<?xml version="1.0" encoding="UTF-8"?><!-- generator="wordpress/2.1.3" -->
<rss version="2.0" 
	xmlns:content="http://purl.org/rss/1.0/modules/content/">
<channel>
	<title>Comments on: Ubuntu Tangobos-GDC package for x86_64</title>
	<link>http://codeblog.palos.ro/2008/05/08/ubuntu-tangobos-gdc-package-for-x86_64/</link>
	<description>Thoughts about high architecture and sheer speed.</description>
	<pubDate>Fri, 30 Jul 2010 10:25:39 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.1.3</generator>

	<item>
		<title>By: zimbatm</title>
		<link>http://codeblog.palos.ro/2008/05/08/ubuntu-tangobos-gdc-package-for-x86_64/#comment-394</link>
		<author>zimbatm</author>
		<pubDate>Thu, 08 May 2008 13:51:31 +0000</pubDate>
		<guid>http://codeblog.palos.ro/2008/05/08/ubuntu-tangobos-gdc-package-for-x86_64/#comment-394</guid>
					<description>Hi, thanks a lot for your packages, it is really the best way to get an usable D environment.

Minor nitpick: ".svn" folders are left in the package. I don't how you package tango, but try using "svn export" instead of "svn co"</description>
		<content:encoded><![CDATA[<p>Hi, thanks a lot for your packages, it is really the best way to get an usable D environment.</p>
<p>Minor nitpick: &#8220;.svn&#8221; folders are left in the package. I don&#8217;t how you package tango, but try using &#8220;svn export&#8221; instead of &#8220;svn co&#8221;</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Valeriu Paloş</title>
		<link>http://codeblog.palos.ro/2008/05/08/ubuntu-tangobos-gdc-package-for-x86_64/#comment-395</link>
		<author>Valeriu Paloş</author>
		<pubDate>Thu, 08 May 2008 13:55:36 +0000</pubDate>
		<guid>http://codeblog.palos.ro/2008/05/08/ubuntu-tangobos-gdc-package-for-x86_64/#comment-395</guid>
					<description>Hmm, I'm afraid I *do* use svn co in my scripts (I'm not very familiar with svn options). I will try and change that in the next release, thanks for the tip! This kind of observation is quite helpful :).</description>
		<content:encoded><![CDATA[<p>Hmm, I&#8217;m afraid I *do* use svn co in my scripts (I&#8217;m not very familiar with svn options). I will try and change that in the next release, thanks for the tip! This kind of observation is quite helpful :).</p>
]]></content:encoded>
				</item>
	<item>
		<title>By: Valeriu Paloş</title>
		<link>http://codeblog.palos.ro/2008/05/08/ubuntu-tangobos-gdc-package-for-x86_64/#comment-396</link>
		<author>Valeriu Paloş</author>
		<pubDate>Thu, 08 May 2008 14:56:53 +0000</pubDate>
		<guid>http://codeblog.palos.ro/2008/05/08/ubuntu-tangobos-gdc-package-for-x86_64/#comment-396</guid>
					<description>Ok zimbatm, I've looked into it and in conclusion this is Tangobos's fault. Somehow, it doesn't properly link the symbols inside the std/c/time.d file with the external C libraries (as Phobos did) when compiled with DMD.

I made a small patch that solves your problem (even though this approach is horrible), however my (strong) advice for you is to switch to Tango if possible (I am aware that this might be hard depending on how much code you have that relies on Phobos).

As a side note, I *am* discussing with Lars the possibility of re-including Phobos into DMD by default, we'll see how that goes...

Here's the patch: 
&lt;pre&gt;
--- time.d	2008-04-06 14:10:25.000000000 +0300
+++ time.d_phobos	2008-02-19 02:22:00.000000000 +0200
@@ -8,7 +8,6 @@
  */
 
 module std.c.time;
-import std.compat;
 
 private import std.c.stddef;
 
@@ -38,10 +37,11 @@ version (Windows)
 else version (linux)
 {
     extern (C) int sysconf(int);
-    clock_t CLK_TCK;
-    static this() {
-        CLK_TCK = cast(clock_t) sysconf(2);
-    }
+    extern clock_t CLK_TCK;
+    /*static this()
+    {
+	CLK_TCK = cast(clock_t) sysconf(2);
+    }*/
 }
 else
 {
&lt;/pre&gt;

You need to save it as "/usr/include/d/tangobos-dmd/std/c/time_patch" and apply it (in the same directory) using:
&lt;pre&gt;
patch -p0 &lt; time_patch
&lt;/pre&gt;

After this, the compilation should work just fine. Tell me how it goes...</description>
		<content:encoded><![CDATA[<p>Ok zimbatm, I&#8217;ve looked into it and in conclusion this is Tangobos&#8217;s fault. Somehow, it doesn&#8217;t properly link the symbols inside the std/c/time.d file with the external C libraries (as Phobos did) when compiled with DMD.</p>
<p>I made a small patch that solves your problem (even though this approach is horrible), however my (strong) advice for you is to switch to Tango if possible (I am aware that this might be hard depending on how much code you have that relies on Phobos).</p>
<p>As a side note, I *am* discussing with Lars the possibility of re-including Phobos into DMD by default, we&#8217;ll see how that goes&#8230;</p>
<p>Here&#8217;s the patch: </p>
<pre>
--- time.d	2008-04-06 14:10:25.000000000 +0300
+++ time.d_phobos	2008-02-19 02:22:00.000000000 +0200
@@ -8,7 +8,6 @@
  */

 module std.c.time;
-import std.compat;

 private import std.c.stddef;

@@ -38,10 +37,11 @@ version (Windows)
 else version (linux)
 {
     extern (C) int sysconf(int);
-    clock_t CLK_TCK;
-    static this() {
-        CLK_TCK = cast(clock_t) sysconf(2);
-    }
+    extern clock_t CLK_TCK;
+    /*static this()
+    {
+	CLK_TCK = cast(clock_t) sysconf(2);
+    }*/
 }
 else
 {
</pre>
<p>You need to save it as &#8220;/usr/include/d/tangobos-dmd/std/c/time_patch&#8221; and apply it (in the same directory) using:</p>
<pre>
patch -p0 < time_patch
</pre>
<p>After this, the compilation should work just fine. Tell me how it goes&#8230;</pre>
]]></content:encoded>
				</item>
</channel>
</rss>
