<?xml version="1.0" encoding="iso-8859-1"?>
<!-- name="generator" content="pyblosxom/1.3.2 2/13/2006" -->
<!DOCTYPE rss PUBLIC "-//Netscape Communications//DTD RSS 0.91//EN" "http://my.netscape.com/publish/formats/rss-0.91.dtd">

<rss version="0.91">
<channel>
<title>Diane's World   </title>
<link>http://woldlab.caltech.edu/~diane/index.cgi</link>
<description>I take the 5th</description>
<language>en</language>
<item>
  <title>#tags test,software</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/pyblosxom-2009-10-30-15-16.html</link>
  <description><![CDATA[
 Testing the tag module
I'm curous how the tag module works for pyblosxom.

]]></description>
</item>

<item>
  <title>SciPy 2006 Day 2</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/scipy2006-08-18-18-00.html</link>
  <description><![CDATA[

<div class="document">
<blockquote>
<ul>
<li><p class="first">AutoLigand AutoDock, Rodney Harris</p>
<ul class="simple">
<li>Tool for &quot;rational&quot; ligand design</li>
<li>computes affinity</li>
</ul>
</li>
<li><p class="first">GpuPy, Benjamin Eitzen, WashU</p>
<ul class="simple">
<li>strengths<ul>
<li>currently improving faster than traditional cpus</li>
</ul>
</li>
<li>weaknesses<ul>
<li>single precision</li>
<li>data must be copied to gpu</li>
</ul>
</li>
<li>programmability<ul>
<li>newer gpus can execute programs in &quot;high level&quot; languages
Cg GLSL HLSL</li>
<li>programs are called shaders</li>
</ul>
</li>
<li>how it works<ul>
<li>GPU executes program once for each pixel that is drawn to screen</li>
<li>overrrides default numpy functions, can  be transparent</li>
<li>some functions are slower on gpu (like simple add or multiply)</li>
<li>some get faster with larger N (pow, arccosh)</li>
<li>some are just faster (cos, sin)</li>
<li>the more complex you can pass over the faster, correlation</li>
<li>not hooked up yet, but have edge detection filter</li>
</ul>
</li>
<li>possible future work,<ul>
<li>autoconfig</li>
</ul>
</li>
<li>simulation of double  with singles hal.ccssd.cnrs.fr/ccsd-00021443</li>
<li>htpp://eecs.wsu.edu/~eitzenb/gpupy</li>
</ul>
</li>
<li><p class="first">Boost Graph Library, Douglas Gregor</p>
<ul>
<li><p class="first">path finding</p>
</li>
<li><p class="first">connectivity &amp; clustering</p>
</li>
<li><p class="first">network flow</p>
</li>
<li><p class="first">coloring</p>
</li>
<li><p class="first">layout &amp; visualization</p>
</li>
<li><p class="first">motivation</p>
<ul class="simple">
<li>python is easier to teach</li>
<li>dynamic features of python</li>
<li>usabiliy (e.g. c++ error messages)</li>
</ul>
</li>
<li><p class="first">Task: network tolerance</p>
<p>from boost.graph import *
g = read_graphviz(&quot;network.dot&quot;)</p>
<p>bicomponent = g.add_edge_property(name='label', type='integer)
art_points = biconnected_components(g, bicomponent)</p>
<dl class="docutils">
<dt>for v in g.vertices:</dt>
<dd><p class="first last">if v in art_points:</p>
</dd>
</dl>
</li>
<li><p class="first">graph types in python</p>
<ul class="simple">
<li>Graph</li>
<li>Digraph</li>
<li>Graph exploration interface</li>
</ul>
</li>
<li><p class="first">vertex and edge properties</p>
<ul class="simple">
<li>attach additional model domain knowledge<ul>
<li>computer net, vertex(host, ip_addr), edge(bandwidth, latency)</li>
</ul>
</li>
</ul>
</li>
<li><p class="first">Properties in the BGL</p>
<ul class="simple">
<li>property map<ul>
<li>implicity, just access property</li>
<li>explicity, allows more specifity, 2 params optional
weight = g.add_edge_property(name='weight', type='float')</li>
</ul>
</li>
<li>using property maps<ul>
<li>similar to dictionaries (weight[e], =0.75</li>
<li>pass to algorithms
dijkstra_shortest_path(g, start_vertex, weight)</li>
</ul>
</li>
<li>has visitor model to do something when you hit interesting
points in the algorithm. (e.g. debugging, new algorithms)</li>
</ul>
</li>
<li><p class="first"><a class="reference" href="http://www.generic-programming.org/~dgregor/bgl-python/">http://www.generic-programming.org/~dgregor/bgl-python/</a></p>
</li>
</ul>
</li>
<li><p class="first">summary of research issues with/Fast multipole Algorithm, Idesbald van den Bosch</p>
<ul class="simple">
<li>ship signature reduction</li>
<li>lots of details about problem and need for optimization</li>
<li>gmsh (mesh generation)</li>
</ul>
</li>
<li><p class="first">Object Oriented Finite Elements, Andrew Reid</p>
<ul class="simple">
<li>OOF2,<ul>
<li>rapid construction of meshes adapted to irregular microstructural
geometriesp</li>
</ul>
</li>
</ul>
</li>
<li><p class="first">Parallel PDE solvers in python, Bill Spotz</p>
<ul class="simple">
<li>trilinos project<ul>
<li>provide central repository for sandia's solver</li>
<li><a class="reference" href="http://softwrae.sandia.gov/trilinos">http://softwrae.sandia.gov/trilinos</a></li>
</ul>
</li>
</ul>
</li>
<li><p class="first">Python imaging tools for reconstructing magnetic resonance images, make trumpis</p>
<ul class="simple">
<li><a class="reference" href="https://cirl.berkeley.edu/view/BIC/ImagingTools">https://cirl.berkeley.edu/view/BIC/ImagingTools</a></li>
</ul>
</li>
<li><p class="first">PyRoot, Wim Lavrijsen (pronounced more like vim)</p>
<ul class="simple">
<li><a class="reference" href="http://cern.ch/wlav/pyroot/">http://cern.ch/wlav/pyroot/</a></li>
<li>ROOT root.cern.ch</li>
<li>reflection for C++ (reflex, <a class="reference" href="http://cern.ch/seal-reflex">http://cern.ch/seal-reflex</a>)</li>
<li>they combine reflection based meta-information to generate python
wrappers on the fly.</li>
<li>also can call python from C++/CINT</li>
<li>very impressive looking</li>
</ul>
</li>
<li><p class="first">QScimpl, Eric Dobbs</p>
<ul class="simple">
<li>personal, dobbse.net</li>
<li>Tech-X Corporation, txcorp (scientific physics stuff)</li>
<li>mostly decided to advertise the advantage of simplicity</li>
</ul>
</li>
<li><p class="first">Genesis SciFlo, Scientific Knowledge Creation on the Grid, Brian Wilson</p>
<ul class="simple">
<li>Problem, terabytes of data (years) distributed at several data centers</li>
<li>iEarth</li>
<li>showed an updating svg diagram showing the execution of their workflow.<ul>
<li>distributed across many machines via web services</li>
</ul>
</li>
<li>data access protocol, retrieve chunks of data over web.</li>
<li>used xpath to extract results from amazon soap results</li>
<li>software shipped<ul>
<li>lxml (XPath 1.0)</li>
<li>twisted, pyldap, openldap,</li>
<li>sleepycat dblxml has xquery and xpath</li>
<li>dojo AJAX library</li>
<li>openDAP, extract cdf, hdf file</li>
<li>globas v4 grid</li>
</ul>
</li>
<li>globus v4 is supposed to be soap webservice</li>
<li>WSRF (web services resource framework)</li>
<li>UDDI universal description discovery and integration, find service</li>
<li>grid workflow,</li>
<li>rest vs soap, each have advantages/disadvantages</li>
<li>UDDI4py from IBM, SOAPpy</li>
<li>recommends more XML microformats for &quot;mash ups&quot;</li>
<li>publish science algorithms as reusable web services</li>
<li>www.opendap.org</li>
</ul>
</li>
<li><p class="first">Python Web &amp; Grid Service Tools, Keith R Jackson</p>
<ul class="simple">
<li>Globus toolkit</li>
<li>decided to simplify system by standardized into webservices</li>
<li>helps organize coordinated usage of resources</li>
<li>his mac crashed</li>
<li>SOAPpy is officially unsported</li>
<li>ZSI is hopefully the &quot;new&quot; better way, though they're more complicated than SOAPpy</li>
<li>XML schema scary complicated</li>
<li>Mapping tool from WSDL to python (wsdl2py)</li>
<li>using m2crypto (they thought it was being supported, unlike pySSL)</li>
<li>LIGO used it to ship 50 terabytes of data through pyglobus</li>
<li>WSRF supposed to define way to hand stateful connections</li>
<li>lifecycle, things will die at some point, though can be refreshed.</li>
<li>WS-N another webservices spec</li>
<li>seems to actually work, java, C, python call all talk.</li>
<li>PyCLST</li>
<li>GridFTP, implement extra features to handle transfer over highperformance network</li>
<li>Visual programming environment, useful for viewing workflow</li>
<li>Userinterface</li>
<li>planner, condor, dagman</li>
<li>exec manager (CBEI)</li>
<li>exec context</li>
<li>physical resources</li>
<li>pyGlobus, <a class="reference" href="http://dsd.lbl.gov/gtg/projects/pyGlobus/">http://dsd.lbl.gov/gtg/projects/pyGlobus/</a></li>
<li>ZSI, <a class="reference" href="http://pywebsvcs.sourceforge.net/zsi.html">http://pywebsvcs.sourceforge.net/zsi.html</a></li>
<li>pyGridWare, <a class="reference" href="http://dsd.lbl.gov/gtg/projects/pyGridWare">http://dsd.lbl.gov/gtg/projects/pyGridWare</a></li>
<li>ViCE, <a class="reference" href="http://dsd.lbl.gov/gtg/projects/vice/">http://dsd.lbl.gov/gtg/projects/vice/</a></li>
<li>PythonCLServiceTool, <a class="reference" href="http://dsd.lbl.gov/gtg/projects/PythonCLServiceTrool">http://dsd.lbl.gov/gtg/projects/PythonCLServiceTrool</a></li>
<li>everything BSD license</li>
<li>eclipse has a nice WSDL interface generator</li>
<li>python to wsdl would require static type annotation.</li>
<li>webservices, text to do control, then binary for shipping data</li>
<li>amazon, uses a very efficient xml binary encoding</li>
</ul>
</li>
<li><p class="first">Seeing through the MIST, Tripp Lilley</p>
<ul class="simple">
<li>interesting point about maliable UIs<ul>
<li>how do you let a user, know theyre changing their world, and revert.</li>
</ul>
</li>
<li>super early project, not even screenshots</li>
</ul>
</li>
</ul>
</blockquote>
</div>

]]></description>
</item>

<item>
  <title>SciPy 2006 Day 1</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/scipy2006-08-17-18-00.html</link>
  <description><![CDATA[

<div class="document">
<blockquote>
<ul>
<li><p class="first">They'll video tape</p>
</li>
<li><p class="first">Cancelations, so some one could move to tomorrow</p>
</li>
<li><p class="first">Python3, Guido's Talk</p>
<ul>
<li><p class="first">Python 2.5, some improvements</p>
</li>
<li><p class="first">Python 3000</p>
<ul class="simple">
<li>Language update</li>
<li>Stroustroup compatibility<ul>
<li>C++ backwords, add a new keyword, was a major incompatibility</li>
</ul>
</li>
<li>Perl 6 other extreme</li>
</ul>
</li>
<li><p class="first">-Q warn about integer errors, // &lt;- integer division</p>
</li>
<li><p class="first">Pep 3100 for info, python.org/dev/peps</p>
</li>
<li><p class="first">guidos blog, artima.com/weblogs/</p>
</li>
<li><p class="first">big changes</p>
<ul class="simple">
<li>kill classic classes</li>
<li>Exception derive from single root</li>
<li>absolute import ( can be turned on in 2.5 )</li>
<li>replacement for d.has_key if x in d</li>
<li>interating over file is better than xreadlines</li>
</ul>
</li>
<li><p class="first">smaller changes</p>
<ul>
<li><p class="first">exec, function again</p>
</li>
<li><p class="first"><cite>x</cite> in favor repr(x)</p>
</li>
<li><p class="first">except x,y,z as err instead except e,err</p>
</li>
<li><p class="first">generator replaces list comprehension
* list comprensions currently leak their name into the surrounding</p>
<div class="system-message">
<p class="system-message-title">System Message: ERROR/3 (<tt class="docutils">&lt;string&gt;</tt>, line 38)</p>
<p>Unexpected indentation.</p>
</div>
<blockquote>
<p>scope</p>
</blockquote>
</li>
<li><p class="first">kill raise E, use raise E(arg) (work away from string exception)</p>
</li>
<li><p class="first">xrange has a problem with long ints</p>
<ul class="simple">
<li>will change range to do xrange things with lon support</li>
</ul>
</li>
</ul>
</li>
<li><p class="first">string type updates</p>
<ul class="simple">
<li>string becomes unicode</li>
<li>also introduce a byte type, all conversions from byte to strings
require specifying conversion</li>
</ul>
</li>
<li><p class="first">new standard i/o stack</p>
<ul class="simple">
<li>replace the c stdio with something a bit more inspired by java
* unbuffered io
* buffered io built on top</li>
<li>see sandbox/sio/sio.py for early prototype</li>
</ul>
</li>
<li><p class="first">print becomes a function</p>
<ul class="simple">
<li>easier to change print to something</li>
<li>2.x/3 compatibility requires only printing single string</li>
</ul>
</li>
<li><p class="first">dict views instead of lists</p>
</li>
<li><p class="first">drop default &lt;, &lt;=, &gt;, &gt;= for classes (currently does memory)</p>
</li>
<li><p class="first">Google python sprint next week</p>
</li>
</ul>
</li>
<li><p class="first">Understanding NumPy, Travis Oliphant</p>
<ul class="simple">
<li>numeric compatibility layer, numpy.oldnumeric</li>
<li>dtype object ( defines what the elements of the NumPy array )<ul>
<li>array scalar ( wraps  element returned from cell )</li>
<li>dtype code '&lt;i4' (byte order, type, size)</li>
<li>isbuiltin, 21 static objects</li>
<li>dt.itemsize</li>
<li>fields, since array is just chunks of memory,
because of alignment issues</li>
<li>a.flags ( properties on arrays )</li>
<li>a.empty((10,),dtype='int,float,a10')
* a.dtype
* a.dtype.fields</li>
<li>arbitrary structure needed to memmap complicated files</li>
</ul>
</li>
</ul>
</li>
<li><p class="first">Extending NumPy, swig, f2py, pyrex, ctypes,</p>
<ul class="simple">
<li>swig, first</li>
<li>weave, embed C code in python string, kinda fortran like in variable names<ul>
<li>compiles on the fly</li>
</ul>
</li>
<li>pyrex<ul>
<li>from c_numpy cimport, ...
import_array()</li>
</ul>
</li>
<li>ctype, calls shared libraries<ul>
<li>.argtypes allow specifying type checking for c calls</li>
</ul>
</li>
</ul>
</li>
<li><p class="first">Python for modern scientific algorithm development, Fernando Perez</p>
</li>
<li><p class="first">Building a Distributed Component Framework (pyre), Michael Aivazis</p>
<ul class="simple">
<li>distributing apps on many machines is hard</li>
<li>user space, ssh, scp, pyre services</li>
<li>pyGridWare (Kieth jackson)</li>
<li>gsl</li>
<li><a class="reference" href="http://www.cacr.caltech.edu/projects/pyre">http://www.cacr.caltech.edu/projects/pyre</a></li>
<li><a class="reference" href="http://pyre.caltech.edu">http://pyre.caltech.edu</a> (soon)</li>
<li>relase early 2007</li>
</ul>
</li>
<li><p class="first">Enthought tool suite</p>
<ul class="simple">
<li>What are traits?<ul>
<li>notification (considered important)</li>
<li>visualization</li>
</ul>
</li>
</ul>
</li>
<li><p class="first">Synthetic Programming with Python</p>
<ul class="simple">
<li>provide low level access to cpu without doing string processing</li>
<li><a class="reference" href="http://www.synthetic-programming.org">http://www.synthetic-programming.org</a></li>
</ul>
</li>
<li><p class="first">3D visualization with TVTK and MayaVi2, Prabhu Ramachandran</p>
<ul class="simple">
<li>Traited VTK</li>
<li>MayaVi2 uses enthought exposes lots of objects and gui tools for
playing with them</li>
</ul>
</li>
<li><p class="first">Realtime Computing with Python, Andrew Straw</p>
<ul class="simple">
<li>introduction to relatime simulation of fly</li>
<li><a class="reference" href="http://dickinson.caltech.edu/Research/Grand_Unified_Fly">http://dickinson.caltech.edu/Research/Grand_Unified_Fly</a></li>
<li>comedi control measurement device interface</li>
</ul>
</li>
<li><p class="first">Lightening Talks</p>
<ul class="simple">
<li>Dude from JPL, Mike Ressler</li>
<li>Future of IPython, Brian Granger, Tech-X corp<ul>
<li>planning on extending to ipython to do distributed programming</li>
<li>control remote supercomputers from laptop</li>
<li>support multiple parallel types</li>
<li>focus on interactive work</li>
<li>distributed programming in python<ul>
<li>multiple processes, threads are limited by GIL</li>
<li>non-blocking sockets</li>
<li>asynchronous error/fault</li>
</ul>
</li>
</ul>
</li>
<li>Travis Oliphant, about PEP<ul>
<li>using trac pages for NumPy
<a class="reference" href="http://projects.scipy.org/scipy/numpy">http://projects.scipy.org/scipy/numpy</a></li>
<li>table &#64; dinner about arrays</li>
</ul>
</li>
<li>Travis Vaught, Enstaller<ul>
<li>egg package installation manger</li>
<li>has post-install scripts</li>
<li>build egg from pointing to a directory</li>
<li>code.enthought.com/enthon/enstaller.shtml</li>
<li>enthought.com/enthought/wiki/Enstaller</li>
</ul>
</li>
<li>The Current State of Vision, Michel Sanner<ul>
<li>Visual programming tools</li>
<li>supports nested graph nodes</li>
<li>looks written in tk</li>
<li>www.scripts.edu/~sanner/software</li>
</ul>
</li>
<li>Quick overview of Chaco, Peter Wang<ul>
<li>chaco visualization tool</li>
<li><a class="reference" href="http://code.enthought.com/chacho">http://code.enthought.com/chacho</a></li>
</ul>
</li>
</ul>
</li>
<li><p class="first">SAGE, Alex Clemesha</p>
<ul class="simple">
<li>good environment for algebra, number theory, geometry</li>
<li>what is it<ul>
<li>Distribution of math software</li>
<li>python library</li>
<li>new interface to math software</li>
</ul>
</li>
<li>where is it: modular.math.washington.edu/sage</li>
<li>their ajax based web notebook looks really impressive</li>
</ul>
</li>
<li><p class="first">Mathematica like plotting for SAGE, Alex Clemesha</p>
</li>
<li><p class="first">BioHub, Diane Trout</p>
</li>
<li><p class="first">Software Carpentry, Greg Wilson</p>
<ul class="simple">
<li>he's a good speaker.</li>
<li>debugger for make</li>
<li>debugger than can trace back and forth between python &amp; c code</li>
<li>Guido pronounced django the web framework.</li>
</ul>
</li>
</ul>
</blockquote>
</div>

]]></description>
</item>

<item>
  <title>Dialout via bluetooth</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/p900-dialout-2006-02-18-17-12.html</link>
  <description><![CDATA[

<div class="document">
<dl class="docutils">
<dt>sometime ago my laptop and phone were paired, but at some point I</dt>
<dd><p class="first">deleted one of the connections, and when I tried dialing out the phone
wanted a passphrase and kde thought there already was one.</p>
<p>I tried to delete /etc/bluetooth/link_key but that didn't work.
Eventually I changed the pin-helper in hcid.conf from the kde one back
to the bluez one.</p>
<p>I they started a connection and rand the pin-helper manually. One copy
of the helper started up, and then magically another one popped up.
I then filled a pin into both the second pin-helper and my phone and was
able to make an outbound connection.</p>
<p>Of course then my ppp settings were wrong.</p>
<p>I commented out any fixed address, and the dns server information, and
then used the /etc/chatscripts/provider which was setup to dialout
to <a href="#id1" name="id2"><span class="problematic" id="id2">*</span></a>99***2# (the second dialup profile on my phone)</p>
<div class="system-message" id="id1">
<p class="system-message-title">System Message: <a name="id1">WARNING/2</a> (<tt class="docutils">&lt;string&gt;</tt>, line 16); <em><a href="#id2">backlink</a></em></p>
Inline emphasis start-string without end-string.</div>
<p>I also added</p>
<dl class="docutils">
<dt>iface ppp0 inet ppp</dt>
<dd>provider p900-blue</dd>
</dl>
<p>to my /etc/network/interfaces but i don't think that was important in this
case.</p>
<p>And now, once again I can connect to tmobile via my phone</p>
<p class="last">I need to make sure i keep my phone better charged.</p>
</dd>
</dl>
</div>

]]></description>
</item>

<item>
  <title>Building GHC6 on Arm</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/ghc-arm-2005-12-25-20-00.html</link>
  <description><![CDATA[

<div class="document">
<dl class="docutils">
<dt>I got talked into darcs and now want to be able to check out my patches on</dt>
<dd><p class="first">my <a class="reference" href="http://www.nslu2-linux.org/">nslu2</a>. Unfortunately darcs requires ghc,
and ghc requires itself to compile. This leads us to the joy of making a new
port.</p>
<p><a class="reference" href="http://www.haskell.org/ghc/docs/latest/html/building/sec-porting-ghc.html">http://www.haskell.org/ghc/docs/latest/html/building/sec-porting-ghc.html</a></p>
<p>I started with the above link and tried to get it to work. So far,
the files they tell you to copy includes/(ghcautoconf.h DeriviedConstants.h
GHCConstants.h) are not created by the configure step so you need to
run make boot in the ghc directory on the target.</p>
<p>Then after copying them over make them read only, as my touches weren't enough
to prevent their makefiles from trying to rebuild those files.</p>
<p class="last">i managed to get it to start building, but it died trying to make the compiler
with a bunch of cast errors.</p>
</dd>
</dl>
</div>

]]></description>
</item>

<item>
  <title>Asterisk Frustration</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/asterisk-2005-12-22-18-00.html</link>
  <description><![CDATA[

<div class="document">
<dl class="docutils">
<dt>I've been trying to setup a VoIP system at home ('cause) and can call</dt>
<dd><p class="first">my cordless phone (plugged into a SPA-3000) from a IP Phone (grandstream
gxp-2000)</p>
<p>That's all good and works fine. I then decided to go further and plug
the SPA-3000's FXO(?) port into my vonage ATA. I can get it to make
outgoing calls, and sometimes even incoming calls.</p>
<p>I can't get authentication to work for outgoing, and I can't get caller id
to work for incoming.</p>
<p>This <a class="reference" href="http://voxilla.com/forum-viewtopic-t-1335.html">thread</a> was rather
helpfull on getting the calls working.</p>
<p>..:</p>
<blockquote>
[spa_outbound]
type=peer
secret=&lt;SECRET&gt;
port=5061
host=dynamic
canreinvite=no
nat=no
qualify=yes
dtmfmod=rfc2833
context=incoming-vonage</blockquote>
<p>is my current outgoing config.</p>
<p>On the SPA I have the PSTN line registering seperately from the ATA line.</p>
<p>If I want to get things to autoforward to the server I need to set
PSTN Ring Thru Line 1 to off and to set a dial plan to
SO&lt; s:&#64;server.host:5060 &gt; (And of course tell the PSTN which dial plan to
use.)</p>
<p>If I set the CID on, asterisk 1.2.1 gets very confused about the source SIP
address and chokes. I tried setting things to insecure=very but to no avail.</p>
<p>I'm thinking I might try openser, as its more sip orientated. (Though there
documentation is quite sparse).</p>
<p class="last">I found some documentation on the <a class="reference" href="http://www.netvoice.ca/sipura/SipuraSPAAdminGuidev2.0.9.pdf">SPA-3000</a></p>
</dd>
</dl>
</div>

]]></description>
</item>

<item>
  <title>Python &amp; Boost</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/programming/python-boost-2005-11-18-23-45.html</link>
  <description><![CDATA[

<div class="document">
<blockquote>
<p>I've wanted to try to use python &amp; boost for a while, but their tutorial
depends on using bjam, which has been a pain to try and figure out. And
of course the boost-python developers think that jam is the best thing
ever and don't bother providing any documentation of how to use boost-python
in a more traditional environment.</p>
<p>Here's my stab.</p>
<pre class="literal-block">
/* hello.cxx */
#include &lt;boost/python.hpp&gt;
using namespace boost::python;

char const* greet()
{
  return &quot;hello, world&quot;;
}

BOOST_PYTHON_MODULE(hello)
{
  def(&quot;greet&quot;, greet);
}
</pre>
<p>I compiled it on an amd64 linux box with the following.</p>
<pre class="literal-block">
$ g++ -c -fpic -I/usr/include/python2.3 hello.cxx -o hello.o
$ g++ -shared -fpic -lpython2.3 -lboost_python hello.o -o hello.so
$ python
&gt;&gt;&gt; import hello
&gt;&gt;&gt; print hello.greet()
</pre>
<p>It turns out the -fpic is required for the AMD64 box, and on an x86 box,
is unnecessary. (Though it doesn't hurt to have leave it in).</p>
<p>Would it have been too hard to show the manual way before extolling
the virtues of their build system? Perhaps they were just being lazy
as coming up with a simple build script on windows would probably be
a lot harder than the above.</p>
</blockquote>
</div>

]]></description>
</item>

<item>
  <title>Pidentd</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/debian/netlink_lookup-2005-11-12-21-22.html</link>
  <description><![CDATA[

<div class="document">
<blockquote>
<p>I was looking at the log files and discovered that pidentd was throwing
error messages about netlink_lookup: write failed: Connection refused</p>
<p><a class="reference" href="http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=305926">http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=305926</a></p>
<p>Seems to suggest that adding the module should make it shut up.</p>
<p>So I added the following to /etc/modules
# tcp_diag may shut pidentd up about netlink_lookup: write failed
tcp_diag</p>
</blockquote>
</div>

]]></description>
</item>

<item>
  <title>Try to figure out power usage</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/power-2005-09-28-13-35.html</link>
  <description><![CDATA[

<div class="document">
<blockquote>
idle measurements
base 20 watts
Via C3 no cards, 44
Amd64 pci-x, 78</blockquote>
</div>

]]></description>
</item>

<item>
  <title>Network</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/network-2005-08-12-01-19.html</link>
  <description><![CDATA[

<div class="document">
<blockquote>
<p>I tried getting chilispot working, but it needed to talk to my webserver/radius server which is the main thing i'm trying to protect. Since I also didn't feel like getting radius properly setup, I just punted and went for the simpler WPA-PSK configuration.</p>
<p>However before I went simple I split the wired and wireless networks into two IP address ranges, and I renumbered my IP address space out of 192.168.1</p>
<p>I followed
wiki.openwrt.org/HotspotOpenvpnHowto
to split the wifi (many nvram commands)
Later I discovered that some of my problems getting chillispot working were actually caused by routing issues so I needed to add
iptables -A FORWARD -i eth1 -o vlan1 -j ACCEPT # wifi to outside
iptables -A FORWARD -i eth1 -o vlan0 -j ACCEPT # wifi to lan
iptables -A FORWARD -i vlan0 -o eth1 -j ACCEPT # lan to wifi</p>
<p>I also added slightly more generalized commands to the firewall.user</p>
<p>Eventually I decided I didn't actually need the complexity of chillispot yet, and just started down the WPA-PSK path</p>
<p><a class="reference" href="http://openwrt.org/OpenWrtDocs/nas">http://openwrt.org/OpenWrtDocs/nas</a>
Gave some information on setting up the propritary wpa config tool.</p>
<p>I also installed wpasupplicant on my linux laptop.
Theoretically it'd be better if I actually got the WPA2/cert mode working, but eh, i'm not ready for that yet.</p>
<p>To get wpa working on the WRT54G I needed to add</p>
<p>nvram set wl0_akm=psk
nvram set wl0_crypto=aes+tkip
nvram_wpa_psk=&lt;password&gt;</p>
<p>I needed both wl0_akm and wl0_crypto for openwrts nas startup script.</p>
<p>Next I need to move the mini out to the living room, and then set up a some kind of client box in my office. Do I try for getting LDAP authentication working so I can unify the passwords between my mac, my windows partition and all my linux boxen?</p>
</blockquote>
</div>

]]></description>
</item>

<item>
  <title>* Repository madness</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/tailor-2005-08-09-18-00.html</link>
  <description><![CDATA[

<div class="document">
<blockquote>
<p>Tailor <a class="reference" href="http://nautilus.homeip.net/~lele/projects/tailor/">http://nautilus.homeip.net/~lele/projects/tailor/</a> allows syncing between multiple source code control systems, and since there's several projects in unpleasant sccs's i thought I'd try converting them to darcs</p>
<ul class="simple">
<li>orange</li>
</ul>
<p>orange certainly looks promising other than the fact they have bugs, don't have doc strings, and lack unit tests. though their tutorials were pretty straightforward.</p>
<p>Unfortunately they're a bunch of OSS windows developers, so their code is sitting in CVS (blech) so I wanted to convert out of it</p>
<p>My first attempts with earlier versions of tailor didn't work so I upgraded to the latest version, and it had problems as well.</p>
<p>What I ended up doing was checking the HEAD out of cvs first, and then telling tailor to convert it to a darcs repo, I don't know if this'll work when they commit something, but it's worth trying.</p>
<p>I used the following to start creating the config file.</p>
<p>$ tailor --verbose -s cvs -R :pserver:cvs&#64;estelle.fri.uni-lj.si:/CVS --module orange -r INITIAL --subdir orange orange &gt; orange.tailor</p>
<p>the file ended up like this</p>
<p>[DEFAULT]
verbose = True</p>
<p>[py]
target = darcs:targetpy
start-revision = HEAD
root-directory = /home/diane/src/orange
state-file = orange.state
source = cvs:sourcepy
subdir = orange.cvs</p>
<p>[darcs:targetpy]</p>
<p>[cvs:sourcepy]
module = orange
repository = :pserver:cvs&#64;estelle.fri.uni-lj.si:/CVS</p>
<p>[csrc]
target = darcs:targetc
start-revision = HEAD
root-directory = /home/diane/src/orange/orange.cvs
state-file = orange.state
source = cvs:sourcec
subdir = source</p>
<p>[darcs:targetc]</p>
<p>[cvs:sourcec]
module = source
repository = :pserver:cvs&#64;estelle.fri.uni-lj.si:/CVS</p>
<ul class="simple">
<li>compclust</li>
</ul>
<p>both chris and I want to try using darcs with compclust but that'll require 2way syncing.</p>
<p><a class="reference" href="http://www.darcs.net/pipermail/darcs-users/2005-January/005070.html">http://www.darcs.net/pipermail/darcs-users/2005-January/005070.html</a></p>
</blockquote>
</div>

]]></description>
</item>

<item>
  <title>There are many different systems for running lightweight clients...</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/diskless-2005-08-09-02-04.html</link>
  <description><![CDATA[

<div class="document">
<p>debian currently has diskless and lessdisks.</p>
<blockquote>
<p>Diskless appears to be an NFS root system where apps run on the client
Lessdisks is a terminal based system that runs software on the server.</p>
<p>First off we want a version of grub that has has pxegrub
apt-get source grub</p>
<p>I then edited the debian/rules file and added the following
--enable-diskless (several net cards)</p>
<p>apt-get install lessdisks</p>
<p>I mostly followed along the lessdisks-doc/install</p>
<p>On the router I added the following to dnsmasq.conf
(I really should back that bugger up)</p>
<p># configure network booting
#dhcp-vendorclass=pxe,PXEClient
dhcp-boot=net:/tftpboot/pxegrub,dara,192.168.1.13</p>
<p>dhcp-option=17,/var/lib/lessdisks  #root-path
dhcp-option=42,192.168.1.1         #ntp-server
dhcp-option=48,192.168.1.13        #font-server
dhcp-option=49,192.168.1.13        #x-display-manager
dhcp-option=66,dara.ghic.org       # tftp-server-name
dhcp-option=pxe,67,pxegrub         #bootfile-name
dhcp-option=69,192.168.1.13        #smtp-server</p>
<p>sigh the pxe client wasn't smart enough to boot off of a different computer than what we started with. So i tried with dara being the dhcp server, i got further but then ran into problems with not being able to load pxegrub from things other than /tftpboot</p>
<p>pxegrub had problems, when i switched to pxelinux.0, installed the default file into pxelinux.cfg (and then copied all of /var/lib/lessdisks/boot into /tftpboot I was able to boot. Unfortunately the kernel they installed doesn't support the ethernet card on my amd64 box.</p>
<p>What do i have to do to get a tftp server to server files off of something other than /tftpboot (or is just the pxeclient that has trouble with it?)</p>
</blockquote>
</div>

]]></description>
</item>

<item>
  <title>* Jury Duty</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/2005-08-04-12-02.html</link>
  <description><![CDATA[

<div class="document">
<dl class="docutils">
<dt>Whee</dt>
<dd><p class="first">Spent some time browsing through the linux howtos
<a class="reference" href="http://fiona/cgi-bin/dwww?type=file&amp;location=/usr/share/doc/HOWTO/en-html/LDAP-HOWTO/sasl.html">http://fiona/cgi-bin/dwww?type=file&amp;location=/usr/share/doc/HOWTO/en-html/LDAP-HOWTO/sasl.html</a></p>
<p>it looks like sasl ships with a number of different auth types with the kerberos stuff split out into sperate modules. The above link documents some information about getting MD5-Digest working with ldap, which should be useful for setting up an ldap auth server.</p>
<p>I forgot to update my xlibs package when upgrading to Xorg so its missing the keymaps necessary to change keyboard settings :(  (no caps lock to control mapping)</p>
<p>Another question is should my home net have the wireless VPNed? and if so which protocol? IPSec or OpenVPN</p>
<p><a class="reference" href="http://openwrt.org/HotspotOpenvpnHowto?highlight=%28Howto%29">http://openwrt.org/HotspotOpenvpnHowto?highlight=%28Howto%29</a>
Talks about setting up openvpn on the wrt54</p>
<p>One interesting thing that I did do was <a class="reference" href="http://thinkwiki.org/wiki/How_to_make_use_of_Graphics_Chips_Power_Management_features">http://thinkwiki.org/wiki/How_to_make_use_of_Graphics_Chips_Power_Management_features</a>
Enable power management of the graphics chips.</p>
<p>Hopefully I the library has free wifi</p>
<dl class="last docutils">
<dt>Home net: what do i want</dt>
<dd>priority routing for my traffic
control / monitering of guest access
securing wireless traffic</dd>
</dl>
</dd>
</dl>
</div>

]]></description>
</item>

<item>
  <title>Extensions Work</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/voip/extensions_work-2005-06-03-11-14.html</link>
  <description><![CDATA[

<div class="document">
<p>So I was able to get extensions working in both asterisk and ser.</p>
<blockquote>
<p>Asterisk is designed as a PBX so all calls are routed through it, SER is designed like the internet and the router is just a location service. Once a SIP call is initiated through SER it's just the two clients talking to each other.</p>
<p>adding a user in asterisk at the simplest involves adding a sip user for authentication and then binding that user identity to an extension.</p>
<p>In sip the user identity is the extension, however if you want to make it easy for phones to dial the user id you can create aliases using the serctl command.</p>
<p>One major difficulty I had with getting ser working was the registration kept failing. It turns out that its really important for SER to know what all the hostnames its using are so it can know if a request is actualy being sent to itself.</p>
<p>Because my dns server reverse name lookup was broken it wasn't able to determine its FQDN. (there's a reason why you should always list names as &quot;host.domain.name.&quot; in the reverse name tables, because if you don't know what you're doing its easy for the default domain to be wrong.</p>
<p>Once I got that fixed SER worked correctly</p>
<p>call transfer worked right too...</p>
</blockquote>
</div>

]]></description>
</item>

<item>
  <title>SoCal Piggies</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/socalpiggies-2005-04-19-21-36.html</link>
  <description><![CDATA[

<div class="document">
<ul>
<li><dl class="first docutils">
<dt>Meta classses for fun and profit</dt>
<dd><p class="first">www.highenergymagic.org/presentations/metaclasses/</p>
<ul class="last simple">
<li>are there issues with using __new__ in metaclass?</li>
<li>__init__ is much more limited than __new__?</li>
<li>do remember to call base class options.</li>
</ul>
</dd>
</dl>
</li>
</ul>
<blockquote>
<ul>
<li><p class="first">Quixote 2.0 tutorial</p>
<p>s = StringForm(&quot;foo&quot;,</p>
<dl class="docutils">
<dt>if request.form:</dt>
<dd><p class="first last">value = s.parse(request)
retuern &quot;value&quot;, value</p>
</dd>
<dt>else:</dt>
<dd><p class="first last">return &quot;&lt;form&gt;%s&lt;/form&gt;&quot; % (s.render()</p>
</dd>
</dl>
</li>
</ul>
</blockquote>
</div>

]]></description>
</item>

<item>
  <title>exim</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/exim4-2005-02-16-10-16.html</link>
  <description><![CDATA[

<div class="document">
<dl class="docutils">
<dt># keywords: exim mail debian</dt>
<dd><p class="first">Debian tried changing exim's group and user from mail:adm to Debian-exim:Debian-exim however my old config isn't setting that properly</p>
<p class="last">To get things working temporarily, I changed /etc/logrotate.d/exim4, the group in /var/log/exim4/* and the keys *.crt *.key in /etc/exim4</p>
</dd>
</dl>
</div>

]]></description>
</item>

<item>
  <title>AMD 64 install</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/amd64-2004-11-27-16-00.html</link>
  <description><![CDATA[

<div class="document">
<dl class="docutils">
<dt>Howto i'm using</dt>
<dd><p class="first"><a class="reference" href="https://alioth.debian.org/docman/view.php/30192/21/debian-amd64-howto.html#id258638">https://alioth.debian.org/docman/view.php/30192/21/debian-amd64-howto.html#id258638</a></p>
<p>I downloaded the netinstall iso, killed my old root partition</p>
<p>I needed to grab nvidia-kernel-common from an x86-32 box that has
access to the all architecture and built that package</p>
<p>i installed using the &quot;desktop system&quot; prompt from the installer, and
then have installed emacs21, vim, and kde (and everything those things
depended on)</p>
<p>installed x-window-system at completion Xprt had a segfault</p>
<p>I built my own nvidia driver, but it wanted to be in 2.6.9 instead of
debians 2.6.9-9-amd64, so I symlinked /lib/modules/2.6.9/nvidia to
/lib/modules/2.6.9-9-amd64/nvidia</p>
<p>installed nvidia-glx nvidia-glx-dev</p>
<p>the nvidia driver cant find my card, first &quot;nv&quot; is not the nvidia
driver, &quot;nvidia&quot; is the nvidia closed soruce drivers</p>
<p>hmm it's also important to make sure that the version of gcc used to
compile the kernel matches that of the version used to build the
module. (I used gcc-3.3 by default and the debian kernel was built
with gcc-3.4)</p>
<p class="last">I went with the ia32 chroot system for running ia32 binaries as
hopefully i can still use apt to manage the packages</p>
</dd>
</dl>
</div>

]]></description>
</item>

<item>
  <title>Installing fglrx drivers</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/t41/fglrx-2004-09-01-12-59.html</link>
  <description><![CDATA[

<ul>
  <li>followed <a href="http://xoomer.virgilio.it/flavio.stanchina/debian/fglrx-installer.html">build instructions</a></li>
  <li>removing xlibmesa-gl1-dri-trunk xserver-xfree86-dri-trunk</li>
</ul>

]]></description>
</item>

<item>
  <title>Trying to get syncml working on opengroupware</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/opengroupware-2004-08-15-00-00.html</link>
  <description><![CDATA[

<p>To get this working you need to source
<code>/usr/lib/opengroupware.org/System/Library/Makefiles/GNUstep.sh</code>
or <code>/usr/lib/skyrix/system/Makefiles/GNUstep.sh</code>. The
former installs into the right locations the latter provides correct
locations for the SaxObjc library.
</p>
<p>
I tried opengroupware's debian packages but couldn't build syncml. I
looked at the packages in debian experimental but they seem old. So
I'm building opengroupware from <a href="http://www.opengroupware.org/en/devs/build/index.html">cvs</a>.
</p>
<p>
After that building syncml went pretty easy. Next step getting the
server to run. I think I should try and get korganizer to talk to
opengroupware, I can at least use it then. Hmm... might want to
restore palm and get that working to fill in my old calendar info.
</p>
<p>though I really do want sync to phone to work. Need to backup p900
first, then I can see how far they've gotten with syncml in
opengroupware. I'm uncertain since the example apps can only download
things.</p>

]]></description>
</item>

<item>
  <title>Mail Merge</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/mail-2004-08-09-18-00.html</link>
  <description><![CDATA[

<p>Time to see how offlineimap works.</p>
<ul>
  <li>chaos: Delete previous Maildir.bkup</li>
  <li>chaos: move current maildir out of the way</li>
  <li>chaos: make place to receive email</li>
  <li>fiona: rsync current mailspool to chaos</li>
  <li>fiona: move current mailspool out of the way</li>
  <li>fiona: try to use offlineimap to download spool</li>
  <li>fiona: it seems to work correctly and is reasonably quick to scan all the folders</li>
  <li>chaos: need to stop dumping all the mail into ~/Maildir</li>
  <li>also I need some way of syncing mail automatically, which requires saving the password. So I'd like to setup pam in such a way that there's a seperate lower quality password for downloading mail.</li>
  <li>I'm also curious as to why the ssh tunnel version had so many problems</li>
  <li>Next is installing an imap web client on chaosl</li>
</ul>

]]></description>
</item>

<item>
  <title>ACPI on the T41</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/t41/acpi-2004-07-04-21-06.html</link>
  <description><![CDATA[


<p>Apparently the DSDT configuration on the T41 is buggy. <a
href="http://www.digriz.org.uk/t40p-linux/#kernel-2.6">
http://www.digriz.org.uk/t40p-linux/#kernel-2.6</a> provided a patched
DSDT.aml file that fixed most of the issues with ACPI.</p>

<p>It requires a patch to load the replacement DSDT configuration
available at <a href="http://gaugusch.at/kernel.shtml">
http://gaugusch.at/kernel.shtml</a>
</p>

<p>Once I had this up and running I was able to turn the bluetooth
adapter on and off without crashing the machine. I was also able to
both suspend and hibernate the machine. Though the atheros wifi card
doesn't come back up after suspend.  </p>

<p>For hibernate I needed to unload & load psmouse to get the
trackpoint working again. </p>

<p>Since bluetooth was working I installed the KDE bluetooth
package. <a href="http://kde-bluetooth.sourceforge.net/">
http://kde-bluetooth.sourceforge.net/</a>. It looks promising, though
I couldn't get it to talk to my phone properly. I'm really tempted by
the it's ties into the notification service and being able to trigger
the screensaver and IM away messages by moving the phone out of
range.</p>


]]></description>
</item>

<item>
  <title>KDE</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/kde-2004-07-01-02-00.html</link>
  <description><![CDATA[

<p>After chris' comments about KDE being easy to configure and to script, I decided to try KDE 3.2. Af first it seemed miserably slow, several operations would start to load and then hang for tens of seconds. For instance clicking on the K menu would hang the panel for quite a while. I eventually traced it down to an attempt to contact the portmapper, stopping and restarting the port mapper fixed the problem.</p>

<p>Next I spent most of the evening trying to get kde working to my satisfaction, and now I think I have something that seems to be the bastardized combination of unix, windows XP, and OS X. I grabbed <a href="http://http://ktown.kde.cl/kooldock">kooldoc</a> which simulates the OS X dock, I also stuck the menu bar up at the top of the screen. (Though I currently have focus follows mouse on, so I really need to use keyboard shortcuts to do anything with the menu. However the window decorations look very XP like, and of course the keybindings are as emacs like as I could get them. (though the editor is quite far from emacs)</p>

<p>one thing that really pushed me toward kde was being able to put sftp:// into any url or file dialog box. </p>

]]></description>
</item>

<item>
  <title>semantic web stuff</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/sw-2004-06-28-17-09.html</link>
  <description><![CDATA[

<p>I'd like to be able to update blog entries with meta data like FOAF
entries and todo entries. The example <a
href="http://www.w3.org/2004/lambda/Sites/index.html"
>http://www.w3.org/2004/lambda/Sites/index.html</a> has some
suggestions of how to embed RDF metadata in html.</p>

<p>I'd also like to get some photo viewier working correctly (with rdf
enhancements of course)</p>

<p>last question is how to divvy up personal entries, work entries,
and private entries.</p>


]]></description>
</item>

<item>
  <title>Blogging technology</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/blog-tech-2004-06-16-13-26.html</link>
  <description><![CDATA[

<p>So I have this clever idea of trying to improve the performance of my blog 
I was thinking of using a makefile to precompute rst to html, but
perhaps I should just turn on pyblosxom's caching service</p>

]]></description>
</item>

<item>
  <title>Matplotlib</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/debian/matplotlib-2004-06-14-17-03.html</link>
  <description><![CDATA[

<p>I managed to feed matplotlib into svn-buildpackage.</p>

<p><code>svn-inject -o</code> loaded it into the archive</p>
<p><code>svn-update</code> worked though it took a try or two for it
to create all the directories that it wanted.</p>
<p><code>svn-buildpackage</code> works reasonably well.</p>

<p>After <code>svn-update</code> I fixed the changelog a little and
then was able to build the package.</p>

<p>lintian complaining about several things. </p>
<ul>
  <li>installing \*.pyc files</li>
  <li>copyright</li>
  <li>dependencies</li>
  <li>description</li>
</ul>

<p>However I managed to fix all of them</p>

<p>Next question should I store the tarball in svn and if so where?
</p>

]]></description>
</item>

<item>
  <title>System Cracked</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/sgvlug-2004-06-10-21-27.html</link>
  <description><![CDATA[

<h1>What to do </h1>
<ol>
  <li>Do nothing</li>
  <li>Recover use as quickly as possible</li>
  <li>determine what happened</li>
  <li>catch the perpetrators</li>
</ol>

<h1>Calif Civl Code 1798.82</h1>

<ul>
  <li>if cracked need to notify users if they've gained access to</li>
  <li>first last name and one of the following</li>
  <ul>
    <li>ssn</li>
    <li>drivers license number</li>
    <li>account number, credit card or debit card numbers</li>
  </ul>
  <li>statue doesn't apply if the data is encrypted?</li>
  <li>doesn't apply to passwords (another weakness as people reuse passwords</li>
</ul>

<h1>Suggestions of how to respond</h1>
<ol>
  <li>use trusted shell from other media</li>
  <li>grab system time to compare access times</li>
  <li>record whose online</li>
</ol>

<p>ssh login by root, perhaps it should keep a log of when I last logged
in so it could check to see there's been a login since I last logged
in.</p> 

<p>recommended timestamp ls commands <code>ls -aluR/</code> or <code>ls -alcR/</code></p>

<h1>some possible tools</h1>
<ul>
  <li>incident response tools for unix part 1/2</li>
  <li>fire.dmzs.com</li>
  <li>http://www.bankersonline.com/idtheft/so_computerintrusion.html</li>
  <li>how to break out of a chroot jail</li>
</ul>


<h1>notes about MS</h1>
<ul>
  <li>WinXP Service Pack 2 will ship with firewall on by default,
  which will break some apps</li>
  <li>wus / sus windows update tools</li>
  <li>Get rid of NTLM, windows 98 clients very week, e.g. samba</li>
</ul>

]]></description>
</item>

<item>
  <title>more charter</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/charter-2004-06-08-11-43.html</link>
  <description><![CDATA[

</p>trying to get tech support they're supposed to come on the 11th (friday) between 1 and 5 to diagnose my problem.</p>

<p>What about jessica? Also should go see about buying a PCI to Cardbus
adapter and a PCI extension widget from frys. Also I should bring home
the 802.11 antenna and see if I can scan for an open network.</pP

]]></description>
</item>

<item>
  <title>More charter woes</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/charter-2004-06-07-22-09.html</link>
  <description><![CDATA[

<p>Had network problems... doesn't look like anything is wrong perhaps
its the modem that is dead? I should call back tomorrow and try to
arrange a modem swap. The next time they can do it is Jun 11, so I'd
need to bug them to see if I can get anything better.</p>

]]></description>
</item>

<item>
  <title>Chater woes</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/charter-2004-06-06-13-10.html</link>
  <description><![CDATA[

</p>Called charter to complain about my net access dropping e.g. </p>
<code>--- 131.215.34.119 ping statistics ---
1193 packets transmitted, 648 received, +1 duplicates, 45% packet loss, time 1194062ms
rtt min/avg/max/mdev = 5.942/593.302/48742.770/4711.795 ms, pipe 49
</code>

<p>They couldn't connect to my cable modem and are generating a ticket
 or their network operations group. Supposed to call back in 4 hours
(around 5)</p>

<p>Ticket 189020</p>

]]></description>
</item>

<item>
  <title>Setup for the p900</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/p900-sync-2004-06-05-16-08.html</link>
  <description><![CDATA[

</p>I setup a DNS server on my laptop and added the configuration from:
<a href="http://www.alfonsomartone.itb.it/bauodg.html#syncml">http://www.alfonsomartone.itb.it/bauodg.html#syncml</a></p>

<p>I'm using this as my ppp/provider script</p>

<code>connect "/usr/sbin/chat -v -f /etc/chatscripts/p900.chat"<br/>
noauth<br/>
user ppp<br/>
crtscts<br/>
lock<br/>
local<br/>
proxyarp<br/>
passive<br/>
115200<br/>
/dev/ttyUSB0<br/>
169.254.1.65:169.254.1.66<br/>
lcp-echo-failure 10<br/>
lcp-echo-interval 86400<br/></code>

]]></description>
</item>

<item>
  <title>TMobile GPRS service</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/tmobile-2004-06-04-23-49.html</link>
  <description><![CDATA[

<p>Wow, that's some lame ping times.</p>

<p>Currently</p>

<p><code>/usr/sbin/pppd /dev/ttyUSB0 115200 crtscts noauth nodetach
defaultroute usepeerdns lcp-echo-failure 10 lcp-echo-interval 86400
connect '/usr/sbin/chat -t 6 -s -v "" ATZ OK ATDT*99***1# CONNECT
\d\c'</code> can bring up a net connection.</p>

<p>I started using wap.voicestream.com but that was filtered on most
conections. (http worked, not sure if imap would've worked or not. I
then looked up in my message spool and found internet3@voicestream.com
that's actually wrong, the proper gprs address for this connection is
"internet3.voicestream.com" which gets me to the full net access (and
is on a real IP address instead of a 10.x address. (Ah, so that's why
it's twice as expensive.)</p>

<p>the sad thing is the ping times are hovering around 1 second. </p>

<p><code>--- 131.215.34.119 ping statistics ---<br/>
259 packets transmitted, 194 received, 25% packet loss, time 258366ms<br/>
rtt min/avg/max/mdev = 800.180/1050.882/1810.805/187.989 ms, pipe 2</code></p>

]]></description>
</item>

<item>
  <title>bluetooth</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/t41/bluetooth-2004-05-29-18-29.html</link>
  <description><![CDATA[

<p>got bluetooth up and running to talk to the p900
didn't really have to do anything other than set the pin.</p>

<p>to xfer to the p900 I needed
<a href="http://www.unrooted.net/openSource/bluez/ussp-push-0.2.tgz">http://www.unrooted.net/openSource/bluez/ussp-push-0.2.tgz</a></p>

<p>which provides a tool obextool which can shove obex requests over bluetooth
<code./obextool push <file> 00:0e:07:c8:9b:68 2</code></p>

<p>Still don't know how to get p3nfs to work yet.</p>

]]></description>
</item>

<item>
  <title>svn homedir</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/svn-home-2004-05-26-18-00.html</link>
  <description><![CDATA[

<h1>ssh</h1>

<p>more efforts on moving stuff around in subversion. Needless to say
fiddling around with your .ssh directory when tunneling through .ssh
is challenging.</p>

<p>It looks like the best solution is to do svn mv svn+ssh://.../.ssh
svn+ssh:/..//etc/ssh ; svn update  and then move by hand the extra
files from .ssh to etc/ssh.</p>

<h1>fetchmail</h1>

<p>fetchmail is grumpy about symlinks to its file I needed to define an
alias to explicitly tell fetchmail where it's config file is located.</p>

<h1>todo</h1>

<p>still need to learn how to stitch things together from multiple
repositories.</p>

<h1>issues</h1>

<p>there was a problem with a lock being left in a directory, to resolve
I checked it out again and replace the .svn directory that was complaining</p>

]]></description>
</item>

<item>
  <title>Home dir in svn</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/svn-home-2004-05-26-02-00.html</link>
  <description><![CDATA[

<p>I've been making progress at loading my home directory into
subversion. So far I've got bin/, .ssh/, and part of src/ loaded. .ssh
showed me that I needed something like cfengine to manage permissions
and symlinks.</p>

<p>I've got the core of cfpython written. it's the first thing from src
that I loaded into svn. I've also written a small script
bin/cf-diane.py that loads the cfpython module and tries to set proper
permissions on .ssh</p>

<p>I do need to update cfpython to handle paths and globs. </p>

<p>Next up is to make an etc/ directory and start moving all of my
configuration files into it. Then update the cf-diane script to stitch
things together correctly with symlinks.</p>

<p>I think I should also have a private/ directory to store things that
shouldn't be replicated like ssh keys and private keys. (Or perhaps
only replicated as encrypted files)</p>

<p>At some point I might want to add a --read-only option to svn so it
checks files out and hard links the copy that's supposed to be
modified. </p>

]]></description>
</item>

<item>
  <title>More WiFi Notes</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/t41/wireless-2004-04-02-15-36.html</link>
  <description><![CDATA[

<p>On <a href="http://www.cs.utexas.edu/users/walter/geek/linux-t40.html#wifi">http://www.cs.utexas.edu/users/walter/geek/linux-t40.html#wifi</a> I found 
a comment that the card doesn't scan until after the interface is
brought up. e.g.</p>

<p><code>ifconfig ath0 up<br/>
iwlist ath0 scanning<br/>
dhclient ath0<br/></code></p>

<p>So how should I turn this into a
<code>/etc/network/interfaces</code> script?</p> 

<p>Also turning on the bluetooth and then turning it off crashed the machine. :(</p>

]]></description>
</item>

<item>
  <title>Installing debian on t41</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/t41/t41-2004-04-02-01-20.html</link>
  <description><![CDATA[

<p>I thought I blew away the windows partition, but later discovered that
the bios prevents you from overwriting the recovery section.</p>

<p>Perhaps I should switch hard disks.</p>

<p>The sarge installer worked fairly well. Though I then upgraded to
a custom 2.6.4 kernel so I could use the pentium M settings. </p>

<p>It nicely autodetected the e1000 hardware. </p>

<h1>power management</h1>

<p>After a suspend to ram cycle the USB controller gets
hosed. <a href="http://www.sowerbutts.com/thinkpad/">http://www.sowerbutts.com/thinkpad/</a> suggests that ACPI causes
this problem. I also saw kernel faults that were tagged "PREEMPT" perhaps
the preemptable kernel was also partially to blame.</p>

<p>I installed powernowd and set it to "SINE" mode which raises and
lowers the CPU frequency depending on use.</p>

<p>I installed acpid <a href="http://erkki.tty0.org/thinkpad.xhtml">http://erkki.tty0.org/thinkpad.xhtml</a> gave some
suggestions on getting the sleep button to work. Though this is
dependent on solving the USB problems.</p>

<p>The second suspend-to-disk option seemed to work slightly
better. (Though I wouldn't claim that it actually worked).</p>

<h1>networking</h1>

<p>As mentioned above the e1000 port worked perfectly. Though I'm not
sure if ifplugd worked or not.</p>

<p>I grabbed madwifi from <a
href="http://madwifi.sf.net">madwifi.sf.net</a> and discovered that
for thinkpads one should build with "export COPTS=-DSOFTLED" to get
the wireless LED indicator working.</p>

<p>make install seemed to get it installed, though there were problems
with the driver, I'm hoping the APIC-less kernel will fix that.</p

<p>bluetooth aparently is connected to the usb bus internally and Fn-F5
connects and disconnects this device. Though it causes the system to
crash when it's disabled</p>

<h1>sound</h1>

<p>alsa with the snd_intel8x0 chip seems to work fine. </p>

<p>the volume controls work regardless of the OS, and can preserve
settings across reboots. (that's why it seemed to stop working). </p>

]]></description>
</item>

<item>
  <title>S Stuff&lt;/h1&gt;

&lt;table&gt;
  &lt;tr&gt;&lt;td&gt;13:40ish&lt;/td&gt;&lt;td&gt;noticed my package was marked &quot;delivered&quot;&lt;/td&gt;&lt;/tr&gt;
  &lt;tr&gt;&lt;td&gt;13:48&lt;/td&gt;&lt;td&gt;Called UPS, Reported package delivered but no
  package at my apartment UPS told me that the package was left by
  the &quot;front door&quot;&lt;/td&gt;&lt;/tr&gt; 
  &lt;tr&gt;&lt;td&gt;14:10&lt;/td&gt;&lt;td&gt;Looked to see if there was a 336 N Madison&lt;/td&gt;&lt;/tr&gt;
  &lt;tr&gt;&lt;td&gt;14:44&lt;/td&gt;&lt;td&gt;Package delivered to &quot;Frt Door&quot;, claims to have
  been signed for&lt;/td&gt;&lt;/tr&gt;
  &lt;tr&gt;&lt;td&gt;18:00&lt;/td&gt;&lt;td&gt;went home in the evening and found package&lt;/td&gt;&lt;/tr&gt;
&lt;/table</title>
  <link>http://woldlab.caltech.edu/~diane/index.cgi/geek/t41/t41-delivery-2004-03-31-14-50.html</link>
  <description><![CDATA[
PS Stuff</h1>

<table>
  <tr><td>13:40ish</td><td>noticed my package was marked "delivered"</td></tr>
  <tr><td>13:48</td><td>Called UPS, Reported package delivered but no
  package at my apartment UPS told me that the package was left by
  the "front door"</td></tr> 
  <tr><td>14:10</td><td>Looked to see if there was a 336 N Madison</td></tr>
  <tr><td>14:44</td><td>Package delivered to "Frt Door", claims to have
  been signed for</td></tr>
  <tr><td>18:00</td><td>went home in the evening and found package</td></tr>
</table
]]></description>
</item>

</channel>
</rss>
