As someone that reads resumes and works with both degree and non-degree individuals, I couldn't disagree with a statement more. The difference in quality of work between those with degrees and those without is significant. It is even possible to distinguish those with some college over those without any.

Like any generality, there are specific counter-examples, but the people that will need to judge you quickly by little pieces of information are very likely to skip over you. That is unless you have some amazing thing on your resume.

I did ok in undergrad (3.4), but I'm finding that as I tackle a masters full-time will working full-time I am rocking... Funny what perspective and motivation working gives you.

What I meant was that even with college, many people utterly fail at life because there aren't as many opportunities as there used to be. I'm sure that if you are evaluating programmers the average programmer with a CS degree will be better than the self-taught person.

I know that not having a degree hurts me severely in a traditional job pool. I don't care. If I need to submit my resume to get a job, I will have failed miserably. I think I would choose street life over working for the man. I probably couldn't get a real job anyways, because I don't have very many skills or qualifications. Despite my lack of skills, qualifications, and especially motivation, I am confident that I have the instinct necessary to manage successful and obscenely profitable companies.

"the people that will need to judge you quickly by little pieces of information are very likely to skip over you. That is unless you have some amazing thing on your resume."

Exactly, so we should be telling people that college is the backup option. Try to do some impressive and interesting stuff, and if you succeed, you've got a free ticket out of college. If not, you can always just do the safe thing and go to college.

I admit this wasn't my plan, and that dropping out of college was an act of desperation in response to the pain it caused me. I just happened to have enough stuff with which to fill out my resume since I had been interested in CS and programming for a long time.

"But it's not true for Generation Y, when college for many people is a meaningless waste of $100,000 that qualifies them to do absolutely nothing."

--Or you can be really smart, and get a free academic ride as I did, or go to schools (state schools) that don't cost too much and are decent.

You can get a DECENT (maybe not the greatest) education for LESS money in this country, you just have to try hard.

Remember, often in undergrad, it is what you do that matters the most. Once you go for Masters/Ph. then the school where you go matters a lot more.

For some wierd reason, the people that I met that didn't go to college but went straight working (usually during 98-2000 boom), even thou are very smart, and did well are somewhat rough on the edges. It is hard to point it out, but they just don't seem well refined...

"they just don't seem well refined..."

I'm not sure if you're talking about all degrees, or just CS... However, speaking specifically about CS, I believe that my lack of refinement comes not from lack of Computer Science courses, but from lack of Math.

There is not much that an undergrad CS education teaches that I have not learned on my own over the past few years. However, I find that when I'm delving into deeper magic, my lack of math comes into sharp perspective. Most recently, this came up while I was hacking together a recommendation engine.

Sure, I can read a few articles and throw something together that works. But, when it comes to fine tuning it... I'm lost.

By the way, I'm a 22 year old hacker who did not go to college (yet). And I'm taking a Linear Algebra class right now to alleviate some of this problem anyway... heh.

Linear algebra will not help you fine tune many algorithms. The best math to know for fine tuning algorithms is how big O notation works, and be able to calculate how long things take to run.

The second thing that will help you is studying data structures algorithms. What is the average case for a search over x items? How long does it take to insert into a linked list? This way you can begin to identify bottlenecks, and get an idea as to how to approach solving problems with performance.