5 ms·
A minor heresy: most valuable programming involves a lot more gluing together libraries than it involves classical algorithms. You can go years in industry wit
by patio11 15y ago
A minor heresy: most valuable programming involves a lot more gluing together libraries than it involves classical algorithms. You can go years in industry without ever seeing anything more complex than a hash table, and you don't even need to know why hash tables work, really. (And sorting? Sorting gets done by calling list.sort(), not by coding a merge sort.)
1) The way to go from "dabbling" to "proficient" is to ship software. It doesn't have to be amazing software. Entire industries are built on very non-amazing software. You will learn more from writing and supporting 1 real application than you will from a hundred tutorials.
You are, from the perspective of 99% of people in the world, a magician. Other people solve their problems by hand, by paper, by process, etc. You can conjure demons, speak arcane words to them, and the problems just go away. Find someone with a problem that they think is a Really Big Deal because they are not on speaking terms with any demons. Speak to demons for them. Get their money. You'll learn a lot from the process -- in particular, how little speaking to demons is actually required of a demon-speaker.
2) Study CS in college. Study something else, too. Your career prospects will not be dominated by your programming ability. Restated for emphasis: your career prospects will not be dominated by your programming ability. If you get really good at one hard thing plus programming, you're pretty much set for life (if that is your goal), since the intersection of the sets "can do This Particular Hard Thing" and "can program" will be only a handful of people.
3) The really useful engineering classes are the ones that teach you how to interact with other people, in particular, how to interact with people who are not engineers. People who are not engineers control most decisions you want made in your favor. Look for courses like Technical Writing or Communication For Geeks or whatever and make a special effort to take those versus learning e.g. how to build microcomputers out of NAND gates.
- auganov 15y agoVery good advice. You could create great software just glueing code, no matter how dumb it sounds, most people are not too good at it and it's very very obvious. Now there are situations where glueing might not be enough and they surely will occur. That's when you need an engineer to solve the problem, CS will help you become one. The question is do you want to be one? If you are more into Internet entrepreneurship studying something that is not CS could give you more advantage. Many great programmers never create any significant software for themselves, their knowledge is too confined to CS. These are people you might want to hire once you have your start-up, tell them what exactly you need your software to do and they will perform. Go to Top Coder's website. All the top schools are in Asia and Eastern Europe. Actually a school from Poland, my country comes, in 3rd. Few years ago it was 1st. I would hear on TV how we have the best programmers in the world. But did they ever create anything significant? No. Most of them now work for Google, MS or some other big corporation that needs people who can write very efficient algorithms fast. So while I could easily attend an university that generates those great programmers (even for free I think, not too sure) I never even thought about it. Knowing the science behind paint does not make you a painter and you don't need it to become one. I'm not in any way against CS, it's definitely very interesting to study and if you truly love programming then by all means go and study it. But beware of what CS is and what it's not.
- cheez 15y agoRegarding the heresy: wow, you're lucky. Every other project, I seem to have to decipher and apply some PhD thesis from the previous year and I didn't even complete university. Sometimes I think I don't get paid enough for this. Edit: I'm not unlucky, I love what I do. But see my responses to a couple of people below :)
- astrofinch 15y agoOr we could call you lucky for getting paid to work on stuff that's cutting edge instead of boring CRUD apps. Seriously, I'm curious where you work.
- cheez 15y agoYeah, I'm definitely lucky but man, sometimes it's stressful. Especially recently, I've had some really difficult problems to get through. I kind of work for myself though it's more accurate to say that I work for a lot of people in various highly technical and often mathematically intensive industries...
- patio11 15y agoFood for thought: if you approached getting paid more like it was a fun new optimization problem to be studied, deconstructed, and hacked, you would already be getting paid more. Instead, you're busy making other people rich. Apologies if I sound like a broken record on this topic. It is largely projection.
- cheez 15y agoHey, you are my inspiration! You actually inspired me to make my own products. The problem is, matching the day job pay is pretty hard because it's pretty high (should get paid more even though I'm paid well.) I think I'll get there eventually, just need to keep plugging away at it.
- dagw 15y agoI have to say that that sounds awesome. I'll happily trade you my job (no serious academic thinking required) and work for 20% less of whatever they're paying you.
- 9ec4c12949a4f3 15y agoUnrelated, I love the imagery you're using. Daemons and wizards.