What To Do Right As a New Programmer?

globeadue writes "My company just tagged me for full time App Dev — I've essentially never coded for money, but the last 3 years of support desk gives me the business sense to know the environment I'll be coding for. Now my company will be training me, so I think the technical side of things will be covered, what I'm looking for is best practices, habits I should/shouldn't develop, etc as I take on my new craft."

Comments

Go with the flow

[Anrego]

Well, I think you'll probably pick up those best practices as part of your "training".

Every shop does things differently.. from simple stuff like naming conventions right up to core design methodologies and team management.

My advice would be to just spend as much time as possible listening and observing. Read through existing code.. pay close attention in meetings to how the brainstorming and final solution tends to evolve.

Some companies take a "we are paying you for your intellegence.. part of your job is to argue your design and beliefs" attitude whilst others take more of a "we are paying you.. so shut up and do it the way we want" approach.

As a side note.. check out the book "Beautiful Code"... It's good mind food. "Pragmatic Progammer" is also good.

Re:Go with the flow

[martinw89]

I agree to an extent. As a new developer I think it's important to listen to more than one source. So listen in training but also be skeptical. Try to find other sources for claims.

And as for my book recommendation, I suggest "Code Complete"

Re:Go with the flow

[Anonymous Coward]

Tip: Get together with your coworkers after work -- hit a bar on Thursday/Friday night, loosen up, relax, talk shit about your managers, etc. Find out what people do outside of work. Ask the girls if they shave their pussy, the guys if they shave their balls. If you're in the bathroom taking a piss together, compliment them on their cock. Little things like that make the work environment easier and the coding more fun.

Re:Go with the flow

[SQLGuru]

Great stuff. Find a mentor. Most technical classes focus on how to use the language, never how to use it right.....at least not until you get to the advanced classes, which as a new dev, you aren't ready for. The best place to learn these things quickly is to figure out who in the group knows what (and is friendly/helpful) and glom on to them. Become their friend (bribe them with caffeinated products) whatever it takes. And absorb everything you can from them. This will usually take more than one expert (best design guy, best coder, best db guy, best politics guy). Just don't be a pain about it....if they explain something once, write it down and don't ask them about it again except for further clarification.

I love taking people under my wing and helping them grow, but if they keep asking the same questions, I see that they aren't trying to learn anything and just trying to take advantage of my knowledge. I still help those people for the good of the team, but usually with "here's your answer, now go away" approach instead of the "here's your answer, oh, and here's a better way that will make you a better programmer" approach.

Layne

Re:Go with the flow

[jd]

On the other hand, there are some universal rules:

• Plan/Design everything
• Document everything
• Version control everything
• Test everything
• Deny everything

Re:Go with the flow

[Whiteox]

On the other hand, there are some universal rules:

        * Plan/Design everything
        * Document everything
        * Version control everything
        * Test everything
        * Deny everything

Don't forget:
        * Lots of GOTOs
        * Delete those annoying REM statements
        * Use weird, arcane variable names like 'ORGO' and 'FTOOMSH'

Good Luck!

Re:Go with the flow

[Anonymous Coward]

For the complete list see how to write unmaintainable code at:

http://mindprod.com/jgloss/unmain.html

Re:Go with the flow

[Nefarious Wheel]

Remember at all times that "Listening" isn't the same as "waiting for your chance to speak". You'll get better mileage by laughing at the customer's jokes than telling your own.

Keep your reading up, and buy those books.

Keep a register of absolutely everything anyone asks you to do. Number and date each item. If you don't have access to a help desk-style problem registry then build your own. Make it quick to use. Notepad works. Spreadsheets a little better. BMC Remedy or HP/Mercury Interactive if you're spending someone else's money. Besides being good organisation, it's a serious cover-your-tail survival bonus.

Make your code clear. You want it to be so easy to read and understand that you can debug it over the phone in a new year's eve party if you have to. (Yes, I've had to.)

Comments first -- if you can't explain it in English, you probably don't have a handle on the problem yet. If you're chasing your tail and can't get that loop to work right, throw the code out and re-write it in English. Then try coding it again. Remember that English, your natal firmware, has been debugged over the course of your entire life and is still the highest bandwidth channel to your brain. (If you can write correct English, of course. If you don't, treat it like an essential coding discipline you have to learn.)

Be careful with your code libraries, and avoid bloat. If you're sorting a customer's web shopping basket, consider writing the sort yourself rather than including 200k worth of otherwise unused code. Even the lowly and properly-maligned bubble sort is better than that.

Conversely, never write your own little calendar applet if you can grab the code off the web. Plenty of work out there without requiring everything to be crafted with your own unique style.

Never, ever code when you're tired, you will break something. Caffiene is for the smiley boost during the work shift, doesn't cure the 10th hour stupids.

Disclaimer -- I've been a programmer for nearly 40 years, so I'm probably a stupid old person now. You're allowed to ignore the above (heh heh).

Re:Go with the flow

[digitig]

Then again, what books to read as programmers can be a whole other post,in fact I think it's come up.

Yes, but the general principle of "keep reading" is a good one. Read code. Read books (online or dead tree). Read articles. Read about new technologies, new approaches to old technologies, innovative solutions, successes, failures, pretty much anything. The person who stops doing background reading once they can do their job is a person who stalls and gets left behind. Do read critically, though, and remember Sturgeon's 2nd Law ("90% of everything is crap") -- there's nothing (well, not much) worse that someone who wants to overhaul the complete development process just because they read a magazine article about a new approach that promises to solve all the problems of everything that went before!

Re:Go with the flow

[TrueJim]

Whenever you write code on future projects, you're often going to start by trying to re-use code from previous projects. So don't just treat the source-code you're writing now as something you know you're going to reuse in the future. Try to think of each useful chunk of code you write as being part of a "package" that includes not only the code itself, but also the associated documentation, the test cases, the performance metrics, use cases, the build files, etc.

You want to write the entire -package- so that it can be re-used on future projects, not just the code itself. Remember that as a developer only a fraction of your time is actually spent writing code: the bulk of your time is spent putting together all that associated "stuff." So make the "stuff" re-usable too.

Re:Go with the flow

[TheRaven64]

Please try to keep up with technology trends. Conditional COMEFROM is now called 'Aspect Oriented Programming.'

Goto is good

[corsec67]

Along with the ? : ternary opp.

Code that is hard to read means job security.

The IOCCC is a good place to learn style.

Sql Injection is a good thing. You don't need to escape user data before send it to the DB, users never do anything bad.

(Go ahead and mod me troll, I can take the hit. Note that this is actually a list of things NOT to do. Except goto is sometimes useful, for breaking out of a few layers of loops/blocks.)

Re:Goto is good

[bigstrat2003]

Hey, I like the ? : construct. You leave it alone!!!

Re:Goto is good

[Rinisari]

?: ternary is fantastic for short clauses, such as $foo = isset($_GET['id']) ? sanitize($_GET['id']) : 0;.

The logical sequence for this is a and b or c, or isset($_GET['id']) && sanitize($_GET['id']) || 0;, but ignore the PHP 'cause PHP won't handle it this way (it'll put a boolean in $foo).

Re:Goto is good

[TheRealMindChild]

http://imgs.xkcd.com/comics/goto.png

Moral of the story: Even when you think a goto is OK, you will still get eaten by a dinosaur.

Re:Goto is good

[mudetroit]

This is one of the times when the saying "there is more than one way to skin a cat" comes to mind.

I work in a shop that has a solid rule of not commenting anything. It carries another hard rule along with it. We write very explicit method, field, parameter, and test names. If the code is in someway not understandable for you than stop and rewrite it so that it is clear.

Comments are a nice concept, but in practice they are rarely kept current. And amazingly enough are rarely correct immediately after they are written.

Sometimes there might be a better way to do things.

Re:Goto is good

[Rei]

I strongly agree. Self-describing code is much better than comments. Comments are only useful, IMHO, when you need to describe a complex situation. If it can be expressed in one sentence or less, it should probably be part of the code itself. I.e., instead of:

std::string mkdec(std::string x) // Converts x, a string representing a hexidecimal number, to a decimal string.

std::string convert_hex_string_to_decimal_string(std::string hex_string)

The latter says the exact same thing, but is far likelier to be maintained properly. Also, if you get in the habit of coding like that, you never have to worry about forgetting to comment. Furthermore, the "comment" is effectively replicated every time the function is used. Hence,

instead of:
hex = "0x" + number_str;
return mkdec(hex);

you see:
hex = "0x" + number_str;
return convert_hex_string_to_decimal_string(hex);

Now, if you had a function that implements a complex algorithm that can't be summed up in short order, then sure, use a comment. But in my experience, 95% of comments in code are like the above "mkdec" comment, and would be better expressed just by using a more descriptive function or variable name. I think a lot of coders are just lazy and don't want to have to type in longer, more descriptive variable and function names.

Re:Goto is good

[dubl-u]

I want to make one point. *any* code is unmanageable if you don't comment

I disagree strongly. I think the important thing is to make your code easy to understand. Comments should be a last resort.

As an analogy, consider your MP3 player. Suppose that every time they struggled with a UI decision, they just decided to put something in a manual or on a sticker on the back. It would be an abomination. The right approach is to make it so that as much as possible, instructions aren't necessary.

When I'm coding, I'll first try to put information in names, like variable and method names. Next I'll try to extract methods or objects, so as to minimize complexity. Then I'll worry about interface and object names. And I put a lot of information in readable unit tests, as that's a kind of documentation that the computer can verify still is valid. Documentation is always a last resort.

I have to deal with a lot of other people's code, and I'll always take beautifully polished code with no docs over mediocre code with lots of docs.

Re:Goto is good

[Almahtar]

I'm pretty much in agreement. The compiler can't check your comments: It's explicitly told to ignore them.

While it's a ridiculous example, someone could easily change

std::string mkdec(std::string x) // Converts x, a string representing a hexidecimal number, to a decimal string.

To

std::string mkdec(std::string x) // Do a little dance, make a little love, get down tonight

And everything would still compile just fine, but all of a sudden new devs on the project are left in the dark as to what mkdec() does without reading through its code. If someone named the function convert_hex_string_to_decimal_string something cryptic or incorrect it would appear in many places (everywhere the function was used) and be much more noticeable and likely get fixed sooner.
Another fun one I've seen in legacy codebases a lot is

//Only used in this place or that place, or for this or that purpose.
int some_func() { blah }

... and that comment was written 5 years ago and is completely wrong. People used doxygen or intellisense or whatever and saw that this function exists and takes the input/output they want but never saw that comment so they used it in 1000 other places.
Or another fun one.

/* function foo()
does this, that, and another thing.
... big long description of inputs, outputs, constants that affect it, whatever ...
*/
void foo() {blah }

... and the description is horribly out of date because it's been changed a million times.

I'll say it again: the compiler can't check your comments, and it's just too easy to change some code that will make a comment wrong and not realize it. So comments have their place but their use should be judicious.

"Self documenting code" (good variable and function names, etc) that the compiler can verify for you are preferable.

Re:Goto is good

[Mongoose Disciple]

I agree that as developers we should strive to write good, self-describing code that doesn't need comments.

Yet I still tend to write a lot of comments.

Well-written code can convey most of the how of what you're doing, but it rarely can convey the why -- and that's important too.

Suppose I look at code someone else (or even myself) wrote two years ago. I see what I think is an error in how the logic applies one of the business rules.

Is it an error?

Or was it right, and the business rule has changed?

Or is it still right, and there's some reason for the way it is that I don't understand?

Did someone in management mandate a change that runs contrary to the last version of the design docs? Was a more obvious solution tried first, but found to fail under testing?

All of these things are useful to know, and as much as I'm in favor of self-describing code, it rarely can communicate them.

Always think about maintenance

[Mad Merlin]

Probably the most important thing you can keep in mind when writing new code is to think about the poor sap who has to maintain that code somewhere down the line. Especially because in a lot of cases, that poor sap will be you. Pretty much everything else follows naturally from there.

Re:Always think about maintenance

[corsec67]

Yep, nothing worse than saying "Who the hell wrote this crap?", running svn blame, and then realizing that I did.

And, if you aren't using a versioning system, like SVN(preferably), CVS, git, that is a very bad thing. SVN and CVS also have the benefit of getting code to a remote computer when you check it in.

Re:Always think about maintenance

[try_anything]

There's a famous epigram (hopefully someone can provide it) to the effect that it's twice as hard to read code as to write it, so if you write your code as cleverly as possible, you will most certainly be unable to read it.

If you're proud of a clever solution and can't wait to show it off, then it's an unsatisfactory solution. The best solution to a difficult problem is the one that is so clear that after reading it, nobody can comprehend that the problem presented any difficulty in the first place. The sign of a really great programmer is that he never gets stuck with any difficult projects. Everything assigned to him turns out to have a simple, easy solution -- what luck!

Re:Always think about maintenance

[enkidu]

Everyone knows that debugging is twice as hard as writing a program in the first place. So if you're as clever as you can be when you write it, how will you ever debug it?
- Brian Kernighan, "The Elements of Programming Style", 2nd edition, chapter 2

Get out now

[Anonymous Coward]

Get out, now, for the love of God, while you still can.

Re:Get out now

[martin-boundary]

Yeah, those book recommendations are cook books!

Document your code

[TheSpoom]

Tab out everything in a code block. This should be obvious, but you'd be surprised how bad some stuff is out there. And try not to put in too many one-line ifs without brackets delimiting the code block... you can easily make the mistake of thinking something should be in the if's scope but isn't becuase there are no delimiters.

Comment. Comments are incredibly, incredibly important. They kinda go along with an overarching "don't be a douche" rule; while you may know what's going on in your own code, if it's at all complicated, tell the reader what it's doing. If you don't, someone is going to be very pissed at you later. If you want to go above and beyond, do Javadoc (or other style appropriate to your language) comments where appropriate; a lot of IDEs actually hook into them so you can highlight a method and see what it's doing.

And try looking at / working on some open source stuff as well. The big apps usually have a coding style they follow throughout and aren't that bad for a reference.

Re:Document your code

[Anonymous Brave Guy]

Unfortunately, test-driven design is not a silver bullet, unless you're lucky enough to have a finite problem space where you can achieve 100% test coverage, which almost no-one does, and you can consistently write perfect test code, which would be surprising if you're worried enough about your normal code to write all those tests in the first place.

In the absence of such unrealistic guarantees, TDD lies somewhere between a useful addition to your coding practices and snake oil, depending on the realism and honesty of the person advocating/implementing it. Either case, it is not a substitute for good documentation and commenting.

Stay away from the vending machine!

[Anonymous Coward]

Eat healthy and exercise. Pack your lunch or buy real food instead of the overpriced over-caffeinated junk in the vending machine. You'll save money and feel better.

Experiment and dabble in other languages

[BrynM]

Don't stick to just one language (the one they expect you to use). Learn how to do some basic things in several languages. This will help you understand "programming" rather than just knowing a language. Many of the same semantics apply in many languages with only the exact syntax changing. Learn the concepts not the implementations. This doesn't mean that you should try to code in many languages for your job, but as you are presented with problems do a general "how to do x" web search before you do a "how to do x using y language". The best coders I know see a particular language as a tool rather than a mandate. If you only stick to one language, you are imposing an artificial limit to your thought process and ability to problem solve.

Listen, think, and listen

[truthsearch]

- Listen to your end users. They're the reason you're writing the software. Even when they ask for something stupid, be sure to listen to their needs.

- Listen to other smart developers. Find the smartest experienced guy in your new team, or other similar teams, and pick up tips and feedback. There is a LOT that can easily be learned from other smart people's experiences. Ask questions, but don't be annoying. Following a few bloggers in your field can be helpful if you find the right ones, but an experienced person on your own team would be best.

- Read up on general best practices. Indent your code consistently, write comments, name variables and functions well, etc.

- Think about your code long term. Code is rarely used just once and never looked at again. Write it so it should last and be relatively easy for you to pick up a year later or for someone else to take over.

- Don't box yourself into one line of thinking. If you become religiously attached to one particular language, for example, you'll eventually stagnate. Learn the best traits of a variety of languages and systems. It'll make you a better all-around programmer.

Advice to increase productivity.

[radarsat1]

Don't read Slashdot at work. :)

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:Always....

[Bill, Shooter of Bul]

yeah, wheels suck. Instead create a circle that can be attached to a vehicle in such a manner that it allows the circle to roll while the vehicle moves in a straight line. And don't forget to give it descriptive name like

circleForVehicleToMoveBetterThanDraggingWouldNormallyAllow

Read Risks Digest

[Jah-Wren Ryel]

Read The Risks Digest -- it ought to be required reading for all software developers because it is fundamentally about how systems fail and if you don't have a good grasp of how the system you are building might fail, then you will probably build it in such a way that it will fail like a house of cards the first time a stiff breeze blows.

It is low volume with pretty high signal-to-noise ratio so it is not a burden to stay current, and when you have some dead time the back issues - going back for more than two decades now - make for great reading too.

A few tips

[mr_mischief]

I'm far from a master programmer myself, but this much I know.

• Don't get attached to your code. Your code sucks as a newbie. Your code will suck a little less with experience. Even the master coders sometimes write a section of code that sucks. Much of the difference between a newbie and a master coder is that the master coder recognizes his own mistakes when he comes back to them and rips his old code apart to be replaced by new code. The quality of the application as a whole is where to take pride if you're going to be proud of something. Being overly proud of a line, a function, a class, or a library will often get in the way of the quality of the application. Your users don't care about the code you write or how clever/inspired/tight/beautiful/special it is. If rewriting part of the code improves the application, then that's what matters.
• Bugs happen. Fix them without blaming or arguing. Don't place blame on the people who wrote them. If blame must be placed, place it only on the code in which they were found. Your job is to make the code work, not to piss people off by pointing fingers. You'll write bugs into code, too, anyway, and you don't want every one of them thrown back in your face.
• Make a habit of promising less and delivering more. It's much better than the other way around.
• If you're doubting how to design or code a section of a program, ask two people whose programming styles differ. Take as much of the advice of both as will fit into one solution. Try to change which two people you ask from one task to another, even if some of them are not the absolute best programmers on the team. You'll learn more this way than attaching yourself to one mentor unless your mentor happens to be a world-class wizard. You'll also keep allegations of cronyism and team splitting down.
• Use source control of some sort. Even if your team doesn't use it overall, use it for your portions. Even if it's something really basic like tarring up your project directory at the end of every work day and keeping the tar files, do it. Try to subtly hint at its benefits for the whole group if they're not already using it.
• Learn a programming language completely different from what you use at work in your spare time. The perspective it gives you can be very helpful. Lisp, Scheme, Haskell, Erlang, and Forth are good candidates for most people to pick up. If you're not exposed to one of Python, Perl, or Ruby at work, pick one and study it at home, too. Any one of those will do, although my personal preference is Perl (after all, it's just a personal preference). JavaScript's object model is interesting, so that wouldn't be a bad choice either.
• Don't read /. and other sites too much when you need to be coding. It's great to take a break and come back to a problem, but don't overdo the break part when a deadline looms. Cramming and hurrying when coding isn't as easy as hurrying up many other kinds of work.
• Get plenty of sleep and drink plenty of fluids. I know it's old advice and it sounds corny. All those tales of lone hackers coding all week on coffee, Jolt cola, cold pizza, pot stickers, and hot and sour soup are romantic and inspiring. They're stories about great people getting stuff done against the odds, though. You need to think clearly to code well. If you can think clearly on 3 hours sleep and cold pizza night after night, then good for you. If not, take care of your body so you can concentrate.
• Set reasonable short-term goals on projects and cross them off one after the other. You don't have to knock the whole project out as one commit two days into the schedule. If you can schedule kind of conservatively and get ahead of schedule, then use that time to improve your code or save it for troubleshooting later in the project. Don't get cocky when one module gets implemented smoothly and tell your boss to shorten the whole schedule. It'll just come back to bite you in the ass if you do.

Columbo

[MarkusQ]

Find and watch episodes of and old cop show called "Columbo".

Whenever you are unsure of anything, act like Peter Falk's character (Columbo). Whenever you are very sure of something, try even harder to act like that. If things don't make sense to you, ask questions, do experiments, use google, use your brain until they do make sense. And if you have a theory (or a plan, or a piece of code) that you are sure is right, put it to the test.

Don't be a know-it-all, don't blindly assume that you know anything. People sometimes get annoyed at developers who take nothing for granted, but that sort of attitude gets results, so they put up with it a lot longer than they put up with developers that assume they already know everything and project that assurance right up to the point where the project goes down in smoldering ruin.

--MarkusQ

Work a year or two doing maintenance

[MikeRT]

There are a few advantages to starting with maintenance work:

1) The majority of the work is probably done for you.
2) You'll have a chance to force yourself to get used to working with someone else's code.
3) If you have good senior software engineers working with you, you'll have people who can show you how things ought to be done/have to be done.

I've been out of college for nearly three years, and most of my experience has been cleaning up the mess that others have made. Usually the projects have been ones written by cheap consultants who got the contract by bleeding themselves dry on their bidding. You'd be amazed at how obviously bad a lot of the work that these do, even though you're just getting out of college.

Write good descriptive comments...

[madhatter256]

Try to right program code comments as much as possible as long as memory permits it (if you do have a memory cap).

It makes your job down the road a lot easier, as well as other people's job easier, too.

Try to have it make sense, too. Overall, doing this helps you in retaining how the code works step by step so that you will almost know it like the back of your hand.

Goals, motivation, and good music

[Shazow]

So far I like mr_mischief's reply best. Aside from that, here's what keeps me on track:

• Triangulate. A common mistake I see that produces bad code is people assuming that their code will be used for one thing and one thing always. That is almost never the case, unless you have the time, resources, and naivety to write code from scratch for every project. Try to see beyond the original assignment; set multiple scenarios and applications to your code, and try to fulfill as many of them as you can. Unification is good, and specialization is good--learn when to pick each end of the spectrum.
• Take pride. It's easy to come up with a 15 if-statement hack, but you'll always save more time in the long run if you spend that extra 15-30 min doing research on how to solve your problem elegantly. If you can't spare the time, do your best to isolate the hacks (such as into their own helper methods) so that you can come back later and replace them with something more sensible. Avoid duplicating code, avoid creating deep chains of method calls, avoid complex undocumented code (if you must write complex code, document what it does so you can keep skimming). In general, try to write beautiful code--something you'd want to paste to your friend, so maybe he can learn a thing or two.
• Music. When you reach a certain level of confidence that lets you build a flow, try immersing yourself in some good music. For me, listening to The Knife or Justice can significantly enhance my productivity and artistic spirit. Find what works for you.

But much more importantly, get enough sleep. I'm at least x2 more productive when I have 8.5 hours of sleep than when I have <7 hours of sleep. That's 1.5 hours that makes the difference of +4 hours of useful work. It's worth it, if you care about your work at all.

- shazow

Re:Some advice I've learned

[mudetroit]

Or you could try working somewhere crazy where they actual value you and:

1.) Asks you how long you believe something will take and then actually listens to you, and more importantly won't punish you if you were wrong. (Though it is important to tell them as soon as you know were wrong.)

2.) Maybe people could actually just volunteer and say "I screwed up." when something comes to light? No need to blame period then.

3.) Overtime pay is completely dependent on what you agree to when you are hired. And if they agreed to it, and still won't pay you, then you don't want to work there.

4.) Or you work at a place where the salesman have a good relationship with the development team. And like above will talk to them about how long things will take and set priorities appropriately.

5.) Don't bring your own code in, but still keep coding in your off time. You learn from it.

6.) Any place you work should give you in writing what they ask to include. If they are out to stab you in the back you don't want to work there.

Not everyplace out there is a hellhole. It took me a while to learn this too, but there are legitimate good places to work. Find one.

Learn by example

[Repton]

1. Read The Daily WTF. 2. Don't do that.

Re:Exceptions!

[nog_lorp]

Lies. Exceptions are not meant for intentional flow control, they are for exceptions. Exceptions are (in almost all implementations) much slower and you would never want to use them in place of a goto in, say, a core loop where the goto case happens a significant portion of the time.

Goto is good, gosub is better

[TiggertheMad]

like ? leave_alone : ;

There, fixed it for you.

Re:Worst commenting i ever saw

[FilterMapReduce]

Actually, I've been told that assembler code ought to be commented in just that way. In well written high-level code, individual lines should have a self-evident purpose, but often this is just not possible in very low-level code. Can you really tell at a glance what "djnz" means? And even if the answer is "yes", can you reasonably expect everyone else to have it memorized?

I don't have enough experience with assembly language to render an opinion myself on whether this is really necessary in a development context, but in school I spent a semester programming ASM and the line-by-line commenting seemed indispensable.

(Or did you mean that the comment was somehow inaccurate? Like I said, who can look at "djnz" and remember what it means? I can't.)

Use a slow machine

[www.sorehands.com]

Use a slow machine with little free memory to test your code. It teaches you to be efficient. That is why 1GB of ram is not enough for an office anymore.

Some tips

[sukotto]

Here are a few tips off the top of my head.

* Learn how to use your company's version control system.. and *use* it

* Comment your code and be smart about it.
That is... Keep the signal/noise ratio of your commenting as high as possible.
Comment the big picture and a description of what tricky bits of code are supposed to be doing (and why). Not commenting trivial things.

* Avoid putting tricky bits in your code :-)

* Never assume that some code you write is temporary. Lots of mission-critical systems started off as a "temporary" project.

* Always try to write the best code you can, even if it's just a little one-off (see above point)

So, that means making sure your function names, variables, etc all have intelligent names. Each block of code does one thing. Each block of code is small (try to keep it on a single printed page, including all whitespace and comments)

Avoid using global variables, gotos, tightly coupled code, code that messes with the internals of objects/data-structures, etc.

* Set up a little svn server on your workstation for all those little snippets of test code you write. You never know when you're going to want to go back and look at that stuff again.

* Read other people's code. Try to figure out if it does what it's supposed to do.

* Get a good IDE and learn how to use it really well. Use the same one as the majority of your dev team (unless it really sucks).

* Make yourself as FAST as possible. If you're really fast/efficient then you have more time to think and solve problems:
- Learn how to type. seriously. get a typing tutor program and do 30 minutes a day until you can touch type as fast as you can speak.

- Learn the hot-key combos for your programming environment. You won't believe how much faster you'll be.

- Stop using your mouse for common tasks.

- Use code templates everywhere you can get away with it. Every time you start a new file, every time you write a new function

- Learn the idioms of whatever language they have you using. You should never have to stop and think about common constructs in your code

* Keep a spellbook. If you learn anything cool, interesting, or elegant. WRITE IT DOWN. By HAND I know it sounds stupid, but it really helps

* Learn how to accurately estimate your time. For everything you're asked to do. Write down how long you think it will take (in hours). At the end of the task, or the end of the workday, track how much time you've currently put in, and how much more you think you'll need. (Never modify your original estimate). Then, when people start asking you to estimate how long a project will take you'll have some historical data to help you come up with a realistic number.

Pro tip... when you're starting out. Multiply all your estimates by 3. Newbies are usually way too optimistic

* Read. lots
Read books on the language your company expects you to learn. Try to also read general books on programming, design, project management, etc. Try to understand the big picture of your project as well as the nitty-gritty of the part you're working on.

Some good books to get you started
- Code Complete
- Pragmatic Programmer: from journeyman to Master
- Programming Pearls
- Joel on Software book
- Mythical man month
- Getting Things Done

* At the end of each project, keep a log of what the project was called, what it was for, who it was for, and what you did to contribute. You can also jot down what language you used, what gotchas sprung on you, your estimate accuracy ratio, etc.

Left-corner design

[steveha]

25 years ago I read a book called Software Tools in Pascal. This had a huge, beneficial impact on me.

The most important single thing I learned from that book is something they called "left-corner design". It goes like this:

Find some small part of your project, preferably something at the lowest design layer. Then code it up and implement it. Make it so brain-dead simple that you can spot all the bugs, and fix them. Now, consider some simple way to make it do something more, something else needed for the final product. Then make it do that, and fix it until the new feature is also working perfectly. Iterate.

As a real-life example, I once made an advanced audio DSP (digital signal processing) engine. It started out as a program that could open a wave audio file, read all the samples, and write them to another file. Then I added a function that could do some simple processing before writing the audio. And then I added some more stuff, and some more, and so on.

As much as possible, make early prototypes that actually do some useful subset of the problem you are trying to solve. If your program will have users, give them early prototypes and see what they say. The Software Tools book had an aphorism that 80% of the problem solved now is usually better than 100% solved later.

You may also find that, as the users try out your prototypes, they may discover surprising things. Perhaps what they originally asked for isn't what they really wanted, and you need to drastically redesign. Perhaps once they start using your prototype, they may invent new features that they really want more than some of the other stuff they asked for.

And perhaps you may get a surprise: "Yeah, we told you you could have 8 months to develop it, but now you only get 4." And three of those months are already used up. The left-corner design hopefully means you will deliver something. And it might just be enough.

The opposite school of design would be to think everything through and plan everything. Hold long rounds of meetings, draw diagrams, that sort of thing. That may actually be appropriate in some industries; if that's how they do things where you work, study it and try to figure out if they have a good reason for it. But even if so, you may need to knock out some sort of handy utility for your own convenience, and left-corner design is the way to do that.

steveha

Read the definitive references

[jthill]

Find the most authoritative reference manual you can find for every tool, every language you come in contact with. The question is not "how do I make it do such-and-so?", the questions are "what is it made to do?" and "what can I make it do?".

Do the onion-skin trick: read -- almost skim -- once, for fast comprehension, don't try to remember everything, just remember the new words and where to find them. Read it again, to remember which parts talk about which other parts. Read it again, to start understanding why those parts talk about those others. Only then should you even start thinking about asking yourself "how do I make this do X?".

Don't trust *anything* in any other book until you can tell what part of the authoritative reference it's talking about. Using C++? Pay the $18 or whatever it is now and read the ISO standard. Every book about C++ you read, tie what it's saying back to a section of the standard, and be sure you understand what both are saying. Using vim? Read the help. All of it. Do it again. Using MS Word? Hit F1 and start reading. Read everything. Using Python? Get the reference manuals and read them. Using TCP/IP? Read the RFCs.

Read The Fucking Manuals. Obsessively. Reread them again a month later. Then again a few months later. The questions are always: What is this made to do? What can I make it do?

Get used to it. I used to tell people I read manuals for a living. Classes are excuses to spend that much time reading the manuals. If you read the manuals, you won't need the classes. There is an *astonishing* amount of crap out there in the help books, useless "simplifications" that obfuscate the point of what they're supposedly explaining.

What worked for me every time I had or could steal the necessary time was, roughly, to overengineer the hell out of it, then boil out all the crap. Antoine de Saint-Exupery's maxim is absolutely dead-on: Perfection is achieved, not when there is nothing left to add, but when there is nothing left to remove.

With a little practice, most of the overengineering and boiling-out parts can be done in your head.

Particularly while you're new, reread your old code, constantly. Go find things you worked on before and read them again.

New Year's Eve Party

[bencollier]

Wouldn't have been the Millennium would it? :-)

Code to read

[univgeek]

Your code is probably going to be read a bazillion times more than edited. So take some more time to write it clearly.

Performance is hardly an issue for most business apps - and you should solve most performance issues by choosing the correct algo/method, rather than optimizing code. 80/20 rules applies, 80% of the time is spent in 20% of the code. And the performance critical parts of the code are probably 5% (for most people).

The compiler is not buggy. Even though gcc might spit insane amounts of errors for a single mistake in STL, the compiler is not buggy. Even though you read your code a hundred times and you can't find a mistake, the compiler is not buggy. It's infinitely more likely that you have screwed up than it is to have a tool error (unless you're doing Verilog/VHDL :)

Learn to use a debugger. Any debugger will make you appreciate what the code does better.

Read Code Complete, Pragmatic Programmer.

10 Learn how to learn better.
20 Keep learning and practicing.
30 Goto 10

Code in C... and nothing else...

[GNUPublicLicense]

Learn C (need to understand the computer memory model), and that's it. All the rest is bullsh... and fashionable stuff.

General Coding Advice from a 20 year programmer

[CFD339]

Taking the question seriously for a minute -- something unusual for this venue --

Whatever rules you use, make sure they are the same as everyone else on the project. That includes the rules around comments, indenting, and bracket locations. I've seen huge arguments over placement of an enclosing bracket on the same line as a declaration or a new line.

Once you have a little comfort in a language, get on an open source team. They have to be very good about practices because they have hundreds of people working in different locations at different time zones around the world. You'll start with no real authority, being allowed to submit small changes specific to small branches of code. The owner of that small branch will be responsible for accepting your changes if they're good enough. Over time, you'll move up the chain. READ THEIR GUIDELINES. OSS teams have, of necessity, very rigid guidelines on code practices that allow hundreds of people to work on code together. They've had huge battles over those guidelines, and very smart people have said very smart things (you can find them hidden in the morass of garbage if you look hard enough) that have gone into those guidelines.

More Specifically:

If you're repeating the same code, put it in a sub or function (a method if you're using an object oriented language)

A subroutine or function should be as fine grained and generalized as you can possibly make it. It should accept as few parameters as it needs, and should return a single value (or perform a single action). Note: Some languages, like C, use the convention of returning a success/failure boolean as the return value, and the result of the function in a buffer passed as a parameter -- that's also a good strategy.

If your routine is longer than a single screen to read, give real thought to how you might break it out into distinct subroutines (or methods or functions)

Avoid overly complex rules for variable naming. I've seen insanely complex variable name rules that are painful to work with and attempt to encapsulate the the data type, scope, and purpose of each variable in its name. That's not necessary or helpful in modern language programming. Most programming environments let you mouse click a variable and instantly view its declaration and often even comments written by that declaration for explanation. If your variable name includes the type and scope, you'll have to refactor it if you change the type or scope (like from integer to long integer, or boolean to enumeration as is quite frequent)

Use an object oriented language where possible

REALLY learn how to use Overloading, Polymorphism, and inheritance in your object oriented language

Avoid global declarations and functions wherever possible. The smaller the scope of any variable or object, the less likely someone or something will step on it later.

draw out -- on paper -- your object model before you build it. Learn to start with an ER (Entity Relationship) diagram so you can understand the relationships between real world objects. Think in terms of "A" is always the parent of one or more instances of "B", "C", or "D", but may be either a parent or a child of "E". Figure that out using real world objects that your code object represents first, on paper. I personally do this on a whiteboard with colored pens first then transfer to software for mapping.

Trust nothing. Your methods, functions, and subroutines should stand on their own regardless of what crap someone else passes to them. In every method, sub, or function, VALIDATE all variable data, always. Check for overflows. Check to see if an object is instantiated (not null or nothing). Check to make sure values are within the expected ranges. These checks are very small in terms of system resources and program run time. Take the time.

Do your declarations, validity checking, and decision making outside your loops. Any loop you make should do as little as possible inside it. Everything you do in a loop ge