How Do I Level Up As A Coder?

[phredd]

I'm hoping the copious coding nerds around here can give me a bit of guidance here.

I would like to improve my programming skills to the point where it would be feasible for me to be hired as a low level code monkey of some stripe. Mostly, I'm thinking about being able to do useful things with GIS and databases, as that kind of thing could easily be leveraged for archaeology projects, which is primarily what I'd like to develop my skills with an eye toward.

My current base of knowledge is this:

• 1980s - high school. Learned BASIC over course of a couple semesters. Never did anything extra-curricular
• 1990s - out of college. Learned PASCAL in a programming class. Taught self Perl. Took algorithm classes. Did a few tiny projects with perl to generate HTML code in the EARLY days of http.
• 2000s - sysadmin @ Columbia. Shell scripting, but never often enough that I got terribly proficient. Also have spent a not insignificant amount of time bashing on source code and compilers to make them work in custom environments, but this is mostly just an issue of finding places in someone else's code where things are pointing to the wrong place on my system and making a quick edit.
• Present - Teaching myself Python (and have dug up a ton of useful resources for this purpose)

I've been around the block a few times with learning basic algorithm issues, am comfortable with recursion, etc. and have learned different programming languages at different times and have some proficiency with them and with general concepts. However, that's about as far as it goes. I've only dealt with command line and file based I/O, no GUI stuff. Haven't ever had to worry about system tuning as I've never had to write anything that's been more than a few hundred lines or had to munch serious buckets of data.

So, let's say I'm a 1st level coder. How should I proceed in order to make progress towards eventually being able to develop my own projects rather than just being able to churn out coding problem sets?

[jenskot]

Do you know what archaeologists use now?

How do they store and reference data? Do they have content management systems? Do they use mobile devices or tablets? Are they just using Excel sheets?

[Pillar]

Phredd and I are in basically the same boat. After a fair amount of investigation, I am convinced the only way to really learn and improve is to do.

I am presently teaching myself Python as well. In order to force myself to learn, I pushed for making a new project we (and by 'we' I mean I will be the only one coding anything) are starting at the lab to be done in Python. I've also been advocating using Python for some other existing tasks. Now I've got deadline pressure forcing me to study, look up stuff, and implement. It's been pretty easy to pick up the basics so far. What I'm producing may not be the best code in the world, but it works and is helping me learn.

I've also considered trying to create other projects for myself to sharpen and enhance these skills. One idea I had was to try to develop a mobile app for the iPhone or Andriod -- just to see if I could do it well. I haven't tried this so far because I've had other self-improvement projects going on that were taking up my free time, but its still something I think about.

There are also competitions like TopCoder that I've thought about entering just to see how much I could learn just by participating. I haven't had the time to investigate them properly yet, though.

Pillar

[phredd]

Do you know what archaeologists use now?

How do they store and reference data? Do they have content management systems? Do they use mobile devices or tablets? Are they just using Excel sheets?

I do have a fairly good idea about what archaeologists use, but I think that focussing on that goes outside the scope of my question. Digging down into such a narrow specialism wouldn't help me develop the kind of proficiency and flexibility I'm looking to achieve. I'm really looking at what I can do outside the context of archaeology to develop useful programming skills such that I could, in theory, be able to actually send my CV out to positions advertised for 'developers' or 'programmers'.

[jenskot]

You could start with the Google App Creator: http://appinventor.googlelabs.com/about/

And then move your way up to: http://developer.android.com/index.html

[littleidiot]

GIS in academia tends to focus on low-level ''this works for my algorithm" languages like PERL and FORTRAN.
GIS outside of academia can still be rather PERL focused which I believe is a huge problem as PERL can get tricky on large projects.
My GIS background in coding has led me down the following path,

- C++
- Java
- Ruby On Rails

Note that we are something of a maverick in pursuing RoR though.

[IngredientX]

I'm in a similar boat, but coming from a different direction. I started with BASIC as a kid, and dabbled with VB after college. At work, I started using VB.NET, and then moved to C# because I liked the syntax better.

It was around then that I started taking programming seriously, and I began to study object-oriented programming. I then moved to Java (with a quick stop in Perl and Ruby), because that's what my company uses.

I'm now proficient enough that my company has officially made me a programmer. But now I have to learn things like unit testing, build systems, and enterprise frameworks. Java is full of add-on technologies that I must pick up.

So I guess it depends on the kind of programming you want to do. I seem to have ensconced myself in Java enterprise technology, but there are so many different directions you can take it (mobile, web, etc.).

One thing that might help is learning basic object-oriented technique. I know Python supports OOP, so it's something that would be useful no matter where you go. After that, you can take that knowledge with you to Objective C, Perl, or any other language that supports it.

[mumblethrax]

Have you tried eating a Saif corpse?

[Pillar]

I'm now proficient enough that my company has officially made me a programmer. But now I have to learn things like unit testing, build systems, and enterprise frameworks. Java is full of add-on technologies that I must pick up.

That reminds me of something. In academia, coding is highly-individualized. What I mean is that a typical academic coder is writing software to solve a particular problem that they have, and often do not give any thought or effort to making the code accessible or user-friendly to others. Beyond this, there is typically very little done in terms of software production systems and all the techniques companies use to make sure software is finished on time and is profitable/effective. As I code and develop in my lab, I am trying to make use of the good aspects of those corporate techniques, both to improve what I produce and make the jump from academia to corporate easier (when/if I decide to make it). However, I haven't found a good set of resources to explain those techniques, probably because companies don't really want to advertise how their coding teams operate too much.

One thing that might help is learning basic object-oriented technique. I know Python supports OOP, so it's something that would be useful no matter where you go. After that, you can take that knowledge with you to Objective C, Perl, or any other language that supports it.

I'd like to say here, for the record, that Perl's OOP is horrid.

Pillar

[littleidiot]

However, I haven't found a good set of resources to explain those techniques, probably because companies don't really want to advertise how their coding teams operate too much.

You've just got to know where to look. There is no secret, it's just wrapped up in comp sci speak.

These techniques are described in various design methodologies and patterns.

So '<your language of choice> design patterns' would be your google search phrase.

It's a steep learning curve though.