PattonCentral

Edit: The following is correct, but this article is better. Also, I was seeing the CIFS shutdown error discussed in [1]-[4]. There is a sophisticated workaround given by [3], but the quick fix in [2,4] seems to work just as well. However, as discussed in [1], both fixes might lead to data corruption problems. (I don’t know). So, I went with my own simple fix, which was simply to unmount the cifs share at shutdown. To do so, download the following script to /etc/init.d/, and symlink to it in /etc/rc0.d and /etc/rc6.d, giving it the prefix K00. The script just does a umount of your share. See [5] if you don’t know what you’re doing.

#! /bin/sh
### BEGIN INIT INFO
# Provides:          unmount_kristen
# Required-Start:
# Required-Stop:
# Should-Stop:
# Default-Start:
# Default-Stop:      0 6
# Short-Description: Just unmounts a samba/cfis share as a bug work around.
# Description:       Just unmounts a samba/cfis share as a bug work around.
### END INIT INFO

PATH=/sbin:/usr/sbin:/bin:/usr/bin
MOUNTPOINT=/path/to/share/mountpoint #e.g., /mnt/myshare

echo "["`date`"] ("`users` ") $1" >> /var/log/unmount_cifs.log

case "$1" in
  start)
  # No-op
  ;;
  restart|reload|force-reload)
  echo "Error: argument '$1' not supported" >&2
  exit 3
  ;;
  stop|"")
  umount $MOUNTPOINT
  ;;
  *)
  echo "Usage: umountnfs.sh [start|stop]" >&2
  exit 3
  ;;
esac

:

Also, I do not automatically mount the CIFS share vi /etc/fstab anymore. Instead, the user just clicks an icon on the Desktop if they want to connect. The Gnome shortcut file is below:

#!/usr/bin/env xdg-open

[Desktop Entry]
Encoding=UTF-8
Version=1.0
Type=Application
Icon[en_US]=gnome-panel-launcher
Terminal=false
Name[en_US]=Mount CIFS Share
Exec=mountshow.sh
Name=Mount CIFS Share
Icon=/usr/share/pixmaps/gnome-computer.png

The applet file (above) calls a script I wrote called mountshow.sh, which can be placed in /usr/bin/. It is pretty self-explanatory.

#!/bin/bash

share=//share_host/share
mountpoint=/path/to/mount/point
options='credentials=/etc/samba/user_name,gid=sambashare,file_mode=0775,dir_mode=0775,users'

# try to mount if not already mounted
if !(mount|grep -q "on ${mountpoint} type"); then
  gksudo "mount -t cifs -o $options $share $mountpoint" --description="mount"
fi

# error if it didn't mount
if !(mount|grep -q "on ${mountpoint} type"); then
  echo Unable to mount $share at $mountpoint using $options. >&2
  exit 1
fi

# open browser if mounted
nautilus --browser $mountpoint 2>/dev/null

[1] bugs.launchpad.net
[2] Ubuntu Forums
[3] Sander Marechal
[4] My Thoughts
[5] Tech Republic

It is easy to edit /etc/fstab so that Samba shares are automatically mounted at startup. However, remember that you need smbfs and smbclient installed. (smbfs is not installed by default under Ubuntu 9.10). Also, you can escape spaces in /etc/fstab using \040 (e.g., My\040Documents). Finally, the mount will be read-only by non-root users unless you specify otherwise. To do so, create a new group (e.g., sambashare) and use the dir_mode option. The /etc/fstab file line will look something like:

//servername/sharename /mountdirectory smbfs username=mywindowsusername,password=mywindowspassword,gid=sambagroup,dir_mode=0775

You can then doing a sudo mount -a to test.

Of course, there are more secure ways, etc. See this article, this article, or just google it.

There are some great web resources that explain the current financial crisis in very accessible and entertaining ways. First there is the video Crisis of Credit, which is very entertaining and easy to follow. Best of all, it is short — less than ten minutes. So, there is no excuse for not watching it! The video’s explanation does have one shortcoming. It doesn’t explain the important role the ratings agencies played in this mess. NOW did an excellent show called Credit and Credibility, which explains this point very well. This American Life did a great show called The Watchmen, and Planet Money has another good show called Secrets Of The Watchmen. Both shows discuss where the blame should be laid for the current credit crisis. This American Life also produced The Giant Pool of Money, which is a very entertaining and informative look at the link between the home owners and wall street. That show helps expose the role of mortgage brokers in the credit fiasco. Finally, what is the problem when banks stop lending? It has to do with our monetary system. On that topic, there is a very accessible short film called Money as Debt. However, I’ve read that the first two thirds of the film are accurate, whereas the last third may not be. Many more videos explaining how debt is created are listed here, but I haven’t yet watched them. Also, a good BBC Documentary here

Some notes on creating portable PDF beamer presentations with embedded multimedia (specifically, movies and GIFs):

The movie15 package looks promising for embedding movies. However, acroread complains about not having the proper plug-in, and the plug-in isn’t available under linux. There is another approach that can be used to embed any kind of media. This approach requires calling external applications, and the resulting documents are not very portable. However, it seems to be the best option right now.

The animate package is very useful for animating a sequence of images, and you can use the package to display GIFs by doing the following.

• To convert a *.gif into a sequence of *.jpg, simply use the convert (imagemagick) command. For example, convert img.gif img.jpg.
• You can get image dimensions using convert img.gif -verbose. Then, use the scale option of \animategraphics to preserve the aspect ratio of your images.
 

If you’re using the git-svn client for version control, what do you do if the location of the svn repository changes? If you were using the svn client, you could just use svn switch , and be done with it. Unfortunately, things are not so simple with git-svn. Read this article for an explanation. Fortunately for me, I am always the one moving the repository. As such, I simply do the following.

1. Complete work to branches and merge to master. (Note: the branches will not survive the remaining steps.)
2. Make sure the working copy is fully commited
3. Move the SVN repository
4. Checkout a new working copy (git clone)

Seriously, that’s all there is to it if you have complete control over the process and you don’t need to maintain any branches. If this is not the case for you, then read:

• http://www.sanityinc.com/articles/relocating-git-svn-repositories
• http://git.or.cz/gitwiki/GitSvnSwitch

A word of warning: I have not tried either of the approaches above, which are themselves a little dated. So, do yourself a favor and Google the problem first. Let me know if you have any more up-to-date advice.

Just a note to myself regarding the creation of DVD backups of my external USB hard drive.

1. Mount The Drive — Plug in the USB Maxtor drive and determine what device it is on using fdisk -l. Then, add the appropriate line to /etc/fstab (/dev/sda1 /media/usbdisk auto rw,user,noauto 0 0). Finally, mount the drive (sudo mount /media/usbdisk).
2. Burn the DVD — See the YoLinux Tutorial: Burning a CD or DVD, especially the bullet “Burning a data backup DVD using more mkisofs commands” in the section titled “Burn a DVD: (using growisofs)”. Note, growisfos failed when I used the -J switch. It worked when this switch was removed. Also, the error messages are not very clear. Make sure you have everything spelled correctly on the command line as growisofs will not tell you if it had a problem interpreting one of the options.

This handy script can be used to generate a BibTeX entry from a SpringerLink citation. You can even save it as a bookmarklet, which is a very handy feature.

Ref: http://www.cs.usyd.edu.au/~niu/cgi-bin/springer.cgi

I spent way too much time trying to install the bbding fonts for TeTeX under Ubuntu. Sadly, the tex-live packages did not work for me. So I had to install the fonts manually. Like everything under linux, this is easy to do if you know what you’re doing. I found a reference on the web [1], but I still had trouble. So, I’m putting my steps up here in hopes you find them useful. Note that these installation steps will make the font available to you only, as opposed to making them available for all users on the system. Also, I use pdflatex and kpdf. I’m not sure if these steps are complete if you use DVI instead of PDF.

1. Download the bbding zip file and unzip.
2. In the bbding directory, run latex bbding.ins and latex bbding.dtx.
3. Copy bbding.sty and Uding.fd to ~/texmf/tex/latex/bbding/. (You will need to make this directory.) Copy bbding10.mf to ~/texmf/fonts/source/, and copy bbding10.tfm to ~/texmf/fonts/tfm/. (Again, you will need to make these directories.)
4. In ~/texmf/, run the command texhash .

That should do it, you should now be able to add \usepackage{bbding} in your LaTeX document, and use the bbding fonts.

[1] http://brneurosci.org/linuxsetup30.html

The Sixth Sense from Pattie Maes’ research group is a wearable device with a projector that paves the way for profound interaction with our environment. This truly inspired engineering research. Check it out here and here.

The booktabs package can help you produce “Publication quality tables in LaTeX”. The xcolor package can be used to easily add color to your tables.

Before an election,

1. Visit your county’s board of elections website .
   1. Make sure you’re registered to vote in your county.
   2. Find your polling location .
   3. Read about the issues
   4. Find out about the candidates
2. Research the issues and the candidates.
   1. Your newspaper should have a guide to the issues.

Before submitting to IEEE, you must make sure your PDF does not require Type 3 fonts, and that all fonts are embedded. You can examine the fonts by viewing the Fonts tab under Document Properties in acroread. You can also use the pdffonts command.

I prefer to configure pdflatex to embed all fonts by default (even base 14 fonts, which are found almost everywhere.) This may make documents unecessarily large, but it makes for a streamlined process. See reference [1] and [2] below for how to embed fonts in PDF figures.

References:
[1] colinm.org
[2] daniel-lemire.com
[3] do.whileloop.org
[4] www.manticmoo.com

This is more of a note to myself than a useful tip:

I use grip to rip CDs straight to mp3. However, the mp3 tags are almost never as useful as I would like them to be. So, I use id3ed to edit them. Also see exfalso for editing m4a tags.

I use Amarok to manage my mp3 collection and podcast subscriptions. Recently, I wanted to move my collection to another computer. So, I installed Amarok 1.4.7 on the new computer, and then copied my mp3 (collection) folder and podcast folder (located at ~/.kde/share/apps/amarok/podcasts) from the old computer to the new one. After pointing Amarok to my new mp3 directory, I could manage my mp3s. However, Amarok would not recognize my podcast directory. I searched the internet, but could not find a way to tell Amarok about this directory. I searched (grepped) for “podcast” in ~/.kde/share/apps/amarok/, and discovered that podcast information is in collection.db. So, I copied the database to my new Amarok install. After a restart, Amarok correctly listed all of my podcast subscriptions.

Some time after an Ubuntu upgrade (I’m running 7.10 at the time of this post), I started receiving the followng error messages when accessing some of my Subversion (SVN) repositories.

svn: PROPFIND request failed on '/matlab_templating'
svn: Could not open the requested SVN filesystem

And I found this in my Apache error logs:

(20014)Internal error: Berkeley DB error for filesystem '/path/to/repos/matlab_templating

Now, I’m not sure at which version this occurred, but at some point, the default Subversion repository file system type became ‘fsfs’ instead of ‘bdb’. And, during some Ubuntu upgrade, I got a new subversion installation. So, I had to “upgrade” my repositories. Here’s how to do it:

1. Find out which repositories are ‘bdb’

   sudo find $PATH_TO_REPOSITORY_ROOT -type f -iname 'fs-type' -exec echo {} \; -exec cat {} \;
   

2. Recovere the BDB repositories.

   sudo svnadmin recover $REPO
   sudo for f in `svnadmin list-unused-dblogs $REPO`; do rm $f; done
   

3. Dump old BDB repository to new FSFS repository.

   mv $REPO $REPO-OLD
   svnadmin create $REPO
   svnadmin dump $REPO-OLD | svnadmin load $REPO
   

   Don’t forget to set the correct file permissions on the repository. Also, run these commands using sudo.

I’ve been getting a lot of emails lately from folks interested in getting started with GNU Radio. So, instead of continually forwarding the same email that I sent to the first student, I decided to post some starting points here.

Read up on GNU Radio to see if it interests you and/or meets your needs:

• A Good Overview on Wikipedia. (See the links therein.)
• Excellent tutorial from Notre Dame
• Visit the GNU Radio Homepage. (See links therein.)
• See an example of a C++ signal processing block
• You probably emailed me because you know I implemented a GNU Radio based radar for my master’s thesis.

If you still have questions:

• Search the mailing list archive. The mailing list serves as the defacto project documentation. So, take this advice seriously.
• I can’t emphasize this next point enough: Subscribe to the GNU Radio mailing list. Just lurk on the list for a while (i.e., read the messages without sending any of your own). You’ll get a good idea of the state of the project from doing this.

If you don’t know Python, you’ll need to learn it:

• Check out this Python tutorial.

If you don’t know Linux, you’ll need to learn it:

• Check out this Linux tutorial.
• If you haven’t already installed a Linux distro. I recommend Ubuntu for use with GNU Radio. Be sure to follow the distribution specific installation instructions on the GNU Radio website.

Good Luck!

Problem:

Intellisense (or auto-complete, or code-completion, whatever…) wasn’t working properly in NetBeans for a free-form project (i.e., a “Java Project with Existing Ant Script”), but it was working for stand-alone applications (i.e., a “Java Application”).

Description:

When typing something like “System.” (notice the dot) in a “Java Application,” a window will pop up with an appropriate list of methods and members for the class “System”. However, in the free-form project (i.e., “Java Project with Existing Ant Script”), this does not happen. Instead, it pops-up with “no suggestions” or a list of miscellaneous words for auto-completion.

Solution:

You must update the project classpath for the free-form project. You can find your classpath by examining the “Classes” tab under Tools->Java Platform Manager. My classpath is:

/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/lib/rt.jar
/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/lib/i18n.jar
/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/lib/sunrsasign.jar
/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/lib/jsse.jar
/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/lib/jce.jar
/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/lib/charsets.jar
/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/classes
/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/lib/ext/sunjce_provider.jar
/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/lib/ext/sunpkcs11.jar
/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/lib/ext/localedata.jar
/usr/lib/jvm/java-1.5.0-sun-1.5.0.11/jre/lib/ext/dnsns.jar

To update the classpath for the free-form project, right-click on the project icon in the projects view pane, and select “Properties”. NetBeans will not let you set the classpath unless you have first explicitly listed “Source Package Folders.” So, click on the “Java Sources” button in the options tree on the left-hand side. Then click the “Add Folder” button next to the “Source Package Folder” list box. Select your Java source directories. Next, click the “Java Source Classpath” button in the options tree (left-hand side again), and click the “Add JAR/Folder” button. Browse to the JAR files/folders that you saw listed in your Java Platform Manager, and select them. They should now be listed in the list box. Click OK.

Hopefully that does it.

There are some radio shows which broadcast for free (either over the air or streaming over the Internet), and I would really like to listen to them. However, I consume all of my audio during my commute, and for all of these shows, I am usually not available to listen. So, mp3 versions of these shows would be super handy.

After some searching, I found some resources that allow one to save an Internet audio stream in mp3 format. I thought I would share these resources here. I am not posting step-by-step instructions. (There are better resources for that.) I am just posting some notes to serve as a starting point:

HOWTO Convert (almost) any audio format (.m4a/.rm/.wma/.mp3) to .wav
How-To rip Real Media RTSP Streams from the web to MP3 using MPlayer

Also, here’s a script that I use to capture installations of The Infinite Mind radio show.

#!/bin/bash
#
# This script accepts the name url to a real-format stream. It then rips
# the (real formatted) audio stream, can converts it to mp3

STREAM=$1
BNAME=`basename $1 .rm`

# get the real formatted stream, saved in real format
mplayer -noframedrop -dumpfile $BNAME.rm -dumpstream $STREAM

# convert the real formatted file to wav format
mplayer -quiet -vo null -vc dummy  -ao pcm:waveheader:file="$BNAME.wav" "$BNAME.rm"

# convert the wav formatted file to mp3
lame -V0 -h -b 160 --vbr-new $BNAME.wav $BNAME.mp3

Note: To find the stream address for a live stream, I usually open the feed with realplayer (or whatever) and find the source URL by examining the stream properties.

I will be maintaining a list of links I find useful when conducting my research. I will add to and edit this list as I have time (and remember to do it).

Very Helpful Software

• Zotero — Make sure to view the demo
• JabRef — I couldn’t do my research without it! Read a good overview of bibtex management tools here.

Patents

• US Patent Office (USPTO)
• Patent Storm
• Google Patents
• Google with the /patent option (Not the same as Google Patents.)

Journals

• IEEE Xplore

  IEEE Xplore’s search is very picky. It will not return any results if you have even one typo in your query. Therefore, I highly recommend using the CrossRef tool that I discuss here. CrossRef is powered by Google. Therefore, it is much less picky about queries. It also searches Wiley, SpringerLink, IOP, ACM, and more. Fore a complete list of publishers that CrossRef indexes see this.

• Elsevier
• SPIE
• SPIE digital library
• Wiley Interscience
• ScienceDirect
• Engineering Villiage
• SpringerLink
• JSTOR
• ACM Portal
• DTIC
• Google Scholar
• Google Books
• IOP Electronic Journals
• Math.Sci.Net
• Scitopia.org
• OhioLink Ebooks

Sites that provide BibTeX output

• IEEE Xplore (via a handy bookmarklet)
• Google Scholar (Instructions here)
• The Smithsonian/NASA Astrophysics Data System
• ACM Portal
• Wiley InterScience (exports EndNote, use bibutils to convert to BibTex)
• Science Direct (exports RIS, use bibutils to convert to BibTex.)
• SpringerLink (export RIS, use bibutils to convert to BibTex. Or, use this handy script.)
• SpringerLink Direct Conversion
• JSTOR

WSU Library/ OhioLink specific

• If you cannot get access to an article, try searching for the journal at the WSU library homepage. The university may have access to full-text (electronic or non-electronic) via a university subscription or through OhioLink.
• You can configure Google Scholar to perform library / OhioLink queries based on your searches.

Checking Cross-References

• Google Scholar has a “Cited by…” link under each result that you can click.
• CiteSeer – never seems to work for me, but….