<kc version="0.1.0"> 
 
<headquote><a href="http://www.cs.utah.edu/projects/flux/mach4/html/">Mach 
4</a> | <a href="http://www.gnu.org/software/hurd/hurd.html">Hurd Servers</a> 
| <a href="http://www.debian.org/ports/hurd/">Debian Hurd 
Home</a> | <a href="http://lists.debian.org/#debian-hurd">debian-hurd 
List Archives</a> | <a 
href="http://mail.gnu.org/pipermail/bug-hurd/">bug-hurd List Archives</a> | 
<a href="http://mail.gnu.org/pipermail/help-hurd/">help-hurd List Archive</a> | 
<a href="http://www.gnu.org/software/hurd/reference-manual.html">Hurd Reference 
Manual</a> | <a href="http://web.walfield.org/pub/people/neal/papers/hurd-installation-guide/english/hurd-install-guide.html">Hurd 
Installation Guide</a> | <a 
href="http://pages.hotbot.com/sf/igorkh/gnumach-cross.txt">Cross-Compiling 
GNUMach</a> | <a 
href="http://www.urbanophile.com/arenn/hacking/hurd/hurd-hardware.html">Hurd 
Hardware Compatibility Guide</a></headquote> 
 
<title>Hurd Traffic</title> 
 
<editor contact="mailto:paule@chem.gla.ac.uk">Paul Emsley</editor> 
 
<issue num="120" date="30 Jul 2002 23:00:00 -0800" /> 

<section
  title="Buildd Status"
  author="Paul Emsley"
  contact="mailto:paule@chem.gla.ac.uk"
  subject="Status Update"
  archive="http://lists.debian.org/debian-hurd/2002/debian-hurd-200207/msg00192.html"
  posts="5"
  startdate="19 Apr 2001 13:28:30 -0800"
  enddate="26 Jul 2002 08:58:06 -0800"
>

<mention>Jeff Bailey</mention>
<mention>Oystein Viggen</mention>

<p>Several weeks ago, Jeff Bailey had said that libc0.3 had been
accepted upstream and that now buildd was working. </p>

<p>Grant Bowman has offered to see if he can take these status reports
and keep <a
href="http://www.vmlinux.org/twiki/bin/view/Hurd/BuilddStatus">part
of the Twiki</a> updated so that there's a continuing list of packages
that are known to fail. </p>
 
<p>Before the release of Woody, the <a
href="http://buildd.debian.org/stats/">stats page</a> only included
release architectures.  As we are now (thankfully) past that stage,
Oystein Viggen reminded Jeff to follow this up so that Hurd package
statistics are also displayed (as of writting, they are not). </p>

</section>

<section
  title="J1 CD images"
  author="Paul Emsley"
  contact="mailto:paule@chem.gla.ac.uk"
  subject="J1 CD images"
  archive="http://lists.debian.org/debian-hurd/2002/debian-hurd-200207/msg00191.html"
  posts="15"
  startdate="21 Apr 2002 02:51:58 -0800"
  enddate="26 Jul 2002 18:14:32 -0800"
>

<p>Phil Charles reported: <quote who="Philip Charles">The J floppies,
debian-cd and baseHurd are all working, although not as well as the H3
series.  I am now at the fiddling stage to get everything working as
well as I can. [...]  If there are no problems I will produce <a href="http://tldp.org/HOWTO/mini/Debian-Jigdo/">jigdo</a>
templates as well.  </quote> </p>

 
<p>Phil went on to say:  </p>

<quote who="Philip Charles">
<p>I am drawing a boundary here.  I will produce the template and let others
work out how to use it.  There are three sources that IMHO that will need
to be used:</p>

<ol>
<li> sid, from an ftp site.  (Frozen at the date of the CD build?)</li>
<li>alpha.gnu.org (when it is rebuilt)</li>
<li>loop mounted images for boot-floppies etc.</li>
</ol>
</quote>

<p>E.O.T. </p>

</section>


<section
  title="Linux on the Hurd"
  author="Paul Emsley"
  contact="mailto:paule@chem.gla.ac.uk"
  subject="UML and the Hurd"
  archive="http://lists.debian.org/debian-hurd/2002/debian-hurd-200207/msg00129.html"
  posts="6"
  startdate="22 Jul 2002 05:29:24 -0800"
  enddate="22 Jul 2002 13:28:40 -0800"
>

<mention>Marcus Brinkmann</mention>

<p>In replying to a question about porting the linux kernel to the
Hurd, Jeff Bailey questioned its utility: <quote who="Jeff Bailey">The
question is: What would you like to get out of doing it?  It's would
probably be an interesting exersize, but wouldn't really further Hurd
development all that much.</quote> </p>

<p>Mention was made of user-mode L4 microkernel.  Marcus Brinkmann
replied that if the Hurd has L4 as a microkernel, then one can also
have a Linux-based single server in parallel with the Hurd on top of
L4. Neil Walfield tidied up saying <quote who="Neil Walfield">Of
course, there are a variety of issues here.  The foremost being
hardware sharing.  Once that problem is solved, it should be possible
with a bit of effort.</quote> </p>

<p>Those interested in the L4 Hurd port should also see <a
href="http://kerneltrap.org/node.php?id=367">this kerneltrap.org
article</a> (and references thereof).</p>




</section>


<section
  title="Documentation for Newcomers"
  author="Paul Emsley"
  contact="mailto:paule@chem.gla.ac.uk"
  subject="Advice"
  archive="http://lists.debian.org/debian-hurd/2002/debian-hurd-200207/msg00143.html"
  posts="10"
  startdate="23 Jul 2002 14:37:29 -0800"
  enddate="24 Jul 2002 04:56:19 -0800"
>
<topic>POSIX</topic>

<p>Replying to the (often asked) question "what documentation should
be read by a newcomer to the Hurd?", Wolfgang J&#228;hrling replied: <quote
who="Wolfgang J&#228;hrling"> <a
href="http://hurd.gnu.org/">http://hurd.gnu.org/</a> and <a
href="http://www.debian.org/ports/hurd/">http://www.debian.org/ports/hurd/</a>.
You will find a lot of useful information there, like how to submit
patches, how to port packages and so on.</quote> </p>

 
<p>Marcus Brinkmann added:</p>
<quote who="Marcus Brinkmann">

 
<p> Unfortunately, the learning curve is very, very steep.  However,
there are many tasks along the road that don't require a full
understanding of everything.  There is very little documentation, I am
sorry.  Large contributions to the Hurd not only require a thorough
understanding of several programming principles, but also a good idea
of what the Hurd design is about (which requires a lot of knowledge
about POSIX, but also other operating system design ideas).</p>

<p>
So, the best way is to take it one step at a time.  In using the Hurd, you
will certainly find programs that have not been ported yet.  You will find
small features that are lacking.  You will wonder about why something was
done in a certain way which you might thing could be done differently.  And
then you do some work, or ask some clever questions.  It's a long way, but
useful contributions are not at the end of that way.  They are along the
way, everywhere a step aside of the learning road. </p>

<p>[Re: CVS access problems] We are in the middle of serious mayhem,
sorry.  Currently the problems are in glibc.  I hope we will soon make
progress on that matter.  </p>

</quote>


</section>

<section
  title="Glibc without installed headers"
  author="Paul Emsley"
  contact="mailto:paule@chem.gla.ac.uk"
  subject="Glibc without installed headers"
  archive="http://mail.gnu.org/pipermail/bug-hurd/2002-July/009753.html"
  posts="10"
  startdate="17 Jul 2002 15:21:17 -0800"
  enddate="23 Jul 2002 04:58:57 -0800"
>

<mention>Jeff Bailey</mention>
<mention>Marcus Brinkmann</mention>

<p>There was discussion between Marcus Brinkmann, Jeff Bailey and
Roland McGrath about building glibc without glibc headers. </p>


 
<p>Marcus said that if you do: </p>

<p>$ make install-headers no_deps=t </p>
 
<p>in the configured Hurd sources, you get errors because certain
header files are not found (bits/utsname.h). </p>

<p>Roland replied: <quote who="Roland McGrath">Ah!  This is new, and
it's my fault.  Previously all headers the Hurd installed were simple
files to be copied.  Now trivfs has some headers that are
mig-generated, and you can't do mig generation without a proper full
set of headers installed.  I think what we should do is have
install-headers, or a differently-named new target, install just the
plain headers and not the generated ones.  Or it might be sufficient
just to do make install-headers in the hurd subdir to build
libc.</quote> </p>


<p>Marcus said that he would try it. </p>


</section>


</kc>
