November 03, 2007

Boiling the Oceans

In the Wikipedia article on the ZFS file system the ZFS project leader Jeff Bonwick is quoted as saying,
"Populating 128-bit file systems would exceed the quantum limits of earth-based storage. You couldn't fill a 128-bit storage pool without boiling the oceans."

What's wonderful about this particular [apparent] hyperbole is that it is not hyperbole. Drawing on quantum mechanics and information theory, he explains—in detail—how storing enough information to overflow a single ZFS file system's capacity would require a storage system that was 136 billion kg of something, converted to pure energy (via E=mc2). To reach that level, you'd need the energy equivalent of converting the Earth's oceans entirely to steam.

Posted by ronlusk at 09:38 AM

January 06, 2005

DBCP and InvocationTargetException

This isn't properly a Linux problem, but in seeking the reason why a DBCP/JOCL test didn't work, and wanted to make another pointer to the solution.
I've been getting an InvocationTargetException while parsing my jocl file, but the error message never contains where the real problem occurs.

1) The commons-pool package uses the commons-collections, but does not document that requirement in any of its documentation.
Other reasons given explain why the error is misleading. But for the moment, make sure that the commons-collections JAR file is in the path as well.

I encountered this running junit from an ant build, trying to get around a LinkageError problem by selectively adding JAR files to the classpath for testing. Turns out my version of ant might have been hosed, too, in some way, but that's another issue.

Posted by ronlusk at 11:17 AM

November 24, 2004

Howto install Python 2.3.3 on SuSE 9.0 by RPMs

longsleep.org - Howto install Python 2.3.3 on SuSE 9.0 by RPMs
A short description howto install Python 2.3.3 from RPM on a vanilla SuSE Linux 9.0 installation.
Posted by ronlusk at 12:27 PM

November 17, 2004

InfoCentral Open Source Church Software

InfoCentral
InfoCentral is an Open Source project aimed at the fully meeting the database needs of churches, schools, and non-profit organizations. Our goal is to create a professional quality, highly-modular application that serves as an all-in-one solution for centralized information management.
Posted by ronlusk at 05:17 PM

October 19, 2004

In Unison, please

Unison File Synchronizer
Unison is a file-synchronization tool for Unix and Windows. It allows two replicas of a collection of files and directories to be stored on different hosts (or different disks on the same host), modified separately, and then brought up to date by propagating the changes in each replica to the other.

I've started using Unison for maintaining directories of lecture notes:

  • I use a small laptop as a notepad at class, using either OpenOffice.org or emacs to edit notes in HTML, or (more recently) using FreeMind to build visual "maps" of the lecture.
  • I do most of my work at home on my "big" laptop, so I use Unison to synchronize the little laptop with the big one. That also gives me class documents (downloaded from the class website) and notes on readings (also in FreeMind) on the little laptop if I need to refer to them in class.
  • For my wife (in the same class) and classmates who need to refer to my notes, I synch the big laptop with a private section of one of my websites. Yes, some other documents are copied there, but I can get to them without routing myself through a server somewhere.

(I also use the little laptop at work, where I synchronize a directory of meeting notes with my workstation/server on my desk.)

This works wonderfully, so far. I've recommended it to a co-worker who uses Windows, since it can work easily across shared directories (if he doesn't want to use ssh).

Posted by ronlusk at 01:11 PM

May 04, 2004

Linux Distro Differences

Wanted to remember this one: The Difference Between GNU/Linux Distributions
Posted by ronlusk at 07:51 AM | Comments (0)

May 03, 2004

Zapping Annoyances

For Linux users, Zap Annoyances on the Web, Part I and its follow-up at Zap Annoyances on the Web, Part II rely on caching and local proxy servers to intercept and get rid of annoying banners.
Posted by ronlusk at 12:49 PM | Comments (0)

April 21, 2004

October 25, 2003

Firebird and Klipper

The code posted in Mozilla and Klipper appears to work with Firebird, too. Set up firebird as an appname and change the script, and it works. Oddly, perhaps, if you have Firebird running, the "Open With Mozilla" script detects it and opens a new tab in Firebird rather than starting Mozilla.
Posted by ronlusk at 01:30 PM | Comments (0)

October 24, 2003

Mozilla and Klipper

Klipper is a clipboard tool that is provided with the KDE Desktop (at least on SuSE Linux, my distribution). It not only holds onto the last n clipboard contents, but also parses the current content. If the content matches a known pattern, Klipper pops up a menu that allows you to execute a command on the selected text.

Klipper comes configured with a command that proposes starting the Konqueror web browser when it encounters an HTTP URL. The command uses the KDE framework to locate a running copy of Konqueror and instructs it to open the page; if Konqueror isn't running, the command starts a fresh copy.

I like to use Mozilla, and wanted the same CPU-saving capability. So after looking around, I found the -remote option for Mozilla, and created the following script, that I added to the Klipper configuration as another action when a URL is encountered.

(mozilla -remote "ping()" && \
mozilla -remote "openUrl($1,new-tab)") || \
mozilla $1 &
Klipper is configured to call this script with the URL as the first parameter, as in
/home/user/scripts/openMozilla %s
The first line “pings” Mozilla to see if it is running already; if so, the second line is executed to tell the running program to open the URL in a new tab. If Mozilla isn't running yet, the third line starts it, pointing to the URL.

I could have included this script as a single line of code in the Klipper configuration, except that Klipper passes the URL as

'http://www.w3.org/'
and Mozilla gets upset by the extra single quotes.
Posted by ronlusk at 12:02 PM | Comments (0)