Complete.Org: Mailing Lists: Archives: freeciv-dev: April 2004:
[Freeciv-Dev] Re: (PR#8164) building out of a builddir: generate_packets
Home

[Freeciv-Dev] Re: (PR#8164) building out of a builddir: generate_packets

[Top] [All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
To: jdorje@xxxxxxxxxxxxxxxxxxxxx
Subject: [Freeciv-Dev] Re: (PR#8164) building out of a builddir: generate_packets.py
From: "Marko Lindqvist" <marko.lindqvist@xxxxxxxxxxx>
Date: Wed, 21 Apr 2004 12:48:39 -0700
Reply-to: rt@xxxxxxxxxxx

<URL: http://rt.freeciv.org/Ticket/Display.html?id=8164 >

Jason Short wrote:

> Raimar Falke wrote:
> 
>>So you are right that the <>-places are searched after the ""-places
>>for ""-includes. But the terms "forward" and "backwards" are wrong or
>>at least very misleading.
> 
> Misleading because it's apparently implementation-defined.  This is 
> unfortunate.  But back-to-front is true for gcc.  For other compilers we

  I seem to be unable to find any mentions about this back-to-front 
behavior of gcc. Can you provide URL?

  Especially, 
"http://gcc.gnu.org/onlinedocs/gcc-3.3.3/cpp/Search-Path.html#Search%20Path";, 
does not mention it.

  I just found out that also what those ""-places for includes actually 
are, is somewhat implementation-defined. It's clear that ""-includes are 
first searched from "current directory" and then from <>-places. Problem 
is that term "current directory" is not defined. I think that usual 
convention is to consider directory containing file with #include 
-directive to be "current directory". But apparently there is 
implementations where "current directory" is working directory -> builddir.
  However, I'm quite sure that no compiler we want to use for compiling 
from separate builddir uses this latter convention.


  - Caz




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