C++ (Passing by Reference) Programmer Insecurity

Sometimes even the smartest, and I mean proper mind blowing good, programmers can have an off day... And I just ran into one.  Now, let me just say I consider myself a good C++ programmer, and an excellent Pascal programmer, I prefer C++ however and am taking every chance I get to use it, and use it cross platform.

Now, the chap I'm talking about, is a bit of a mega brain, he is really good, but he's just had a massive problem on his hands, and I don't think he liked my input...

First of all, he was working with "int", on a 32bit machine... But he was using this "int" as an unsigned value... So why not use "unsigned int"... apparently, he is having the kind of day where he doesn't want to type "unsigned " everywhere, so I suggested, replace "int" with "uint" and do "typedef unsigned int uint;" the silence was palpable...

He did it, and his code started to work, but it seemed he really wanted to just continue having an off day, rather then have me intereject with a solution...

Secondly, when he was done, he was getting a lot of memory used, he reckoned his code was fast and not creating copies... "int ValueReturn (int _lhs, int _rhs)".... A second suggestion, which I gave him via another member of staff "void ValueReturn (const uint& _lhs, const uint& _rhs, uint& _retVal)" to cut down the number of parameter copies and return copies he makes also went down like a tonne of bricks, he's stormed off, moaning that "that lot are a bunch of know it alls"

I can relate to his frustration, he's had a major problem, and we've effectively solved it for him in a few seconds, with what could be assumed to be obscure pieces of C++ code, but these practices of passing by reference, and even constant reference, and if you want to avoid creating more memory returning a reference to an [in/out] parameter is all normal fair.  Scott Meyer points this out a lot in his books.

I think I grew out of this tantrum "my code is right, the system is wrong" attitude about three years ago, it feels longer, but I know it takes a long time for a programmer to be comfortable enough with their ability to sit back and say "yeah, I got that wrong" and fix it, rather than think admitting fault lowers their value to everyone.  I certainly know from experience that those programmers who sit on their pedestal and are never challenged about shoddy code they do (even code which gets the job at hand done) will never progress and improve.

Anyway, something funny...