What are the worst programming habits?

[PIEBALDconsult]

Eddy Vluggen wrote:

Both are private

No; the class is internal. Trying to make it private yields: Error 1 Elements defined in a namespace cannot be explicitly declared as private, protected, or protected internal F:\Projects\ConsoleApplication1\Program.cs 9 17 ConsoleApplication1

You'll never get very far if all you do is follow instructions.

[Lost User]

PIEBALDconsult wrote:

No; the class is internal.

Yes, have just been reminded (a few posts below) that classes are internal by default, only members are private. ..which does obvious make a bit more sense.

Bastard Programmer from Hell :suss: If you can't read my code, try converting it here[^]

[richard_k]

I like this list.. more to add: 1. Making all member variables public (usually in the context of unit testing.. but also to increase coupling because the original OO design wasn't decomposed correctly) 2. Making all member methods public 3. lack of use of auto_ptr and other more modern mechanisms for properly handling pointer lifetime management. 4. lack of understanding of exceptions, especially their side effects on locking and memory management. (years of my life has been spent on fixing these types of issues..) 5. misunderstanding by value semantics when passing instances of classes across method interfaces. 6. passing container instances by value (eek!). Note these are ALL things I've directly observed/fixed.

[Lost User]

PIEBALDconsult wrote:

X| Yuck, filth, "littering code without adding ANY value whatsoever". X|

:) There's also a link somewhere in these posts to an article from Joel Spolsky. There have been various ways of indicating that something is a member, as opposed to a local variable. How do YOU differentiate between the two?

class X
{
private int i;
public int I { get { return i; }
public X(int someI)
{
this.i = someI;

// further down the road
int i = I;

}
}

Looks better than prefixing the stuff, but there's hardly any hint that it is a private member or a local variable. It is obvious that I must be something public, either a field or a property. It'd also cause trouble with translating to VB. A simple workaround is using this to indicate the member;

class X
{
private int i;
public int I { get { return this.i; }
public X(int someI)
{
this.i = someI;

// further down the road
int i = I;

}
}

The prefix is merely there to indicate that it is a member that is being referenced. I'd be writing the same as below;;

class X
{
int _i;
public int I { get { return _i; }
public X(int someI)
{
_i = someI;

// further down the road
int i = I;

}
}

No confusion, and the minimal amount of symbols to convey all the information I want; everything private is recognizable by the underscore, everything public starts with a capital (similar as the recommendations) and everything local start with a non-capital. The value that it adds is that it makes the naming predictable and consistent, makes it easier to spot errors and read the code. It is something that I found that does not cost time, but saves me time.

Bastard Programmer from Hell :suss: If you can't read my code, try converting it here[^]

[Lost User]

Dave Kreskowiak wrote:

I think the entire "private as default" or whatever modifier is default is a Band-Aid on a bigger problem.

With the argumentation that one needs to save the developer from himself. I'm sorry, but that only flies if you don't have any using-clause in your files. Be explicit in the class that you use, or accept the default behaviour.

Dave Kreskowiak wrote:

I don't believe that the problem is with VB. I

Not? You just explained why it is :)

Dave Kreskowiak wrote:

I've has more than few degreed grads that couldn't tell me the difference between public and private. I've also heard most of those same grads say they've never written an API, to which I call bullsh*t since every application contains it's own API, usually for the sole consumer being the application itself.

Knowing the difference between the access-modifiers would be kinda basic knowledge. And no, unless it is intended to be consumed by the public, it is not a public API. If there's no SDK to download, then you're not building a framework.

Bastard Programmer from Hell :suss: If you can't read my code, try converting it here[^]

[jschell]

Chris Maunder wrote:

What's your list?

1. Building something because the developer wants it rather than because a customer wants it. 2. Assuming that because something is new that it better. 3. Assuming that because something is new is it without fault and requires no time learn to use well.

[RafagaX]

I think i'm guilty of the first one, although I usually only comment when it's not clear what the code does or it may have some unintended side effects.

CEO at: - Rafaga Systems - Para Facturas - Modern Components for the moment...

[R Erasmus]

Using too many if-statements where better suitable mechanism could be used instead.

"Program testing can be used to show the presence of bugs, but never to show their absence." << please vote!! >>

[CHill60]

Quote:

9. Swallowing Exceptions

My number 1 pet hate. Should be in capitals. //BUT THAT IS SOMETHING ELSE THAT'S ANNOYING IN COMMENTS Plz 4giv me shtng :-D

[moonwalker72067]

ctx - it is assumed that it should point to some context structure, it is(should; may) not variable in usual sense but rather - the function parameter so it may be "pure functional".

[Naoya Yamaguchi]

I have nothing to say against use of comments. They may not be necessary for some, but usually do no harm to anyone. I believe programmers are free to choose any name for a variable so long as they communicate well to people you work with. Using "o" is fine. If it's a problem, change your font. Bad formatting used to a bad practice, but nowadays, we have apps and services to make them neat.

[Lost User]

I'm not sure what you mean - the timings were all on the same machine.

PooperPig - Coming Soon

[Jorgen Andersson]

I was just curious on the performance of SingleOrDefault vs FirstOrDefault as per Petes suggestion.

Wrong is evil and must be defeated. - Jeff Ello[^]

[Lost User]

Oh - I see. I didn't do that - the figures were from a test I did some time ago when someone kept going through our code base changing all the .where(predicate).something() to .something(predicate) because they said it was more efficient - which I didn't think was the case.

PooperPig - Coming Soon

[GSN CP]

same goes for me!

.:>GSN<:.

[Ravi Bhavnani]

Thanks for this week's survey - it's great! :thumbsup: /ravi

My new year resolution: 2048 x 1536 Home | Articles | My .NET bits | Freeware ravib(at)ravib(dot)com

[Luiz Monad]

Too long? I want to kill you if you don't delete old code. That's why we have (I have, but it's hard to convince people to use) version control. There should be 0 old code commented. One time I just deleted all comments that where not proper "human" language, except for code examples, because they're the only exception to having code in comments.

[Richard Andrew x64]

Luiz Felipe Stangarlin wrote:

I want to kill you if you don't delete old code.

Maybe you should cut down on the caffeine? ;P

The difficult we do right away... ...the impossible takes slightly longer.