<?xml version="1.0"?>
<rss version="2.0">
  <channel>
    <title>Advogato blog for torsten</title>
    <link>http://www.advogato.org/person/torsten/</link>
    <description>Advogato blog for torsten</description>
    <language>en-us</language>
    <generator>mod_virgule</generator>
    <pubDate>Fri, 22 Aug 2008 03:22:53 GMT</pubDate>
    <item>
      <pubDate>Thu, 13 Dec 2001 01:34:28 GMT</pubDate>
      <title>13 Dec 2001</title>
      <link>http://www.advogato.org/person/torsten/diary.html?start=8</link>
      <guid>http://www.advogato.org/person/torsten/diary.html?start=8</guid>
      <description>Spent half the night to fix the oustanding bugs of 
my python-gnome Debian package and uploaded a new 
release of gs-pdfencrypt as well which did not nearly
take as much time. Refreshing to do something practical
instead of wasting time with studying computer science.
</description>
    </item>
    <item>
      <pubDate>Tue, 20 Nov 2001 01:33:38 GMT</pubDate>
      <title>20 Nov 2001</title>
      <link>http://www.advogato.org/person/torsten/diary.html?start=7</link>
      <guid>http://www.advogato.org/person/torsten/diary.html?start=7</guid>
      <description>In spite having no time for unimportant stuff like that I 
am still to curious if the updated toolchain on hppa 
automagically fixes bug &lt;a href="http://bugs.debian.org/107437" &gt;#107437&lt;/a&gt; so I logged 
into sarti.debian.org and tried to build gs there. 
libgimpprint-dev is not installed so the build will not be 
complete but anyway it should suffice to test basic 
functionality.&lt;p&gt;
Okay, now it does not even build anymore. It is stuck in a 
tool that tries to detect the characteristics of the build 
system. 0% cpu usage in a memset loop!? Never mind...
</description>
    </item>
    <item>
      <pubDate>Wed, 14 Nov 2001 07:01:55 GMT</pubDate>
      <title>14 Nov 2001</title>
      <link>http://www.advogato.org/person/torsten/diary.html?start=6</link>
      <guid>http://www.advogato.org/person/torsten/diary.html?start=6</guid>
      <description>Updated the gimp-python package to support python 2.1
closing a serious bug. And rebuilt python-gnome with the
next libgtkhtml and libgal (this is getting annoying).</description>
    </item>
    <item>
      <pubDate>Thu, 8 Nov 2001 16:32:26 GMT</pubDate>
      <title>8 Nov 2001</title>
      <link>http://www.advogato.org/person/torsten/diary.html?start=5</link>
      <guid>http://www.advogato.org/person/torsten/diary.html?start=5</guid>
      <description>Rebuilt python-gnome again to fix the latest libgal 
dependency breakage :-( and uploaded the result.
</description>
    </item>
    <item>
      <pubDate>Tue, 6 Nov 2001 09:04:16 GMT</pubDate>
      <title>6 Nov 2001</title>
      <link>http://www.advogato.org/person/torsten/diary.html?start=4</link>
      <guid>http://www.advogato.org/person/torsten/diary.html?start=4</guid>
      <description>Matthias Klose did an NMU of python-opengl so I closed &lt;a href="http://bugs.debian.org/118298" &gt;bug#118298&lt;/a&gt; to allow 
installation of python-gnome into testing.</description>
    </item>
    <item>
      <pubDate>Sun, 4 Nov 2001 23:29:05 GMT</pubDate>
      <title>4 Nov 2001</title>
      <link>http://www.advogato.org/person/torsten/diary.html?start=3</link>
      <guid>http://www.advogato.org/person/torsten/diary.html?start=3</guid>
      <description>Rebuilt python-gnome with python2.1 and uploaded the 
result. Had to hack configure to add the include dirs for 
gdk-pixbuf and gtkhtml - very ugly :-( 

&lt;p&gt; A problem remains: python-gtkglarea depends on 
python-opengl which is not yet built for python2.1. Wrote 
a bug report to ask the maintainer to do it. Enough for 
today, tired.</description>
    </item>
    <item>
      <pubDate>Sun, 4 Nov 2001 01:28:29 GMT</pubDate>
      <title>4 Nov 2001</title>
      <link>http://www.advogato.org/person/torsten/diary.html?start=2</link>
      <guid>http://www.advogato.org/person/torsten/diary.html?start=2</guid>
      <description>Worked on repackaging &lt;a
href="http://packages.debian.org/python-gnome"&gt;python-gnome&lt;/a&gt;
for the new Python package in Debian. Problem is that I want
to build for both python1.5 and python2.1. dh_movefiles does
not like to move from different install trees so it seems I
have to craft something by hand. Deferred for now. Building
works so far but the result needs to get installed into the
packages trees.

&lt;p&gt; Had a look at &lt;a
href="http://bugs.debian.org/torsten@debian.org"&gt;my bug
reports&lt;/a&gt;: 
&lt;ul&gt;
&lt;li&gt; &lt;a href="http://bugs.debian.org/115308" &gt;115308&lt;/a&gt;:
Looks like I built a new python-gnome against libgal11 but
never uploaded it. Silly me!
&lt;li&gt; &lt;a href="http://bugs.debian.org/108039" &gt;108039&lt;/a&gt;:
Found out that the installed shared libraries are only
needed for special SWIG wrappers with runtime type
information. I am going to split those libraries out. Not
sure what to do about versioning though.
&lt;li&gt; &lt;a href="http://bugs.debian.org/112011" &gt;112011&lt;/a&gt;:
Seems like gsfonts in testing is too new for the gs
available in the same distribution. I guess we can only wait
until the current gs gets into testing.
&lt;/ul&gt;

&lt;p&gt; Checked &lt;a
href="http://ftp-master.debian.org/testing/update_excuses.html"&gt;update_excuses&lt;/a&gt;
for why gs is not current in testing. Seems like the
autobuild failed due to libgimpprint-dev being unavailable
on mipsel :-((  Wrote an email to debian-{devel,mips}
explaining the situation. </description>
    </item>
    <item>
      <pubDate>Wed, 19 Sep 2001 21:32:30 GMT</pubDate>
      <title>19 Sep 2001</title>
      <link>http://www.advogato.org/person/torsten/diary.html?start=1</link>
      <guid>http://www.advogato.org/person/torsten/diary.html?start=1</guid>
      <description>&lt;p&gt;Spent most of the day for university and related stuff. 
Gave back a book which was I wanted to keep but which loan 
period I could not extend because somebody subscribed it 
already. Kind of silly since there were about 30 more 
copies available so I got it back from there. But I have to 
pay 6,10 DM because they sent me a reminder already. 
Morons! Guess I have to rewrite their online booking 
software.

&lt;p&gt; &lt;p&gt;Accordingly I did not have much time for free software. 
At least I managed to upload a new &lt;a href="http://gnofin.sourceforge.net/" &gt;gnofin&lt;/a&gt; package 
which fixes the build failure on some architectures.

&lt;p&gt; &lt;p&gt;Did not even read the mailing lists or slashdot. Perhaps 
tomorrow. At least I managed to run a dist-upgrade and 
that way uninstalled parts of my kde. Left the fixing for 
tomorrow.
</description>
    </item>
    <item>
      <pubDate>Tue, 18 Sep 2001 15:37:22 GMT</pubDate>
      <title>18 Sep 2001</title>
      <link>http://www.advogato.org/person/torsten/diary.html?start=0</link>
      <guid>http://www.advogato.org/person/torsten/diary.html?start=0</guid>
      <description>Today I spent my time looking at more of the &lt;a
href="http://bugs.debian.org/gs"&gt;bug reports&lt;/a&gt; of the &lt;a
href="http://www.ghostscript.com/"&gt;Ghostscript&lt;/a&gt; packages
I maintain for Debian. 

&lt;p&gt; &lt;p&gt; I decided I should improve my contact with the upstream
developers and subscribed to some mailing lists. I have to
work out what is going on on the &lt;a
href="http://www.debian.org/ports/hppa/"&gt;PA-RISC port&lt;/a&gt;
regarding Ghostscript. Somehow the &lt;a
href="http://bugs.debian.org/107437"&gt;ps2pdf tool is failing
on the doxygen documentation&lt;/a&gt;. I confirmed that the
problem persists with my &lt;a
href="http://people.debian.org/~torsten/gs/"&gt;preview
package&lt;/a&gt; of the 6.51 upstream version. 

&lt;p&gt; &lt;p&gt; The idea that the Postscript file generated on PA-RISC
is
wrong was disproved by copying the .ps file to my local i386
machine and converting it with ps2pdf which went without a
problem. So the Ghostscript interpreter seems to be defect
on PA-RISC.

&lt;p&gt; &lt;p&gt; Hmpf, seems like the &lt;a
href="http://bugs.debian.org/112668"&gt;python-gtk packages
does not install anymore&lt;/a&gt; in Debian unstable. Looks like
the interface of the compileall script in python-base was
changed. Do not have the time now to check, this will have
to wait for tomorrow.</description>
    </item>
  </channel>
</rss>
