I've had quite a few problems with naming conventions in the past (as I’m sure a lot of you had) and now i'm trying to find the best set of rules i can always follow to be more consistent.

The problem is not that big when just using the same language like when using Java it is absolutely clear to me that the best practice is to use camelCase naming system. But what about when read a value from a field in a database. The field name is probably going to be in lower case separated by _ (field_name). Therefore, it kinda makes sense to call a variable that gets that value "field_name". I used this logic for my PHP scripts, until i started using some classes, and using phpDoc so it works much better using Java naming conventions.

There was also something I’ve seen in quite a few programing books I read, called the Hungarian notation. I found this to work great with visual c++ for user interface components, giving them a name such as lblFirstName and txtFirstName is very nice when you want to validate a form for example and highlight any problems with the data . This principle also works well in other languages such as Java and is also good for HTML/Javascript form validation.

But again there is a problem to how far to actually take this. The following rules seem to work for the most part:
1. Data related components(file names, database design, XML, HTML tags) should use a lower case names separated by an underscore (_).
2. Data components related to programming(programming languages) should use camelCase.
3. GUI components that will be manipulated by the program(HTML and GUI classes in some programing languages) should use Hungarian notation.

However, I still find myself breaking them every once and a while, sometimes they don't make much of a sense.

So I guess my question to you is how do you deal with this? What set of rules do you follow?
Do you believe that every language or every situation needs to adhere to different naming standards? or a universal standard needs to be created?

P.S.

I didn't even talk about abbreviations and too wordy names.

Here is some information I found in Wikipedia: http://en.wikipedia.org/wiki/Identifier_naming_convention.

I have such a hard time with naming conventions as well, especially when it's just a project I'm working solo. I almost always use camelcase when I'm programming in something like C/C++/C#/Java and then switch over to underscores when I'm using PHP/Ruby. But honestly I mix and match, and sometimes I will use hungarian notation and sometimes I won't. Can be frustrating. I would really like a universal naming standard between languages just so I could be consistent with myself.

something i was sent by my boss

Programming standards

Well from what I understand, most programming projects/jobs will have coding conventions you must follow that dictate things like this.

That said, when coding Java I usually stick to their naming style (itLooksLikeThis).

In almost anything else I use the C-style, with_underscores_like_this. It's just more readable to me.

Naming conventions are just something that you get used to.

Hungarian notation is fading in my opinion since languages are strongly typed nowadays.

(For those who don't know Hungarian notation is putting a letter in the front of the variable name to indicate what it is i.e.



This post has been edited by KYA: 22 Jun, 2008 - 12:21 AM

There are still a few though that are prevalent, especially popular ones like PHP/Javascript/Perl where it can be useful to use. Even though I don't so much anymore...

@rgfirefly24- Thanks for the link, looks interesting. I would like to see some more of these documents on the web tho, particularly for web development.

@tom- Well the underscore thing works great when it comes to procedural programming, but how often do you really see that today? (except like in javascript). So for OOP I think the best choice is camelCase to differentiate between classes and everything else. I actually used the CSomeClassName thing that was a standard in one book and recommended by Microsoft in the old MFC age. But nowadays that looks pretty silly to me.
I know that every project has its' own conventions set up for it, but I was hoping to set some for my personal projects. Maybe even set some guide lines for </D.I.C> members to have more consistent tutorials and snippets.

@KYA- well Hungarian can be more than one letter can also be 3. I use it as I mentioned before in HTML forms when I need to identify a label and a text field or another component to notify the user of any mistakes he/she as made in entering the information. Is there a better way to do that?

Sounds like a good application of that naming convention

Procedural programming is I guess what I do the most of then.

I don't think you'll find any one standard convention though. As you've pointed out, different languages and programming paradigms call for different naming conventions. I say that if it's just for your personal projects, use whatever you like the most.

I've never heard of this "hungarian" style before, but it sounds like just making descriptive sounding variable names. I usually use one or two letter variables to be completely honest, I just get so sick of retyping long names. If a long name really is necessary, then I'll use preprocessor macros to shorten it up a bit. Otherwise, I try to keep my functions short (so that whoever is reading it can see the declarations and the body at the same time), and I try to write comments that explain what the #%!^!# is going on.

It's Hungarian, because it's a nation, and I'm one of them, so watch your capitalization
On the other side I don't use it, because I mostly program in Java, and it has it's own recommended convention, that I find pretty clear and fine.
A big advantage of the Hungarian notation is that since you can guess the type of the variable from it's name, you don't have to look up it's declaration when you land in the middle of a class or function, and this makes reading and understanding the code easier. Sure, a modern IDE will show it in a hint as well, so it's not that important nowadays. Historically the source of Windows was using this notation.

I'm usually pretty good about capitalization, proper nouns, all that jazz. I'm not really a morning person though.

Wow, I knew barely anything about these naming conventions yet I've been following them almost perfectly.

Must have been all the code I've been reading.