To start eldoc mode, add the following to your .emacs file

;;Turn on documentation in elisp mode
(add-hook 'emacs-lisp-mode-hook
          '(lambda ()
	     (turn-on-eldoc-mode)))

c-eldoc mode is eldoc mode for c. It displays the arguments of C functions while one is programming.

To add c-eldoc mode, get the latest c-eldoc.el, copy it in your load-path and then add then following to your .emacs file.

(add-hook 'c-mode-hook 'c-turn-on-eldoc-mode)

C-c C-a pkgbuild-tar
C-c C-b pkgbuild-makepkg
C-c C-e pkgbuild-etags
C-c RET pkgbuild-update-md5sums-line
C-c C-r pkgbuild-increase-release-tag
C-c C-u pkgbuild-browse-url

I especially like pkgbuild-update-md5sums-line – very useful.

It also has some other cool functions. Check it out here.

My fork of the same is available here.

1) This requires the XMPP console so start that plugin. It comes by default in Pidgin and can be enabled via Tools->Plugins.

2) Start the XMPP console. Tools -> XMPP Console.

For more information on the upcoming steps, check the documentation linked above.

3) Send the following XML in the XMPP console.

<iq type='get' to='gmail.com'>
  <query xmlns='http://jabber.org/protocol/disco#info'/>
</iq>

4) Now get your current status list. Change username to your gmail username before sending

<iq type='get' to='username@gmail.com' id='ss-1' >
  <query xmlns='google:shared-status' version='2'/>
</iq>

This list can be pretty big if you have many saved messages. The output should be something like.

<iq to='username@gmail.com/89089080' from='username@gmail.com' id='ss-1' type='result'>
	<query xmlns='google:shared-status' status-min-ver='2' status-max='512' status-list-max='3' status-list-contents-max='5'>
		<status/>
		<show>dnd</show>
		<invisible value='false'/>
	</query>
</iq>

Observe the invisible value there. That is what we will be manipulating to the our status to invisible.

5) Now, to set your status to invisible, send the following in the xmpp console.

<iq type='set' to='username@gmail.com' id='ss-2'>
  	<query xmlns='google:shared-status' version='2'>
		<invisible value='true'/>
       </query>
</iq>

Although, the pidgin status does not change, you are infact invisible. You can check in gmail or ask someone on your chat list

6) To get out of invisibility, send the following.

<iq type='set' to='username@gmail.com' id='ss-2'>
  <query xmlns='google:shared-status' version='2'>
    <status/>
	<show>dnd</show>
	<invisible value='false'/>
</query>
</iq>

Here the invisible line is important. Other lines are a copy paste of the result of your status list with some minor changes.

This can be a bit tedious, but if anyone is ready to make a plugin out of it, the currently non working one linked above would be a good start.

Enjoy the power of invisibility.

Special thanks to Amod for helping me to test this.

1) Group Blocker

This plugin allows you to block an entire group in plugin. Nice if you have a group of Irritating people in your chat list.

An Arch package for this is available on AUR here.

2) Away on Lock

Pidgin has a feature to set your status to away after a while if idle. However, this feature is known to be quite buggy especially on the Windows version of Pidgin.

The following plugins can be used for the same –

For Linux – Away on Lock. An arch package for the same can be found on the AUR here.

For Windows – Lock ‘n’ Roll

Lock ‘n’ Roll does not change the status if it is busy, invisible or already away. I see this as a bug in the busy case. The code is quite simple to understand and edit and also has documentation. If anyone has a pidgin windows devel environment, that can be changed easily.

Other Third Party Pidgin Plugins.

Since this is the first post of this year, Happy New Year people

One or two days after writing the article about amarok1 installation on Arch, a new version of amarok1 with ruby1.9 support was uploaded. As most of the ruby scripts for amarok1 still work only with ruby1.8, I decided to keep my old version. Today, while I was looking through a few bash scripts, I realized that using #! in the needed ruby scripts was a better option. So, now I use the  AUR package and just change the hashbang lines in my scripts to point to ruby1.8 exec. If only I had realized this earlier

Anyways, more info about hashbang here.

wget has four levels of verbosity which are described as below in its man page –

-d --debug
Turn on debug output, meaning various information important to the developers of Wget if it does not work properly. Your system administrator may have chosen to compile Wget without debug support, in which case -d will not work. Please note that compiling with debug support is always safe---Wget compiled with the debug support will not print any debug info unless requested with -d.

-q --quiet
Turn off Wget's output.

-v --verbose
Turn on verbose output, with all the available data. The default output is verbose.

-nv –no-verbose

Turn off verbose without being completely quiet (use -q for that), which means that error messages and basic information still get
printed.

The problem with these is that the verbose mode prints too much info, quiet prints nothing and non-verbose prints info which is not useful in any way.

By default the non-verbose mode prints the output as follows –

$ wget -nv ftp://ftp.gnu.org/gnu/wget/wget-1.12.tar.bz2
2009-11-26 22:16:40 URL: ftp://ftp.gnu.org/gnu/wget/wget-1.12.tar.bz2 [1609032] -> "wget-1.12.tar.bz2" [1]

I wanted the non-verbose mode to print a progress bar for reasons which are similar to those mentioned in this bug report. The only difference being that instead of Gentoo, I use Arch and pacman. So, I took the diff provided in the bug report modified it a bit to work with v1.12 and also made a PKGBUILD for my Arch use. I am attaching both of them below. After using this patch the output with the -nv option is as follows –

$ wget -nv ftp://ftp.gnu.org/gnu/wget/wget-1.12.tar.bz2
38% [=============> ] 614,880 79.5K/s eta 13s

Problem solved

Patch — Save this as wget_nv_pbar.diff in the same folder as the PKGBUILD if you are building in Arch using the PKGBUILD given below. If you use normal patch, you can do it however you want

#Created by Chinmay Kamat <chinmaykamat@gmail.com>
--- wget-1.12.orig/src/main.c	2009-09-22 08:33:11.000000000 +0530
+++ wget-1.12/src/main.c	2009-11-26 12:21:15.156859000 +0530
@@ -1140,7 +1140,7 @@

   /* Initialize progress.  Have to do this after the options are
      processed so we know where the log file is.  */
-  if (opt.verbose)
+  if (!opt.quiet)
     set_progress_implementation (opt.progress_type);

   /* Fill in the arguments.  */

--- wget-1.12.orig/src/progress.c	2009-09-22 08:19:32.000000000 +0530
+++ wget-1.12/src/progress.c	2009-11-26 12:53:11.309295000 +0530
@@ -291,7 +291,7 @@
          been retrieved.  12.8% will round to 12% because the 13% mark
          has not yet been reached.  100% is only shown when done.  */
       int percentage = 100.0 * bytes_displayed / dp->total_length;
-      logprintf (LOG_VERBOSE, "%3d%%", percentage);
+      logprintf (LOG_NOTQUIET, "%3d%%", percentage);
     }

   {
@@ -325,7 +325,7 @@
           wgint bytes_sofar = bytes_displayed - dp->initial_length;
           double eta = dltime * bytes_remaining / bytes_sofar;
           if (eta < INT_MAX - 1)
-            logprintf (LOG_VERBOSE, " %s",
+            logprintf (LOG_NOTQUIET, " %s",
                        eta_to_human_short ((int) (eta + 0.5), true));
         }
     }
@@ -333,10 +333,10 @@
     {
       /* When done, print the total download time */
       if (dltime >= 10)
-        logprintf (LOG_VERBOSE, "=%s",
+        logprintf (LOG_NOTQUIET, "=%s",
                    eta_to_human_short ((int) (dltime + 0.5), true));
       else
-        logprintf (LOG_VERBOSE, "=%ss", print_decimal (dltime));
+        logprintf (LOG_NOTQUIET, "=%ss", print_decimal (dltime));
     }
 }

@@ -652,7 +652,7 @@
   create_image (bp, dltime, true);
   display_image (bp->buffer);

-  logputs (LOG_VERBOSE, "\n\n");
+  logputs (LOG_NOTQUIET, "\n\n");

   xfree (bp->buffer);
   xfree (bp);
@@ -1071,8 +1071,8 @@
 display_image (char *buf)
 {
   bool old = log_set_save_context (false);
-  logputs (LOG_VERBOSE, "\r");
-  logputs (LOG_VERBOSE, buf);
+  logputs (LOG_NOTQUIET, "\r");
+  logputs (LOG_NOTQUIET, buf);
   log_set_save_context (old);
 }

--- wget-1.12.orig/src/retr.c	2009-09-04 22:01:54.000000000 +0530
+++ wget-1.12/src/retr.c	2009-11-26 12:21:15.167860000 +0530
@@ -234,7 +234,7 @@
   if (flags & rb_skip_startpos)
     skip = startpos;

-  if (opt.verbose)
+  if (!opt.quiet)
     {
       /* If we're skipping STARTPOS bytes, pass 0 as the INITIAL
          argument to progress_create because the indicator doesn't

PKGBUILD

# $Id: PKGBUILD 52882 2009-09-23 03:39:20Z allan $
# Maintainer: Allan McRae <allan@archlinux.org>
# Contributor: Judd Vinet <jvinet@zeroflux.org>
# Modified By:  Chinmay Kamat <chinmaykamat@gmail.com>
pkgname=wget
pkgver=1.12
pkgrel=1
pkgdesc="A network utility to retrieve files from the Web"
arch=('i686' 'x86_64')
url="http://www.gnu.org/software/wget/wget.html"
license=('GPL3')
groups=('base')
depends=('glibc' 'openssl')
optdepends=('ca-certificates: HTTPS downloads')
backup=('etc/wgetrc')
install=wget.install
source=(ftp://ftp.gnu.org/gnu/$pkgname/$pkgname-$pkgver.tar.gz)
md5sums=('141461b9c04e454dc8933c9d1f2abf83')

build() {
  patch -Np0 -i $startdir/wget_nv_pbar.diff || return 1
  cd ${srcdir}/$pkgname-$pkgver
  ./configure --prefix=/usr --sysconfdir=/etc
  make || return 1
  make DESTDIR=${pkgdir} install
}

Get the wget.install from here.

yaourt -S amarok1

and you are done. This will get Amarok 1.4 up and running on your system. But making it work is a bit of a problem. As the binaries are installed in a different directory viz. /opt/amarok/ you will need to add that path to your PATH varible. This will make sure amarok starts without the amarokapp error but the collection scanner does not seem to work. And without a collection there is no way I am going to use amarok. I tried to find various ways around this problem (I had Amarok 2 installed the first time this was done, tried installing without it on the system, adding the /opt/ path at the start of $PATH ) but they did not work so I decided to edit the PKGBUILD file provided on the AUR page. Few of the other problems that I have solved here and my assumptions are mentioned below –

• I removed the /opt/amarok prefix so now the amarok1 pkg will conflict amarok 2.2 — only one can be installed at a time.
• I found that ‘ruby’ is included as a dependency in the PKGBUILD where as ruby1.8 is a build dependency. Amarok uses ruby1.8 so I modified the PKGBUILD to set ruby1.8 (which is compiled from AUR) as a dependency. (EDIT — Just as I finished this article, it seems the AUR PKGBUILD has added the patches to compile Amarok1 with ruby1.9 … but the scripts still require ruby1.8 … you always have a choice )
• I also modified the ruby1.8 PKGBUILD to remove the postfixes used there. This assumes that you do not use ruby 1.9. All Amarok scripts that I know are either ruby1.8 based or pyqt3 (handled later). So without this, those scripts fail.
• Due to this bug in the current pyqt3 package, I had to to install pyqt3 from source. ( This is insanely simple due to ABS).

I am attaching the modified PKGBUILD below. Use with care

• Amarok1 PKGBUILD (With my modifications)

# Contributor: Juanma Hernández <juanmah@gmail.com>
# Modified by Chinmay Kamat <chinmaykamat@gmail.com>
pkgname=amarok1
_origname=amarok
pkgver=1.4.10
pkgrel=7
pkgdesc="amaroK - a media player for KDE - v1.4.10 xine engine with wikipedia patch (single package)"
arch=("i686" "x86_64")
url="http://amarok.kde.org"
license=('GPL')
depends=('xine-lib>=1.1.3' 'libmysqlclient>=5.0.45' 'postgresql-libs>=8.2.4' 'ruby1.8' 'tunepimp>=0.5.3' 'kdelibs3>=3.5.7' 'libgpod>=0.6.0'  'libifp>=1.0.0.2' 'libnjb>=2.2.5' 'libvisual>=0.4.0')
makedepends=('pkgconfig' 'kdelibs3' 'libmtp>=0.2.6.1' 'xine-lib>=1.1.8' 'sdl>=1.2.12' 'ruby1.8')
conflicts=('amarok')
provides=('amarok1')
install=amarok.install
source=("ftp://ftp.solnet.ch/mirror/KDE/stable/amarok/${pkgver}/src/${_origname}-${pkgver}.tar.bz2"
	"amarok_addaspodcast.desktop"
        "wikipedia.patch"
        "libmtp.patch"
        "gcc44.patch"
        "amarok-1.4.10.coverfetcher.diff")
options=(!makeflags)

build() {
  patch -Np0 -i $srcdir/wikipedia.patch || return 1
  patch -Np0 -i $srcdir/libmtp.patch || return 1
  patch -Np0 -i $srcdir/gcc44.patch || return 1
  patch -Np0 -i $srcdir/amarok-1.4.10.coverfetcher.diff || return 1

  cd ${startdir}/src/${_origname}-${pkgver}
  . /etc/profile.d/qt3.sh
  . /etc/profile.d/kde3.sh

  # prepare build
  RUBY=/opt/ruby1.8/bin/ruby  ./configure \
  	--with-gnu-ld \
	--enable-mysql \
	--enable-postgresql \
	--with-mp4v2 \
	--with-ifp \
	--with-libnjb \
	--with-libmtp \
	--with-libgpod \
	--without-arts \
	--without-gstreamer \
	--with-xine \
	--without-nmm \
	--without-mas \
	--with-libvisual \
	--disable-debug \
	--enable-debug=no \
	--without-xmms || return 1

  # build
  sed -i -e 's/konquisidebar//g' amarok/src/Makefile
  make || return 1
  make DESTDIR=${startdir}/pkg install || return 1

  mkdir -p ${startdir}/pkg/opt/kde/share/apps/konqueror/servicemenus/
  install -m644 ${startdir}/src/amarok_addaspodcast.desktop ${startdir}/pkg/opt/kde/share/apps/konqueror/servicemenus/
}
md5sums=('3bdbf26181bf5e5925f48968caba7ac2'
         '7736a7e3e22b15818180df3cee7e4a79'
         'a93eb0dad2078afa045a69f38256b369'
         '02eb286127594b9427d27da7eb01f0de'
         'cfa9310dec03c33048e5a067a0334266'
	 'f4e8754cde659342bdfd3fca36612bb5')

• Ruby1.8 PKGBUILD (With my modifications)

# Contributor: v01d
# Modified by Chinmay Kamat <chinmaykamat@gmail.com>
pkgname='ruby1.8'
pkgver=1.8.7_p174
_pkgver=1.8.7-p174
pkgrel=1
pkgdesc="An object-oriented language for quick and easy programming"
arch=('i686' 'x86_64')
license=('custom')
url="http://www.ruby-lang.org/en/"
depends=('gdbm>=1.8.3' 'db>=4.7.25' 'openssl>=0.9.8k' 'zlib>=1.2.3.3' 'readline>=5.2.013')
makedepends=('tk>=8.5')
provides=('ruby=1.8.7')
source=(ftp://ftp.ruby-lang.org/pub/ruby/stable/ruby-${_pkgver}.tar.bz2)
options=('!emptydirs')
depends=('gdbm>=1.8.3' 'db>=4.7.25' 'openssl>=0.9.8k' 'zlib>=1.2.3.3' 'readline>=5.2.013')
optdepends=('tk: for Ruby/TK')

build() {
  cd "${srcdir}/ruby-${_pkgver}"
  ./configure --prefix=/opt/ruby1.8 --enable-shared --enable-pthread || return 1
  make || return 1
  make DESTDIR="${pkgdir}" install || return 1

  install -Dm644 COPYING "${pkgdir}/usr/share/licenses/${pkgname}/COPYING" || return 1

  mkdir -p $pkgdir/usr/bin
  for i in erb irb rdoc ri ruby testrb; do
    ln -s /opt/ruby1.8/bin/$i $pkgdir/usr/bin/$i
  done
}
md5sums=('88c45aaf627b4404e5e4273cb03ba2ee')

• Finally, the steps to install pyqt3 from source –

Get the PKGBUILD for it from here. Then to install do the following

$makepkg -s PKGBUILD

This should work if you have sudo. If you use su, do the following

#makepkg --asroot -s PKGBUILD

This creates the pyqt3*.tar.gz pacakge to
To install the package –

#pacman -U pyqt3*.tar.gz

And you are done. With all this, I was able to use the wiki-lyrics, copycover,lyricsID3 etc scripts — Amarok 1.4 will all bells and whistles

<RANT> — After using Amarok 2,  Sonata+MPD, qtmpc+MPD, Banshee, Exaile, Listen etc etc … I still feel that Amarok 1.4 is the best player available. Amarok 2.2 is good and Amarok 2.3 will be better. I will shift full time to the 2.x series soon… until then Amarok 1.4 FTW !!! </RANT>

The Why – Well, after using the same distro for about 3 years, I wanted to try something different. With so many distros around, there is hell lotta choice – Debian, Suse, Fedora, Arch etc. I went for Arch coz it has all the things that I was looking for – stability, option to go bleeding edge, a big user community with an active forum, a good package manager and something I have not tried before and coz one of my friends just would not let me rest without trying it . Although, all this might seem stupid , the bottom line is I am now an Arch Linux user.

The Review - How is Arch different from / better or worse than Ubuntu ?

The Basics – The Arch Way vs The Ubuntu way

While Ubuntu is a linux distro with a release every six months, Arch is a rolling release. Ubuntu’s six month cycle has had its advantages but sometimes it also has had major disadvantages but keeping it in sync with the GNOME release cycle has been good for it. There has been times however when I wanted to move to a new package but no matter what I did, I could not satisfy the dependencies ppa or no ppa. I ended up using beta/alpha repos of the next release (case in point – my amarok 2.2 install guide).  Keeping packages of 2 version in Ubuntu is not very pleasant . With Arch, being a rolling release, this is avoided as packages are available as they are available upstream either in the main repository or in testing.

However, this is not the major difference between Arch and Ubuntu. Ubuntu believes in giving a working desktop out of the box ( and yet there have been times when LTS releases fail to start the X server ). On the other hand, the Arch way says that you only get a bare bones system and you need to make it the way you want. This translates to installing / configuring everything including X, GNOME (or any other desktop environment). This is where Arch’s user base is decided. If you want to go into the details, Arch is for you. If you don’t – stay away. I think with average experience with Linux, it should not take more than 6/7 hrs to get Arch up and running like Ubuntu is after install. Afer the first 6/7 hrs (which may or may not be a pain) what you get is a great distro. The 6/7 hrs are usually pretty easy to get through if you follow the Arch wiki and the Beginner’s guide. It took me around 4 hrs to get things up and running including some experimentation with available ati drivers.

The package management –

The one thing that I really liked about Ubuntu was APT – the package manager that Ubuntu uses. With its power and flexibility, it is definitely one of the best package management systems out there. Combine it with PPA’s and you have a bleeding edge system with ease. If I was going to move to Arch as my primary distro, I needed something similar and the combination of pacman and AUR is as good as if not better than APT + PPA. Whatever the other differences may be, pacman does seem faster than apt. Its wicked fast.

Another thing that I appreciate about Arch is the way packages are in the repository — they are vanialla and modular for eg. GNOME is packaged without pulseaudio \m/ and even gnome-termnial for that matter (please install it before starting gnome or you might just end up with system with no access to the root terminal ). In fact all you get when you install gnome are the following packages -

#pacman -S gnome
gnome package not found, searching for group...
:: group gnome (including ignored packages):
epiphany gnome-applets gnome-backgrounds gnome-control-center
gnome-desktop gnome-icon-theme gnome-media gnome-mime-data gnome-mount
gnome-panel gnome-screensaver gnome-session gnome-settings-daemon
gnome-themes gnome2-user-docs libgail-gnome metacity nautilus
notification-daemon yelp

And you are still given a choice to install the entire group or select each individual package.

If you are unhappy with the modularity of KDE, there is the KDEmod project  which is KDE built especially for Arch  . If you are a person who installs a lot of packages from source, then ABS and yaourt are definitely for you. It builds pacakges from source but integrates them with pacman – nice.

The great thing here is – I do not miss APT.

Help available – The great thing about Ubuntu is the help avaiable on the forums. There usually aren’t many question to which you cant find answers to on the ubuntuforums. On the Arch side, the wiki and the forums to-gether are all the help you need.

Who should use what – As I said earlier, if you do not want to do much and want something that runs out of the box, Ubuntu is still the best choice. I will still say that for new linux users Ubuntu rules. But if you are ready to do a bit of config in CLI and/or build your system exactly the way you want, Arch is for you.

Conclusion – Arch is a great distro. Besides the first 4 hrs (which weren’t really a pain but more of a great learning curve) , I have really liked the way this distro operates. Pacman + yaourt are awesome and AUR is quite vast. Arch Linux is definitely going to be my first choice distro for at least the next year . That said, Ubuntu is still the best distro for beginners.

PS – If you are wondering how to get to the terminal if you are in GNOME without gnome-terminal and Ctrl+Alt+F* are not working, add the following to your GRUB kernel line –

# (0) Arch Linux Single-user
title Arch Linux Single-user
root (hd0,0)
kernel /vmlinuz26 root=/dev/sda1 ro 1 #The 1 is the runlevel
initrd /kernel26.img

More info here.

$sudo apt-get build-dep <source-package-name>

I have posted this directly from the APT HOWTO . Check it out for a better understanding of APT’s power.

Installing all the build-deps is easy enough but they are (usually) useless once the comilation job is done. So, can they be removed as easily?

There are few options to do this -

• /etc/apt/apt.conf

APT
{
    // Options for apt-get
  Get
  {
     Build-Dep-Automatic "true";
  };
};

Now whatever packages are installed using apt-get build-dep will be removed when you execute apt-get autoremove.

/usr/share/doc/apt/examples/configure-index.gz is a configuration file showing example values for all possible options.

• Command line option

If you want to/don’t want to remove the build-deps of a particular package, pass the following option while using the build-deps command.

sudo apt-get build-dep -o APT::Get::Build-Dep-Automatic=true

True – To be removable by autoremove, False – otherwise.

• If you want to do it after having used build-dep

Well, so far I have been unable to find anything that can reverse the effect after you have already installed the packages. There are some suggestions here and here but I found that they remove cross dependencies as well ( not good )

1. Using plugins (generally not available for free)
2. Using online generator such as those described here. (Usually give css code which cant be used in free mode of WordPress)
3. Htmlize in Emacs

Emacs has an addon called “htmlize” which can be used to convert any file into an htmlized version. It provides 3 output types – css, inline-css and font. Emacs FTW again .

PS :: This is also available in the emacs-goodies.el packages in Debian and Ubuntu but the version of htmlize.el in those is outdated and does not work with Emacs 23.x

EDIT :: It seems that WordPress internally supports syntax highlighting with the use of SyntaxHighlighter. For more info check here. List of all the supported languages is here. For blogs other than WordPress, htmlize still seems a very good option.