I have been in the same position as you, and I chose that 'screw college' road you speak of. I had a love for software development, a C++ hobby on top of a basic HS programming course, and dreams. Now I am a professional developer, so I'll give you my experience.
After going to college for 1 year (I had a full scholarship for technical theatre), I figured out that I liked software more than set building.
Year 1 - I started my 'own thing' which consisted of desktop support to pay rent, and developing. Developing anything I could make, for anybody that wanted it, at a fraction of the price. Looking back I was probably doing $20,000 applications for $1,000. Starting out on your own really sucks because even if you did have the experience to know it's a $20K app, you don't have the credibility to ask for it. And worst of all, I have no idea what I don't know, and no other developers around me. I created applications that were maintenance nightmares. I had no skill in architecture or design patterns, so I basically made things that blew up and did network support to pay bills. Lots of Taco Bell, mixed with "well, at least I'm not working for the man". I've got some dreams of apps to write and get out to the world, but that has to come after bills, ya'know?
Year 2 - Becoming slightly better developer by learning what not to do and watching things blow up in my face. Barely getting by on desktop support, learning servers and making web sites. It must be easier than this working for the man, but I have no real portfolio so, press on.
Year 3 - Starting to get the hang of this. When I hit File > New Project, I have some vague idea of where I want to go and how to build things. Still choosing the wrong architectures, web services seem kinda cool, so why not build EVERYTHING with those? Need a calculator desktop app? I'll build a web service! Starting to pick up a few clients and being the IT guy and some software projects along the way. One thing I did do was create a Offsite Backup service using Web Services, so my dream was to be a 'Mozy' while everybody was still swapping tapes. Broadband was just becoming commonplace so I was ahead of the curve, and this was going to be my million-dollar idea. But the service had problems (due to my lack of architecture skills), I had no connections in the industry so no one ever heard about it except for the couple clients I signed up.
Year 4 - Finally, a customer believes in me for a long-term project. I manage to do it without screwing up badly; the code isn't great but it works. Starting to get caught up on bills, I get to work with a few other developers (fake it till you make it, right?) and even answering a few Experts Exchange questions. Oh yeah.
Year 5 - If you hadn't noticed by now, those dreams in Year 1 still aren't written, so that's starting to get a little depressing. I have a decent portfolio of stuff I've written successfully, got some decent momentum, and a respectable client base. Still don't really know what I don't know, and breaking even.
Years 5 - 8 - I'll combine these since it's more of the same of "do a project, learn a little on each, bring that experience to the next one". Today is in the middle of Year 8, and it's only in the last year or two that I've become a good developer. Those dreams in Year 1 have already been invented many times by someone else. In case you hadn't guessed, I didn't create Mozy.
Along the way I've had new dreams and new ideas, and some have been good, some have been horrible. I now have the skills to make them happen, and some of them are happening, and it's exciting. However, I have a feeling if I would have done things differently I could have shortened this journey quite a bit.
I can't speak to how differently college changes this journey; I'll leave that to others on this thread. But the pieces of advice I will give:
- You need to work with other developers. I didn't realize how important this was. You don't know what you don't know until you look at someone else's code or get a horrible code review.
- Fail before you have major responsibilities. If you really want to go out on your own, try to do it before you get married, have a house payment, kids, etc. You will fail and you will fail many times. Get used to it and value it as it's the best experience ever. But when your killer app that you just spent all your time and money on doesn't have a single customer, it's a lot easier to recover when it's just you.
- There is absolutely nothing wrong with bootstrapping. If you've got network skills, go work in a Network Operations Center or help desk (something within the realm of IT), and work on becoming a better developer off-hours and on the weekends. Make connections with people at real jobs. You'll need them later.
- Be 125% sure that you LOVE software development. The passion for software comes before the 'millionaire' part, not the other way around. If you don't have a passion for this, or your heart doesn't start beating a little faster when you hit New Project, go do something else and keep this as a hobby.
I'm sure I could go on, but the funny thing is I saw this question while working on one of those dreams and had to answer this one. :) Good luck.