Turning into a hacker aged 27
pragmaticstartup.wordpress.com
pragmaticstartup.wordpress.com
The exception to that is if you have a lot of cash, and are willing to hire someone to build something for you. That's very expensive of course, since you're competing with companies offering them $80k - $150k+ /yr. jobs plus benefits. You can find cheaper labor, but you'll get what you pay for.
Have lots of fun learning! Programming can be extremely rewarding! :)
Where it does help to know code is when you need to understand the programmer. If he or she can communicate clearly, then hang onto that person by any means necessary.
Within the law. No need to get all Buffalo Bill on him or her.
This isn't meant to be a personal attack on you by any means. I've noticed a huge movement toward "mentorship" lately, and I'm just not seeing the purpose. Even worse, those who do not get these critical mentorships wither away and blame others for their shortcomings and failures. Obviously you were fortunate enough to find a group who "took you in", but what would happen if you didn't?
Downvote if you want, but I'm honestly curious as to what's going on with this mentorship trend.
I've been hacking on rails for a while but only after I joined a company where they had guys with 4 years of professional rails experience did I come to know of all my bad habits and also learned how to do things in a neater faster way.
Basically being around people who are good at what they do will make you good at that skill.
The book "The Talent Code" explains this in detail. Read this book - it'll significantly help you understand what you need to do get good at a skill fast.
Venkatesh Rao wrote a blog post on this as well which is a very worthy read. He called it "The Crucible Effect". http://www.ribbonfarm.com/2009/07/21/the-crucible-effect-and...
And this goes for any skill from windsurfing, soccer to coding to running a startup.
1) Environment. Having the right people and drive around you striving for the same goal has a positive impact on your own actions. Makes you more motivated to do stuff (at least for me)
2) Make less mistakes. I made a ton of mistakes the last couple of years and I am now VERY receptive to not learning things the hard way.
In the event that I did not, the plan was to get a regular 9-5 job and devote nights and weekends to improve my technical chops.
There are many ways to do this but having a mentor or a right environment to do it certainly helps an already difficult journey.
Before the internet was pervasive, a self-study of Programming was for the curious tinkerer. Now that it is incredibly mainstream, we are seeing more people who are approaching programming as a means to an end, for whom the solitary pursuit of arcane knowledge is not an end unto itself.
For people who are not auto-didacts, a little more structure is needed. From a (Vygotskian) pedagogical perspective, most learners require an Expert to create "scaffolding" (support structures) that allow the learner to increase their "zone of proximal development" -- that is, they require someone else who has mastery to provide a little help or guidance in order to help the to tackle tasks the learner does not yet fully master. the completion of tasks with help then increases the learner's level of mastery and they can then take on ever-more-challenging tasks.
Tl;Dr: most people prefer to not teach themselves hard things.
for a direct example, compare the README.md in many GitHub projects to the pages on SourceForge.
We used to take for granted that you'd know how to get cvs, configure make and install for your system (and would understand architecture flags and so forth...) nowadays library authors (who in many ways are the "first educators" for their tech) are putting in more effort to make things as easy as "copy and paste this into your terminal to install."
That all said, it is kind of terrifying how many people are developing networked applications with only the most rudimentary understanding of networking...
Now we're seeing more HR/management training people teaching that mentoring and coaching new hires (esp. those just out of school) is a better way to go. It gets them up and running faster and you can brainwash^H^H^H^H I mean teach them they way you want them to do things.
Personally, I think it's a good thing.
Thanks for this comment. I started coding about 8 months ago and am enjoying it thoroughly (I just turned 30 and have a background in business). I was actually thinking about doing the same thing this guy's done - get a developer intern position somewhere since reading this post earlier this morning. I'm sure I would learn tons that I wouldn't otherwise learn (I'm thinking specifically architectural and security type stuff, among other stuff) but am also learning tons on my own.
Just last night I was talking with a good friend that has been coding for 15 years and he suggested something kind of similar: fix the bugs and maintain some code that a more experienced developer can check - run errands so to speak. That way I'm not breaking something or pushing insecure, buggy code into production. I think a combination of the two would be optimal.
Then share your experiences with HN!
Why is it a difficult rut to get out of? Because when people build teams to build systems from scratch, they look for other people who have built things from scratch. Then when they need someone to maintain it, they look for someone who has maintained other piles of buggy crap.
When we were formally taught Java at university, there were 10 hours of tutored labs available. I certainly didn't need any help to do the coursework, yet I went. I soon found some real hackers, who happily ripped my coding attempts to pieces (in a good way, mind).
When I started work on my first attempt at architecture, I got some guidance from some of the same people. Whilst design patterns were around, these people helped me understand which ones to look at, why, and what overkill was.
When I graduated and started working, I was already a "better" programmer than some of my peers, as I had been shown GDB (and hence JDB) by my mentors. However two years in I worked on a project with some talented developers, and again I learned about IOC etc.
Mentors save you a lot of time and energy. They are present in almost all trades where you have apprenticeships, and I for one know I thrive in that sort of situation - the one where I am challenged by someone who has been there before, who forces me to back up my designs/arguments by reading directed texts.
If I hadn't had them, I'd have found my own path. But I'd have spent a lot of time wondering somewhat aimlessly, reading semi-useful books cover to cover (yawn), and screwing up projects...
Look up Google Summer of Code. You won't get into the program if you're not a student, but it's a great jumping-off point for finding communities that are willing to nurture newbies.
-- They help weed out the bad and the WTF to bring forward the good (in my case, okay) stuff.
-- Their discussion of design issues can help you figure out your framework, and show directions you hadn't known.
-- They give you some sense of what you've accomplished, what you haven't and what you must improve.
I learned Python first, then learned how to use linux and emacs, played with firefox add-ons to learn Javascript better, then started building web apps in PHP/MySQL, and now my plan is to spend my days learning and building stuff in Ruby and my nights working through these four books: http://stackoverflow.com/questions/414779/what-should-a-self...
Best resource ever is from @yegg, I always check this for guidance before learning anything new: http://www.gabrielweinberg.com/startupswiki/Ask_YC_Archive
I've come a long way since I first set up my server on Slicehost. Good luck on your journey too.
Question: why are you hopping around different languages?
The two languages I'm really focusing on now is Ruby and PHP. I want to get good with both. And if you want to do anything on the web, you gotta throw Javascript in there as well.
Keep it up!
My tips:
1) Study and practice. Pick a topic you don't know, choose a book that's well reviewed, and learn it.
2) It's better to ask a dumb question than to never know the answer.
3) Do a bit (not too much) of reading on things like HN and JoelOnSoftware, etc, to help you pick what to study next. (Like when I read Spolsky's argument for why distributed version control is better than centralized, I eventually decided to learn Git. Or when I saw all the buzz about Rails, it pointed me that direction.)
3) Learn the crap out of your tools. I started out with a Windows HTML editor. A couple years later, I was using Vim in Unix and I'm way more productive. But I'm still learning new stuff about it every day. I intend to know it backwards and forwards eventually.
4) You'll never know everything - don't get discouraged.
I knew HTML and got a job doing tech support for cell phones. My employer had no internal web site and I saw that they needed a page to share links to reference sites, so I made a simple HTML page, just dropped onto a shared file server, not even a web server. Then I started adding pages of training material, etc.
I had just heard (this was 2007) that CSS could be used for layout - gasp! - so I bought Eric Meyer's CSS: The Definitive Guide and studied during lunch breaks and evenings. I spiffed up the site. Next I got Head First Javascript to add some animated menus. Soon after, somebody told me about jQuery, and eventually I caught onto that.
As I added more pages, it sucked to have to copy and paste my common menu, so a friend recommended PHP. I had to set up Apache for that, though, so I used XAMPP. Now I could do includes! But wouldn't it be nice to have some database interaction? I read a PHP book and added that too. Eventually the system was pretty complex, with object-oriented database interactions, permissions controls, AJAX widgets, etc.
All this time I was getting advice from my best friend (we were roommates in college), who is a truly awesome coder with famous companies on his resume. And I was discovering that I really enjoyed this stuff and had a knack for it. Also, all this time, I was still doing tech support: I could be knee-deep in a coding problem and get a phone call, and I had to pick it up, and it might take 2 hours.
But by the time I left, I had a good enough resume to get a full-time programming job. And I continue to learn. I've made the jump to Ruby and Rails (Ruby is a lovely language), test-driven development, etc. I'll never be DHH or Linus Torvalds, but I hope to do solid work and keep having fun.
The problem is, like the rest of America, junior positions seem to have vanished. At least on job boards and hiring pages.
Can anyone point us to a good resource to find junior positions or even internships in front-end coding or UX design?
I'm not asking for any company to completely teach me stuff. The ideal position would require a portfolio but maybe one that isn't completely cross-browser, production-quality code.
I started coding HTML/CSS/Javascript when I was 15. When I was 18 I was working on a $14M project doing HTML & jQuery. 19 and 20 rolled around and I thought I was hot shit and could get away without knowing much about object oriented programming and database design. Boy, did experience beat me with a heavy stick.
After blowing through a lot of money paying other people, and not being able to evaluate what they were doing, I started learning to program myself a little over a year ago (December 2009). It has been, by far, the best investment I've ever made.
You HAVE to have mentors, though, bookwork and experience alone don't cut it. Some of my mentors, didn't do anything but talk to me - and ingrain in me the importance of fundamentals - and some basic vocabulary. Some stuff that didn't make sense when I heard it, but produced those "Aha" moments later. Other stuff, like the importance of pre-planning, prototyping, code readability, and NEVER using code you don't fully understand, I was told from the beginning were very important.
Plus, mentors can be different types of people in different types of situations. One of my mentors was an experienced programmer who I lived with for a few months. Others were dudes on stackexchange and serverfault (Yes, that counts), and others were those I read.
The bottom line is, you need to see yourself as an apprentice, always looking for a master to learn from.
Being in a room with smarter/wiser people than you can only make you move forwards. And do that enough, I hope to actually join the club someday. Heh.
And yea, SO rocks but I think a lot of newbies are taking advantage of it and getting shot down advice to wannabe hackers: do your homework first or get flamed on SO!
All night debug sessions? Check. Several complete rewrites from scratch? Check. Using a language feature considered "unsafe" because it was the only way to get done what one needed? Check. Figuring out what "unsafe" meant? Check.
Once people are writing all of this code many want to know exactly how their code gets translated into what the computer does and they go off into their 'compiler' stage (some stay there for ever :-). This is where folks write their own compilers, by first writing their own lexical analyzers, then needing to parse those, and then needing to take those parse trees and generate new code with them.
There is a lot of exploration to do in 10,000 hrs. That's coding 6 hrs a day (and by that I mean coding, not checking Facebook or reading HN :-) for 5 years. There aren't a lot of short cuts. Sure there are CS programs which will show you what other people learned in their 10K hrs and help distill the concepts down for you but, like playing Piano, its not "known" until its in your finger bones.
Mentors are great for helping you get unstuck (since being stuck means you aren't coding which means you're not clocking your 10K hrs) and they can sometimes give you a heads up on road your travelling down, but mostly I think they are best for just sharing your enthusiasm and adding energy.
Decided about a year ago to switch careers, and I've been getting familiar w/ a few low level languages (C, C++, Java), and working my way through a few books like SICP, Cormen's and Skiena's algorithms books, Code Complete, etc.
The OP seems to have landed an internship already, whereas I have been failing miserably at this. I've been cold emailing companies, applying for internships, etc. I've written a few basic beginner type programs, nothing large scale or real world yet.
I have a ways to go, but it sure would be nice to have some sort of mentoring and/or get involved with real projects of some kind. I'm doing great self-teaching, but there's obviously value in working with more experienced developers. I know having code to show is important (github, opensource), but what else should I be doing to land that first software job?
From the code perspective, a lot of companies noticed my open source contributions. Granted they were mostly minor utility code, there's always that little spark of joy from their voices whenever we go over that topic.
So go join an open source project! I know it's reiterated over and over again, but a good project can act as proxy real-world experience in helping you discover what it takes to write production quality code.
I guess if there was a single tip I could give, just be highly visible in committed code, web presence (as a programmer, not as a party animal) in websites and social networks, and face-to-face networking. That would raise your chances of being noticed.
If you have an undergrad degree (of any type) already, you could look around for a MS program with an advisor who is doing HPC research. You have a shot at getting into a program which will probably require you to take or place out of undergrad coursework (SICP and CLR are probably going to make that workout just fine for you).
If you want to go all self-schooled, maybe you could pick up one of the new CUDA books and do some decent re-implementations of hard HPC stuff for that environment. Side projects like that would make me look at a resume twice if I had a need for a HPC hacker . . .
After you've set up a development environment (on your home machine / laptop will do), pick a project that's been on your mind and just start. Everytime you run into something you don't know how to do, hit stackoverflow, read through APIs, and figure it out. Piece by piece you'll become a programmer.
Good luck!
Overall I doubt I will ever be a programmer, but at least now when I think of an idea, I do not think "Who can build this? For how much?", i think "Do I like this idea enough to build a working prototype?".
And the best part about being a newbie is not knowing your limitations.
But dare I say in most people's eyes (e.g mine), you're definitely a programmer.
Now I wish I had taken the leap. python at age 29 is no fun. But I am looking forward to learning it and becoming comfortable, much like when you move to a new city...or at least I hope.
I had a great time with it at 32. Of course, to fully appreciate Python it helps to have spent years using Java and Perl.
18 months later, I still suck, but most importantly, my mentors taught me how to learn on my own, where to look for documentation, what sorts of problems to look for. I'm definitely not a 'hacker' in the way you're talking about, and every time I think I know something, I get slapped back to reality, but the REAL satisfaction is being able to just keep making forward progress. To be able to sit at a terminal and feel like your startup's destiny, YOUR destiny is in your hands is immensely empowering and a far cry from waiting and hoping your developer is on the same page as you.
One piece of advice, which I still struggle with: don't try to bite off too much. There's a compulsion to understand the entire codebase before you commit to that host startup's project. Don't. Figure out heat you want to get done and just try to understand a tiny corner of that codebase in isolation, then fix it. Keep doing that and you'll eventually have covered the whole codebase.
It's a long journey going forwards, but you'll look back and it'll seem like time has blasted by. Having good mentors in your startup will make all the difference.
Yeah, I'm starting off with bitesized problems. :)
I'm 23 and I never got any compsci/programming training in college. I'm now working part-time and learning programming (Java, .NET, SQL, web scripting) full-time.
Why? 3 reasons.
1) I am sick and tired of not knowing how to create solutions that add real, measurable value to almost any human task imaginable. In 2 years, I'll be able to do so much more than I ever imagined in any office setting.
2) Good programmers have the job I dream of: being able to use their minds all day to create cool stuff that other people depend on. Top talent gets paid pretty well.
3) The big secret that no one tells you in college is that if you want any decent office job in which you'll be using your mind instead of bullshit "management" or "leadership" (i.e. wielding social prowess to get what you want) skills, you're going to be programming. Want to make a spreadsheet? Guess what, John Walkenbach, the maker of the Excel Bible, says that even if you're just making a basic spreadsheet, it helps to think about it as object-oriented programming. And he's right. Especially if you're a power user of Excel, Excel == Programming. Period. Same with just about any other MS Office program.
The bottom line is that if you use a computer at work, like it or not, your freaking livelihood hinges on your ability to create and manage software and/or hardware. You might as well know what the hell you're doing.
Would anyone hire a freshly-minted 43-year-old developer? Or is this a fool's errand?
Once this business is chugging along, I'll do something more scalable. But that's just as likely to involve publishing, where I have passion and experience, as it is to involve a startup, where passion alone would need to develop experience.
So it's entirely possible that I'll be sitting here in 10 years, having never learnt more than some basic HTML and CSS, because I keep creating more immediately valuable opportunities. Though I have noticed, as more of my friends have kids (we don't, yet), that 'long term' things like learning a new skill, going back to college, or even committing to a local football team, suddenly seem smaller and more realistically achievable - ask me now if I want to commit 10,000 hours to something new and I won't have time, but with a backdrop of 18+ years of child-minding in front of you, it doesn't seem nearly as large. Of course, as I understand it, kids take time too!
eg: http://www.tourdust.com/products/844-atlas-mountain-sahara-d...
I need to get that and write it to a database with its fields like description, photo, cost, number of people etc.
Let me know if your interested. In return I can answer guide you becoming a better hacker ie, what you should read, questions on C,Python etc ...
But one warning - in the end its just about sitting in front a computer building something till its done. :)
I am considering the jump as well (from i-banking), and have been putting in a considerable amount of work, but part of me is worried the growth we have seen will come crashing down soon and opportunities will disappear. It is hard to give up a nice salary and job security, especially when a similar position is very difficult to return to if things don't work.
I'm experiencing the excruciating pain of not being able to get your hands dirty... having great devs is great.. but the whole process of turning your vision to reality through code is something I'm embarking on as well.