Comments on: Mozilla C++ portability guide http://briancarper.net/2007/07/18/mozilla-c-portability-guide/ Wed, 07 Jan 2009 03:11:01 +0000 http://wordpress.org/?v=2.3.2 By: Ranjith Thai Valappil http://briancarper.net/2007/07/18/mozilla-c-portability-guide/#comment-5790 Ranjith Thai Valappil Thu, 13 Sep 2007 14:07:28 +0000 http://briancarper.net/2007/07/18/mozilla-c-portability-guide/#comment-5790 I will be using new C++ compilers, which will be supporting RTTI. Is there still any problem in using RTTI(dynamic_cast all those things)? I will be using new C++ compilers, which will be supporting RTTI. Is there still any problem in using RTTI(dynamic_cast all those things)?

]]> By: Brian http://briancarper.net/2007/07/18/mozilla-c-portability-guide/#comment-4487 Brian Sun, 22 Jul 2007 23:47:42 +0000 http://briancarper.net/2007/07/18/mozilla-c-portability-guide/#comment-4487 Awesome. Thanks for the advice and clarification. Awesome. Thanks for the advice and clarification.

]]> By: Syd http://briancarper.net/2007/07/18/mozilla-c-portability-guide/#comment-4480 Syd Sun, 22 Jul 2007 07:29:19 +0000 http://briancarper.net/2007/07/18/mozilla-c-portability-guide/#comment-4480 I happened to work at Netscape at the time this document was produced. Many of the items are dated; I don't, for example, agree today that avoid RTTI is a reasonable guideline. I'm currently preparing a book on C++ portability that will be released by Prentice Hall, based on my experiences at Netscape. Of the items mentioned by Mozilla in that document, when it comes down to it, the only ones I mention are 1) take care regarding the sign of char types 2) use of NSPR (using an abstraction library is definitely a good thing and 3) fix your compiler warnings. Beyond that, I think anything that Stroustrup claims is safe (especially if portability means working with Visual C++, and GCC on Mac and Linux -- which is the case for a great many people). After reading Stroustrup, I'd move on to the Effective C++ book by Scott Meyers. Read that book back to front, and then do it again twice more. Use Stroustrup as a reference, but read (and practice) what Meyers has to say until it is second nature. You'll be a better C++ programmer for it. I happened to work at Netscape at the time this document was produced. Many of the items are dated; I don't, for example, agree today that avoid RTTI is a reasonable guideline.

I'm currently preparing a book on C++ portability that will be released by Prentice Hall, based on my experiences at Netscape. Of the items mentioned by Mozilla in that document, when it comes down to it, the only ones I mention are 1) take care regarding the sign of char types 2) use of NSPR (using an abstraction library is definitely a good thing and 3) fix your compiler warnings. Beyond that, I think anything that Stroustrup claims is safe (especially if portability means working with Visual C++, and GCC on Mac and Linux — which is the case for a great many people).

After reading Stroustrup, I'd move on to the Effective C++ book by Scott Meyers. Read that book back to front, and then do it again twice more. Use Stroustrup as a reference, but read (and practice) what Meyers has to say until it is second nature. You'll be a better C++ programmer for it.

]]>