermediate steps, etc.

Those are just a few examples I thought of already, but there must be more I hope you could give me a hand with, also, I don't know how to get all this information out of it easily. There must be a way to automate it, otherwise it's not viable for our project (and also scalable, since the example here with 6 options will eventually be expanded to 100s).

Does this make it more clear what I'm asking? I hope so that I can get some insight from you

Here's a "dunno how to google for this" type of question, maybe one of you has run into this problem before:
How do I make sure that a brandname/website name I want to use isn't offensive in another language? Surely there has to be a database for this on the web somewhere

Guess I could click through google translate

Good question! I'd be interested in that as well.

TDD (test driven development) blows monkey nuts, confusing as hell for a beginning programmer

It's really awesome for anything you want to ship or change though because a bunch of tests usually means that you find simple issues rather quickly.

If it's too extreme you can switch to document-first development. Write a docstring for each method before writing the method (and you can add doctests, too )

TDD is amazingly powerful, although it is difficult to force yourself to use it properly.

Probably difficult because you have to think about what your doing before you do it, instead of as you are doing it, which is what a lot of starting developers do.

It takes experience to have that kind of foresight about a project.

If I understand your project correctly, you first need to decide how to programaticaly model your data first so you can do analysis on it. I think if you had a really well designed database (key point here is it has to be designed really well) it might be the easiest path to running reports and discovering information about the models involved. Apart from the design of the database, the queries themselves should be relatively easy to write.

And discipline. It's one thing to go back and half-ass some documentation after something is done since the docs don't integrate with the code, not the same with test cases that have to be passed.

G'day, Programming LC thread!

I'm completely out of date in the realm of things programming (my last brush-up was about 10 years ago), and I'm curious what people are using to develop apps for a device like the iPad? A quick search says Objective-C, but I didn't know if anyone here had any experience in this sort of thing.

Many thanks, and many happy returns! ^_^

heres the thing about TDD

you probably already do a lot of testing already, but don't realize it.

i'm a hobby programmer with half of a CS degree before i dropped out to play poker. so most of my coding hasn't been professional, but rather hacked together personal projects. right now i'm working on my first public software release

here's what my typical development work flow goes like:

- have an idea
- write some code to implement the idea
- execute the program to see if idea works
- program doesn't work
- find buggy code, throw in a messagebox to confirm this is the bug area
- re-execute the program to confirm bug
- attempt to fix the buggy code
- re-execute the program to confirm bugfix
- etc.

as you can see, i'm constantly executing my code to test if it works, to test if it fails unexpectedly, or to test if it fails expectedly., when i'm done with my process, i mark the idea complete or the bug fixed, and move on.

HOWEVER, since i ran all those tests manually, i can never reproduce them without a lot of time and effort. had i just written test code at the same time, i could re-run those tests at any point, to have peace of mind that my code works. if i ever rewrite a part of that code to add a feature, i could just run my tests and ensure that everything is good. but since i didnt, i have to run all my tests manually.

now, with all that said, i still DON'T write unit tests. why? because i'm lazy. because mentally i think that the time is not worth the effort, even though i've just written this post which argues against the workings of my mind. ill have to re-read this and let it sink in for myself

I have no idea how we could even write test-code for our programs

Objective C is a great language. They used at NeXT and Jobs brought it over for Mac OS X, and then it came to iOS when AppKit because UIKit.

I write unit tests for **** that I am not sure will work all the time. Kent Beck (the father of TDD) said it best:

http://stackoverflow.com/questions/1.../153565#153565

I can definitely see the power in it if you get used to doing it ^^

It's true I am doing my own 'testing' by manually testing code and looking at the website. I think once I can at least produce something semi decent, learning TDD will become much much easier. Like you guys said it takes some experience and foresight in what you ultimately want the program to do

Contrast this with Bob Martin, who's turned in to a 100% test coverage TDD shill.

ONE HUNDRED PERCENT TEST COVERAGE GODDAMNIT

I worked for a company that turned an agile PHP-driven platform into a Java-based enterprise-level waterfall bull**** model protected by program managers and senior devs entrenched in keeping the status quo. They believed in 100% coverage and SDETs who were infalliable.

I am happy to say this company is eating **** right now after the switch to a more "mature" language.

This is primarily why I hate Java FWIW

ETA: "nowadays it is irresponsible for a developer to ship a line of code he has not executed in a unit test."

Yeah sorry I don't test my this->load->models in my constructors that I never get wrong.

It's not Java's fault that your manager/decision makers were stupid. Waterfall is a bad/unrealistic design model no matter what language you use. It is when you loose sight of the true intentions and worry about hitting all the measurable stats.

Oh, I am aware that it's not reasonable to hate Java because of that. But it was my first corporate introduction to Java and the vast majority of the startups/mid-sized companies I interview with in Seattle know about that story. They always bring it up at interviews - "Hey, were you at [redacted] when they switched from PHP to Java?" "Yeah." "lol what a ****-up that was, hoo boy. Anyway our platform is in Ruby and not that garbage."

There's not too many attractive companies to work for here (aside from large corporations which I'm not interested in) that develop in Java. Most of them have a similar view as I do, though tempered.

Funny that it was big enough news that other people in the area knew about it.

wait your problem is with TDD or an anal over-use of it?

because that quote is from bob martin who is awesome and TDD, used properly, is the best friend of an agile process, not its enemy.

yeah, i have no idea either, especially since its all window based

but i recently got a testing framework set up and did this exercise to attempt to learn TDD

http://osherove.com/tdd-kata-1/
i messaged him my solution, he was surprised and posted it heh

It's a well-known story. Pretty big failure.

BDD > TDD IMO but 100% test coverage and testing for "everything" is a big waste of resources. In my experience as a consultant and developer, TDD is far, far more likely to be interpreted by program managers and senior devs as CYA bull**** rather than actually improving the quality of the code.

Objective-C is the supported language for iOS. Check out the apple developer pages, there is a lot of info there, but essentially, you are going to be programming in Objective-C using XCode.