Thread: How properly inherit from template?

Concerning error-proneness, I guess Elysia is saying that

Code:

sort(vector.begin(), vector.end();

has more potential for making typos than

Code:

vector.sort();

However, these functions have a different level of usability. One lets you sort part of the container, and the other sorts all of it. If that is problem for you, nothing can stop you from writing a free function

Code:

template <class WritableRangeT>
void sort(WritableRangeT& r)
{
    sort(r.begin(), r.end());
}

Now, that would be no more error-prone.

Code:

sort(vector);

Which is exactly what I was hinting at that member functions (or free functions) are less error-prone than algorithms and you should be preferred when possible.

Originally Posted by Elysia

Which is exactly what I was hinting at that member functions (or free functions) are less error-prone than algorithms and you should be preferred when possible.

Ok, but how do you write the free functions (or member functions). Would you implement your own sort, because std::sort is so "error-prone"?

Most likely, I would try to encapsulate std::sort via different member functions in objects, yes. Because it gives the illusion or whatever you want to call it that the object is sorting itself.

I fail to see why a vector should be able to sort itself. This kind of thinking makes not even a tiny bit of sense to me. Even Java doesn't have a sort() member function in ArrayList. Instead, there's the static method Collections.sort(). This is Java's way of making it a free function. So even a language like Java, which doesn't even have free functions, is built without the idea that an array should sort itself.

If the vector is an object, then naturally you should be able to tell it to sort itself.
But I'm actually getting tired of these arguments, so I'm only going to respond to sized down posts that I find curious and easy to respond to.
And phantomotap, I have no idea what that algorithm does or how it sorts data or how it's called, and that's why I prefer to avoid dealing with algorithms.

Originally Posted by Elysia

If the vector is an object, then naturally you should be able to tell it to sort itself.

That's not an argument. What's natural about telling a vector to sort itself? Do you feel that sorting themselves is something that comes naturally to vectors? I don't. I think sorting vectors is very much an external operation.

Originally Posted by Elysia

But I'm actually getting tired of these arguments, so I'm only going to respond to sized down posts that I find curious and easy to respond to.

Right. Life is not a post count contest. You could try not posting at all, and just accept what people who know more than you are telling you. Being stubborn doesn't make anyone right.

How about responding to everyone's arguments or satisfying their curiousness? Post count means nothing to me.
I can accept how everything is, but that does not mean I have to do it nor like it. Like it or not, I'm only doing it for fun and not for work.

Originally Posted by CornedBee

That's not an argument. What's natural about telling a vector to sort itself? Do you feel that sorting themselves is something that comes naturally to vectors? I don't. I think sorting vectors is very much an external operation.

If you tell the vector to be an object, which you are, then it should sort itself with what you give it. That's my opinion.

hmm... you are right, the debate has spiralled to a dead end.

Incidentally, I was trying to tell the thread to close itself, but eventually I realised I had to use an external operation to do it.