Debian Package of the Day (static archived copy)

Article submitted by James Cameron. Please help DPOTD by submitting good articles about software you like!

If you’ve been reading debaday for a while, chances are you now have a heap of packages installed that you’ve tried out, some of which you want to keep installed, and some of which you have forgotten.

Those forgotten packages can cost you. They might contain setuid binaries or running daemons that could compromise your security. Future upgrades take longer, and there will be more to download. Your backups have grown. You are doing backups, right? That’s what duplicity was for.

If you’ve been using aptitude to install packages, then the autoclean command will remove a few things you don’t want.

An example using deborphan on a Debian Etch system:

# deborphan
libssl0.9.7
libstdc++5

In this example, deborphan discovered two fairly small packages that were superfluous, and so they can be removed. Since deborphan gives you only package names in this mode, they can be used in a command:

# aptitude purge `deborphan`

Or if you are in the habit of working at a level below aptitude:

# dpkg --purge `deborphan`

Recursive Deborphan

What you remove may reveal something else as newly orphaned. You
might even want to do this recursively:

#!/bin/sh
while [ -n "`deborphan`" ]; do
    deborphan
    echo
    aptitude purge `deborphan`
done

You might call that a deborphan-recursive script.

Not Just Libraries

Deborphan gives you a list of packages installed on your system that no other package depends on. By default this will only show library packages. But it can also be told to give you a list of all such packages using the -a flag … consider this compound command:

# dpkg-query -W --showformat='${Installed-Size} ${Package}\n' \
`deborphan -a | awk '{print $2}'` | sort -rn

You might name this deborphan-by-size. What it does is give you a sorted list of packages that nothing depends on, in descending order by package size. Work down from the top, choosing what you do not want installed. But remember that every time you remove a package, the list might change because the removed package might have dependencies that are now orphaned.

What’s the worst that could happen? You might remove some package that you will have to reinstall and reconfigure. You have backups. [If you have installed any applications without using the Debian packaging system, it's a different story. You will have to manually make sure that the packages deborphan removes are not required for them to operate. Ed.]

Or how about a small system where you have removed as much as possible so that you have the most space available for your own stuff? Deborphan can be used step by step to challenge you to remove packages you don’t really need. Eventually dpkg, apt-get, or aptitude will warn you about removing a package it considers essential. That’s where you stop, unless you really want some trouble.

Sometimes library packages are installed just to satisfy some dependency of a package in development. Developers will find deborphan’s decisions cause them to reinstall packages they need for software development … an answer to that is to either create a metapackage that refers to the packages you need, or to do your package building in a pbuilder.

Deborphan is available in Debian and Ubuntu since before time began.

Article submitted by Chris Camacho. We are running out of articles! Please help DPOTD and submit good articles about software you like!

Wget is so flexible you’ve probably been using it for years without knowing it, many scripts use it because it is a boilerplate method of grabbing files, it will even automatically retry under certain circumstances…

Probably the best compliment I can pay to it, is that script writers can use it and then forget about it.

It’s one of those great tools that makes *nix so great, its simple, it does what it says on the tin, and like many other pieces of unix “glue”, it’s robust.

Another cool thing about wget is that it’s non-interactive, which means you can start a new download, disconnect from your current session and find your downloads the next time you connect again.

How to use it

The most simple way to invoke wget is by typing wget URL/fileName

$ wget https://debaday.debian.net/contribute/

If you typed in this command you now have a file called index.html in the directory you where in while typing. This file contains the contribute page of the Debian package of the day blog. Read It, DPOTD needs you ;)

Get a directory hierarchy

To get the full content of a directory and his subdirectories you will need to specify wget that it should download your URL recursively. To do this you will need to add the -r option:

$ wget -r https://debaday.debian.net/

This command will generate a local mirror of the debaday blog. Note that wget respects the robots.txt file by default, if it exists. This means it will not download directories and files excluded by the robots.txt file.

Multiple URLs

Wget supports multiple URLs. You can either specify them in a file (one URL per line) or specify them in the command line (space separated).

$ wget url1 url2 ... urlN

or specify the URL container with the -i option

$ wget -i filePathAndName

Other options

Wget has a lot more options, you can for instance use:

• -l for how deep the recursive download should go, the default depth is 5.
• -c is invaluable as it allows you to continue an interrupted download
• -O let’s you specify a target output file (-O fileName)

There are plenty other options in wget, the best way to know them is to read it’s rich man page. For those vista refugees amongst you try typing the following into a terminal ;)

$ man wget

Availability

As wget is part of the GNU project we assume it’s part of most Linux distribution. Never the less, official Debian and Ubuntu package are available:

• Debian: stable, old stable, testing and unstable
• Ubuntu: dapper, edgy, feisty and gutsy.

Community & developers

GNU wget is currently being maintained by Micah Cowan. The original author of GNU Wget is Hrvoje Nikšić.

Links

• Introduction to GNU wget
• Documentation
• Wget wiki
• Wget on the Wikipedia

Article submitted by François Ingelrest. We are running out of articles! Please help DPOTD and submit good articles about software you like!

What is an Audio Player?

Aren’t you tired of those audio players with billions of useless features that clutter up their graphical interface? I am. Most of the time, the player looks good on paper, but when I’m faced with the interface, I don’t even know where to start in order to play my music. There are a lot of buttons, lists, combo boxes a bit everywhere and their usage is not so intuitive.

Above all, the worst thing about these features is that I really don’t need them. Do you really rate your tracks? I don’t. I know my music collection, I know what I like and what I don’t. Do you really need those fancy smart playlists? I don’t. When I want to listen to my music, I want to listen to an entire album, not just a track from here and another one from there.

A Solution Exists

Decibel Audio Player is a simple and nice audio player for the GNOME desktop. GNOME was precisely designed to be simple and nice, so why bother with a complex player and a cluttered graphical interface? Decibel Audio Player follows as closely as possible the GNOME Human Interface Guidelines, here is how it looks like:

Another key aspect of this player is its modular design that allows users to enable only the features they want. The core of the player is very small and simple, and features are implemented in their own module that can be enabled/disabled as needed through the preferences:

So if you don’t want a feature, you simply disable the corresponding module and it gets out of your way. Decibel Audio Player is built on top of GStreamer, and currently supports MP3, Ogg Vorbis and FLAC formats. Its interface is intuitive, and tracks may be dragged, reordered, removed… as needed. Although it is still a young project, current features are correctly implemented and simply work as they should.

Availability

Decibel Audio Player is available in Debian unstable/testing and Ubuntu Gutsy Gibbon repositories (package decibel-audio-player). A more recent version, with music library support, is also available from the website. If you’re looking for a simple audio player, give it a try. You won’t be disappointed.

Links

• Homepage
• More screenshots

Article submitted by Markus Grunwald. We are running out of articles! Please help DPOTD and submit good articles about software you like!

Debian currently offers 84 packages that mention “calculator” (apt-cache search calculator | wc -l). An alternative to the more eye-candy-heavy calculators on offer is bc. bc has no fancy GUI, but is instead console-based and allows a simple mathematical language, making it really fast and programmable.

Everyday use

By default, bc behaves quite unhelpfully: zero digits and no predefined functions such as sin, cos, log etc. You can remedy this by calling bc with the -l parameter, so that it uses the standard math library. Instead of always typing “bc -l”, you can use the fact that bc reads the environment variable BC_ENV_ARGS. So for all the bash guys: add an export BC_ENV_ARGS=-l to your ~/.bashrc.

Let’s try the everyday use first: some simple calculations. Just type your expression and complete it with the enter key:

$ bc
bc 1.06
Copyright 1991-1994, 1997, 1998, 2000 Free Software Foundation, Inc.
This is free software with ABSOLUTELY NO WARRANTY.
For details type `warranty'.
850*77.1        <enter>
65535.0

That’s better than Excel 2007! And all of you terminal people will get this answer quicker than every KDE or GNOME user who starts their GUI calculator.

Custom scale and functions

The scale (total number of decimal digits after the decimal point) of bc’s results is 20 when you run it with “-l”, but you can set it up to a maximum of 2147483647. So to get a quite good estimate of π, just type:

scale=200  <enter>
4*a(1)     <enter>
3.141592653589793238462643383279502884197169399375105820974944592307\
81640628620899862803482534211706798214808651328230664709384460955058\
223172535940812848111745028410270193852110555964462294895493038196

Impressive. But what’s “a(1)”? bc uses very short names for the trigonometric functions. a(x) is the arctangent of x. For all of you who like longer names, you can define your own function and write it in a file. Add this file to your BC_ENV_ARGS and it will be read every time you start bc. The best thing to do is to get extensions.bc from http://x-bc.sourceforge.net/extensions_bc.html.

My BC_ENV_ARGS looks like this:

$ echo $BC_ENV_ARGS
-q -l /home/gru/.bc/extensions.bc

extensions.bc defines several functions by the names people normally know them, for example:

define sin(x)
{
        return (s(x))
}

This way, you can define your own function. I have to convert ratios to decibel and use this quite often:

define db(x,y)
{
        return 20*log(x/y);
}

Converting to other bases

bc can convert from and to arbitrary number bases. Say you want to convert from decimal to hexadecimal. Set the output base obase to 16:

obase=16 <enter>
10
A

And of course, you can calculate with these numbers too:

10/7
1.6DB6DB6DB6DB6DB6B

You can set the input base too:

ibase=16 <enter>
A/2
5
sin(AFFE) <enter>
-.B1F4021654E454E72

I leave the interpretation to the user ;)

Use in shell scripts

Bash is a nice shell, but when it comes to complex calculations $(()) won’t do the job. bc can help:

$ a=$( echo "l(1024)/l(2)" | bc )
$ echo $a
10

Conclusion

I searched quite long for the ideal calculator for me. After trying several GUI calculators, I didn’t find mine. Then I tried bc and it’s perfect for me. I hope you will like it too.

bc is available from Oldstable to Unstable in Debian and Ubuntu Dapper, Edgy, Feisty and Gutsy.

Links

• Homepage
• Online manual
• extensions.bc
• Wikipedia entry for bc programming language
• X-Bc – a graphical user interface to Bc

Article submitted by Matt Clare. We are running out of articles! Please help DPOTD and submit good articles about software you like!

Synopsis

hpodder is a command-line podcast client that just works. The command arguments are simple to master and allow for flexibility in downloading.

In depth

So you’ve have been desperate for a podcast client that rivals the ease and efficiency of iTunes (but have found Linux clients wanting) then hpodder may be the answer to your quest. Having found other clients to be a little jack-of-all-trades for your needs or else, just plain buggy, hpodder is a breath of fresh air.

Once installed, typing hpodder in the terminal will get things under way. The initial run will allow you to make a directory for access your podcasts from and give you the option to subscribe to the sample feeds. If you pass and just want to get onto subscribing your own feeds type:

$ hpodder add http://yourfavouritesite.com/podcastfeed.xml

This will add the feed to the hpodder database and assign an ID that allows for easy specification and manipulation of a feed later on. Hpodder will handle any valid podcast feed, even if it is not the traditional .xml address —including feeds extracted from an OPML export of iTunes podcasts. Once you have your list of podcasts you wish to download, simply type:

$ hpodder update
$ hpodder download

Or, if you want to make life easy:

$ hpodder fetch

As this does effectively the same job as the two commands combined. If you are using hpodder to listen to a podcast that you have perhaps been listening to on another machine, and want to download only the most recent episode, you can do this by typing:

$ hpodder catchup
$ hpodder fetch

If you want to only download or update specific podcasts, you can do so by specifying the podcast’s ID after the command (update, download, fetch, catchup). If you want to delete a podcast type:

$ hpodder rm ID

Where ID is the podcast’s ID number (else you risk deleting them all!).

Downloading the actual episode of your podcast is handled by cURL and is fast, efficient and if interrupted for whatever reason, can be picked up later on without having to start the download from the beginning.

If you want to make hpodder fetch on a regular basis, you can add it to your cron jobs and set it to run as frequently as you like.

If you want a full list of the commands and more subtle features available to hpodder users you can download the manual in PDF format, or read it online.

Any downsides?

The only real downside is that you need to use a second program to listen to your downloaded podcast episodes, but then most people have a favourite media player (kaffeine, mplayer, totem, vlc et al) that they use given the chance.

hpodder is available in Debian since Etch, and in Ubuntu since Feisty