I'd never thought of blogging it before because, well, I just assumed everyone did this...

Jeff Peters said he has a "test" directory in his local development environmnt that he uses to test... no, not code... hypotheses, ideas, concepts. He was talking about how everyone should do this and be able to go "Hmm, I wonder what would happen if I did X?" and actually trying it out (instead of perhaps just asking a mailing list).

Since he highlighted this, I figured that maybe not everyone actually does do this and therefore it's worth mentioning.

Seriously, create a "test" or "scratch" directory in your local development environment. Any time you find yourself wondering "what would happen if...?" or "can I do Y?" or "how does framework Z behave if I do this?", just write some code and try it out! Before you post to a mailing list, before you just simply give up because you "don't know the answer". Just. Try. It!

You'll be amazed: you'll find out the answer on your own and you'll be able to post to a mailing list and tell other people what you learned.

I'm not going to pick on anyone by name but I will say this post is actually inspired by a recent raft of posts on some mailing lists asking questions - and often sending somewhat irate follow-up posts a few hours later because no one responded - when a five minute test would have answered the question.

As a community we need to stop being so lazy. If you don't know how something works, don't just fire off a lame question to a mailing list! Those hundreds of other folks are busy with real work too - they're not there to do your job for you! Try it out for yourself. If it works, post the question and the answer. If it doesn't work, then post the question - along with all the details about what you've tried that didn't work.

I really (really!) don't like seeing follow up posts complaining that no one has answered a post when a simple bit of personal detective work could have answered the question in far less time that was spent waiting for an answer.

Damien

I've always had a 'test' directory - and more often than not, projects I have released (Transfer being the big one) came out of vague ideas that I started playing with in there.

I don't now how I could actually physically work without a test directory, to tell you the truth.

Good post.

What I have done is create a folder in my {drive}:\inetpub\ appropriately named skunkworks. I then create a virtual mapping in the root for every website in my development environment pointing to that same folder and allow directory browsing in it. In Eclipse I create a project also named the same that is always open and available for writing these test scripts. This gives me a central repository for these scripts regardless which project I'm working on.

Great Post!

It's so helpful to isolate a problem into it's most basic form in order to understand why something isn't working in a more complex system.

I also try and file away some of the more useful test files for future reference or starting points.