How much defense do you put in an object with a single responsibility, keeping in mind that you're not even allowed to throw exceptions that are not thrown by the base class?
The examples on the wiki are ridiculous; parameter-checking is a general best practice. Yes, sometimes you implement a strategy-pattern as a way of "extending" the application later on.
I'm following an idea from PHP; do, or die. If the code encounters anything exceptional, an exception is raised. If it's an unexpected exception, the app will terminate. It has to, since I can't guarantee that it's data isn't corrupted at that point.
Almost no defensive programming... except if it makes my code crash sooner. (so I can find the problem more easily)
I also add some defensive programming if a poor crash description wasted my time once.
I'm more reactive that preemptive. But I always make sure that nor me, nor someone else will waste time again on the same problem.
As the in-house coder for my firm, I get asked to do a lot of ad hoc stuff, once-off applications meant to fulfil an immediate need and then never used again.
In those cases, I don't bother writing defensively. I will, however, put in an expiration date: if the application is run after that date, the app turns into nagware that reminds the user that it has expired. That is usually enough to get the user to rethink using the app, and maybe asking me to redesign it for long-term use.
I've seen screwdrivers being used as hammers so setting arbitrary limits on things that don't need limits ensures something will fail catastrophically at some future date.
It happens no matter how well you try to document the functionality. If a tool is there, it will be used in grossly inappropriate ways at some point in its existence thanks to project managers who don't know what they need and programmers who can't program[^].
Same for me and in C++, I also use compile time checking in some cases. For exemple, if the code is hardcoded for the case a constant has a given value (maybe to be more efficient), I would do a compile time check so that if the constant is changed, it would raise a compile time error...
Programming defensively for stupid users is just tiresome.
In India, you have people trying hard to break any system you give them.
Thay are not stupid. They are intelligent.
It is their way of rebelling against the system.
There are three kinds of systems: fool-proof, idiot-proof and elephant-up proof.
India is the world's laboratory for elephant-up proofing anything. Sort of like the Underwriters Laboratory in the US.
Here in Chennai, the Outsourcing Capital of the World, the Municipal Corporation computerized the property payment system.
Inexplicably, the computerized system assigned number 162-0000-025 to my friend's house. The previous number was 162-0000-000.
The guy who took in the payment of the property tax, struck out the 025 and put in 000 - the old number - and tried to apply the payment. The system moved it to "Suspense Account". This went on for 4 years. It has taken countless visits to an overcrowded central office to try to resolve the issue. My friend is out at least one tax payment.
It is a good thing that 000 was not assigned to someone else. Then my friend would have been paying somebody else's property tax and the Municipal Corporation would cheerfully have told her to contact the other person and get the money back!
A few people weren't too sure what it means, so to clarify:
Like defensive driving - always assume the worst is waiting to happen. Yes, it makes you slower, and it may be less fun, but you arrive in one piece with a low adrenalin level. I never bent a single fender, whereas the spousal unit has scrapped three (THREE!) cars because of his aggressive driving style.
here in Asia "Agile Development" pretty much equates to "Broken underfunded projects delivered in 2 months" and we're very god damn agile here in Asia, can't really afford *defensive* things. Embrace risk, code offensively! Amen!
Built-in: i choose to use C#. The choice is defensive, and works. Think: Visual Studio.
OTOH, a current project in VBA required me to build all KINDS of my own fences, error processing, and tracing mechanisms merely to maintain a sane environment. Given a choice, i will choose anything other than VBA, up to and including refusing the project.
Grace + Peace
Peter N Roth, President
Ah! See, I was thinking defensive programming would protect ME, whereas other responders are looking to protect the SOFTWARE. To hell with it, I say; it’s just a machine… but choosing the right language (whichever one keeps you most safe) is crucial. Type safety, garbage collection, Object Oriented, etc.
Grace + Peace
Peter N Roth, President
The choice of language is, to an extent, irrelevant in my
Consider what happens if you ask a user there age and they
This is not entirely true. Your example actually should show you this.
If the UI has a numeric control they can not input "Apples", only a numeric. The framework allowed it to be captured as such.
Even in the case of packaged libraries this is true. You as the user must use my IAcceptableInterface or it will not compile. I then can guarantee your object meets the constraints that I need with in my library during my compile time even though I have no idea how you are using it.
Ahhh the beauty of generic constraints and interfaces. Allows for very defensive programming and is framework specific.
Computers have been intelligent for a long time now. It just so happens that the program writers are about as effective as a room full of monkeys trying to crank out a copy of Hamlet.
I guess I don't quite know what defensive programming is then...
A few things that I think are good defensive programming practices are...
1. NULL checks on pointers.
2. Limited use of stack buffers, and bounds checking when used...
3. Use the const operator for reference passing, to avoid accidental writes to objects.
4. Comments when the operation of a method is opaque.
5. Make data members protected or private unless there is a good reason not to.
6. Make all unsafe methods protected or private. If they need an interface, make a public one that provides some validation of the inputs.
I've been criticized by some in using const references to pass objects. I'm told it's an optimization. Instead, I should copy objects on the stack by default. But that seems to me, to be bloatization, as a default reference point. It certainly can bring a server to it's knees if it's used as the default practice.
Kinda surprised at the people that say they don't know how the code will be used or writing secure code.
I see defensive coding as following a set pattern of tests before accessing an object:
Is it null when it shouldn't be?
Is it of the proper format expected?
Am I trapping possible exceptions?
Am I logging those exceptions and returning a message?
Basically defensive programming is making sure that in spite of usage that nothing will cause the unexpected exception window to pop up and stop the application from running.
From the wikipedia entry it seems "defensive programming" is more about avoiding security holes (buffer overflows, etc.). "continuing function of a piece of software in spite of unforeseeable usage of said software" is a bit vague. To me that sounds like using a spreadsheet program to do video editing.
Ok, we're talking about software here aren't we?
So what can be less foreseeable than developing a piece of code?
At the time of development you may have an absolute idea of what it is meant to do, and with a strict data input but it will most certainly change.
It may take a day, a month, a year but if the software itself lives, your code will have to handle some "unexpected" scenarios, and this is where your "defense" will be put to the test.
Business change, people change, everything change, so every piece of code you do must adapt even if by just handling the errors and reporting them correctly to the IT dep.
To be prepared for errors won't consume more development time if its implemented by design. Think about it right from the beginning and it will feel natural to use, not an hack.
Hmm, if you do this for a living you work to a written and agreed specification. If the customer changes his mind or wants more (they always do) he has to write it down, agree the details with you and, of course, give you more money.
You can never guess in advance all the things that can go wrong, though after a few years you get a pretty good idea, and do things like checking input data even if you sent it yourself from the previous module. If you know your customer you can anticipate future changes and put comments in - "comment this bit back in for southern hemisphere".
Like you said, customers change their mind, either because they just do or because the business itself did, but things change despite its written on a 1000 sheet spec or not.
At the end of the day the customer wants it changed, even agreeing to pay for it, it's up to you to decide how hard it will be for you .
So what I think is that the software we write must be prepared to handle change in more way than just the user input, we also must prepare it for business change, for scalability, extensibility, etc...
Usually this doesn't mean a whole bunch of extra work, specially if it's design that way right from the beginning.
If the design is right you can start to think about software changes as a way of earning extra money instead of just loosing it!
Our customers expect our SW to grow with them, so it is always in the spec that it has to be extensible and so on. They don't specify what that means, or how you test it (you have to be able to test any requirement). We always have that in mind, and it really helps if you assume it from the start.
Anyway, we are supposed to be product based now, so we have to plan from the start the possibility of selling it to someone else, so things like putting button labels in a file instead of hard-coding them, so you can change language.
(I ported a project to Arabic once, that was such fun!)
What I consider to be defensive programming seems somewhat different from some descriptions floating around the internet. If I had to give a short description of what I consider defensive programming I would say "distrust, check, and if checks fail they do so loudly".
I think of defensive programming as two rules to follow while programming:
- A block of code (e.g. function) should never handle external data (e.g. function parameters, data returned from other functions) without strictly checking that the data it receives are know good values. If the checks fail they should fail loudly and the block of code should not touch the data at all.
- A block of code should have its behaviour checked (e.g. check return/result data, unit testing). If the checks fail again they should do so loudly.
I'm a strict follower of the first rule, but I confess that I'm a bit lax in relation to the second, especially regarding unit testing.
We make safety critical stuff so we have to, and it gets checked. Having said that, you can still write rubbish, but it comes out in the test and integration.
I like to try and write foolproof code. It's a bit of added fun to what could otherwise be boring. I am a badass and dreaded tester, because I have a rough idea where the bodies are going to be buried, having interred a few myself over the last decades. And yes, I have found a few nasties in my own code ... oops
The theory is there, but full application of said theory is impossible, IMHO.
"the meat from that butcher is just the dogs danglies, absolutely amazing cuts of beef." - DaveAuld (2011) "No, that is just the earthly manifestation of the Great God Retardon." - Nagy Vilmos (2011)
"It is the celestial scrotum of good luck!" - Nagy Vilmos (2011)
"But you probably have the smoothest scrotum of any grown man" - Pete O'Hanlon (2012)
I always try to work out everything the user could possibly do wrong, and code to catch it.
I consider that approach to be bad, security wise. Trying to prevent wrong behaviour/data is a flawed strategy and far too frequently fails. A better approach is to only accept know good behaviour/data. Yes, sometimes some good behaviour/data may be blocked but bad behaviour/data will almost certainly also be blocked.
I use defensive programming;
-by proper unit testing while developing a particular module.
-by proper comment lines (for behavior of respective function/ method) and regions.
and of course, meaningful naming conventions (by which one should easily come to know what's its use) all over in application.