If I\'ll begin developing cross platform applications, The complied file speed i

Question

If I'll begin developing cross platform applications, The complied file speed is very important for me .

I have research before than and find same result about IDE :

(I ignored eclipse because very java base, but not complete)

for IDE :

NetBeans C/C++ IDE .
CodeBlocks .
CodeLite .
or others .
And for complier || framework || toolkit :

QT toolkit or UnitTest++ .
gnu g++ .
gnu gcc .
wxWidgets .
cygwin .
VC++ .
or others.
But I watch to some QT apps are very dependent to QT's libs :( , It's not good and decrease speed.

Now my real question if I want start project what is the best IDE with debugger ,what is the best compiler that give fast speed native code to me, or I force use multi-compiler for any OS ?

& please say your reason why that IDE & why that complier.

Explanation / Answer

The GNU C++ compiler has been worked on for many years to the point that nowadays it can compile binaries for a diverse range of target platforms in terms of operating system / host environment, processor and even binary format.

You can use GNU C++ compiler along with GDB debugger on many platforms already - Linux, Mac, Windows and others. There are a multitude of IDEs that also can use g++ with GDB - you already list a good selection. Most of these are highly capable, but I think that Eclipse stands out as it is both cross-platform and highly configurable/extensible. This means that there is a vast array of additional features available through plugins.

Although aome people may say that Eclipse is a bit slow to ue, C++ developers really care about runtime performance rather than edit-compile-debug time, so I'll assume you mean that the fastest native runtime rather than native compile time.

This is where it gets more complex. While the G++ compiler front end is shared across platforms, the backend code generators can vary. If you're asking for x86 or ARM processors, it's likely that large companies such as Apple have helped tune the back end here. .

But..

No amount of tuning of the generated code can make up for poor algorithm choice.

That you're asking about the performance this early seems wrong - surely you want productivity and accuracy/stability first?

Make it work, then make it fast.

Related Questions

Navigate

• Browse (All)
• Browse I
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.