Thread: What makes you a C++ expert?

This question is for anyone, really, but I'm only really asking it because I saw phantom say that he had 15 years experience with C++ and was still learning new nuances of the language. So this got me wondering, if I know this much after about 2 - 3 months of C++, how far will I be if I keep doing this 15 years down the line?

All you other C++ experts, how do you define your expertise with the language?

What are the things you've learned over the years that make you an 'expert' and by 'expert' I mean, acquainted with the subtleties of the language that allow you to write code with superior throughput, as that is nearly the only thing that matters now. You can always buy more RAM, you can't buy more time.

If anything, I'd say that I'm not particularly good at solving a broad range of problems but I am really good at using C++ to solve a particular problem.

But I'd say expertise grows because the C++ standard and STL change constantly but I think some people are stuck in 1998. Or was it 99?

What phantom is saying here is just a bit aggrandized. With that many years of programming experience, I am sure he is quite knowledge about most aspects of C++, even the nuances. He's only emphasizing the hold provervb of 'there is still much one can learn from something old.' In the simplest terms, I believe an expert is someone who knows nearly every aspect of what they are said to be expert in. But since most knowledge is not static, an expert can still be learning, even though they are an expert. If an expert does not continue to learn, they will eventually no longer be considered an expert. Being expert is a full time job. I'm sorry I couldn't give you the definite answer you might have sought, but hopefully, I answered a part of your question.

I remember how I felt when I realised that there are not one living person who has skills worth mentioning in all fields of mathematics...

The issue with code languages, take C/C++ for instance, is that you really need to be very skilled in the implemented field. You need to know the algorithms used, how the develop reasoning suitable for code etc in your field to be of use as a coder. A plain coder used by an other scientist is never good. Then you have a situation like in the biology/bio-informatics. I've been working with biologists and bio-inmormatist and there are so many issues I'd be sitting here all day if I tried to put them down in this post.

Well that's just my... humble thoughts.

Most programmers end up specializing in some type of programming and may end up becoming experts in their specific field of programming, but only use a subset of all the features of the C++ language which includes the standard template library and a standard library function set. It's more likely that most of the C++ "experts" can be found in the field of education, and the people that maintain and update the language standards for C++.

I just am.

Originally Posted by grumpy

If I ever become an expert, I'll let you know what it takes to become one. Be patient, because it will be a long wait.

I knew one person who I considered to be an expert. She denied it.

Of course she would because she doesn't consider herself as one ; she also struggles with stuff.

Originally Posted by Aslaville

Of course she would because she doesn't consider herself as one ; she also struggles with stuff.

I think this idea was well summarized by brewbuck: "...expertise is a quality that other people ascribe to you, it should never be something you ascribe to yourself."

Being knowledgeable in a complicated field means having a good sense of how deep the subject matter has been collectively explored, thus allowing you to gauge your ignorance level relative to the potentially "complete" knowledge no individual can ever achieve themselves.

Those looking in from the outside, with little or no experience in that field (or a specific branch of that field), don't have a full sense of that depth - they just see a level of absolute knowledge which to them might make that person appear to be an expert.

Even to those who do have a sense of depth of the field, someone who appears to have a greater knowledge (i.e. closer to the point of "complete" knowledge) can appear to be an expert.

However, someone who is intelligent and honest clearly sees the gap between their current knowledge and the theoretical "complete" knowledge, which makes them hesitant to call themselves an expert.

In short - it's all relative.

Alright, well thank you guys for the responses.

I've seem some job applications that have requested a C++ 'expert' or 'guru'. Basically, if you had to get a programming job, what would you tell the interviewer that makes you so good? Especially, at C++?

One thing I learned from multithreading was writing cache coherent code or at least, consciously trying to. I've also started consciously using the stack over dynamic allocation, now that I've learned the full cost of memory look-up. By this, I don't mean that the stack isn't in the RAM but rather, malloc() has to search for a validly sized region to write to while stack allocations, you pretty much automatically know where you'll allocating because it's always the address of the top of the stack.

These are just a couple of the big things to have radically changed my style and basically woke me up into paying attention to what I'm doing. And I would never go so far as to say that these traits alone make me an expert but they are definitely something I didn't know like 9 months ago. At 23, it's hard to even imagine myself 15 years down the road. I'd be like 38? Yeah... That's waaaaaaaaaay too far away for me but I guess it actually happens and that just freaks me out so I was trying to peek at the end and ask about what I might be like 15 years from now 'cause I am never going to stop programming.