discuss-gnustep
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: how to compile a .m


From: David Chisnall
Subject: Re: how to compile a .m
Date: Wed, 14 Oct 2009 09:28:13 +0100

A couple of other people have already mentioned GNUstep Make. For short programs, gnustep-config is also an option. You can compile a simple Objective-C program like this:

gcc `gnustep-config --objc-flags --base-libs` hello.m

This will generate an a.out file linked against GNUstep Base (Foundation). Substitute --gui-libs if you want to link AppKit (but, generally, if you are linking against AppKit you will want to make a bundle and then it's much easier to use GNUstep Make). For both, you will need to source the GNUstep.sh file first, I believe (I'm not 100% sure if gnustep-config needs this).

On 14 Oct 2009, at 00:30, Jean-Loïc Mauduy wrote:

#import <stdio.h>


This is wrong. A few Objective-C tutorials make this mistake, and tell you to just use #import instead of #include in Objective-C programs, but this is terrible advice. #include is a trivial preprocessor directive that just inserts the contents of the specified file at this point. #import is a bit more clever, and ensures that the file is only ever inserted once.

Objective-C headers are, generally, designed to be used with #import. A lot of C (and C++) headers, however, are not. They will protect themselves from multiple inclusion with macros and may be designed to work differently if included more than once in a compilation unit. If you get into the habit of using #import with C headers, then you are going to end up with something breaking eventually, and you are going to be very confused about why. Only use #import with Objective-C headers; stick with #include for C headers. This also provides a clue to people reading your code about what kind of header you are including.

David

-- Sent from my Apple II





reply via email to

[Prev in Thread] Current Thread [Next in Thread]